Republic Of South Africa 18-K/A Annual report (foreign governments) (amended)

Exhibit 99.E

Quarterly Bulletin 6HSWHPEHU 2016 South African Reserve Bank

Quarterly Bulletin September 2016 No. 281

© South African Reserve Bank All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without fully acknowledging the Quarterly Bulletin of the South African Reserve Bank as the source. The contents of this publication are intended for general information only and are not intended to serve as financial or other advice. While every precaution is taken to ensure the accuracy of information, the South African Reserve Bank shall not be liable to any person for inaccurate information or opinions contained in this publication. Enquiries relating to this Bulletin should be addressed to: Head: Economic Research and Statistics Department South African Reserve Bank P O Box 427 Pretoria 0001 Tel. +27 12 313 3668/3676 http://www.reservebank.co.za ISSN 0038-2620 Quarterly Bulletin September 2016

Contents Quarterly Economic Review Introduction ................................................................................................................................ Domestic economic developments ............................................................................................ Domestic output ................................................................................................................ Real gross domestic expenditure ....................................................................................... Final consumption expenditure by households................................................................... Final consumption expenditure by general government ..................................................... 1 4 4 8 9 11 Fixed capital formation ....................................................................................................... 12 Inventory investment .......................................................................................................... 13 Factor income ................................................................................................................... 13 Gross saving ...................................................................................................................... 14 Employment ....................................................................................................................... 15 Labour cost and productivity in the formal non-agricultural sector ..................................... 19 Prices ................................................................................................................................. 20 Foreign trade and payments ...................................................................................................... 27 International economic developments ................................................................................ 27 Current account ............................................................................................................... 31 Financial account ............................................................................................................... 37 International investment position ........................................................................................ 40 International reserves and liquidity ..................................................................................... 41 Exchange rates .................................................................................................................. 41 Turnover in the South African foreign-exchange market ..................................................... 43 Monetary developments, interest rates and financial markets .................................................... 44 Money supply..................................................................................................................... 44 Credit extension ................................................................................................................. 47 Interest rates and yields ..................................................................................................... 50 Money market .................................................................................................................... 54 Bond market ...................................................................................................................... 56 Share market...................................................................................................................... 57 Market for exchange-traded derivatives ............................................................................. 60 Real-estate market ............................................................................................................. 61 Non-bank financial intermediaries ...................................................................................... 61 Flow of funds...................................................................................................................... 62 Public finance............................................................................................................................. 65 Non-financial public-sector borrowing requirement ............................................................ 65 Budget comparable analysis of national government finance ............................................. 68 Statement issued by Lesetja Kganyago, Governor of the South African Reserve Bank 21 July 2016 ............................................................................................................................... 75 Statement issued by the South African Reserve Bank ..................................................... 79 Note Note on unit labour cost measurement in South Africa .............................................................. 84 Notes to tables .......................................................................................................................... 89 Statistical tables Contents ................................................................................................................................ S–0 Statistical tables ..................................................................................................................... S–2 Key information ...................................................................................................................... S–146 Quarterly Bulletin September 2016

' ··.. .• .. . .. I· .• ·· ·. .•.-·' I .... . .,..

Quarterly Economic Review Introduction Global economic growth moderated in the second quarter of 2016, following a slowdown in activity in both advanced and emerging-market economies. This outcome is broadly aligned with the International Monetary Fund’s (IMF) marginally lower global growth forecast of 3,1 per cent for 2016 in its July 2016 World Economic Outlook Update. Among advanced economies, the sluggish economic performance of Japan and the euro area was notable. The outcome of the Brexit referendum and potential banking stresses in some European countries further negatively affected the outlook for the United Kingdom (UK) and the euro area. Output growth in emerging-market economies decelerated on account of slower growth in India, a contraction in economic activity in Russia and recessions in several major Latin American economies, with the outlook remaining mixed. The fragile environment prompted the authorities in the United States (US) to keep the federal funds rate unchanged so far this year. More forcefully, central banks in Japan, the euro area and the UK eased policy further in response to heightened global uncertainty and disappointing growth. Expanded quantitative easing programmes and lower policy rates, some even below zero, boosted demand for emerging-market bonds. World trade volumes contracted in the three months to June 2016, with export volumes of both emerging-market and advanced economies declining. However, most international commodity prices recovered somewhat from recent lows registered in January 2016 after an extended period of decline. The price of crude oil, in particular, increased significantly although it remained volatile around the somewhat higher levels amid reduced exploration and uncertainty around potential cooperation among producers in managing crude supplies to the market. In South Africa economic growth turned positive with real gross domestic product (GDP) rising at a brisk annualised pace of 3,3 per cent in the second quarter of 2016, following a contraction in the first quarter. The variations in economic growth from quarter to quarter were mostly the result of fluctuations in the primary and secondary sectors. Real production in the primary sector increased in the second quarter of 2016, following sustained contractions since the second quarter of 2015. The real value added by the mining sector rebounded as platinum production rose substantially alongside more modest increases in manganese ore, iron ore and coal output. By contrast, agricultural production inched lower to record a sixth consecutive quarter of contraction as the drought left its scars on most field crops. In livestock farming real value added was supported by drought-induced culling. Real value added by the secondary sector also accelerated in the second quarter of 2016, as the volume of manufacturing production picked up significantly. Increased production of vehicles buoyed activity as exports benefited from the more competitive exchange value of the rand. In the construction sector activity essentially moved sideways, adversely affected by the steady rise in interest rates, escalating building costs and a lack of new projects. Growth in real value added by the tertiary sector accelerated in the second quarter of 2016. The pace of activity in the commerce sector was supported by an increase in wholesale trade and firm, albeit slower, growth in retail trade. As real disposable income of households continued to edge higher, retail trade held up fairly well despite low consumer confidence, high levels of consumer debt, interest rate increases and poor employment prospects. Freight transport benefited from rising exports and the deployment of new rolling stock by Transnet, while activity in the finance sector accelerated somewhat as banks’ fee income picked up in the second quarter. Viewed from the expenditure side, growth in real GDP in the second quarter emanated primarily from increased real net exports. The contraction in real gross domestic expenditure in the first quarter of 2016 persisted into the second quarter. The increase in real final consumption 1 Quarterly Bulletin September 2016

expenditure by households, after having contracted in the first quarter of 2016, and the slower pace of contraction in real gross fixed capital formation assisted expenditure in the second quarter. Real final consumption expenditure by general government continued to increase at a pedestrian pace, consistent with government’s commitment to restrain current expenditure. The turnaround in real final consumption expenditure by households in the second quarter of 2016 reflected stronger growth in real expenditure on semi-durable goods, non-durable goods and services. Having contracted in the first quarter of 2016, growth in semi-durable goods and non-durable goods consumption turned positive in the second quarter. By contrast, expenditure on durable goods contracted for the sixth consecutive quarter. This turnaround in real final consumption expenditure by households was partly related to growth in real disposable income of households which edged higher, following a marginally faster pace of increase in compensation of employees. Encouragingly, the ratio of household debt to disposable income edged lower. However, as the increases in interest rates of the first quarter worked through the debt-service cost ratio rose to its highest level since the first quarter of 2010. Real gross fixed capital formation registered a third successive quarter of contraction in the second quarter of 2016. Capital outlays by private business enterprises, which account for almost two-thirds of total capital formation, contracted at a slower pace, impeded by low business confidence, excess capacity and subdued demand. Real capital spending by public corporations turned negative in the second quarter of 2016, held back by postponed capital projects and unplanned delays, while government’s capital expenditure contracted for the third consecutive quarter. Real inventories contracted markedly in the second quarter of 2016 as production for the export market was augmented by running down stock levels and the smaller grain crops were reflected in inventories. The further increase in formal non-agricultural employment in the first quarter of 2016 resulted entirely from temporary appointments in the public sector by the Independent Electoral Commission (IEC). Overall employment growth in the public sector remained aligned to government’s commitment to fiscal consolidation. Consistent with the contraction in domestic economic activity, the private sector shed jobs in the first quarter of 2016. Employment losses were broad-based in both the goods-producing and services sectors. The official unemployment rate amounted to 26,6 per cent in the second quarter of 2016, with the youth unemployment rate twice as high. The pace of increase in nominal remuneration per worker remained unchanged at a year-on-year rate of 6,1 per cent in the first quarter of 2016, as an acceleration in the private sector was offset by a moderation in the public sector. Headline consumer price inflation breached the upper limit of the inflation target range of 3–6 per cent from January to June 2016. The marked acceleration in headline consumer price inflation to 7,0 per cent in February 2016 – its highest level in almost seven years – reflected the quickening of food price inflation in response to the severe drought conditions and exchange rate pass-through from the depreciation of the exchange value of the rand. The subsequent moderation in consumer price inflation to 6,0 per cent in July mainly resulted from favourable base effects related to petrol price changes and somewhat lower services price inflation. Given concerns about the expected medium-term inflation trajectory, the Monetary Policy Committee (MPC) increased the policy interest rate in January and again in March 2016. A moderation in inflationary pressures and generally weak domestic economic activity resulted in the policy rate being left unchanged at 7,0 per cent at the subsequent meetings of the MPC. Money-market interest rates remained generally aligned with the policy stance while being influenced by movements in the exchange value of the rand, inflation outcomes and interest rate expectations. Notwithstanding the moderation in global economic activity, merchandise export volumes increased markedly in the second quarter of 2016, benefiting from the lagged effect of the depreciation of the rand and increased external demand. Export values surged across all major categories, especially vehicles and platinum group metals, along with a broad-based increase in international commodity prices. Simultaneously, weak growth in domestic demand and a 2 Quarterly Bulletin September 2016

sustained contraction in capital spending led to a decline in merchandise import volumes, although import values in rand terms increased marginally as the prices of oil and other imported commodities firmed. South Africa’s trade balance consequently switched to a surplus, and coupled with the marginally smaller deficit on the services, income and transfer account, the deficit on the current account of the balance of payments narrowed significantly to 3,1 per cent of GDP in the second quarter of 2016. This smaller shortfall on the current account was financed through net portfolio and direct investment inflows, while other investment registered a net outflow. While South Africa’s foreign assets continued to exceed its foreign liabilities, the net international investment position receded to 13,3 per cent of GDP at the end of March 2016, as the market value of the country’s foreign liabilities increased marginally while that of foreign assets declined. South Africa’s gross gold and foreign-exchange reserves amounted to US$45,7 billion at the end of August 2016, and covered on average 5,3 months’ worth of imports of goods and services in the second quarter of 2016. The external value of the rand on balance strengthened slightly further in the second quarter of 2016 and continued to appreciate markedly in July and early August 2016 along with other high-yielding currencies. Towards the end of August the rand, however, depreciated abruptly in the wake of domestic political events. Year-on-year growth in money supply (M3) decelerated significantly in the first half of 2016. This deceleration was driven by pedestrian growth in corporate deposits, in contrast with consistent double-digit growth in household deposits. Growth in bank credit extension to the domestic private sector moderated somewhat in the second quarter of 2016 with the corporate sector’s demand for credit remaining relatively strong but that of households exceptionally weak. The yield curve flattened during 2016 along with the general decline in South African bond yields. The downward trend was interrupted in May and late-August, mainly by negative perceptions related to political developments which also led to rand depreciation. Share price movements on the JSE Limited (JSE) remained volatile, but on balance the overall price index recorded an increase thus far in 2016. In the first eight months of 2016, turnover in the share and bond markets reached record-high levels. National government’s budgeted plans seemed to be on track, with revenue increasing at a faster pace than expenditure in the first quarter of fiscal 2016/17. Consistent with subdued domestic economic activity, corporate income tax collections increased sluggishly, but to the contrary personal income tax collections increased at a double-digit rate. Revenue from import duties also increased strongly, buoyed by the relatively weaker domestic currency which raised the rand value of imports. The primary balance recorded a deficit of 0,7 per cent of GDP in the first quarter of fiscal 2016/17, demonstrating progress with fiscal consolidation when compared with the value of 1,7 per cent of GDP registered in the first quarter of fiscal 2015/16. Total gross loan debt of national government seemed to be levelling off as a ratio of GDP, while the non-financial public-sector borrowing requirement also narrowed somewhat in the first quarter of fiscal 2016/17. 3 Quarterly Bulletin September 2016

Domestic economic developments 1 Domestic output Growth in real gross domestic product (GDP) in South Africa turned positive in the second quarter of 2016, accelerating to an annualised rate of 3,3 per cent following a contraction of 1,2 per cent in the first quarter. The variations in economic growth between quarters mainly reflected fluctuations in the real value added by the primary and secondary sectors, contrary to that in the tertiary sector that continued at a more even pace over the period. The level of real output in the second quarter of 2016 was 0,6 per cent higher than in the corresponding period in 2015. By contrast, the level of nominal GDP at market prices was 7,3 per cent higher over the same period. Adjusted for inflation, real GDP growth of 3,3 per cent in the second quarter of 2016 translates into nominal GDP growth of 7,9 per cent. 1 The quarter-to-quarter growth rates referred to in this section are based on seasonally adjusted annualised data. Growth in gross domestic product Percentage change 14 quarter to quarter* 12 over four quarters 10 8 6 4 2 0 -2 -4 2011 2012 2013 2014 2015 2016 Excluding the contribution of the usually more volatile primary sector, growth in real GDP picked up from 0,7 per cent in the first quarter of 2016 to 2,7 per cent in the second quarter. Real gross domestic product Quarter-to-quarter percentage change at seasonally adjusted annualised rates 2015 2016 Sector 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Primary sector ................................. Agriculture ................................... Mining.......................................... Secondary sector ............................ Manufacturing.............................. Tertiary sector.................................. Non-primary sector ......................... Total ................................................ 6,3 -11,3 12,7 -0,4 -2,1 1,7 1,2 2,0 -10,9 -20,4 -7,8 -4,9 -6,3 0,8 -0,5 -2,0 -10,8 -11,8 -10,5 2,5 4,7 1,5 1,7 0,3 -0,5 -6,7 1,4 -1,3 -2,5 1,4 0,8 0,4 0,9 -5,9 3,2 0,0 -0,3 1,6 1,3 1,3 -15,5 -6,5 -18,1 0,1 0,6 0,8 0,7 -1,2 8,8 -0,8 11,8 5,3 8,1 2,0 2,7 3,3 Growth in the real value added by the primary sector turned positive in the second quarter of 2016 – the first expansion in real output since the first quarter of 2015. The improved growth performance in the second quarter of 2016 can mainly be attributed to a robust increase in 4 Quarterly Bulletin September 2016 Real GDP Real GDP over f Nominal GDP our quarters * Seasonally adjusted annualised rates

the real value added by the mining sector which more than offset a further contraction in the agricultural sector over the period. The real value added by the agricultural sector declined for the sixth consecutive quarter, contracting at a slower pace of 0,8 per cent in the second quarter of 2016. Adverse weather conditions during the 2015/16 season resulted in lower production of most field crops. By contrast, the real value added by livestock farming held up well, reflecting increased culling throughout the quarter. Commercial maize crop estimates Crop (Million tons) Area planted (Million hectares) 2014/15: Final ...................................................................................... 9,9 2,7 2015/16: Eighth production forecast ..................................................... 7,3 1,9 Source: Crop Estimates Committee of the Department of Agriculture, Forestry and Fisheries Total mining production rose notably in the second quarter of 2016. Having contracted markedly at a rate of 18,1 per cent in the first quarter of 2016, real value added by the mining sector rebounded in the second quarter, rising at an annualised rate of 11,8 per cent and contributing 0,8 percentage points to overall growth in GDP. This notable improvement was predominantly brought about by a substantial increase in platinum production. Simultaneously, production volumes of manganese ore, iron ore and coal increased at a moderate pace, while the production of gold, diamonds and other non-metallic minerals declined over the period. Despite the significant increase in mining production in the second quarter of 2016, the level of real output in the first half of 2016 was 6,5 per cent lower than in the corresponding period in 2015. Activity in the mining sector continues to be restrained by, among other factors, the growing frequency of section 54 safety stoppages resulting in part from deep mining excavation, global oversupply of various mineral groups, low demand, policy uncertainty and an escalation in input costs. Real value added by the mining sector R billions, 2010 prices Percentage change 15 245 240 10 235 230 5 225 220 0 215 210 205 200 -5 -10 195 2011 * Seasonally adjusted 2012 2013 2014 2015 2016 Subsequent to an increase of 0,1 per cent in the first quarter of 2016, growth in the real value added by the secondary sector accelerated to 5,3 per cent in the second quarter. Activity in the manufacturing sector picked up significantly in the second quarter of 2016, while real output of the construction subsector increased only marginally. By contrast, real value added by the electricity, gas and water subsector contracted over the period. 5 Quarterly Bulletin September 2016 Total* Over four quarters (right-hand scale)

Real gross domestic product by main sector Indices: 2011/01 = 100 120 115 110 Secondary ector 105 100 Primary ector 95 90 2011 2012 2013 2014 2015 2016 Real output of the manufacturing sector advanced in the second quarter of 2016. Having increased by 0,6 per cent in the first quarter of 2016, growth in real value added by the manufacturing sector accelerated to a rate of 8,1 per cent in the second quarter, adding 1,0 percentage point to overall GDP growth in the quarter. The resurgence in real manufacturing output came especially from increased production of motor vehicles, parts and accessories and other transport equipment; petroleum, chemical, rubber and plastic products; and food and beverages. The increase in the production of motor vehicles was probably underpinned by firm export demand following the more competitive exchange value of the rand. Increased production in the petroleum and chemicals sector could, inter alia, be linked to the relatively low price of crude oil. The utilisation of production capacity in the manufacturing sector remained low and receded marginally from 82,2 per cent in February 2016 to 82,0 per cent in May – the primary reason for under utilisation being insufficient demand for especially durable manufactured goods. Measured over a year, manufacturing production increased by 3,6 per cent – the highest year-on-year increase recorded since the first quarter of 2011. Physical volume of manufacturing production Indices: 2011/01 = 100 112 110 108 106 otal 104 102 100 Durable goods 98 96 94 2011 2012 2013 2014 2015 2016 Following a contraction of 2,8 per cent in the first quarter of 2016, the real value added by the electricity, gas and water sector declined by 1,8 per cent in the second quarter. Electricity consumption remained lacklustre, affected by relatively mild winter temperatures, higher 6 Quarterly Bulletin September 2016 Seasonally a Non-d urable goods T djusted T ertiary sector s s Seasonally a djusted

electricity tariffs, the generally weak economic environment, a reduction of electricity-intensive production of ferrochromium and aluminium, and the use of alternative energy sources. However, production levels stabilised during the quarter concerned, supported by improved plant performance, the addition of new production capacity and fewer unplanned breakdowns in transmission. Growth in the real value added by the construction sector remained subdued. Real output of the construction sector rose by only 0,1 per cent following a 0,4 per cent increase in the preceding quarter. Activity in the civil construction sector declined, constrained by a lack of new projects while the demand for residential and non-residential buildings picked up somewhat over the period. Rising production costs generally weighed on the construction sector in the second quarter of 2016. Following a slowdown in activity from the third quarter of 2015, growth in the real value added by the tertiary sector accelerated from 0,8 per cent in the first quarter of 2016 to 2,0 per cent in the second quarter. Activity in all four services sectors picked up pace over the period. Supported by an increase in wholesale trade and firm, albeit slower, growth in retail trade, the real value added by the commerce sector made a positive contribution to growth in GDP, adding 0,2 percentage points to overall growth in GDP in the second quarter of 2016. Growth in real output increased from 1,3 per cent in the first quarter of 2016 to 1,4 per cent in the second quarter. Wholesale trade sales in the second quarter of 2016 were supported by higher sales of fuel. Sales of agricultural raw materials and livestock remained lively, whereas sales of consumer-related products were subdued. Despite low consumer confidence, rising inflationary pressures and subdued growth in disposable income of households, retail trade held up fairly well over the period. Output of the catering and accommodation subsector benefited from an uptick in tourism activity over the period. New and used vehicle sales volumes Indices: 2011/01 = 100 240 220 Used vehicle sales 200 180 160 140 Commercial vehicle sales* 120 100 New passenger vehicle sales* 80 Seasonally adjusted 60 2011 Source: NAAMSA* 2012 2013 2014 2015 2016 Affected by the weak sales of especially passenger vehicles, total vehicle sales decreased in the second quarter of 2016. Sales of commercial and used vehicles, however, remained relatively buoyant over the period. As indicated by the accompanying graph, sales of used vehicles, which accounted for roughly 40 per cent of total vehicle sales, increased at a much faster pace over time compared to other components of motor trade sales. This development may be attributed to consumers acting in a more conservative manner due to, among other factors, relatively low levels of economic activity, pressure on household balance sheets, high levels of consumer debt in certain income groups, lower finance approvals, interest rate increases, high fuel prices, above-inflation new vehicle price increases and poor employment prospects. 7 Quarterly Bulletin September 2016

Supported by the weak exchange rate of the rand, vehicle exports recorded firm growth on a quarter-to-quarter basis. After contracting by 2,7 per cent in the first quarter of 2016, growth in real output of the transport, storage and communication sector accelerated to 2,9 per cent in the second quarter, mainly due to increased activity in the freight transport subsector. In the second quarter of 2016, freight transport reflected rising exports and was supported by the deployment of new locomotives and wagons by Transnet, thereby improving traction effort, efficiency and reliability. Passenger transport activity, however, declined over the period. Activity in the finance, insurance, real-estate and business services sector picked up with growth in the sector’s real value added accelerating from 1,9 per cent in the first quarter of 2016 to 2,9 per cent in the second quarter. The stronger performance reflected a recovery in commercial banking activity, particularly fee income, alongside an increase in trading activity in the commodity derivatives market. Following an increase of 1,1 per cent in the first quarter of 2016, growth in the real value added by general government accelerated marginally to 1,2 per cent in the second quarter. The increase in real value added in the second quarter of 2016 could largely be attributed to increased spending on temporary workers employed to assist with preparations for the local government elections. Real gross domestic expenditure Real gross domestic expenditure continued to contract in the second quarter of 2016. Subsequent to a decline of 2,2 per cent in the first quarter of 2016, real gross domestic expenditure receded by 3,3 per cent in the second quarter. The contraction mainly stemmed from a decrease in real gross fixed capital formation and a de-accumulation in real inventory holdings. Real final consumption expenditure by households increased in the second quarter of 2016, having contracted in the first quarter while growth in real final consumption expenditure by general government accelerated over the period. Real gross domestic product and expenditure Percentage change from quarter to quarter 12 10 8 6 4 2 0 -2 -4 -6 -8 -10 2011 2012 2013 2014 2015 2016 The largest contribution to growth in real GDP in the second quarter of 2016 emanated from real net exports, adding 6,7 percentage points to total economic growth. Real gross fixed capital formation subtracted 1,0 percentage point and the de-accumulation in real inventories subtracted 3,2 percentage points from growth in real GDP over the period. 8 Quarterly Bulletin September 2016 Gross domestic product Gross domestic expenditure Seasonally adjusted annualised rates

Real gross domestic expenditure Quarter-to-quarter percentage change at seasonally adjusted annualised rates 2015 2016 Component 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Final consumption expenditure Households ................................................. General government.................................... Gross fixed capital formation .......................... Domestic final demand ................................. Change in inventories (R billions)* ................... Gross domestic expenditure** ...................... 2,0 -1,7 2,6 1,4 68,4 9,2 0,3 1,6 -0,9 0,3 -10,6 -8,5 2,4 0,8 4,6 2,5 -16,9 1,7 2,1 2,6 -2,8 1,2 -4,2 1,9 1,7 0,2 2,5 1,6 9,2 1,7 -1,7 1,2 -10,0 -2,8 1,2 -2,2 1,0 1,3 -4,6 -0,1 -22,7 -3,3 * At constant 2010 prices ** Including the residual Contribution of expenditure components to growth in real gross domestic product Percentage points 2015 2016 Component 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Final consumption expenditure Households .................................................. General government..................................... Gross fixed capital formation ........................... Change in inventories ...................................... Net exports ..................................................... Residual .......................................................... Gross domestic product ................................ 1,2 -0,4 0,5 7,0 -7,0 0,6 2,0 0,2 0,3 -0,2 -10,2 6,9 1,0 -2,0 1,4 0,2 1,0 -0,8 -1,5 0,0 0,3 1,3 0,5 -0,6 1,7 -1,5 -0,9 0,4 1,0 0,0 0,5 0,0 -0,4 0,1 1,3 -1,0 0,2 -2,2 0,7 1,0 0,0 -1,2 0,6 0,3 -1,0 -3,2 6,7 -0,2 3,3 Final consumption expenditure by households Following a contraction of 1,7 per cent in the first quarter of 2016, growth in real final consumption expenditure by households accelerated to an annualised rate of 1,0 per cent in the second quarter. This turnaround reflected stronger growth in real expenditure on semi-durable goods, non-durable goods as well as services. By contrast, real spending on durable goods declined further over the period. Real final consumption expenditure by households Percentage change 6 5 4 3 2 1 0 -1 -2 2011 2012 2013 2014 2015 2016 9 Quarterly Bulletin September 2016 Quarter to quarter * Over four quarters * Seasonally adjusted annualised rates

Real final consumption expenditure by households on durable goods shrunk for the sixth consecutive quarter, contracting at a rate of 7,3 per cent in the second quarter of 2016. The second-quarter contraction was mainly due to declines in expenditure on furniture and household appliances; personal transport equipment; and recreation and entertainment goods. Households, however, stepped up their spending on personal computers and software over the same period. Real outlays in the durable goods category were probably hampered by the steady increase in household indebtedness, amid a somewhat higher interest rate environment and an acceleration in durable goods inflation. Real durable goods and transport equipment purchases by households Indices: 2011/01 = 100 140 130 120 110 100 90 80 2011 2012 2013 2014 2015 2016 Growth in real spending on semi-durable goods accelerated from a decline of 1,2 per cent in the first quarter of 2016 to 2,1 per cent in the second quarter. Apart from a decline in real outlays on household textiles, furnishings and glassware, spending on all other categories of semi-durable goods rose over the period. Real final consumption expenditure by households Quarter-to-quarter percentage change at seasonally adjusted annualised rates 2015 2016 Category 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Durable goods................................................ Semi-durable goods ....................................... Non-durable goods ........................................ Services ......................................................... Total ............................................................... -1,1 8,0 1,5 1,9 2,0 -11,0 -0,5 1,3 2,4 0,3 -3,6 8,9 1,1 3,6 2,4 -5,5 11,1 3,1 1,1 2,1 -2,1 4,0 2,2 1,8 1,7 -12,5 -1,2 -0,9 0,1 -1,7 -7,3 2,1 0,4 3,2 1,0 Having contracted in the first quarter of 2016, real spending on non-durable goods increased by 0,4 per cent in the second quarter. The turnaround was mainly due to accelerated expenditure on household consumer goods, medical and pharmaceutical products, and petroleum products. Declines were, however, recorded in spending by households on food, beverages and tobacco and household fuel and power over the same period. 10 Quarterly Bulletin September 2016 Seasonally a Total durable goods Transport equipment djusted

Real spending by households on services advanced in the second quarter of 2016. Spending in most subcategories increased at a faster pace, with the exception of spending on medical services which moderated and real outlays on transport and household services which declined in the second quarter of 2016. Real expenditure by South African residents abroad increased over the period, probably enticed by the appreciation of the external value of the rand during the quarter. Following a marginally faster pace of increase in compensation of employees, growth in real disposable income of households edged higher from an annualised rate of 0,4 per cent in the first quarter of 2016 to 0,7 per cent in the second quarter. Households continued to incur debt in the second quarter of 2016. However, since quarter-to-quarter growth in nominal disposable income outpaced concomitant growth in household debt, the ratio of household debt to disposable income decreased from 75,7 per cent in the first quarter of 2016 to 75,1 per cent in the second quarter. Consistent with the increases in the prime lending rate in the first half of 2016, the cost of servicing household debt advanced from 9,7 per cent of disposable income in the first quarter of 2016 to 9,8 per cent in the second quarter – the highest debt-service ratio since the first quarter of 2010. Debt-service cost Percentage of household disposable income 18 16 14 12 10 8 6 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 * In addition to interest payments on household debt, required capital repayments on mortgage advances and instalment sale and leasing finance were also taken into account. Final consumption expenditure by general government Having increased at an annualised rate of 1,2 per cent in the first quarter of 2016, growth in real final consumption expenditure by general government inched higher to 1,3 per cent in the second quarter. Real spending on both compensation of employees and non-wage goods and services increased marginally over the period, consistent with government’s commitment to restrain current expenditure. Additional salaries and other expenses were mainly associated with increased spending by the Independent Electoral Commission (IEC) as the local government elections approached. As a result, the level of real spending by general government in the first six months of 2016 was 1,7 per cent higher than in the corresponding period in 2015. 11 Quarterly Bulletin September 2016 Including capital repayments* Debt-service cost Seasonally adjusted

Fixed capital formation Having contracted at annualised rates of 2,8 per cent in the fourth quarter of 2015 and 10,0 per cent in the first quarter of 2016, real gross fixed capital formation contracted by a further 4,6 per cent in the second quarter. Capital spending by private business enterprises and general government contracted at a slower pace in the second quarter of 2016. Real capital investment by public corporations turned negative over the period, having increased in the first quarter of 2016. The level of real capital outlays in the first six months of 2016 was 2,6 per cent lower than in the corresponding period in 2015. Real gross fixed capital formation Quarter-to-quarter percentage change at seasonally adjusted annualised rates 2015 2016 Sector 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Private business enterprises .......... Public corporations....................... General government ..................... Total ............................................. -3,4 4,4 29,5 2,6 -9,0 0,1 36,3 -0,9 0,0 -2,9 33,5 4,6 -4,4 4,1 -4,4 -2,8 -0,6 3,5 14,6 2,5 -13,3 3,5 -12,3 -10,0 -3,1 -5,4 -8,9 -4,6 Comprising nearly two-thirds of total gross fixed capital formation, real fixed capital outlays by private business enterprises continued to be affected by low business confidence levels, excess capacity and subdued demand. Following a contraction of 13,3 per cent in the first quarter of 2016, real capital outlays by private business enterprises contracted by 3,1 per cent in the second quarter. Increased investment in buildings and transport equipment partly neutralised the quarterly decline in capital outlays on machinery and other equipment over the period. Notwithstanding the drought conditions, capital outlays by the agricultural sector remained relatively firm, probably due to recent favourable movements in commodity prices and an uptick in foreign demand. Real gross fixed capital formation Annualised percentage change from quarter to quarter Index 20 60 16 55 12 8 50 4 0 -4 45 40 -8 -12 35 -16 30 -20 2011 2012 2013 2014 2015 2016 Source: Bureau for Economic Research, Stellenbosch University* 12 Quarterly Bulletin September 2016 Total Private business enterprises Business Confidence Index (right-hand scale)* Neutral 50 level Seasonally adjusted

Having increased at an annualised rate of 3,5 per cent in the first quarter of 2016, real fixed capital expenditure by public corporations contracted by 5,4 per cent in the second quarter. Real capital outlays by the transport sector declined as some of the leading public entities, more especially Transnet, postponed capital projects. The lower capital spending by public corporations could partly be attributed to ongoing sluggish demand in a generally weak economic environment. In line with the tightening of fiscal policy announced in the 2016/17 Budget which included the deferment of some large planned capital projects, capital outlays by general government retracted for the third consecutive quarter in the second quarter of 2016. Increased capital expenditure by provincial government departments was largely offset by lower capital investment by other levels of government. Capital spending by provincial governments was particularly focused on roads and transport, utilities, human settlements, education and health. Inventory investment Following an accumulation of R1,2 billion in the first quarter of 2016, real inventory holdings contracted by R22,7 billion (at 2010 prices and annualised) in the second quarter mainly due to destocking in the mining and manufacturing sectors. The de-accumulation of inventories in the mining sector mainly occurred in the platinum mining subsector following firm export demand. Inventory holdings in other mining categories remained subdued. Change in inventories R billions, 2010 prices 80 60 40 20 0 -20 Seasonally adjusted and annualised -40 2011 2012 2013 2014 2015 2016 Factor income The year-on-year rate of growth in total nominal factor income increased from 6,5 per cent in the first quarter of 2016 to 7,2 per cent in the second quarter, reflecting an acceleration in gross operating surpluses in the economy. Remuneration of employees increased at a slower pace over the period. Measured over four quarters, growth in total compensation of employees moderated to 8,0 per cent in the second quarter of 2016, slower than the rate of increase of 8,6 per cent in the preceding quarter. Growth in salaries and wages, especially of the transport and general government sectors, moderated over the period. According to the Wage Settlement Survey conducted by Andrew Levy Employment Publications, the average wage settlement rate in the first half of 2016 amounted to 7,7 per cent, slightly lower than the settlement rate of 7,8 per cent agreed upon in the corresponding period in 2015. 13 Quarterly Bulletin September 2016

Growth in the total gross operating surpluses, measured over four quarters, accelerated from 4,1 per cent in the first quarter of 2016 to 6,4 per cent in the second quarter, mainly due to increases in gross operating surpluses in the mining and manufacturing sectors. Higher mineral sales, especially in the platinum-mining subsector, as well as a moderate recovery in metal prices in recent months boosted operating surpluses in the mining sector. Higher profits in the manufacturing sector reflected the effect of increased production and lower imported input costs following the recent strengthening in the trade-weighted value of the rand. Consequently, the contribution of gross operating surpluses to total factor income increased from 45,6 per cent in the first quarter of 2016 to 46,1 per cent in the second quarter. Gross saving Aggregate saving in South Africa improved in the second quarter of 2016. Measured as a percentage of GDP, gross saving increased from 14,9 per cent in the first quarter of 2016 to 16,5 per cent in the second quarter. Higher saving by general government and business enterprises boosted the total savings performance, whereas gross saving by the household sector decreased somewhat over the period. Gross saving as percentage of gross domestic product Per cent 18 17 16 15 14 13 12 3 2 useholds 1 0 -1 -2 2011 2012 2013 2014 2015 2016 Gross saving by the corporate sector as a percentage of GDP rose from 13,4 per cent in the first quarter of 2016 to 13,9 per cent in the second quarter. Quarter-to-quarter growth in operating surpluses of business enterprises increased at a faster pace in the second quarter of 2016 notwithstanding steady dividend and tax payments by the corporate sector. Despite an increase in the second quarter, the corporate saving rate was still lower than the 14,4 per cent registered in the corresponding period in 2015. The saving rate of general government increased from 0,3 per cent of GDP in the first quarter of 2016 to 1,5 per cent in the second quarter. Government remained committed to fiscal prudence in restraining current expenditure, and managed to raise tax revenue at a stronger pace than current outlays in the quarter under review, despite the sluggish performance of the economy. 14 Quarterly Bulletin September 2016 Ho Total Corporate sector Government Seasonally adjusted

The household saving rate fell back marginally from 1,2 per cent in the first quarter of 2016 to 1,1 per cent in the second quarter. The increase in household consumption expenditure marginally exceeded the increase in household disposable income over the period. Employment Total formal non-agricultural employment advanced further in the first quarter of 2016, following the appointment of around 49 000 temporary public-sector workers by the IEC. However, the private sector continued to shed jobs, consistent with the contraction in domestic economic activity in that quarter. According to the Quarterly Employment Statistics (QES) survey by Statistics South Africa (Stats SA), the number of people formally employed in the non-agricultural sector of the economy increased by 1,3 per cent on a seasonally adjusted and annualised basis in the first quarter of 2016, including the temporary employment opportunities. This represented a gain of 29 600 job opportunities and raised the level of formal non-agricultural employment to 2 The QES data reported in this section are seasonally adjusted, unless stated to the contrary. 2 an estimated 9,3 million. Formal non-agricultural employment Number (millions) 2,4 7,1 2,3 7,0 2,2 6,9 2,1 6,8 2,0 6,7 1,9 6,6 2010 2011 2012 2013 2014 2015 2016 The March 2016 QES survey was based on an upgraded sample drawn from the 2015 Business Sampling Frame. The new sample, which was introduced from the second quarter of 2015, resulted in an upward adjustment of roughly 0,2 million in the level of total formal non-agricultural employment. The pickup in formal non-agricultural employment in the first quarter of 2016, as mentioned, resulted entirely from a 7,7 per cent increase in public-sector employment, following the appointment of the temporary workers by the IEC to assist with the 2016 local government elections. In addition, national departments also increased their staff complement, while employment levels at all other public-sector tiers decreased in the first quarter of 2016. This is in line with government’s commitment to fiscal consolidation as set out in the National Budget. 15 Quarterly Bulletin September 2016 Private sector (right-hand scale) Public sector (adjusted for election-related outliers) Seasonally adjusted

Formal non-agricultural private-sector employment Number (millions) 2,4 4,9 4,8 2,3 4,7 2,2 4,6 2,1 4,5 2010 2011 2012 2013 2014 2015 2016 Following a notable upward revision in the fourth quarter of 2015, private-sector employment receded in the first quarter of 2016 as employment levels decreased in both the goods-producing and services sectors of the economy. Employment losses were broad-based among various sub-sectors in the private sector, with labour paring occurring in the transport, storage and communication sector; the mining sector; the manufacturing sector; the finance, insurance, real-estate and business services sector; and the community, social and personal services sector. The construction sector and the trade, catering and accommodation services sector were the only sectors to have recorded employment gains in the first quarter of 2016. Formal private-sector employment opportunities increased by around 194 000 between the second quarter of 2010 and the first quarter of 2014, as indicated in the accompanying table. This period represents the most recent upward phase in the employment cycle. Encouragingly, only 3 300 private-sector jobs were shed during the ensuing downward phase up to the first quarter of 2016. However, notable job losses occurred in the goods-producing sectors, while the services sectors continued to record employment gains over this period. In the mining, manufacturing and construction sectors combined, a total of 90 700 jobs were shed between the second quarter of 2014 and the first quarter of 2016 as the economy moved into a downward phase of the business cycle. Throughout the period, the domestic economy experienced lacklustre global demand, relatively low business confidence, cutbacks in capital expenditure, escalating operating costs, an uncertain policy environment and protracted labour-market disruptions – all of which had put a damper on momentum in private-sector employment growth. Employment in the mining-sector decreased for the sixth successive quarter in the first quarter of 2016. The sector shed a cumulative 79 400 employment opportunities between a recent peak in its employment numbers in the second quarter of 2012 and the first quarter of 2016, exacerbated by the marked decline in international commodity prices since 2011. Fortunately for domestic mining companies, the sustained depreciation in the exchange rate of the rand over this period cushioned the effect of the sharp fall in commodity prices to some extent. Although international commodity prices have recovered somewhat since the beginning of 2016, activity in the mining sector continued to be affected by relatively subdued global demand rather than a typical cyclical downturn. Employment levels in the gold-mining sector decreased marginally in the first quarter of 2016, having increased in the previous quarter. Encouragingly, the pace of job shedding in the non-gold mining sector decelerated meaningfully in the first quarter of 2016, following two quarters of substantial employment losses in the sector owing to ongoing restructuring and downscaling of some mining operations. In addition, wage negotiations are currently underway in the platinum-mining sector, with the Association of Construction and Mineworkers Union (AMCU) demanding double-digit wage increases and a one-year wage agreement, contrary to the typical three-year arrangement. 16 Quarterly Bulletin September 2016 Tertiary sector (right-hand scale) Mining, manufacturing and construction sectors Seasonally adjusted

Change in enterprise-surveyed formal non-agricultural employment by sector* Change over one quarter 1st qr 2016 Change over four quarters to 1st qr 2016 Cumulative job losses (-) gains (+) Sector 2nd qr 2010 to 1st qr 2014 2nd qr 2014 to 1st qr 2016 Per cent annualised Number Number Per cent Total mining ..................................... Gold mining.................................. -4 300 -100 -3,7 -0,3 -34 800 900 -7,1 0,7 600 -41 100 -36 400 -2 700 Other mining ................................ Manufacturing ................................. Construction.................................... -4 200 -7 500 2 600 -4,8 -2,6 2,0 -35 700 -16 500 -4 400 -9,5 -1,4 -0,8 41 700 -46 500 10 600 -33 800 -29 400 -24 900 Trade, catering and accommodation services ........................................... Private transport, storage and communication services .................. Finance, insurance, real-estate and business services ............................ Community, social and personal services ........................................... Total private sector......................... National departments ...................... Provinces ........................................ Local governments .......................... 8 800 1,8 35 800 1,9 68 700 58 000 -3 600 -4,3 1 800 0,6 5 900 -7 800 -8 700 -1,6 43 100 2,1 132 400 36 200 -100 -13 000 4 900 -10 700 -3 200 -0,1 -0,7 4,2 -3,9 -3,9 -1 000 24 000 37 700 -30 400 3 100 -0,2 0,3 8,4 -2,8 1,0 22 400 194 100 44 000 109 600 72 000 900 -3 300 31 900 -42 800 4 500 Public transport, storage and communication services .................. Other public-sector enterprises, including electricity and IEC** .......... Total public sector .......................... -1 900 -5,8 -9 500 -7,2 26 700 -7 500 53 400 42 600 114,1 7,7 64 400 65 300 26,4 2,9 19 400 271 700 70 200 56 300 Grand total ..................................... 29 600 1,3 89 300 1,0 465 800 53 000 * Seasonally adjusted. Components may not add to totals due to rounding. ** Independent Electoral Commission Source: Statistics South Africa, Quarterly Employment Statistics (QES) survey Mining-sector employment and commodity prices Indices: First quarter 2010 = 100 150 110 108 140 106 130 104 120 102 110 100 100 98 90 80 96 70 94 60 92 2010 2011 2012 2013 2014 2015 2016 *Seasonally adjusted 17 Quarterly Bulletin September 2016 South African rand export commodity price index South African US dollar export commodity price index Employment in the mining sector* (right-hand scale)

Disconcertingly, employment in the manufacturing sector decreased further in the first quarter of 2016, with the pace of labour paring accelerating somewhat. The manufacturing sector has been struggling for a prolonged period, evidenced by stagnant output volumes and depressed sentiment levels. The Manufacturing Survey of the Bureau for Economic Research (BER) indicated that business confidence among manufacturers rose from 18 index points in the first quarter of 2016 to 23 index points in the second quarter. Despite this increase, manufacturing business confidence has persistently remained below the neutral level of 50 since 2008, barring the first two quarters of 2011. According to the BER, subdued manufacturing production has been a global phenomenon post the 2007/08 financial crisis. While manufacturing business confidence increased somewhat in the second quarter of 2016, this outcome occurred despite a deterioration in the BER’s underlying activity indicators compared with the first quarter. Softer domestic demand continued to weigh on confidence, while exports provided little additional support to overall demand. With demand remaining weak, output growth slowed further as employment indicators fell to multi-year lows. Owing to a persistent slack in capacity in the sector, manufacturers continued to report a decline in current and planned fixed investment levels. Despite a marginal increase in the first quarter of 2016, employment in the construction sector continued to trend downwards, consistent with slower growth in this sector’s real value added. Job shedding in the sector can be attributed to weak residential and non-residential building activity. As such, the First National Bank (FNB)/BER Building Confidence Index fell for a second consecutive quarter, from 39 index points in the first quarter of 2016 to 34 index points in the second quarter. Building confidence declined in all six sub-sectors. Conversely, the FNB/ BER Civil Confidence Index rebounded from 28 index points to 41 index points over the same period following increased activity in the civil construction segment. However, the percentage of respondents rating insufficient demand for new projects as a constraint to business operations rose from 74 per cent in the first quarter of 2016 to 85 per cent in the second quarter. Employment growth in the services sector of the economy decelerated somewhat in the first quarter of 2016, following a notable upward revision in the final quarter of 2015. Employment in the finance, insurance, real-estate and business services sector decreased in the first quarter of 2016, following two consecutive quarters of employment gains. By contrast, the trade, catering and accommodation services sector continued to record meaningful employment gains, generating a cumulative 47 600 job opportunities in the three quarters up to the first quarter of 2016. According to the Ernest & Young (EY)/BER Retail Survey for the second quarter of 2016, business conditions deteriorated notably and confidence dropped in the trade sector, as weak sales volume growth, rising input costs and the inability of business enterprises to raise selling prices to the same extent weakened profit margins and business sentiment. With consumer demand and consequently overall consumer spending expected to deteriorate further, the pace of services-sector employment growth could moderate further in the remainder of 2016. According to the Quarterly Labour Force Survey (QLFS) conducted by Stats SA, the number of persons employed in South Africa decreased by 129 000 from the first quarter of 2016 to the second quarter, lowering the level of total employment to roughly 15,54 million. In addition, total employment decreased by 112 000 in the year to the second quarter of 2016, with the year-on-year growth rate switching notably from 1,4 per cent in the first quarter of 2016 to -0,7 in the second quarter as the base numbers normalised after the implementation of the new master sample a year earlier. Job losses occurred in the informal and in the agricultural sector, as employment decreased by 154 000 and 44 000 respectively. Disappointingly, only 82 000 employment opportunities were created in the formal sector of the economy in the year to the second quarter of 2016. The number of unemployed persons decreased by 90 000 from the first to the second quarter of 2016, but increased by 403 000 in the year to the second quarter of 2016, bringing the total number of unemployed South Africans to around 5,63 million. Disconcertingly, the number of discouraged job seekers increased by 92 000 in the year to the second quarter of 2016, resulting in the total number of discouraged job seekers amounting to a record number of 2,53 million. 18 Quarterly Bulletin September 2016

As such, the official unemployment rate decreased marginally from 26,7 per cent in the first quarter of 2016 to 26,6 per cent in the second quarter, notably higher than the comparable rate of 25,0 per cent a year earlier. Seasonally adjusted, the unemployment rate decreased from 26,6 per cent in the first quarter of 2016 to 26,3 per cent in the second quarter. However, the youth unemployment rate, which relates to people aged from 15 to 24, rose to 53,7 per cent in the second quarter of 2016, up from 49,9 per cent a year earlier. Unemployment rate Per cent 27 26 25 24 23 2010 2011 2012 2013 2014 2015 2016 Labour cost and productivity in the formal non-agricultural sector The pace of increase in nominal remuneration per worker remained unchanged at a year-on-year rate of 6,1 per cent in the first quarter of 2016, as an acceleration in the private sector was offset by a moderation in the public sector. Conversely, growth in average real salaries and wages slowed from 1,7 per cent in the fourth quarter of 2015 to 0,1 per cent in the first quarter of 2016, as consumer price inflation accelerated markedly over the period. Remuneration growth* and wage settlement rates** Percentage change over four quarters Per cent 18 18 16 14 16 14 12 12 10 10 8 8 6 6 4 4 2 2 0 0 -2 -2 2010 2011 2012 2013 2014 2015 2016 * Adjusted for election-related outliers ** Quarterly estimate based on cumulative data provided by Andrew Levy Employment Publications 19 Quarterly Bulletin September 2016 Nominal wages per worker Real wages per worker Wage settlement rate (right-hand scale) Seasonally adjusted unemployment rate Official unemployment rate

The pace of increase in private-sector nominal remuneration per worker accelerated from a year-on-year rate of 5,6 per cent in the fourth quarter of 2015 to 6,2 per cent in the first quarter of 2016. Remuneration growth quickened in the gold-mining sector (17,4 per cent); the finance, insurance, real-estate and business services sector (7,7 per cent); the community, social and personal services sector (6,5 per cent); the manufacturing sector (4,9 per cent); and in the transport, storage and communication services sector (4,6 per cent). However, wage growth per worker moderated to 9,6 per cent in the non-gold mining sector; to 4,1 per cent in the trade, catering and accommodation services sector; and to 3,2 per cent in the construction sector. By contrast, the pace of increase in average public-sector salaries and wages per worker decelerated from a year-on-year rate of 7,4 per cent in the fourth quarter of 2015 to 5,5 per cent in the first quarter of 2016, pushed down by the large number of short-duration employment opportunities created by the IEC that only received remuneration for a portion of the quarter. Nominal wage growth slowed at local government level and in other public-sector enterprises, while quickening at provincial and national departmental levels in the first quarter of 2016. According to Andrew Levy Employment Publications, the average wage settlement rate in collective bargaining agreements moderated marginally from 7,8 per cent in the first half of 2015 to 7,7 per cent in the first half of 2016. By contrast, the number of working days lost due to strike action advanced from 176 000 in the first half of 2015 – the lowest half-year number since 2002 – to 200 000 in the first half of 2016. Having remained broadly unchanged in the fourth quarter of 2015, labour productivity contracted by 1,2 per cent in the year to the first quarter of 2016, as year-on-year output growth contracted while employment growth accelerated marginally in the first quarter. Owing to a marginal year-on-year acceleration in total remuneration growth in the first quarter of 2016 coupled with a year-on-year output contraction, growth in nominal unit labour cost quickened from 6,1 per cent in the fourth quarter of 2015 to 7,3 per cent in the first quarter of 2016, exceeding the upper limit of the inflation target range for a second successive quarter. Labour productivity, nominal unit labour cost and consumer price infiation Percentage change over four quarters 15 12 9 6 3 0 -3 2010 2011 2012 2013 2014 2015 2016 Prices Headline consumer price inflation accelerated from a low of 3,9 per cent in February 2015 to 3 All rates mentioned reflect year-on-year changes, unless stated to the contrary. 3 7,0 per cent a year later – its highest level in almost seven years. The marked acceleration in inflationary pressures in the opening months of 2016 largely reflected a notable quickening in food price inflation in response to severe drought conditions as well as exchange rate pass-through following the depreciation in the exchange rate of the rand throughout 2015 and early 2016. Consumer price inflation then moderated to 6,0 per cent in July 2016 owing mainly to 20 Quarterly Bulletin September 2016 Unit labour cost Consumer price index Productivity (adjusted for election-related outliers)

favourable base effects related to petrol price changes and lower-than-anticipated services price inflation. Headline consumer price inflation breached the upper limit of the inflation target range of 3–6 per cent for six months between January and June 2016, after having remained within the range for 16 consecutive months since September 2014. Producer and consumer price inflation Percentage change over twelve months 9 8 7 6 5 4 3 2 1 0 2011 2013 2014 2015 2016 2012 Following a path similar to that of headline consumer price inflation, producer price inflation for final manufactured goods accelerated to 8,1 per cent in February 2016 before slowing to 6,5 per cent in May; price inflation moderated across a fairly broad range of product categories over the period. Headline producer price inflation thereafter accelerated to 7,4 per cent in July 2016 mainly due to a quickening in price inflation of food, beverages and tobacco products. Intermediate manufactured producer goods price inflation accelerated for six consecutive months, picking up from 1,6 per cent in December 2015 to 9,2 per cent in June 2016 – its highest reading in more than two years – before slowing somewhat to 8,5 per cent in July. The acceleration indicates that the steep depreciation in the external value of the rand between August 2015 and February 2016 is still filtering through the price formation process. The main contributor to the quickening in intermediate manufactured producer price inflation in June 2016 was basic and fabricated metals, which accelerated to a year-on-year rate of 11,3 per cent following a rebound in iron ore prices. Producer price inflation for mining products accelerated notably in the first seven months of 2016, from -0,7 per cent in December 2015 to 16,0 per cent in July 2016, as the prices of gold and other metal ores rose steeply. Similarly, producer price inflation for agriculture, forestry and fishing products accelerated markedly from a recent low of 5,2 per cent in September 2015 to 24,9 per cent in February 2016, before moderating somewhat to 18,8 per cent in July. The majority of agricultural producer product prices rose sharply over this period as a consequence of the prolonged drought experienced in large parts of the country. Producer price inflation for electricity and water remained elevated in the first half of 2016, fluctuating between 11,5 per cent and 12,6 per cent over this period, before moderating to 8,4 per cent in July due to slowing electricity price inflation. Recent outcomes in headline consumer price inflation were largely shaped by movements in consumer goods price inflation, in particular that of non-durable goods. Consumer goods price inflation moderated from a recent peak of 7,9 per cent in February 2016 to 6,5 per cent in July. This moderation resulted almost entirely from a marked slowdown in non-durable goods price inflation, which slowed from 9,6 per cent to 6,7 per cent over the same period. The moderation in non-durable goods price inflation was largely driven by a marked slowdown in petrol price inflation due to the high price base in the preceding year. 21 Quarterly Bulletin September 2016 Producer price index for final manufactured goods Headline consumer price index

Consumer food and petrol price inflation Percentage change over twelve months 12 32 9 24 6 16 8 3 0 0 -8 -3 -6 -16 -9 -24 -12 -32 2011 2012 2013 2014 2015 2016 By contrast, durable and semi-durable goods price inflation accelerated in recent months, largely in response to the depreciation in the exchange rate of the rand throughout 2015 and early 2016. Semi-durable goods price inflation quickened from 3,4 per cent in December 2015 to 5,0 per cent in July 2016 as price inflation for clothing and footwear picked up. Over the same period durable goods price inflation accelerated from 3,0 per cent to 7,1 per cent – its highest level in seven years – following a quickening in price inflation of new motor vehicles and household appliances as a result of earlier exchange rate depreciation. Headline consumer price inflation Percentage change over twelve months 2016 Weights May Jun Jul Goods ........................................................................... Non-durable goods ................................................... Food..................................................................... Petrol.................................................................... Semi-durable goods .................................................. Clothing................................................................ Footwear .............................................................. Durable goods ........................................................... New vehicles ........................................................ Household appliances .......................................... Services ........................................................................ Total ............................................................................. 49,86 35,52 14,20 5,68 6,32 2,76 1,31 8,02 5,28 0,60 50,14 100,00 6,6 7,1 10,8 -1,4 4,7 5,3 4,1 6,0 9,0 5,2 5,7 6,1 6,7 7,2 11,0 -1,0 4,7 5,2 4,1 6,1 8,9 6,0 5,8 6,3 6,5 6,7 11,5 -3,4 5,0 5,0 3,8 7,1 10,5 6,2 5,7 6,0 Consumer services price inflation moderated from 6,1 per cent in February 2016 to 5,7 per cent in March, following a slowdown in dwelling rental inflation and inflation in education fees, the latter reflecting unchanged tertiary education fees. Thereafter, the rate of increase accelerated somewhat to 5,8 per cent in June, before slowing again to 5,7 per cent in July 2016. The United Nations Food and Agriculture Organization (FAO) US dollar denominated international food price index increased for the fifth consecutive month, posting the largest monthly increase during the past four years in June 2016, before receding somewhat in July. The year-on-year change, however, remained negative despite accelerating from -16,5 per cent in January 2016 to -1,4 per cent in July. Expressed in rand terms, the FAO food price index increased in the first 22 Quarterly Bulletin September 2016 Goods prices Food prices Petrol prices (right-hand scale)

half of 2016 due to the depreciation in the exchange rate of the rand over the period. Changes in international cereals prices remained in deflation on account of favourable supplies, high inventory levels, increased competition and an improved outlook for maize plantings in the US. However, largely due to exchange rate movements, the year-on-year change in the FAO international cereals price index (expressed in rand terms) remained elevated at 21,8 per cent in May 2016, before slowing notably to 3,0 per cent in July. Extreme drought conditions resulted in a marked acceleration in agricultural producer food price inflation to 27,2 per cent in February 2016, before slowing to 17,0 per cent in April. This rate of increase, however, quickened again to 18,8 per cent in July. Likewise, producer price inflation for cereals and other crops receded from a high of 79,2 per cent in January 2016 to 42,4 per cent in April, before accelerating marginally in the months to July. Producer food price inflation at the manufactured level accelerated from 7,8 per cent in January 2016 to 12,6 per cent in July. The recent price increases were broad-based, with double-digit year-on-year rates of increase being recorded in eight of the ten subcategories in July 2016. Producer and consumer food price inflation Percentage change over twelve months 15 12 9 6 3 0 2011 2012 2013 2014 2015 2016 Consumer food price inflation responded to the acceleration in producer food price inflation in a similar fashion, quickening from 7,0 per cent in January 2016 to 11,5 per cent in July – its highest level in 55 months. Similar to agricultural producer food price inflation, the quickening in unprocessed consumer food price inflation occurred fairly early and was more pronounced than that of processed food price inflation. Unprocessed consumer food price inflation picked up pace and amounted to 14,1 per cent in April 2016, while processed consumer food price inflation accelerated to a more subdued 8,3 per cent in the same month. Unprocessed consumer food price inflation, however, turned around and moderated for the third consecutive month in July 2016, amounting to 12,5 per cent. Processed consumer food price inflation, on the other hand, accelerated from 6,3 per cent in January 2016 to 10,4 per cent in July, signifying that the rise in processed food price inflation was broadly offset by lower unprocessed food price inflation. Consumer food price inflation accelerated in eight categories and moderated in the remaining one of the nine subcategories contributing to the consumer food price basket in July 2016, leading to the renewed pickup in consumer food price inflation in that month. After moderating throughout 2015, most measures of underlying inflation trended higher in the first half of 2016, moving closer to the upper limit of the inflation target range. Subtracting the impact of food, non-alcoholic beverages and petrol prices from the calculation of headline consumer price inflation, the resultant underlying measure of inflation moderated from 5,9 per cent in February 2016 to 5,7 per cent in May, before accelerating once again to 5,8 per cent in July. When also excluding electricity prices, underlying inflation slowed from 5,7 per cent in February 2015 to 5,4 per cent in March 2016 due to a moderation in services price inflation. 23 Quarterly Bulletin September 2016 Final manufactured producer food prices Consumer food prices

Underlying inflation thereafter accelerated to 5,7 per cent in July, following the outcome of scheduled quarterly surveys, especially in the rent and transport services product categories, as well as a quickening in durable and semi-durable goods price inflation. This measure of core inflation has been moving in a narrow band of between 5,1 per cent and 5,9 per cent over the past three years, with the latest movement reflecting only a marginal increase in underlying inflationary pressures. Headline and underlying measures of consumer price inflation Percentage change over twelve months 8 7 6 5 4 3 2 2011 2012 2013 2014 2015 2016 *Food includes non-alcoholic beverages After decelerating for three consecutive months from a recent high in February 2016, price changes based on the classification of individual consumption by purpose (COICOP) categories suggested a slight acceleration in inflationary pressures in June and July 2016. Headline consumer price inflation in COICOP categories Percentage change over twelve months 2016 Weights May Jun Jul Food and non-alcoholic beverages................................ Miscellaneous goods and services ................................ Restaurants and hotels.................................................. Recreation and culture .................................................. Housing and utilities ...................................................... Health............................................................................ Alcoholic beverages and tobacco.................................. Clothing and footwear ................................................... Education ...................................................................... Household content and services ................................... Transport ....................................................................... Communication ............................................................. All items headline consumer price index ..................... 15,41 14,72 3,50 4,09 24,52 1,46 5,43 4,07 2,95 4,79 16,43 2,63 100,00 10,5 7,0 5,6 6,2 6,5 4,9 5,4 4,9 4,6 3,8 3,1 -0,4 6,1 10,8 7,2 6,2 6,5 6,5 5,2 5,2 4,7 4,6 3,9 3,3 -0,4 6,3 11,3 7,0 6,9 6,3 5,8 5,3 5,2 4,7 4,6 4,1 3,1 -0,1 6,0 24 Quarterly Bulletin September 2016 Targeted measure of inflation Headline consumer price index excluding food* and petrol Headline consumer price index excluding food*, petrol and electricity Trimmed mean consumer price index

Consumer price inflation accelerated in five of the twelve COICOP categories, moderated in four categories and remained unchanged in three categories in July 2016. The twelve-month rates of increase in four of the COICOP categories exceeded the upper limit of the inflation target range of 6 per cent in July 2016, while seven categories recorded price increases within the inflation target range and only one category registered a rate of change lower than 3 per cent. The food and non-alcoholic beverages category recorded the highest rate of increase in the year to July 2016, followed by the category for miscellaneous goods and services. By contrast, the category for communication recorded deflation in July 2016. Following a marked acceleration in petrol price inflation to 20,7 per cent in February 2016, administered price inflation quickened to 11,5 per cent over the same period. A subsequent moderation in petrol price inflation to -3,4 per cent in July 2016, along with a slowdown in the cost of education, resulted in administered price inflation moderating to 3,1 per cent in July. When the effect of petrol prices is excluded from the calculation, administered price inflation moderated from 8,5 per cent in January and February 2016 to 7,3 per cent in March and remained at this level up to June, before slowing further to 5,7 per cent in July. When further excluding electricity prices from the calculation, the rate of increase moderated from 7,1 per cent in February 2016 to 5,2 per cent in May, before accelerating to 5,5 per cent in June largely due to a quickening in public transport prices. Headline consumer price inflation expectations Per cent, as surveyed in the second quarter of 2016 Financial analysts Business representatives Trade union representatives All surveyed participants Average inflation expected for: 2016.......................................................... 2017.......................................................... 2018.......................................................... The next five years .................................... 6,7 6,1 5,5 5,6 6,2 6,3 6,3 6,3 6,1 6,3 5,9 5,8 6,3 6,2 5,9 5,9 Source: Bureau for Economic Research, Stellenbosch University Average headline consumer price inflation expectations of analysts, business representatives and trade union representatives remained virtually unchanged in the second quarter of 2016 when compared with the first quarter of 2016. On average, inflation expectations amounted to 6,3 per cent for 2016, 6,2 per cent for 2017 and 5,9 per cent for 2018. Survey of inflation expectations Percentage change over twelve months 7,5 7,0 6,5 6,0 5,5 5,0 4,5 4,0 consumer price inflation ahead ahead d 3,5 3,0 2,5 2011 2012 2013 2014 2015 2016 Source: Bureau for Economic Research, Stellenbosch University 25 Quarterly Bulletin September 2016 Headline Five years Two years One year ahea Current year

Having remained unchanged at 6,1 per cent in the first quarter of 2016, five-year inflation expectations receded, on average, to 5,9 per cent in the second quarter. All social groups lowered their five-year forecasts relative to the previous survey. The results for the second quarter of 2016 confirmed that average inflation expectations remained anchored, albeit at the top end of the inflation target range. After breaching the upper limit of the target range in 2016 and 2017, respondents expect inflation to revert to within the target range in 2018. 26 Quarterly Bulletin September 2016

Foreign trade and payments International economic developments Growth in real global economic activity moderated to an annualised rate of 2,7 per cent in the second quarter of 2016, down from 3,2 per cent in the first quarter. This moderation was observed in both advanced and emerging-market economies. Real output growth in advanced economies slowed to 1,2 per cent in the second quarter, mainly due to the sluggish economic performance of Japan and the euro area. At the same time, real output growth in emerging-market economies decelerated to 4,0 per cent primarily on account of slower growth in India and a contraction in economic activity in Russia. In addition, several Latin American economies experienced recessions. Real global output growth and contributions from advanced and emerging-market economies Percentage change from quarter to quarter Percentage points 5 5 Advanced economies (right-hand scale) Emerging economies (right-hand scale) 4 4 3 3 2 2 1 1 0 0 -1 -1 2011 2012 2013 2014 2015 2016 Seasonally adjusted annualised rates Sources: Bloomberg, Haver Analytics, International Monetary Fund, JPMorgan, national statistical offices and staff calculations In its July 2016 World Economic Outlook Update, the International Monetary Fund (IMF) lowered its global growth forecast for both 2016 and 2017 by 0,1 percentage points to 3,1 per cent and 3,4 per cent respectively. The growth forecasts for emerging-market economies remained unchanged, while those for advanced economies were pared back by 0,1 percentage points for 2016 and 0,2 percentage points for 2017. The outlook for the United Kingdom (UK) and the euro area was adversely affected by the outcome of the Brexit referendum in which the UK electorate voted to leave the European Union, together with potential banking stresses in some European countries. Brexit is projected to have a muted impact on the United States (US), but lower-than-expected first quarter growth resulted in a downward revision of the 2016 growth forecast. By contrast, the outlook for emerging-market economies remains mixed. Growth prospects for Russia have for instance improved following the uptick in oil prices since the beginning of the year, while the outlook for Nigeria has deteriorated significantly due to a shortage of foreign currency, insufficient power generation and low investor confidence levels. Disappointing global growth so far in 2016 and the Brexit shock have prompted greater policy easing in advanced economies. Japan and the euro area have followed the examples of Sweden, Switzerland and Denmark in lowering some of their policy rates below zero. At the same time, these countries have expanded their quantitative easing programmes, joined recently by the UK. As a result, demand for bonds and equities of emerging-market economies, which yield positive returns partly due to higher risk, also rebounded. 27 Quarterly Bulletin September 2016 Globa l growth

Government bond yields in selected advanced economies Per cent 6 United States 5 Japan Switzerland 4 3 2 1 0 Ten-year maturity -1 2004 2006 2008 2010 2012 2014 2016 Source: Bloomberg Real output in the US grew by 1,1 per cent in the second quarter of 2016, a modest acceleration from the 0,8 per cent rate of increase in the first quarter, but well below market expectations. The firmer real GDP growth in the second quarter was mainly driven by increased household spending, which was partly offset by a negative contribution from private inventory investment. The Federal Open Market Committee (FOMC) left the federal funds rate unchanged at its July 2016 meeting, citing uncertainty about global economic conditions and weak inflationary pressures as its main reasons. The deflator for personal consumption expenditure, the Federal Reserve’s preferred inflation measure, has moderated from 1,0 per cent in May 2016 to 0,8 per cent in July. The FOMC expects inflation to accelerate to 2 per cent as the effects of lower energy prices fade. Real output growth in selected advanced economies Percentage change at seasonally adjusted annualised rates 2014 2015 2016 Country/region 3rd qr 4th qr Year 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr United States................. Japan ............................ Euro area....................... United Kingdom ............ Canada ......................... Australia ........................ New Zealand ................. Advanced economies... 5,0 -2,4 1,4 3,3 2,1 1,6 3,8 2,7 2,3 2,2 1,7 3,4 3,4 2,0 3,5 2,2 2,4 0,0 1,1 3,1 2,5 2,7 3,0 1,9 2,0 5,0 3,3 1,1 -1,0 3,7 1,0 2,7 2,6 -1,9 1,4 1,7 -0,5 0,9 1,3 1,3 2,0 2,1 1,5 1,8 2,2 3,9 3,3 2,0 0,9 -1,7 1,8 2,8 0,5 2,8 3,6 1,2 2,6 0,5 2,0 2,2 1,1 2,4 3,0 2,1 0,8 2,1 2,1 1,8 2,5 4,3 3,0 1,6 1,1 0,7 1,2 2,4 -1,6 2,1 1,6 1,2 Underlined numbers indicate projections. Some regional totals include countries with forecasted data. Sources: Bloomberg, national statistical offices and staff calculations Economic growth in Japan slowed sharply to 0,7 per cent in the second quarter of 2016 from 2,1 per cent in the first quarter. Against the background of heightened global uncertainty, partly due to Brexit, the Bank of Japan decided at its July 2016 meeting to expand monetary easing in support of domestic economy activity. To this end it has increased its purchases of exchange-traded funds, expanded its lending programme and increased the amount of collateral that can be pledged under its US dollar funds-supplying operations. Furthermore, the Japanese government announced additional fiscal stimulus measures in August 2016 in support of the Japanese economy. 28 Quarterly Bulletin September 2016 United Kingdom Euro area

Real GDP growth in the euro area decelerated from 2,1 per cent in the first quarter of 2016 to 1,2 per cent in the second quarter. Much of this weakness stemmed from disappointing growth in Germany and a contraction in output in France. By contrast, real output growth surprised positively in Greece. Meanwhile, euro area consumer inflation turned positive in recent months, although still remaining muted. The European Central Bank (ECB) expects inflation to pick up towards the end of 2016 due to base effects, and to accelerate further in 2017 and 2018. In March 2016 the ECB lowered several key policy rates and expanded its bond-buying programme from €60 billion to €80 billion per month to stimulate economic activity and address deflationary risks. The ECB has since maintained its policy stance and confirmed that quantitative easing is set to continue at least until March 2017. Real economic growth in the UK accelerated from 1,8 per cent in the first quarter of 2016 to 2,4 per cent in the second quarter. However, the Brexit referendum result has cast substantial uncertainty over economic prospects for the UK economy. Recent business surveys covering manufacturing, construction and services have already indicated a significant decline in activity. Although the Bank of England (BoE) still expects the UK economy to avoid a recession, it has lowered its 2017 growth projection from 2,3 per cent to 0,8 per cent. Meanwhile, consumer price inflation accelerated to 0,6 per cent in July 2016. Even though inflation is well below the BoE’s inflation target of 2,0 per cent, a sharp pickup in inflation is expected over the next six months following the depreciation of the British pound. At its August 2016 Monetary Policy Committee meeting, the central bank adopted measures to support economic growth by reducing its policy rate to 0,25 per cent, buying corporate bonds amounting to £10 billion and increasing the purchases of government bonds by £60 billion to £435 billion. Economic growth in emerging Asia decelerated somewhat to 6,4 per cent in the second quarter of 2016. Real GDP growth in China accelerated to 7,1 per cent in the second quarter from 6,5 per cent in the first quarter. Although growth in China has moderated significantly from the double-digit rates achieved in 2011, the pace of expansion still remains elevated when compared to most other countries. Progress towards rebalancing the economy towards increased consumption has been slow and the country is facing a number of challenges, including high and rising corporate debt. Consumer price inflation in China moderated from 2,3 per cent in April 2016 to 1,3 per cent in August, while producer price deflation has eased following higher commodity prices. Real output growth in selected emerging-market economies Percentage change at seasonally adjusted annualised rates 2014 2015 2016 Country/area 3rd qr 4th qr Year 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr China........................... India ............................ Indonesia..................... Emerging Asia ............ Russia ......................... Turkey ......................... Poland......................... Emerging Europe........ Brazil ........................... Mexico ........................ Argentina ..................... Latin America ............. Emerging economies.. 7,2 8,3 5,0 7,0 -1,8 1,3 2,4 0,2 0,1 1,8 -3,8 0,9 4,8 7,0 3,3 4,8 5,9 -1,1 5,5 3,2 1,5 1,5 2,6 3,2 1,9 4,4 7,3 7,2 5,0 6,8 0,7 3,0 3,3 1,8 0,1 2,2 -2,6 1,0 4,6 6,7 7,0 4,1 6,3 -10,1 4,3 5,7 -2,9 -3,8 1,7 6,9 -0,2 3,7 7,2 10,4 4,8 7,5 -4,1 5,5 1,6 -0,6 -8,8 3,4 5,7 -1,8 4,6 6,7 7,9 5,2 6,7 0,4 4,1 3,2 2,1 -6,1 2,8 -0,5 -1,2 4,5 6,5 4,8 5,8 5,9 -1,6 4,7 5,3 1,4 -5,0 1,8 -1,5 -1,3 3,9 6,9 7,6 4,8 6,7 -3,7 4,0 3,6 -0,2 -3,8 2,5 2,4 -0,1 4,0 6,5 9,6 3,9 6,9 1,2 2,6 -0,4 1,5 -1,7 2,0 -2,7 -0,7 4,8 7,1 5,5 6,0 6,4 -2,0 1,2 3,6 0,3 -2,2 -0,7 -5,9 -2,5 4,0 Underlined numbers indicate projections. Some regional totals include countries with forecasted data. Sources: Bloomberg, Haver Analytics, JPMorgan, national statistical offices and staff calculations 29 Quarterly Bulletin September 2016

The pace of economic growth in India decelerated significantly to 5,5 per cent in the second quarter of 2016 following growth of 9,6 per cent in the previous quarter. By contrast, output in Indonesia expanded by 6,0 per cent in the second quarter of 2016 from 3,9 per cent in the first quarter. Consumer price inflation in India accelerated from 4,8 per cent in March 2016 to 6,1 per cent in July, while inflation in Indonesia has moderated from 8,4 per cent in December 2014 to 2,8 per cent in August 2016. After reducing its reference rate by a cumulative 100 basis points earlier this year, the Bank of Indonesia adopted the seven-day reverse repurchase rate in August as its new policy rate to strengthen the monetary policy framework. The Reserve Bank of India has kept its policy rate unchanged at 6,5 per cent since April 2016. More importantly, the Indian government confirmed an official inflation target of 4 per cent in August, within a range of plus or minus 2 per cent, in line with the central bank’s recommendation. Real output growth in emerging Europe decelerated to an annualised rate of 0,3 per cent in the second quarter of 2016, mainly due to output contracting by 2,0 per cent in Russia and a moderation in economic activity in Turkey. Real output in Poland, however, grew at a rate of 3,6 per cent in the second quarter of 2016, following a contraction in the first quarter. Sluggish economic activity in emerging Europe has largely suppressed inflationary pressures in the region, although currency depreciation has mitigated this effect in Turkey. Consumer price inflation in Russia has decelerated notably from 16,9 per cent in March 2015 to 6,9 per cent in August 2016, while deflation continued in Poland with prices falling by 0,8 per cent in August. By contrast, inflation in Turkey accelerated from 6,6 per cent in May to a five-month high of 8,8 per cent in July, before slowing somewhat to 8,1 per cent in August. The policy rate was kept unchanged in recent months in Poland, while Turkey and Russia reduced their policy rates by 25 and 50 basis points respectively. Real output in the Latin American region has contracted for six consecutive quarters, declining at an annualised rate of 2,5 per cent in the second quarter of 2016. A number of major economies, including Argentina, Brazil and Venezuela, remain in recession. Second-quarter output also declined in Chile and Mexico. Inflation in Brazil has decelerated from 10,7 per cent in January 2016 to 8,7 per cent in July partly due to the appreciation of the real. However, the central bank of Brazil has maintained its policy rate at 14,25 per cent since mid-2015, citing the slower-than-expected disinflationary process. Meanwhile, the central banks of Mexico and Colombia have tightened their policy rates cumulatively by 100 and 200 basis points respectively since the beginning of the year to contain rising inflationary pressures. According to the CPB Netherlands Bureau for Economic Policy Analysis, world trade volumes (as measured by three-month moving average of world exports) contracted by 2,7 per cent in June 2016. Export volumes of advanced economies decreased by 2,9 per cent in June 2016, compared with a contraction of 3,2 per cent in the preceding month. In emerging-market economies, export volumes contracted by 2,5 per cent. After nearly six years of declines (two years in the case of energy prices), most commodity prices have rebounded somewhat from the low levels registered in January 2016. Energy prices increased rapidly by 42 per cent up to August, followed by precious metals (25 per cent), metals and minerals (16 per cent) and agricultural products (9 per cent). The price of Brent crude oil (which constitutes 84,5 per cent of the weight in the World Bank’s energy index) rebounded by more than 90 per cent from mid-January 2016 to US$50 per barrel in early June. The increase in crude oil prices largely resulted from the depreciation in the US dollar and falling US oil production. However, oil prices declined below US$42 in early August, before rising again to levels of around US$50 per barrel in mid-August amid speculation of a possible agreement between Saudi Arabia, Russia and other major producers to limit production. 30 Quarterly Bulletin September 2016

International commodity prices Indices: 2010 = 100 170 150 130 110 90 70 50 Precious metals 30 2010 2011 2012 2013 2014 2015 2016 Source: World Bank Crude oil prices have declined by more than 10 per cent to US$45 per barrel towards the end of August as doubts about the possible agreement intensified. However, oil prices rebounded again in early September as US inventories dropped sharply following import and production disruptions due to a hurricane in the Gulf of Mexico. Brent crude oil futures prices, for delivery in the first and second quarters of 2017, were trading at around US$51 per barrel and US$52 per barrel respectively in early September 2016. Volatile oil prices have had a major impact on oil producing nations such as the members of the Organization of the Petroleum Exporting Countries (OPEC); the considerable reduction since 2015 in their combined nominal GDP is shown in the graph below. International crude oil prices US$ per barrel US$ billions 4 600 140 September 2016) 120 (current US dollars) 4 400 scale) 100 4 200 80 4 000 60 3 800 40 3 600 20 2012 2013 2014 2015 2016 2017 Sources: Bloomberg, International Monetary Fund and staff calculations Current account 3 Notwithstanding the moderation in global economic activity, South Africa’s merchandise export volumes rallied strongly in the second quarter of 2016, benefiting from the lagged effect of the depreciation in the exchange rate of the rand. Simultaneously, weak growth in domestic demand led to a decline in merchandise import volumes. South Africa’s trade balance consequently 3 Unless stated to the contrary, the current-account transaction flows referred to in this section are all seasonally adjusted and annualised. 31 Quarterly Bulletin September 2016 Brent crude spo Futures prices t price (on 9 OPEC GDP (right-hand Projection Projection Energy Agriculture Metals and minerals

switched from a deficit of R48 billion in the first quarter of 2016 to a surplus of R33 billion in the second quarter. Over the period, South Africa’s trade balances with Asia and Europe improved notably, largely reflecting higher exports to these regions. Balance on current account R billions, seasonally adjusted and annualised 2015 2016 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Merchandise exports ........................................ Net gold exports............................................... Merchandise imports ........................................ Trade balance .................................................. Net service, income and current transfer payments............................................. Balance on current account ............................ As percentage of gross domestic product .... 940 63 -1 070 -68 984 71 -1 050 5 984 65 -1 082 -34 988 72 -1 101 -41 974 68 -1 076 -34 1 005 52 1 097 55 -1 105 -1 118 -48 33 -134 -202 -5,1 -128 -123 -3,1 -148 -182 -4,5 -150 -191 -4,6 -140 -174 -4,3 -174 -221 -5,3 -168 -134 -3,1 Components may not add up to totals due to rounding The shortfall on the services, income and current transfer account shrank somewhat in the second quarter of 2016 as an increase in net income and services payments to non-residents was more than neutralised by a marked decline in net current transfer payments to the rest of the world. The improvement in the trade balance coupled with the marginally smaller deficit on the services, income and current transfer account significantly narrowed the deficit on the current account of the balance of payments to 3,1 per cent of GDP in the second quarter of 2016; on average, this ratio amounted to 4,4 per cent since the third quarter of 2015. Balance of payments: Current account R billions 100 50 0 -50 -100 -150 -200 R billions Per cent 0 0 -1 -2 -3 -4 -5 -6 -7 -8 -50 -100 -150 -200 -250 -300 2011 2012 2013 2014 2015 2016 Seasonally adjusted and annualised 32 Quarterly Bulletin September 2016 Balance on current account Percentage of gross domestic product (right-hand scale) Trade balance Balance on services, income and current transfer account

The value of merchandise exports which had advanced by 1,8 per cent in the first quarter of 2016 surged by 9,1 per cent in the second quarter following firm increases in all major export categories, namely manufacturing, mining and agriculture. The value of manufactured exports rose by no less than 9,8 per cent in the second quarter, underpinned by especially an increase in the value of vehicles and transport equipment. The number of both passenger and commercial vehicles exported surged over the period – passenger vehicles were largely destined for the United Arab Emirates and Angola while the export of commercial vehicles was mainly earmarked for the United Kingdom, Germany and the African region. In addition, South African manufacturers sold a number of container and tank wagons to Swaziland and six locomotives to Côte d’Ivoire. As shown in the accompanying graph, the deficit for trade in vehicles and transport equipment accordingly narrowed to R39 billion in the second quarter of 2016, considerably less than the quarterly average of R58 billion recorded during the period 2010 to the first quarter of 2016. Trade in vehicles and transport equipment R billions 220 200 180 160 140 120 100 80 60 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 2011 2012 2013 2014 2015 2016 Seasonally adjusted and annualised Mining export proceeds rose by 10,3 per cent in the second quarter of 2016, largely benefiting from the sharp increases in the export value of platinum-group metals, base metals and articles thereof, more in particular ferrochrome destined for China, as well as iron ore. Furthermore, no crude oil was exported in the second quarter of 2016 following sales of South Africa’s strategic crude oil reserves in the preceding quarter. The value of agricultural exports advanced in the second quarter of 2016, benefiting largely from a sharp rise in the export value of citrus products. Merchandise export volumes shrank by 0,9 per cent in the first quarter of 2016. However, the steady pace of increase in global economic activity and the lagged effect of the depreciation in the exchange rate of the rand created additional demand for domestically produced goods, lifting the volume of merchandise exports by 5,6 per cent in the second quarter of 2016. Consistent with higher export volumes, the volume of bulk and break bulk cargo handled at national ports rose by 11,0 per cent from the first quarter of 2016 to the second quarter. The dollar price of a basket of South African-produced non-gold export commodities increased by 7,0 per cent in the second quarter of 2016. Though price increases were broad based during the period, pronounced increases were observed in the prices of iron ore and platinum, 33 Quarterly Bulletin September 2016 Net exports of vehicles and transport equipment I mports Exports

increasing by 14,9 per cent and 9,8 per cent respectively. The international price of nickel rose amid prospects of an environmental review of the mining sector in the Philippines, the world’s largest producer of nickel. After rising by 2,7 per cent in the first quarter of 2016, the rand price of merchandise exports increased further by 3,3 per cent in the second quarter. Volume of merchandise exports R billions, 2010 prices 400 375 350 325 300 275 250 225 ding gold) g adjusted and annualised Million metric tonnes 190 180 170 160 150 140 bulk, annualised 130 2011 2012 2013 2014 2015 2016 Source: SARB and National Ports Authority The London fixing price of gold which averaged US$1 180 per fine ounce in the first quarter of 2016 rose by 6,6 per cent to US$1 258 per fine ounce in the second quarter. The price of gold was supported by, among other factors, low to negative bond yields in advanced economies and increased demand for the metal as a safe haven asset amid increased uncertainty following the outcome of the UK referendum to leave the European Union on 23 June 2016 (Brexit announcement). The dollar price of gold rose by 4,0 per cent to US$1 315 per fine ounce immediately after the announcement of the result and traded at US$1 318 per fine ounce during the last week of June. In July 2016, the average fixing price of gold on the London market came to US$1 337 per fine ounce. The strengthening of the exchange rate of the rand in the second quarter of 2016, however, resulted in the average realised rand price of gold moving broadly sideways over the period. Nonetheless, a rise in the volume of net gold exports in the second quarter of 2016 resulted in an increase of 5,0 per cent in the value of net gold exports. A statistical release by the World Gold Council in August 2016 showed that the central banks of Russia, China and Kazakhstan, among others, accumulated gold reserves of 38,4 tonnes, 25,8 tonnes and 9,8 tonnes respectively over the period. Notwithstanding the decline in gross domestic expenditure, the value of merchandise imports rose by 1,2 per cent in the second quarter of 2016, following an increase of 0,4 per cent in the preceding quarter. Firm increases were registered in crude oil and agricultural imports, whereas the value of manufactured imports shrank in the second quarter. Declines were fairly broad based among the subcategories for manufactured imports; increases were mainly registered in the imports of chemical products, optical and professional equipment and prepared foodstuffs, beverages and tobacco. An uptick in the wholesale prices of South Africa’s main trading partner countries and higher international oil prices gave rise to higher rand prices in the second quarter of 2016, notwithstanding an increase in the nominal effective exchange rate of the rand. 34 Quarterly Bulletin September 2016 Cargo handled at ports Bulk and break Mining (exclu Manufacturin Seasonally

International prices of selected South African export commodities US dollar indices: 2010 = 100 Export values of selected mining products R billions, annualised 150 100 90 125 80 100 70 60 75 50 50 40 25 30 2011 2012 2013 2014 2015 Copper Coal 2016 2011 2012 2013 2014 2015 2016 Gold Iron ore Nickel Platinum Gold Iron ore Coal Platinum Real merchandise imports R billions, 2010 prices 900 troleum products and refined pertroleum products 800 700 600 500 400 2011 2012 2013 2014 2015 2016 Percentage of gross domestic expenditure 28 26 24 22 2011 2012 2013 2014 2015 2016 Seasonally adjusted and annualised 35 Quarterly Bulletin September 2016 Import penetration ratio Crude oil and refined pe Total excluding crude oil

Merchandise import volumes, which had declined by 2,6 per cent in the first quarter of 2016, decreased by a further 0,9 per cent in the second quarter. The physical quantity of manufactured imports declined in the second quarter of 2016 in line with the sustained contraction in capital spending since the fourth quarter of 2015. The import volume of machinery and electrical equipment decreased visibly as the independent power producers curtailed their imports of solar components alongside a decline in vehicles and transport equipment. At the same time, the volume of crude oil imports increased by 9,5 per cent. South Africa’s import penetration ratio remained elevated at 26,2 per cent in the second quarter of 2016, unchanged from the first quarter of 2016. The country’s terms of trade (including gold) improved in the second quarter of 2016 as export prices rose at a somewhat faster pace than import prices. Having widened by R24 billion or 16 per cent in the first quarter of 2016, the shortfall on the services, income and current transfer account narrowed by about R6 billion in the second quarter. Increases in net income payments to foreign investors alongside marginally higher net payments for services were more than neutralised by a decline in net current transfer payments to the rest of the world over the period. Net interest payments to non-residents increased in the second quarter of 2016, partly due to higher domestic interest rates, while an increase in dividends received by South African investors on their offshore investments – the first increase since the first quarter of 2015 – was more than offset by a further increase in gross dividend payments to non-residents. The relative size of the deficit on South Africa’s income account compares favourably with that of some developing countries. The accompanying graph depicts the balances on the income accounts of a selection of countries as an average ratio of the respective countries’ GDP for the period 2010–2015. Advanced economies mainly registered surpluses on their income accounts while most of the other countries recorded deficits. Selected countries: Net income balance* as percentage of gross domestic product for the period 2010–2015 France** Germany** Sweden** United States** Netherlands** Philippines United Kingdom** China Turkey India Argentina Brazil Mexico South Africa Australia** Indonesia Poland Russia Malaysia Colombia Hungary New Zealand Chile Thailand Peru 4 -8 -6 -4 -2 0 2 * Negative sign refers to net payments whereas positive sign refers to net receipts ** Advanced economies Source: IMF and own calculations 36 Quarterly Bulletin September 2016 2,4 2,2 2,0 1,3 1,0 0,4 -0,3 -0,5 -1,0 -1,1 -2,2 -2,3 -2,3 -2,5 -2,5 -3,1 -3,1 -3,1 -3,3 -3,6 -4,1 -4,2 -4,3 -4,5 -5,9

Owing largely to a small decrease in gross travel receipts from abroad, net payments for services increased somewhat in the second quarter of 2016. Travel receipts, which gained significant momentum in the second half of 2015 following the relaxation of South African visa requirements, were nevertheless 14 per cent higher in the first half of 2016 when compared with the corresponding period in 2015. Net current transfer payments by South Africa declined notably in the second quarter of 2016. This resulted from a reduction in the amounts payable to South Africa’s partner countries in the Southern African Customs Union (SACU) in 2016/17, mainly driven by lower import duties collected in an earlier year. Financial account The South African economy continued to attract inward foreign investment on a net basis in the second quarter of 2016, albeit at a slower pace than in the first quarter. The net inflow of capital on the financial account of the balance of payments (including reserve assets but excluding unrecorded transactions) shrank from R31,7 billion in the first quarter of 2016 to R5,1 billion in the second quarter. On a net basis, the categories for direct and portfolio investment recorded capital inflows in the second quarter while other investment registered capital outflows. Capital inflows over the period were supported by indications of a possible delay in the tightening of monetary policy in the US and the relatively high return offered on domestic debt securities. Global financial markets experienced some turbulence towards the end of June 2016 following the Brexit announcement. Some aversion towards risky financial assets, declines in the prices of UK-linked equities and the depreciation of emerging-market currencies were noted during the last week of June. Despite the subsequent apparent return to calmness in global financial markets, uncertainty around the British exit plan is expected to remain a concern during the next two years. Unease lingers among large recipients of foreign direct investment and other capital funding from the UK. Net financial transactions R billions 2015 2016 1st qr 2nd qr 3rd qr 4th qr Year 1st qr 2nd qr Change in liabilities Direct investment ................................ Portfolio investment ............................ Financial derivatives ............................ Other investment ................................ Change in assets Direct investment ................................ Portfolio investment ............................ Financial derivatives ............................ Other investment ................................ Reserve assets ................................... Total identified financial transactions*... As percentage of gross domestic product............................................... -13,9 39,3 -72,6 39,6 6,9 54,8 -74,3 -20,6 15,9 11,8 -71,0 6,2 13,7 -0,3 -103,0 44,6 22,6 105,6 -320,9 69,8 9,9 13,5 -149,6 2,2 9,7 33,0 -103,4 -20,9 -11,3 -5,3 73,3 -18,3 12,3 43,0 -4,9 -10,0 70,9 -20,0 1,4 4,1 -14,7 -24,2 78,3 38,8 0,5 41,7 -37,4 -10,6 103,3 48,3 -5,1 53,5 -68,3 -50,1 325,8 48,8 9,1 142,3 -20,5 8,0 148,5 15,5 4,2 31,7 -7,6 -2,4 94,4 1,1 1,2 5,1 4,4 0,4 4,1 5,2 3,5 3,0 0,5 * Including reserve assets but excluding unrecorded transactions Inflow (+)/Outflow (-) Foreign-owned assets in South Africa South Africa’s direct investment liabilities increased further in the second quarter of 2016 as foreign investors with subsidiaries in the domestic market expanded their shareholding in the domestic entities. Non-resident investors acquired direct investment assets to the value of R9,7 billion in the second quarter, having invested R9,9 billion in the first quarter of 2016. A foreign parent company, among other transactions, sold part of its shareholding in a South African bank, while another foreign parent company increased its equity holdings in a South African company in the food and beverages industry during the second quarter. 37 Quarterly Bulletin September 2016

Inward portfolio investment into South Africa increased notably in the second quarter of 2016 following relatively minor capital inflows in the second half of 2015 and in the first quarter of 2016. Non-residents were net buyers of South African debt and equity securities to the value of R33,0 billion in the second quarter of 2016 following an inflow of R13,5 billion in the preceding quarter. The bulk of these inflows reflected the acquisition of domestic debt securities by foreign investors searching for higher yields in emerging markets. The attractiveness of domestic debt securities was further buoyed by international rating agencies’ decision to maintain South Africa’s sovereign rating at investment grade, leading to a decline in domestic bond yields relative to that of other emerging-market economies. The acquisition of domestic debt securities in the second quarter was furthermore underpinned by the issuance of an international bond by National Treasury to the value of US$1,25 billion, which more than compensated for the redemption of a €750 million international bond. Over the same period, non-resident investors also sold less domestic equity securities compared to the first quarter of 2016. The disposal of South African equities amounted to R1,6 billion in the second quarter of 2016 following net sales of R7,7 billion in the first quarter. Inward portfolio investment R billions 60 50 40 30 20 10 0 -10 -20 -30 -40 2011 2012 2013 2014 2015 2016 Other investment flows into South Africa switched to a capital outflow of R20,9 billion in the second quarter of 2016, following inflows of R44,6 billion and R2,2 billion in the fourth quarter of 2015 and the first quarter of 2016 respectively. The outflow of other investment capital can largely be attributed to the repayment of loans by the domestic banking sector as well as the redemption of long-term loans by the private sector and public corporations. 38 Quarterly Bulletin September 2016 Debt Equity Box 1 The impact of the revision of JSE statistics on inward portfolio investment flows On 24 July 2016 the JSE announced a revision of its data regarding the purchases and sales of South African equity securities by non-resident investors during the period 31 May to 20 July 2016. The revision will not affect the inward portfolio investment statistics published in the June 2016 Quarterly Bulletin as these statistics covered only the first quarter; second-quarter estimates of inward portfolio investment flows are released for the first time in this issue of the Quarterly Bulletin. The South African Reserve Bank (the Bank) published a brief note in the June 2014 Quarterly Bulletin explaining its revised methodology for the calculation of the portfolio investment liabilities in the financial account of the balance of payments. In addition to the statistics obtained from the JSE, the Bank also uses other relevant and available data sources, including data on stock positions acquired from South Africa’s Central Securities Depository (STRATE), to estimate inward portfolio flows.

South African-owned assets abroad South African entities acquired a smaller amount of foreign direct investment assets in the second quarter of 2016 – outward direct investment flows amounted to R7,6 billion in the second quarter compared with outflows of R20,5 billion in the first quarter. Increased volatility in global financial markets probably caused South African entities to be somewhat more cautious in acquiring offshore assets. South African residents acquired foreign portfolio assets worth R2,4 billion in the second quarter of 2016 following the disposal of portfolio assets to the value of R8,0 billion in the first quarter. South African institutional investors purchased foreign portfolio assets to the value of R50,1 billion in 2015. Other outward investment flows moderated from an inflow of R15,5 billion in the first quarter of 2016 to an inflow of R1,1 billion in the second quarter. The repatriation of part of the domestic banking sector’s deposits with foreign banks mainly accounted for the latter inflow of capital. Foreign debt South Africa’s total outstanding foreign debt increased from US$124,1 billion at the end of the fourth quarter of 2015 to US$129,2 billion at the end of the first quarter of 2016, buoyed by a rise in rand-denominated debt. Expressed in rand terms, however, the country’s gross external debt declined from R1 933 billion to R1 922 billion over the same period largely due to the strengthening of the rand against the US dollar. External debt of South Africa US$ billions Per cent 180 50 45 160 40 140 35 120 30 25 100 80 20 60 15 40 10 20 5 0 0 2011 2012 2013 2014 2015 2016 Foreign currency-denominated external debt declined from US$69,3 billion at the end of the fourth quarter of 2015 to US$67,6 billion at the end of the first quarter of 2016. The decrease of foreign currency-denominated external debt was mainly brought about by a decline in non-residents’ deposits with the domestic banking sector as well as the redemption of short-and long-term loans by this sector. The redemption of a US dollar-denominated international bond by a parastatal amplified the decline in foreign currency-denominated external debt. South Africa’s short-term debt (i.e. debt with an original maturity of less than one year) improved marginally to 29 per cent of total foreign debt at the end of the first quarter of 2016, slightly less than the ratio of 30 per cent recorded at the end of the fourth quarter of 2015, as the increase in external debt mainly took the form of long-term debt securities. Rand-denominated external debt expressed in US dollar terms increased for the first time since the end of June 2014, from US$54,8 billion at the end of the fourth quarter of 2015 to US$61,5 billion at the end of the first quarter of 2016. The substantial increase in rand-denominated debt expressed in US dollars resulted from further net purchases of domestic 39 Quarterly Bulletin September 2016 Rand-denominated Foreign currency-denominated Ratio of external debt to GDP (right-hand scale)

government bonds by non-residents along with a rise in the market value of these bonds amid the strengthening in the external value of the rand. This increase was exacerbated by non-residents’ rand-denominated deposits with and short-term loans extended to the domestic banking sector. South Africa’s foreign debt US$ billions at end of period 2014 2015 2016 4th qr 1st qr 2nd qr 3rd qr 4th qr 1st qr Foreign currency-denominated debt............... Debt securities ................................................. Other ................................................................ Public sector ................................................ Banking sector ............................................. Non-monetary private sector ........................ Rand-denominated debt ................................. Debt securities.............................................. Other ............................................................ Total foreign debt ............................................ As percentage of gross domestic product .... As percentage of total export earnings ......... 67,5 22,7 44,8 8,2 18,4 18,2 77,6 41,1 36,4 145,1 41,3 123,8 67,0 23,6 43,3 7,3 17,7 18,3 77,4 39,4 38,0 144,4 41,3 124,6 65,7 23,9 41,8 7,7 16,3 17,8 76,0 38,8 37,2 141,7 41,4 123,8 66,6 23,8 42,8 7,8 16,5 18,5 67,9 34,5 33,4 134,5 40,8 121,7 69,3 23,3 46,0 7,7 19,9 18,4 54,8 28,7 26,1 124,1 39,4 118,9 67,6 22,6 45,0 8,1 17,9 19,0 61,5 32,8 28,8 129,2 43,6 133,2 South Africa’s external debt as a ratio of annualised GDP increased from 39,4 per cent in December 2015 to a record high of 43,6 per cent in March 2016 – the ratio amounted to, on average, 40,4 per cent during the two-year period 2014–2015. Relative to the country’s export proceeds, South Africa’s external debt rose from 118,9 per cent to 133,2 per cent over the same period with the latter value representing a new record high. International investment position South Africa’s positive net international investment position narrowed from R700 billion at the end of December 2015 to R544 billion at the end of March 2016 due to an increase in the value of South Africa’s foreign liabilities and a slight decline in the value of foreign assets. Having increased by 7,7 per cent in the fourth quarter of 2015, the market value of South Africa’s foreign liabilities (arising from inward investment) advanced at a significantly slower pace of 2,3 per cent in the first quarter of 2016 (i.e. from R5 563 billion at the end of December 2015 to R5 694 billion at the end of March 2016). The increase in foreign liabilities can mainly be attributed to an increase in equity values of both direct and portfolio investment which in turn rose largely due to valuation effects. The market value of South Africa’s foreign assets (outward investment) declined marginally during the first quarter of 2016 to an amount of R6 238 billion at the end of March 2016 compared with R6 263 billion at the end of December 2015. The quarterly rate of decrease in the value of foreign assets came to 0,4 per cent in the first quarter of 2016 compared with an increase of 18,7 per cent in the preceding quarter. Marginal advances in direct and portfolio investment assets were more than offset by declines for financial derivatives and reserve assets. As a ratio of South Africa’s GDP, the country’s foreign liabilities increased marginally from 138,6 per cent at the end of December 2015 to 139,3 per cent at the end of March 2016 while the country’s foreign assets declined from 156,1 per cent to 152,7 per cent over the same period. This resulted in the country’s positive net international investment position receding to 13,3 per cent of GDP at the end of March 2016 from 17,4 per cent at the end of December 2015. 40 Quarterly Bulletin September 2016

International reserves and liquidity Subsequent to a decrease of R4,2 billion in the first quarter of 2016, South Africa’s international reserves declined by R1,2 billion in the second quarter. Measured in US dollar, the value of South Africa’s gross gold and other foreign reserves (i.e. the international reserves of the Bank before accounting for reserve-related liabilities) decreased from US$46,8 billion at the end of March 2016 to US$46,4 billion at the end of June as the impact of the higher international price of gold was more than offset by currency revaluation effects due to an appreciation of the US dollar. The country’s gross reserves declined further to US$45,7 billion at the end of August 2016. The level of import cover (i.e. the value of gross international reserves relative to the value of imports of goods and services and income payments) decreased from 5,4 months at the end of March 2016 to 5,3 months at the end of June. International reserves US$ billions Months 60 7 6 55 5 50 4 45 3 40 2 35 1 0 30 2010 2011 2012 2013 2014 2015 2016 The Bank’s international liquidity position displayed moderate fluctuations, decreasing from US$41,2 billion at the end of March 2016 to US$40,8 billion at the end of June, remaining broadly unchanged up to the end of August. Exchange rates The external value of the rand on balance strengthened slightly further in the second quarter of 2016, appreciating by 0,8 per cent against the US dollar and by 2,1 per cent on a trade-weighted basis. The domestic currency started the second quarter of 2016 on a firm footing; the trade-weighted exchange rate of the rand increased, on balance, by 4,0 per cent in April 2016 supported by a weaker US dollar. It depreciated markedly by 8,3 per cent in May 2016 as hawkish comments made by some US Federal Reserve officials revived expectations of an early US interest rate increase. Exchange rates of the rand Percentage change 30 Sep 2015 to 31 Dec 2015 31 Dec 2015 to 31 Mar 2016 31 Mar 2016 to 30 Jun 2016 30 Jun 2016 to 7 Sep 2016 Weighted average* .............................. Euro ..................................................... US dollar .............................................. Chinese yuan........................................ British pound ........................................ Japanese yen ....................................... -10,0 -8,8 -11,3 -9,4 -9,5 -11,1 2,5 0,7 4,7 4,2 8,0 -2,2 2,1 2,9 0,8 3,6 7,7 -7,9 4,7 4,1 5,3 5,6 5,8 4,1 * Trade-weighted exchange rate against a basket of 20 currencies (nominal effective exchange rate) 41 Quarterly Bulletin September 2016 Gross reserves International liquidity position Import cover ratio (right-hand scale)

On a trade-weighted basis the exchange value of the rand increased by 7,7 per cent from the end of May to 23 June 2016, buoyed by the confirmation of South Africa’s investment grade credit rating by credit rating agencies. However, in the three days following the surprise outcome of the British vote to leave the European Union, the exchange rate of the rand depreciated by 2,8 per cent on renewed risk aversion amid heightened uncertainty in the global financial markets. Overall, the domestic currency appreciated, on balance, by 7,1 per cent from the end of May 2016 to the end of June. The South African rand, along with other high-yielding currencies, continued to appreciate markedly in July and August 2016, while the British pound depreciated considerably against most currencies. High-yielding currencies were supported by the notion that major central banks could add liquidity to support markets, as well as expectations that the US Federal Reserve could refrain from raising interest rates this year. Towards the end of August the rand, however, depreciated abruptly in the wake of domestic political events. Selected exchange rates: Foreign currency units per South African rand Indices: 31 December 2015 = 100 140 130 120 110 100 90 80 Jan Feb Mar Apr May 2016 Jun Jul Aug Sep Nominal effective exchange rate of the rand US dollar Brazilian real British pound Euro Japanese yen Turkish lira The real effective exchange rate of the rand increased by 1,5 per cent from December 2015 to June 2016. In January 2016 the real effective exchange rate of the rand reached its lowest level since December 2001. In the subsequent months improved economic fundamentals, such as a stronger trade balance, relatively high domestic bond yields and a recovery in commodity prices contributed to upward pressure on the currency and a reduction in its undervaluation in purchasing power parity terms. 42 Quarterly Bulletin September 2016 Rand appreciation Rand depreciation

Real effective exchange rate of the rand Index: 2010 = 100 120 110 100 90 80 70 60 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Turnover in the South African foreign-exchange market The net average daily turnover in the South African market for foreign exchange has trended lower over the past two years, and decreased slightly from US$21,2 billion in the first quarter of 2016 to US$20,8 billion in the second quarter. The net average daily turnover in the second quarter of 2016 was 6,2 per cent lower when compared with the level in the corresponding quarter of 2015. Net average daily turnover in the foreign-exchange market US$ billions 30 ransactions against the rand ransactions in third currencies otal 25 20 15 10 5 0 2011 2012 2013 2014 2015 2016 Turnover in the domestic market for foreign currency against the rand fell from US$15,9 billion in the first quarter of 2016 to US$15,6 billion in the second quarter, while turnover in the market for third currencies receded from US$5,3 billion to US$5,2 billion over the same period. Transactions in the swap market accounted for roughly 73 per cent of the net average daily turnover in the rand market in the second quarter of 2016, with non-resident participation representing 61 per cent of swap transactions. 43 Quarterly Bulletin September 2016 T T T Stronger rand = higher Weaker rand = lower

Monetary developments, interest rates and financial markets Money supply Following two years of gradual acceleration, twelve-month growth in the broadly defined money supply (M3) decelerated significantly in the first half of 2016. By the second quarter of 2016 growth in M3 fell below that of nominal GDP for the first time since mid-2014. The deceleration in M3 growth was concentrated in the deposit holdings of the corporate sector; factors which may have contributed to the slower growth include a lower transactions demand for money arising from lower capital expenditure, drawdowns of corporate deposits due to the fact that June is a provisional tax month, and the spending by local authorities of some of the equitable share transfers they received earlier in the year. Growth in household sector deposits maintained its resilience, consistently recording double-digit rates since January 2014. Individuals continued to contribute significantly to growth in this deposit category, taking advantage of higher interest rates and attractive depository savings and investment products offered by banks. M3 and gross domestic product Year-on-year percentage change 14 12 10 8 6 4 2 0 2010 2011 2012 2013 2014 2015 2016 The monetary statistics were affected to a moderate extent by a methodological change introduced in June 2016 which affected mostly corporate deposits. An explanation of this technical issue is provided in the accompanying box. 44 Quarterly Bulletin September 2016 Box 2 Revisions to money supply data With the release of the monetary aggregates for June 2016, the Bank refined its measurement of deposits included in the broadly defined money supply (M3). This box discusses the background to and nature of the revision and its impact on the monetary aggregates released by the Bank. During the first half of 2016, the Money and Banking Division of the Economic Research and Statistics Department reviewed the financial instruments included in deposits with banks. This was done to ensure that the calculation of the monetary aggregates is firstly, aligned with international standards, secondly, is treated consistently across all reporting institutions and lastly, provides a fair reflection of monetary conditions. Certain types of short-term debt instruments issued by banks, especially those that can be converted into currency or transferable deposits at short notice without incurring a significant loss on the amount initially invested, meet the internationally accepted definition of broad money. Examples include negotiable certificates of deposit (NCDs), negotiable promissory notes (NPNs) and floating and fixed rate notes (FRNs). The characteristics of certain short-term debt instruments exclude them from compliance with this definition. M3 M3: comparable estimates Nominal gross domestic product

With slower growth in M3 and a firmer growth rate in nominal GDP registered in the second quarter of 2016, the income velocity of circulation of M3 rose to 1,42 in the second quarter of 2016, slightly higher than the value of 1,41 recorded in the first quarter of the year. Viewed from a maturity perspective, growth in long-term deposits gained momentum in the second half of 2015 and maintained its thrust in the first half of 2016, recording twelve-month rates of increase of 10,2 per cent in June and 9,2 per cent in July. Demand for long-term deposits was supported by the current less accommodative monetary policy stance resulting in higher interest rates and banks offering more attractive saving and investment products to address the composition of their funding. Growth in transactional balances – notes, coin, cheque and transmission deposits – also remained high over the past year, whereas other short-and medium-term deposits stagnated. As shown in the graph on the following page, individuals continued to take advantage of higher interest rates on deposits in the wake of uncertain and volatile financial market conditions, with year-on-year growth in the deposit holdings of the household sector consistently exceeding 11 per cent since September 2014. By contrast, corporates experienced moderation in the rate of increase of their M3 deposits in the first seven months of 2016. Within the corporate sector, growth in deposits of non-financial corporates decelerated somewhat but the sharpest decline related to the deposit holdings of financial companies, and was mostly a reflection of the methodological improvements to the measurement of M3 deposits, as discussed in Box 2. Growth over twelve months in corporate sector deposits remained pedestrian and averaged 5,2 per cent in the first half of 2016, but declined to 1,3 per cent in July. 45 Quarterly Bulletin September 2016 Banks were therefore requested in their balance sheet reporting to move these instruments, such as structured and credit-linked notes, from deposit to debt securities as they fall outside the definition of M3. The affected banks revised their balance sheet data from January to May 2016; it was assessed that further backward revision was not feasible. The reclassification resulted in a reduction in the level of M3 of approximately R40 billion (about 1,3 per cent of M3). In terms of the monetary analysis, it is the ‘net other assets and liabilities’ category that is affected as statistical counterpart of M3. The impact of the revision on the level of the different monetary aggregate categories is presented in the table below. Analysed by maturity, all the deposit categories were influenced, with long-term deposits being the most affected category. Impact of revision on monetary aggregates R millions M1 M2 M3 Before After Before After Before After revision revision revision revision revision revision 2016: Jan ........................................... 1 447 146 1 445 289 2 424 559 2 420 841 2 982 664 2 939 393 Feb........................................... 1 419 057 1 416 194 2 417 003 2 411 088 2 991 454 2 945 963 Mar........................................... 1 492 421 1 490 172 2 458 235 2 459 285 3 066 328 3 023 076 Apr ........................................... 1 451 919 1 449 869 2 439 000 2 446 136 3 070 402 3 021 453 May .......................................... 1 439 352 1 436 987 2 437 093 2 431 670 3 078 284 3 031 641 The revision to the level of non-bank domestic private-sector deposits also affected the income velocity of circulation of M3. The revised value for the first quarter of 2016 is 1,41 against an original value of 1,39.

M3 holdings of households and companies Percentage change over twelve months 20 15 10 5 0 -5 2010 2011 2012 2013 2014 2015 2016 M3 holdings of households and companies Quarter-on-quarter change R billions Percentage of total M3 deposit holdings 2015 2016 2015 2016 3rd qr 4th qr 1st qr 2nd qr 4th qr 2nd qr Households ........................................ Companies: Total................................ Of which: Financial ......................... Non-financial................... Total M3 deposits .............................. 27,4 41,4 10,1 31,2 68,7 38,5 6,7 -26,1 32,8 45,2 11,5 35,9 11,5 24,4 47,4 31,4 -29,7 -0,1 -29,5 1,7 32,1 67,9 37,6 30,3 100,0 33,0 67,0 37,3 29,7 100,0 In a statistical sense, growth in M3 during the second quarter of 2016 could be attributed to growth in claims on the government sector, an increase in net other assets of the monetary sector, as well as an increase in net foreign assets. However, this was largely offset by a reduction in claims on the private sector. Net claims on the government sector increased due to a rise in the holdings of government securities by the banking sector in the second quarter of 2016. Net foreign assets increased due to a decline in the foreign liabilities of the monetary sector, which exceeded the simultaneous decline in foreign assets. The gold and foreign-exchange assets of the monetary sector were devalued as a result of the appreciation of the exchange value of the rand over the period. Liabilities in respect of derivative instruments entered into with the foreign sector and foreign-sector funding were also impacted by the appreciation of the exchange value of the rand. Claims on the private sector contracted in the second quarter of 2016, partly as a result of a decline in demand for credit by the corporate sector. Statistical counterparts of change in M3 R billions 2015 2016 3rd qr 4th qr 1st qr 2nd qr Net foreign assets ........................................................ Net claims on the government sector ........................... Claims on the private sector ......................................... Net other assets ........................................................... Total change in M3 ...................................................... 48,1 40,9 65,1 -85,3 68,7 48,0 -13,1 87,8 -77,5 45,2 -24,7 -0,4 78,8 -6,3 47,4 7,8 9,1 -19,0 3,8 1,7 46 Quarterly Bulletin September 2016 Households Financial companies Non-financial companies

Credit extension Year-on-year growth in bank credit extended to the domestic private sector moderated somewhat in the second quarter of 2016 to an average of about 7 per cent, following stable and benign growth of around 8 per cent since 2012. While there was some moderation of demand for credit by the corporate sector during this period, the slowdown was also partly affected 4 In April 2016 the newly restructured African Bank was included in the banking sector surveys while the old entity, of which the data had been held constant since July 2014, was removed. Due to, among other things, the repayment of loans by clients and the transfer of selected assets to another entity, the balance sheet of the new bank differs substantially from that of the old entity. 4 by a technical correction in the data. Growth in credit extension to the household sector has been exceptionally weak since 2014 and even after adjusting for the technical change, remains negative in real terms. Amid a weak economic growth environment the cumulative increase of 200 basis points in the repurchase rate since early 2014 and concomitant increases in lending rates have added to the pressure on both households and corporates, especially those exposed to high debt levels. Financial strain on households has been further exacerbated by rising prices of food and electricity as well as weak employment prospects. Bank loans and gross domestic product Year-on-year percentage change 18 16 14 12 10 8 6 4 2 0 -2 -4 -6 2010 2011 2012 2013 2014 2015 2016 Growth in total loans and advances moderated in the second quarter of 2016 while that of nominal GDP firmed somewhat. As a consequence, the ratio of credit to GDP declined in the second quarter of 2016. Already weak consumer confidence slumped further in the second quarter of 2016, with consumer concerns about their finances translating into subdued demand for credit. At the same time business confidence also fell to its lowest level since the fourth quarter of 2009. However, corporate credit demand remained relatively strong, probably reflecting a mix of pockets of buoyancy, the continuation of projects already embarked upon, expectations of improved future conditions, and strain from current weak economic activity. Total loans and advances to the domestic private sector decreased by R23,7 billion in the second quarter of 2016, as opposed to an increase of R94,8 billion recorded in the first quarter of the year. The decline in the second quarter of 2016 was largely due to the incorporation of the newly available data on the restructured African Bank in the credit aggregates. This technical correction of the data had its largest impact on credit extension to the household sector, which plummeted in the second quarter of 2016. This was also mirrored in the quarter-to-quarter seasonally adjusted and annualised growth rate in total loans and advances to the private sector, which declined from 8,0 per cent in the first quarter of 2016 to 3,7 per cent in the second quarter. With the balance sheet of the ‘old’ African Bank held constant in compiling the monetary and credit aggregates, twelve-month growth in general loans to the household sector gradually edged higher from a lower turning point of virtually no growth in July 2014. Then in April 2016 the data for the newly restructured African Bank was incorporated into the analyses while the old entity was removed. Due to, among other things, the repayment of loans by clients and 47 Quarterly Bulletin September 2016 Loans to companies Loans to households Nominal gross domestic product Adjustment for African Bank phased in from September 2014

the transfer of selected assets to a non-bank entity, the balance sheet of the new bank was smaller and differed substantially from that of the old entity. In the data reported by the new bank a lower value was especially evident in general loans to households, as it reflected client repayments over the period as well as an amount in bad debt transferred to a non-bank residual company. When the correction of the data is phased in from September 2014, it indicates that growth in general loans to households was negative in 2015 and only barely turned positive in early 2016. Using the adjusted value, growth in general loans to households would have measured 0,5 per cent in July 2016. On balance, the uptake of bank credit by households remained muted in recent months. General loans to the household sector Percentage change over twelve months 40 35 30 25 20 15 10 5 0 -5 -10 2010 2011 2012 2013 2014 2015 2016 Growth in mortgage advances levelled off in the first half of 2016, after accelerating somewhat in 2015. Twelve-month growth increased from 4,4 per cent in February 2015 to 6,0 per cent in September 2015, subsequently remaining at around that level to June 2016, with a slight moderation to 5,7 per cent in July. Growth continued to be biased towards commercial property, while growth in mortgage advances on residential property was more subdued. While capital repayments on residential property tend to curtail actual growth in mortgage advances, these repayments have been lower in the first six months of 2016 when compared with the same period in 2015. This slowdown is partly explained by the impact of rising interest rates which places greater strain on consumers’ ability to service their debt and repay more capital than contractually required. The rate of expansion in instalment sale credit and leasing finance, which mainly represents the financing of new and second-hand vehicles, slowed further in the first half of 2016. Growth over twelve months decelerated from 6,5 per cent in January 2015 to 2,4 per cent in May 2016, consistent with the lower level of sales of especially new vehicles. Growth decelerated further to 1,0 per cent in July but was affected by a securitisation transaction of R6,7 billion. Instalment sale financing of new passenger vehicles has been contracting since the second half of 2015, whereas the financing of used passenger vehicles fluctuated broadly sideways over the same period. Not unexpectedly, recent surveys showed the business confidence of new vehicle dealers at depressed levels while the confidence among used vehicle traders also declined in the second quarter. Corporate use of bank credit remained fairly buoyant over the year to July 2016, but with a softer undertone, as shown in the accompanying graph. The moderation occurred against the backdrop of rising interest rates and weak economic activity. 48 Quarterly Bulletin September 2016 Adjustment for African Bank phased in from September 2014

Selected loans and advances extended to corporate sector Percentage change over twelve months 35 30 25 20 15 10 5 0 -5 -10 -15 -20 2010 2011 2012 2013 2014 2015 2016 Credit extension to the household sector remained fragile in the first half of 2016, although slightly higher than in the previous year. After adjusting for the technical effect caused by the restructured African Bank, year-on-year growth in loans to households picked up from 2,3 per cent in May 2015 to 3,5 per cent in February 2016, before moderating to 2,6 per cent in July. Growth in residential mortgage advances inched higher, while the rate of increase in overdrafts accelerated notably but from a low base. Growth in general loans moved from negative to slightly positive. Consumers remained constrained by rising interest rates and sluggish income growth in an environment of suppressed employment prospects and inflationary pressures. Tighter national credit regulations, such as greater emphasis on affordability assessment, also constrained credit extension to households. Selected loans and advances extended to household sector Percentage change over twelve months 50 40 30 20 Instalment sale finance 10 0 advances -10 -20 2010 2011 2012 2013 2014 2015 2016 49 Quarterly Bulletin September 2016 and General loans (adjusted for African Bank) Overdrafts credit leasing Mortgage General loans Mortgage advances Instalment sale credit and leasing finance Overdrafts

Viewed by economic sector, rail and transport companies, together with companies involved in mobile telecommunications and networks continued to increase their demand for loans in the second quarter of 2016. The electricity sector’s demand for financing also increased as this sector benefited from the various phases of the renewable energy projects initiated by government. The real-estate sector continued to benefit from the expansion of portfolios by listed property funds. Growth in bank credit to the manufacturing sector has edged higher since the final quarter of 2015, probably indicating the impact of more lucrative export opportunities, higher selling prices and expectations of an improvement in demand for manufactured goods. By contrast, alongside subdued mining production and following several years of declining international prices of commodities, the mining sector recorded negative growth in both the first and second quarters of 2016. Growth in bank credit by economic sector over the year to June 2016 Percentage of total credit extension Sector Percentage change Transport, storage and communication ................................................. Electricity, gas and water....................................................................... Real estate ............................................................................................ Manufacturing ....................................................................................... Finance and insurance .......................................................................... Agriculture, forestry and fishing.............................................................. Construction.......................................................................................... Wholesale and retail trade ..................................................................... Community, social and personal services .............................................. Business services .................................................................................. Mining and quarrying ............................................................................. Households ........................................................................................... Other ..................................................................................................... Total ...................................................................................................... 29 27 17 16 15 11 9 7 7 5 -19 5 3 9 3,3 1,7 9,8 4,7 19,6 2,1 1,0 4,9 8,4 3,1 2,1 34,6 4,6 100,0 Interest rates and yields The MPC kept the repurchase rate unchanged at 7,0 per cent per annum at its July 2016 meeting. The committee remained concerned about the weak economic growth outlook and the medium-term inflation trajectory which was expected to remain above the target range of 3–6 per cent until the second half of 2017. The July meeting expected inflation to average 6,6 per cent in 2016 and 6,0 per cent in 2017, compared with 6,7 per cent and 6,2 per cent projected at the May MPC meeting. The domestic economic growth outlook was seen as extremely challenging, following the contraction in real GDP in the first quarter of 2016. Although this was anticipated to have been the lower turning point in the cycle, the recovery from that low base was expected to be weak. The Bank’s latest forecast is for zero per cent growth in 2016, compared with 0,6 per cent previously expected. For the complete July 2016 MPC statement, see page 75 of this Quarterly Bulletin. Money-market interest rates stabilised after adjusting to the repurchase rate increase announced at the March 2016 MPC meeting, remaining generally aligned with the unchanged policy stance of the subsequent period. For example, the three-month Johannesburg Interbank Average Rate (Jibar) increased by only 14 basis points from 7,23 per cent on 18 March to 7,37 per cent on 27 June. The rate declined slightly to 7,35 per cent on 13 July 2016 and has remained range-bound since then. By 7 September, the rate stood at 7,36 per cent. 50 Quarterly Bulletin September 2016

Money-market rates Per cent 9,0 8,5 8,0 7,5 7,0 6,5 6,0 5,5 5,0 bills 4,5 Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul 2016 Sep Nov 2014 2015 Over the same period the twelve-month Jibar initially increased by 23 basis points from 8,58 per cent on 18 March 2016 to 8,81 per cent on 19 May. The rate then edged lower alongside the appreciation of the exchange value of the rand, reaching a low of 8,40 per cent on 18 August. Thereafter, the rate fluctuated higher to 8,54 per cent on 7 September. The tender rate on 91-day Treasury bills exhibited similar behaviour when it increased by 28 basis points from 7,11 per cent in late March 2016 to 7,39 per cent on 8 July, before declining to 7,28 per cent on 12 August. In early September the rate increased to 7,36 per cent. Overall, the rate on 91-day Treasury bills had increased by 50 basis points since late January 2016, significantly less than the 75 basis points increase in the repurchase rate over the same period. However, 91-day Treasury bills had already increased in mid-December 2015, when it discounted a potential increase in the repurchase rate. The South African Benchmark Overnight Rate on Deposits (Sabor) displayed limited volatility from April to early September 2016, varying by only about 10 basis points between 6,88 and 6,98 per cent and remaining well within the standing facility band. This stability reflected orderly conditions prevailing in the interbank market. The implied rate on one-day rand funding in the foreign-exchange swap market (overnight FX rate) displayed ongoing bouts of volatility. After briefly declining to below 7 per cent in early August 2016 as the exchange value of the rand appreciated, the overnight FX rate fluctuated between 7,15 and 7,58 per cent from 11 August to 7 September. Benchmark overnight rates Per cent 12 10 8 6 4 2 2014 2015 2016 51 Quarterly Bulletin September 2016 Standing facility upper band Overnight FX rate Sabor Standing facility lower band 91-day Treasury 12-month Jibar 6-month Jibar 3-month Jibar Repurchase rate

Rates on longer-term forward rate agreements (FRAs) trended lower in recent months. This reflects that market expectations regarding the pace and magnitude of future interest rate increases has moderated. Forward rates benefited when international rating agencies kept South Africa’s investment grade credit rating unchanged towards the middle of 2016, and were further aided by the appreciation of the exchange value of the rand and somewhat more favourable inflation outcomes recently. In addition, there has been a general improvement in investor appetite for emerging-market assets following growing expectations that global monetary policy will remain accommodative for longer. Prominent movements were recorded by the 9x12 month FRA which declined from 8,13 per cent on 18 May 2016 to a recent low of 7,40 per cent on 18 August. Subsequently, the rate increased to 7,60 per cent on 7 September, following the recent depreciation of the exchange value of the rand. The shorter-dated forward rates trended broadly sideways between April and early September 2016. Forward rate agreements Per cent 9,0 8,5 8,0 7,5 7,0 6,5 6,0 5,5 5,0 4,5 Jan Mar May Jul 2014 Sep Nov Jan Mar May Jul 2015 Sep Nov Jan Mar May Jul Sep Nov 2016 1x4 months 3x6 months 6x9 months 9x12 months The prime lending rate and the predominant rate on mortgage loans remained at 10,50 per cent since late March 2016, consistent with the unchanged repurchase rate. Rates on the different loan categories of private-sector banks also remained fairly stable during the period. Despite the increases in the policy rate in 2016, the inflation-adjusted or real repurchase rate of the Bank remained relatively low; it declined to 0,9 per cent in July, down from a recent high of 1,8 per cent recorded in February 2015. Interest rates on RSA fixed-rate retail bonds, being priced-off the South African government bond yield curve declined by between 100 and 125 basis points from February to August 2016. This decline brought the rate on the 2-year bond to 8,0 per cent, the 3-year bond to 8,50 per cent and the 5-year bond to 8,75 per cent in August 2016. Having decreased since mid-2012, the amount in issue of all retail bonds rebounded from R7,0 billion in December 2015 to R9,1 billion in July 2016 as government started investing accumulated housing subsidies of employees, in the interim, in retail bonds. 52 Quarterly Bulletin September 2016

Amount in issue of South African retail bonds R billions 12 11 10 9 8 7 6 5 4 3 2016 2010 2011 2012 2013 2014 2015 South African government bond yields trended generally lower in the first eight months of 2016, although abrupt increases were observed in May and late-August. While financial markets were unsettled by Britain’s decision to exit from the European Union, lower domestic bond yields were supported by the following factors: – – – – – the appreciation of the exchange value of the rand; various monetary stimulus packages being implemented by advanced economy central banks; global investor’s support of emerging-market bonds in a search for yield; generally lower consumer price inflation since March 2016; and continued weak economic growth. The increases in yields, in May and late-August 2016, were mainly related to negative perceptions arising from political developments in South Africa. As shown in the accompanying graph, these developments simultaneously led to rand depreciation in the foreign-currency market. Government bond yield and the exchange rate Per cent Rand per US dollar 17 16 15 14 13 12 11 10 9 8 7 6 11 10 9 8 7 6 2010 2011 2012 2013 2014 2015 2016 The daily closing yield of South African ten-year government bonds declined by 172 basis points from a high of 10,30 per cent on 11 December 2015 to 8,58 per cent on 7 September 2016. This was in line with the trend in bond yields in many international markets which reached record-lows during 2016. For example, the daily closing yield of US 10-year government bonds decreased by 77 basis points from 2,31 per cent on 30 December 2015 to 1,54 per cent on 7 September 2016. 53 Quarterly Bulletin September 2016 Exchange rate of the rand (right-hand scale) South African ten-year government bond yield

Break-even inflation rates, measured as the difference between the nominal yield on conventional government bonds and the real yield on inflation-linked government bonds, have been on a downward trend throughout most of 2016. This reflected a positive inflation outlook mainly as a result of the generally somewhat stronger exchange value of the rand. However, in May and late-August 2016 the downward trend in break-even inflation rates was briefly interrupted largely due to rand depreciation, political developments and rising international oil prices. Subsequently, break-even inflation rates declined once more. The real yield on the R197 inflation-linked bond declined by only 42 basis points from 11 December 2015 to 7 September 2016, whereas the nominal yield on the conventional R2023 bond declined by 138 basis points over the same period. The break-even inflation rate in the seven-year maturity range accordingly declined from 7,36 per cent on 11 December 2015 to 6,44 per cent on 7 September 2016. The yield curve flattened during 2016 as it moved lower for most maturities while the extreme short-end increased in line with the higher repurchase rate. This caused the yield gap, measured as the difference between yields at the extreme long and short ends of the curve, to narrow significantly as the differential more than halved from 500 basis points on 11 December 2015 to 236 basis points on 7 September 2016. The yield on South African US dollar-denominated bonds followed the same trend as rand-denominated government bond yields, but declined more during 2016, driven by international 5 The differential between the South African government bond yield on rand-denominated debt issued in the domestic market and the yield on South African government US dollar-denominated bonds issued in the United States, both in the nine-to-ten-year maturity range. 5 investors’ search for yield. This contributed to the widening of the currency risk premium on South African government bonds from 387 basis points in January 2016 to 476 basis points in August. 6 The JPMorgan Emerging Markets Bond Index Plus (EMBI+) yield spread above US government bonds narrowed from 443 basis points in January 2016 to 350 basis points in August. This narrowing was mainly supported by investors’ appetite for higher returns on US dollar-denominated emerging-market debt as weak US economic data signalled possible delays in interest rate increases in the near-term, coupled with monetary stimulus by advanced 6 EMBI+ measures total returns on US dollar-denominated debt instruments of emerging-market economies. 7 economy central banks. Similarly, the sovereign risk premium on South African government US dollar-denominated bonds in the nine-year maturity range narrowed from 349 basis points to 228 basis points over the same period. Money market The actual daily liquidity requirement of private-sector banks varied between a low of R44,5 billion and a high of R57,6 billion during the second quarter of 2016, somewhat higher and wider than the range of between R47,0 billion and R54,3 billion in the first quarter. The gradual longer-term rise in the actual daily liquidity requirement of the private-sector banks conforms to the strategy introduced in recent years to improve the broad alignment of the money-market shortage with the underlying conditions in the money market that largely arise from the secular upward trends in notes and coin in circulation and in the banks’ required cash reserve holdings. This approach is aimed at enhancing the robustness of the monetary policy implementation framework. The current framework in South Africa is based on creating a structural shortage, which results in the banks having to refinance their operations through the Bank at the repurchase rate. Since the introduction of the new liquidity management strategy in August 2013, the actual average liquidity requirement has increased from R22,8 billion to levels of just below R60,0 billion in recent months. The increased use of foreign-exchange swaps by the Bank coincided with the implementation of the strategy, as a measure to manage the impact of the short-term swings in notes and coin on the shortage. 7 The differential between the yield on South African government US dollar-denominated bonds and that on US dollar-denominated bonds of the US government. Liquidity to the net amount of R4,6 billion was drained from the money market during the second quarter of 2016, compared with a net drainage of R0,4 billion in the first quarter. The drainage was partly related to banks increasing their cash reserve balances by R2,2 billion while government deposits placed with the Bank, related to maturing swaps of R3,3 billion, also served to tighten market conditions. Spot sales of foreign exchange by the Bank, amounting to R0,1 billion, had a minor contracting impact on money-market liquidity. 54 Quarterly Bulletin September 2016

Liquidity requirement R billions 80 150 140 70 130 60 120 50 110 100 40 90 30 80 20 70 10 60 Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov 2014 2015 2016 A contraction of notes and coin in circulation outside the Bank injected liquidity to an amount of R2,5 billion in the second quarter of 2016. To offset the factors that impacted on money-market liquidity conditions, the Bank drained a net amount of R3,2 billion through liquidity management instruments during this period. This included liquidity draining foreign-exchange swaps of R3,7 billion. In July and August 2016, liquidity management operations served to inject R7,4 billion into the market, while overall money-market conditions were characterised by a net expansion of R3,0 billion. Factors influencing money-market liquidity Notes and coin in circulation Change in cash reserve accounts Foreign-exchange transactions Government deposits with SARB Reverse repurchase transactions SARB debentures Tightening Easing Forward position (swaps) CPD call deposits Other Liquidity provided to the banking system -15 R billions -10 -5 0 5 15 10 January–March 2016 April–June 2016 55 Quarterly Bulletin September 2016 7,5 2,5 -1,4 -2,2 -0,2 -0,1 -9,3 -3,3 -2,5 -1,3 -0,9 0,9 -3,7 -3,7 4,4 0,9 5,5 1,7 -0,4 -4,6 Estimated weekly liquidity requirement Actual daily liquidity requirement Notes and coin in circulation (right-hand scale)

Supplementary auctions were initially introduced by the Bank in April 2001 as daily square-off operations to be utilised at the Bank’s discretion, in exceptional circumstances. The use of these square-off operations was discontinued in 2010 but reintroduced from 11 March 2016 to provide for instances where the banking sector ends up having a long or short cash position at the end-of-day square-off, due to factors beyond its control. Factors considered by the Bank in exercising its discretion on whether to offer a supplementary auction include instances of substantial over-or underestimation of notes and coin in circulation, or of the required cash reserve balances of the banks with the Bank, and unexpected movements in the funds of the Corporation for Public Deposits (CPD). Capital redemption payments and scheduled coupon interest payments on various government bonds amounting to R41,6 billion were effected from the government tax and loan accounts from June to August 2016, with only R311 million of this amount accruing to the Bank. Bond market Debt issuances in the domestic bond market by national government, public corporations and banks continued steadily, while that of non-bank private companies and securitised issuances trended lower. Securitisation never regained favour after the financial crisis in 2008, while the curatorship of African Bank Limited during the third quarter of 2014 was an important reason for the subsequent stagnation in bond issuances by non-bank companies. Subsequent to African Bank’s curatorship some companies have re-intermediated to bank funding as the cost of corporate debt funding in the bond market has risen compared with bank lending. Bank lending also offer borrowers more flexible tenures and shorter-term funding. Bond amount in issue by issuer type R billions 350 300 250 200 150 100 R billions 120 110 companies 100 90 80 70 60 50 2010 2011 2012 2013 2014 2015 2016 Bond prices and trading volumes were boosted by positive investor sentiment. This filtered through to a record-high value of turnover in the secondary bond market in the first eight months of 2016. The daily average turnover of R121 billion in the eight months to August 2016 was 27 per cent more than in the corresponding period of 2015, supported by a strong performance of the All-Bond Index. 56 Quarterly Bulletin September 2016 Non-bank private Securitisation Banks

Bond yields in Europe and Japan were pushed lower into negative territory in recent months following continued bond purchases by their central banks. As the global search for yield continued, issuances of rand-nominated bonds in both the European and Japanese Uridashi bond markets amounted to R29 billion in the first eight months of 2016, more than the R22 billion issued in the same period of 2015. Allowing for redemptions, net issues of such bonds amounted to around R11 billion in the eight months to August 2016. Rand-denominated bond issuances in international markets R billions 7 6 5 4 3 2 1 0 2016 2010 2011 2012 2013 2014 2015 Non-residents’ sentiment towards the domestic secondary bond market remained positive in 2016. This is reflected by exchange-reported net purchases of domestic debt securities to the value of R16 billion in the first eight months of 2016, compared with net sales of R9 billion in the same period of 2015. Some of the factors that supported the switch from net sales to net purchases of domestic bonds by foreign investors in 2016 were the search for higher yielding emerging-market assets, concerns over global growth prospects, monetary stimulus by major central banks, and the US Federal Reserve’s decision to keep policy interest rates unchanged. Share market Equity capital-raising activity by companies listed on the JSE in the domestic and international primary share markets remained on the back foot in the first eight months of 2016. Equity funding on the JSE declined by as much as 62 per cent to R52 billion during the period under review, compared with the corresponding period in 2015. This decline can mainly be attributed to sluggish domestic economic growth and uncertainty in global equity markets following Britain’s exit vote (Brexit) from the European Union. Subdued equity funding was also consistent with a reduction in the number of companies listed on the JSE as delistings exceeded new listings in the eight months to August 2016. Listed companies on the main board amounted to 325 at the x end of August 2016, with 61 companies listed on the Alternative Exchange (Alt ) and 2 on the development and venture capital boards. Despite an increase in the number of delistings, turnover in the secondary share market of the JSE remained high thus far in 2016. Turnover was supported by on average higher volatility along with a higher volume of shares traded. The daily average turnover of R24 billion traded in the first eight months to August 2016 was 24 per cent higher than in the corresponding period of 2015. Alongside relatively volatile share prices, the market capitalisation of the JSE remained at a level of R15 trillion during most of 2016. 57 Quarterly Bulletin September 2016 European bond markets Japanese Uridashi bond market

JSE share turnover and share prices R billions Index 30 60 000 25 50 000 20 40 000 15 10 30 000 5 0 20 000 2010 2011 2012 2013 2014 2015 2016 Exchange-reported data showed net sales of domestic shares by non-residents of R17 billion in the first quarter and R57 billion in the second quarter of 2016. This contributed to exchange-reported cumulative net sales of domestic shares by non-residents of R80 billion in the eight months to August 2016, compared with net purchases of R39 billion in the corresponding period of 2015. Non-residents’ loss of interest in the domestic share market may have been fuelled by, among other things, continued unfavourable and uncertain global sentiment, a weak global growth outlook and subdued domestic economic growth. The FTSE/JSE All-Share Price Index (Alsi) followed an upward trajectory in the first five months of 2016, supported by higher international commodity prices. The Alsi subsequently declined th noticeably during June 2016 to a low of 50 087 index points on the 27 of the month, following the United Kingdom’s decision to exit from the European Union as uncertainty continued to dominate global equity markets. However, the Alsi subsequently recovered, increasing by 7 per cent to 53 456 index points on 7 September as policymakers from the G-20 countries reassured the financial community that they intend to work towards supporting global growth. Furthermore, the decisions by the South African Reserve Bank and the US Federal Reserve to leave interest rates unchanged supported the rally in the domestic equity market. Share price indices in local currency Indices: 4 January 2010 = 100 200 South Africa: 175 150 125 100 75 50 2010 2011 2012 2013 2014 2015 2016 58 Quarterly Bulletin September 2016 All-Share Price Euro area: Euro Stoxx 50 United States: S&P 500 Composite China: Shanghai A Share Daily average share turnover JSE All-Share Price Index (right-hand scale)

Box 3 Price-earnings ratio of all ordinary shares listed on the JSE on the JSE. Over recent months, the reported JSE historical (trailing) PE ratio for the overall equity market 10 59 Quarterly Bulletin September 2016 1 The price-earnings (PE) valuation ratio measures how expensive a company’s share is relative to its earnings. It is calculated as the current market price of a company’s share divided by its earnings per share. Earnings per share, in turn, is an indicator of profitability, capturing the after-tax profits of the company. The PE ratio indicates the number of years an investor will have to hold the share to recover the current share price at current earnings. A higher PE ratio can be interpreted to suggest that investors anticipate higher future earnings. As a multiple of earnings, it is thus an indicator of over-or undervaluation when compared with historical averages. As stated in the footnote, it is better suited for interpretation on an individual company basis compared with its peers or peer sectors than for the market as a whole. However, the PE ratio still renders interesting valuation insights when calculated for all ordinary shares listed 2 has increased significantly from a multiple of 19,3 in October 2015 to an all-time high of 56,8 in June 2016, as shown in the accompanying graph. This resulted from a broadly sideways trend in share prices over this period with simultaneously a sharp decline in company profits. The marked contraction in earnings growth over twelve months rendered a PE ratio well in excess of its historical average of 18,25 from 2007 to date. While total year-on-year earnings contracted by 31 per cent from the end of 2015 to July 2016, share prices have simultaneously been buoyed by, in particular, quantitative easing by central banks in advanced economies and the weaker exchange value of the rand. The contraction in earnings growth was especially significant for companies in the resources sector, whose performance has been adversely affected by weak global demand for commodities and a marked decline in commodity prices, especially due to the slowdown in the Chinese economy. The mining sector was also negatively impacted by factors related to labour unrest, closures and electricity constraints. Price-earnings ratio and earnings growth RatioPercentage change over twelve months 60200 150 50 100 4050 0 30 -50 20-100 -150 -200 0-250 2007200820092010201120122013201420152016 Sources: JSE and own calculations According to the latest Bloomberg forecast, the PE ratio is expected to fall below its historical average level by the end of 2017. However, this will depend on investor demand for shares and confidence in the realisation of expected earnings growth in the resources sector in particular. A decline in the PE ratio, for the market as a whole back to its historical average, will largely depend on a recovery in resource sector earnings. This recovery in resource sector earnings started in August 2016 and contributed to a turnaround in total earnings growth and a significant decline in the PE ratio to 32,4. 1 The price-earnings ratio is best suited for comparing a company to its peers individually or peer sector. The ratio for the market as a whole should be interpreted with caution. 2 Current share price divided by earnings per share for the past twelve months. Resources earnings growth (right-hand scale) Total earnings growth (right-hand scale) Historical average PE ratio from 2007 Price-earnings ratio for all listed ordinary shares

Market for exchange-traded derivatives Domestic agricultural commodity prices generally declined during 2016. This is largely consistent with the appreciation in the exchange value of the rand and the downward trend in international agricultural prices arising from adequate global supplies. The prices of domestic grains temporarily increased in May 2016, before declining again in the ensuing three months to August. The spot price of white maize, the country’s main staple, remained in high territory despite a decline of 23 per cent from a record-high of R5 280 per ton on 20 January 2016 to R4 092 per ton on 7 September. Apart from the impact of the exchange rate, lower futures prices of some agricultural commodities may partly be aligned with a possible improvement in global weather conditions associated with a La Nina weather pattern, which tends to produce above-average rainfall and cooler temperatures. South African grain prices Rand per ton 8 500 7 500 6 500 5 500 4 500 3 500 2 500 1 500 500 2010 2011 2012 2013 2014 2015 2016 2017 Dots represent Safex futures prices as on 7 September 2016 In July 2016, the JSE introduced a Quanto Soybean Calendar Spread Can-do futures contract which is traded and settled in rand on the Commodity Derivatives Market. This contract is based on the price differential between two contract months of the reference Chicago Board of Trade soybean contract. The contract aims to provide a more precise hedge against adverse movements in price spreads in the soybean market. Turnover in the commodity derivatives market of the JSE increased briskly during the eight months to August 2016 when compared with the corresponding period of 2015 – reaching an all-time high of R118 billion in June 2016. Along with lower bond yields and exchange rate fluctuations, the value of interest rate and currency derivatives traded rose noticeably thus far in 2016, compared with the same period of 2015. However, equity derivatives trading continued to dominate overall derivatives trading on the JSE, as shown in the accompanying table. Derivatives turnover on the JSE, January to August 2016 Value (R billions) Change over one year (Per cent) Type of derivative Equity ............................................................................................. Warrants...................................................................................... Commodity.................................................................................. Interest rate ................................................................................. Currency...................................................................................... 4 403 0,4 704 770 474 7 13 45 44 44 60 Quarterly Bulletin September 2016 Sunflower seed Wheat Yellow maize White maize

Real-estate market The year-on-year rate of increase in the nominal prices of residential property remained fairly stable, ranging between 4 per cent and 7 per cent across the different measures in August 2016. The moderate pace of increase in nominal house prices was consistent with slow economic growth, weak employment and consumer confidence as well as tight household finances. The price of large houses increased at the slowest rate on a year-on-year basis in August 2016. Simultaneously, the average number of weeks that residential properties remained on the market increased from 11 weeks in the first quarter of 2016 to 13 weeks in the second quarter. However, according to First National Bank, some accelerated price growth was evident in the Western Cape Province with a year-on-year rate of increase in house prices of 12 per cent in June 2016. This increase probably resulted mainly from sound economic prospects in the region and strong net inward migration. Nominal house prices Percentage change over twelve months 50 Bank Lightstone 40 30 20 10 0 -10 Nominal house prices by province Percentage change over twelve months 40 Gauteng KwaZulu-Natal 30 20 10 0 -10 -20 2004 2006 2008 2010 2012 2014 2016 Non-bank financial intermediaries 8 The consolidated balance sheet of non-bank financial institutions expanded further in the second quarter of 2016. Total assets of these institutions increased by 2 per cent from the first quarter of 2016 to R8,5 trillion in the second quarter. This represented an increase of 8 per cent from the second quarter of 2015. Underpinned by higher asset prices, the asset base of the Public Investment Corporation and unit trusts increased the most, respectively rising by 3 per cent and 2 per cent in the second quarter of 2016. 8 These consist of unit trusts, the Public Investment Corporation, long-and short-term insurance companies, public and private pension funds, participation bond schemes, finance companies, and non-monetary public financial corporations. The asset allocation of non-bank financial institutions shows unchanged equity holdings at 57 per cent of total assets in the first and second quarters of 2016. The current equity exposure is some 3 percentage points higher than the 10-year average, suggesting that appetite for a substantial further increase in the allocation of funds into this asset class may diminish going forward. 61 Quarterly Bulletin September 2016 Oth er provinces Western Cape Eastern Cape Absa First National Standard Bank

Asset holdings of non-bank financial institutions Per cent of total assets R billions 60 9 000 8 000 50 7 000 40 6 000 5 000 30 4 000 3 000 20 2 000 10 1 000 0 0 2010 2011 2012 2013 2014 2015 2016 Shares Cash and deposits Total assets (right-hand scale) Interest-bearing securities Other assets However, factors that could favour equity weightings include an increase in share prices and a decline in equity volatility, relative to that of interest-bearing securities and other asset classes. Holdings of interest-bearing securities by these institutions increased by 1 percentage point from the first quarter of 2016 to 29 per cent of total assets in the second quarter – the highest since the third quarter of 2015. The increase in this asset class was buoyed by a decline in bond yields, with the All-Bond Index returning 4 per cent in the second quarter of 2016, following 7 per cent in the first quarter. The holding of interest-bearing securities increased by 1 percentage point in the year to date. The proportion of funds invested in cash and deposits fell by 1 percentage point from the first quarter of 2016 to 6 per cent of total assets in the second quarter. Investment in cash and deposits moderated alongside a decline in risk aversion in financial markets. Meanwhile, money-9 Gross inflows minus gross outflows. 9 market funds recorded net inflows of R11 billion in the second quarter of 2016, compared with R8 billion in the first quarter. Credit extension by non-bank financial institutions was subdued in the second quarter of 2016 with the value of loans extended by these institutions remaining unchanged at 5 per cent of total assets in both the first and second quarter of 2016. This reflected slow activity by finance companies. The change in the asset allocation of non-bank financial institutions was also influenced by an increase in the funds available for investment in the second quarter of 2016. However, weak economic conditions, job losses and a deterioration in the financial position of households had a negative impact on net inflows to long-term insurers and pension and provident funds. The wealth effects associated with the increase in the balance sheet of non-bank financial intermediaries in the second quarter of 2016 are likely to positively impact the financial assets of households. Flow of funds Capital inflows from non-residents were largely related to the purchases of interest-bearing securities and loans during the first quarter of 2016. Non-residents purchased interest-bearing securities amounting to R21 billion due to relatively more favourable domestic bond yields, while they extended loans of R44 billion. These purchase-related transaction inflows were augmented by those that emanated from the disposal and maturity of financial assets abroad by residents over the same period. The overall net capital inflow amounted to R63 billion in the first quarter of 2016, higher than the R45 billion recorded in the final quarter of 2015. 62 Quarterly Bulletin September 2016

Financial intermediaries raised funding of R62 billion from deposits and R28 billion from interest in retirement and life funds in the first quarter of 2016. In turn, financial intermediaries purchased interest-bearing securities worth R20 billion, thereby helping to finance some of the deficit of general government. An amount of R67 billion was also used to acquire equity during the first quarter of 2016, while R45 billion was extended to the economic sectors through loans. At this level, total loans extended by financial intermediaries were higher than the amount recorded during the final quarter of 2015, but still below the average level since 2010, consistent with subdued domestic economic activity. The intermediaries’ extension of bank loans and advances to households and non-financial corporate businesses amounted to R11 billion and R47 billion in the first quarter of 2016, respectively. Financial intermediation R billions 350 300 Equity 250 200 150 100 50 0 -50 2010 2011 2012 2013 2014 2015 2016 Consistent with the general government sector’s tight fiscal space and the general slowdown in domestic output, capital spending declined by 1 per cent in the first quarter of 2016 compared with the level in the final quarter of 2015. Thus, the general government’s financing shortfall narrowed from R35 billion in the fourth quarter of 2015 to R18 billion in the first quarter of 2016, allowing the fiscal authorities to incur lesser liability flows. This shortfall was mainly financed in the domestic market through long-term bond issuances amounting to R41 billion. Most of the bonds issued were purchased by non-residents and insurers and pension funds. The moderation in domestic economic activity was also reflected in the corporate business enterprises sector, as its savings slowed down while inflows from loans increased in the first quarter of 2016. This resulted in the net incurrence of financial liability flows of R220 billion, from R154 billion in the preceding quarter. Consequently, a wider deficit of R17 billion was recorded during the same period from R3 billion in the final quarter of 2015. The corporate business enterprises’ overall capital spending slowed from R136 billion in the fourth quarter of 2015 to R127 billion in the first quarter of 2016. A slower growth rate was registered in both the public and private corporations’ capital outlays. At R38 billion, public enterprises accounted for 30 per cent of the total corporate business enterprises’ capital expenditure while the private corporates represented 70 per cent in the first quarter of 2016. 63 Quarterly Bulletin September 2016 Loans Interest-bearing securities

Capital outlays of non-financial corporate business enterprises R billions 200 160 120 80 40 0 2010 2011 2012 2013 2014 2015 2016 The household sector’s flows moderated during the first quarter of 2016, weighted down by ongoing poor employment prospects, lower consumer confidence levels, rising living costs and higher interest rates. Evidently, households’ credit uptake remained hesitant in the first quarter of 2016, with bank loans and advances and mortgage borrowing registering R23 billion in the first quarter of 2016 from R20 billion in the final quarter of 2015. Over the same period, households’ deposit flows declined to R23 billion in the first quarter of 2016 from R43 billion during the preceding quarter. Consequently, the sector recorded a wider financing deficit of R65 billion in the first quarter of 2016, from R35 billion in the final quarter of 2015. 64 Quarterly Bulletin September 2016 Private companies Public corporations

Public finance10 Non-financial public-sector borrowing requirement11 The non-financial public-sector borrowing requirement amounted to R51,4 billion in the first quarter of fiscal 2016/17 (April–June 2016). This was R18,4 billion less than in the same period of the previous fiscal year. The lower non-financial public-sector borrowing requirement could primarily be attributed to the lower cash deficit of non-financial public enterprises and corporations. The consolidated general government also recorded a lower deficit in the period under review, driven mainly by the narrowing in the deficits of national and local governments. Extra-budgetary institutions, social security funds and provincial governments all recorded surpluses in April–June 2016. 10 In this section, unless stated otherwise, year-on-year rates of increase compare fiscal 2015/16 to fiscal 2016/17. 11 Calculated as the cash deficit/surplus of the consolidated central, provincial and local governments, and non-financial public enterprises and corporations. Non-financial public-sector borrowing requirement R billions Level of government Apr–Jun 2015* Apr–Jun 2016* Consolidated general government ......................................................... National government.......................................................................... Extra-budgetary institutions ............................................................... Social security funds.......................................................................... Provincial governments...................................................................... Local governments ............................................................................ Non-financial public enterprises and corporations ................................. Total** ................................................................................................... As percentage of gross domestic product ......................................... 39,9 32,7 -3,3 -5,6 -8,1 24,1 29,9 69,8 ** 7,0 34,0 24,6 -4,7 -4,7 -3,5 22,4 17,4 51,4** 4,8 * Deficit +, surplus – ** Components may not add up to totals due to rounding As a ratio of GDP, the non-financial public-sector borrowing requirement amounted to 4,8 per cent in the first quarter of fiscal 2016/17, lower than the ratio of 7,0 per cent in the same period of the previous fiscal year. Non-financial public-sector borrowing requirement R billions Per cent 250 10 200 8 150 6 4 100 50 2 0 0 2011/12 Fiscal years 2012/13 2013/14 2014/15 2015/16 2016/17* * April–June 2016, annualised Consolidated general government Non-financial public enterprises and corporations Percentage of gross domestic product (right-hand scale) 65 Quarterly Bulletin September 2016

The borrowing requirement of the consolidated general government amounted to R34,0 billion in the first quarter of fiscal 2016/17, or R5,9 billion less than in the same period of the previous fiscal year. As a ratio of GDP, the consolidated general government borrowing requirement amounted to 3,2 per cent in the first quarter of fiscal 2016/17 – lower than the 4,0 per cent in the same period of the previous fiscal year. The non-financial public enterprises and corporations showed a preliminary borrowing requirement of R17,4 billion in the first quarter of fiscal 2016/17. This was R12,5 billion less than in the same period of the previous fiscal year. Notably, both revenue and expenditure increased in comparison with the first quarter of fiscal 2015/16, with the increase in revenue of 18,8 per cent significantly more than the increase of 2,7 per cent in expenditure. Financial activities of non-financial public enterprises and corporations R billions 150 125 100 75 50 25 0 -25 -50 2012/13 Fiscal years 2013/14 2014/15 2015/16 2016/17 The total net investment in non-financial assets by non-financial state-owned enterprises and corporations amounted to R22,7 billion in the first three months of fiscal 2016/17 – R2,8 billion less than in the same period a year earlier. Among other reasons, this contraction may be attributable to a decline in infrastructure-related spending following the completion of key infrastructure building programmes. Delays in the rolling out of some capital projects undertaken by major state-owned companies also held back investment expenditure. Against the backdrop of subdued domestic and global economic activity, Transnet also announced its intentions to defer some of its capital projects outlined in its Market Demand Strategy to a later stage. Net investment in non-financial assets continued to drive the cash deficit of non-financial state-owned enterprises and corporations, notwithstanding the improvement in operating cash balances. During the first quarter of fiscal 2016/17, the operating cash flow coverage ratio of receipts to payments remained relatively stable at around 1,1 per cent. The major non-financial state-owned companies continued to source funding from the capital markets during the period from 1 April to 30 June 2016. During this period, net issuance of domestic marketable bonds by these entities amounted to R6,6 billion, some R2,7 billion more than in the same period of the previous fiscal year. Total outstanding listed debt of both financial and non-financial state-owned enterprises and corporations increased from R271 billion as at 30 June 2015 to R288 billion as at the end of June 2016, with non-financial entities accounting for 86,8 per cent of the latter total. Uncertainty over possible credit rating downgrades and the downward revisions of economic growth projections, among others, pose downside risks to capital investment and borrowing programmes of state-owned companies. Preliminary statistics of national government showed that cash receipts from operating activities amounted to R271 billion in the first quarter of fiscal 2016/17 – 9,7 per cent more than in the first 66 Quarterly Bulletin September 2016 Total operating receipts Total operating expenses Net investment in non-financial assets Cash deficit

quarter of fiscal 2015/16. The increase in national government revenue was boosted by brisk collection of taxes on international trade and transactions as well as non-tax revenue, alongside marginal to moderate increases in all other major tax categories. Cash payments for operating activities of national government amounted to R294 billion in April–June 2016, some 6,0 per cent more than in April–June 2015. Similarly, compensation of employees amounted to R35,5 billion, representing a year-on-year rate of increase of 14,0 per cent over the same period. In April–June 2016 grants paid by national government to other levels of general government amounted to R166 billion – accounting for 56,4 per cent of total cash payments for operating activities. National government’s net cash flow from operating activities, along with net investment in non-financial assets, resulted in a cash deficit of R24,6 billion in the first quarter of fiscal 2016/17 – some R8,2 billion less than in the same period of the preceding fiscal year. Preliminary data of provincial governments’ financial activities indicated a cash surplus of R3,5 billion, some R4,6 billion less than the cash surplus in the first quarter of fiscal 2015/16. The reduced cash surplus resulted from higher spending on compensation of employees in the first quarter of fiscal 2016/17. For the fiscal year 2016/17 as a whole, the Budget Review 2016 estimated that provincial governments would record a cash deficit of R1,0 billion. During April–June 2016, total provincial government cash receipts from operating activities amounted to R130 billion, representing a year-on-year rate of increase of 7,2 per cent. Grants from national government – the primary source of cash receipts of provincial governments – amounted to R126 billion or 96,8 per cent of their total cash receipts in the first quarter of fiscal 2016/17. These grants comprised of equitable share transfers and conditional grants earmarked for specific purposes. The larger portion of the equitable share, about 40,8 per cent, is distributed to the KwaZulu-Natal and Gauteng provinces. Provincial government expenditure – cash payments for operating activities together with net investment in non-financial assets – amounted to R126 billion in the first quarter of fiscal 2016/17, or 11,7 per cent more than in the same period of the previous fiscal year. Compensation of employees – which comprises the largest share of provincial government expenditure at 61,2 per cent of total spending in April–June 2016 – amounted to R77,4 billion. This represented a year-on-year rate of increase of around 10,6 per cent. The higher increase in compensation of employees resulted primarily from the implementation of a cost-of-living adjustment together with a housing allowance for eligible employees. The education and health sectors combined accounted for the largest share (around 87 per cent) of provincial government spending on compensation of employees in the first quarter of fiscal 2016/17. Provincial governments’ deposits with the end of March 2016 to R20,4 billion at the provincial government deposits with private CPD increased from R19,0 billion at the end of June 2016. Over the same period, banks also increased from R17,3 billion to R18,6 billion, while their overall indebtedness to banks decreased from R0,7 billion to R0,4 billion. The increase in the provincial governments’ deposits with the CPD and private banks was in line with the cash surplus recorded in April–June 2016. During April–June 2016, preliminary local government financial statistics indicated a cash deficit of R22,4 billion. This was R1,7 billion less than the cash deficit in April–June 2015. The decline in the local government cash deficit can be attributed to the rate of increase in cash receipts from operating activities outpacing that of total expenditure. Cash receipts from operating activities of local government amounted to R63,2 billion in April– June 2016, representing a year-on-year increase of 12,2 per cent – well above the general rate of inflation. Sales of goods and services continued to be the main driver of the increase in total cash receipts, notably sales of electricity and gas and to a lesser extent the sales of water during the period under review. Cash payments for operating activities of local government amounted to R73,4 billion in the first quarter of fiscal 2016/17, resulting in a year-on-year increase of 11,5 per cent. This increase was 67 Quarterly Bulletin September 2016

due to higher purchases of goods and services – particularly purchases of water and electricity that increased due to higher prices being charged by the relevant wholesale suppliers. Net investment in non-financial assets amounted to R12,2 billion, registering a year-on-year decrease of 16,6 per cent in April–June 2016. Composition of municipal operating revenue Preliminary estimates of financial statistics for extra-budgetary institutions indicated a cash surplus of R4,7 billion in April–June 2016 – more than the cash surplus of R3,3 billion in the same period of the previous fiscal year. The increase in the cash surplus could be attributed to higher growth in transfers from national government, which boosted cash receipts. Social security funds recorded a preliminary estimated cash surplus of R4,7 billion in the first quarter of the current fiscal year – slightly less than the cash surplus of R5,6 billion during the same period of the previous fiscal year. The decrease in the cash surplus could be attributed to a marked increase in social benefit payouts, which amounted to R9,4 billion in April–June 2016 compared to R7,6 billion in the same quarter a year earlier. Budget comparable analysis of national government finance National government expenditure increased at a much slower pace than revenue in the first quarter of fiscal 2016/17. This resulted in a narrowing of the cash-book deficit when compared with that in the first quarter of fiscal 2015/16. National government finances: key statistics, 2016/17 Year-on-year percentage change* Budgeted** Full 2016/17 Actual Apr–Jun 2016 Expenditure ...................................................................................... Revenue ........................................................................................... 6,0 8,7 7,5 11,7 Memo: cash-book deficit ................................................................ R156 billion R30 billion * Fiscal 2015/16 to fiscal 2016/17 ** Budget Review 2016 68 Quarterly Bulletin September 2016 April–June 2015April–June 2016 11,8% 13,6% 1,4% 4,3% 5,8% 18,8% 59,8% 22,1% 58,6% Service chargesProperty rates GrantsProperty income Other

National government expenditure amounted to R291 billion in April–June 2016, or 7,5 per cent more than in April–June 2015. Relative to GDP, national government expenditure reached 27,2 per cent in the first quarter of fiscal 2016/17 – broadly unchanged when compared with the same period a year earlier. The Budget Review 2016 projected that total national government expenditure would amount to R1 318 billion in fiscal 2016/17, representing a year-on-year rate of increase of 6,0 per cent. National government expenditure in fiscal 2016/17 Originally budgeted Full 2016/17 Actual Apr–Jun 2016 Expenditure item Percentage change* Percentage change* R billions R billions Voted amounts ................................................... Of which: Current payments............................... Transfers and subsidies ....................... Payments for capital assets ................ Statutory amounts** ........................................... Of which: Interest on debt............................... Total expenditure ............................................... 721,1 208,6 493,3 14,4 597,2 147,6 1 318,3 3,2 7,5 7,6 -16,6 9,6 14,6 6,0 161,6 49,3 109,4 2,2 129,5 22,3 291,1 7,0 14,2 3,1 20,0 8,3 14,1 7,5 * Fiscal 2015/16 to fiscal 2016/17 ** Including extraordinary payments The year-on-year increase in national government expenditure in the first quarter of fiscal 2016/17 was underpinned by higher voted expenditure and interest payments on debt. Voted expenditure consists mostly of current payments and transfers and subsidies, which accounted for about 98 per cent of total voted spending in the first quarter of fiscal 2016/17. The remaining 2 per cent of voted expenditure comprise payments for capital and financial assets. Composition of national government expenditure in April–June 2016 0,8% Current payments increased markedly at a year-on-year rate of 14,2 per cent in the first quarter of fiscal 2016/17. For fiscal 2016/17 as a whole, these payments are expected to increase by 7,5 per cent as indicated in the Budget Review 2016. The departments of Police, Defence and Military Veterans, and Correctional Services, which all form part of the Justice, Crime Prevention and Security cluster, contributed 66,5 per cent to total current payments. 69 Quarterly Bulletin September 2016 1,5% 16,9% 35,3% 37,6% 7,7% 0,3% Current paymentsTransfers and subsidies Payments for capital assetsPayments for financial assets Interest paymentsEquitable share transfers to provinces Other expenditure

Transfers and subsidies increased at a year-on-year rate of 3,1 per cent in the first quarter of fiscal 2016/17. The Budget Review 2016 projected that this spending category would record a year-on-year rate of increase of 7,6 per cent. Transfers and subsidies were mainly to the departments of Education, Health, Social Development and Transport, which accounted for 75,7 per cent of total transfers and subsidies. Payments for capital assets increased at a solid year-on-year rate of 20,0 per cent in April–June 2016. Notwithstanding this increase, this category of expenditure is expected to record an annual decline of 16,6 per cent, according to the Budget Review 2016. Payments for financial assets increased significantly during the period under review, although this spending category only accounted for 0,5 per cent of voted expenditure and 0,3 per cent of total expenditure. Interest paid on national government debt amounted to R22,3 billion in the first three months of fiscal 2016/17, some 14,1 per cent more than in the first three months of fiscal 2015/16, and consistent with the rising level of government debt. For the whole fiscal 2016/17, the Budget Review 2016 projections are for the interest payments on debt to amount to R148 billion, representing a year-on-year rate of increase of 14,6 per cent. Equitable share transfers to provinces totalled R103 billion in the first quarter of fiscal 2016/17, or 7,3 per cent more than in the corresponding period of the preceding fiscal year. In the Budget Review 2016, equitable share transfers to provinces are projected to amount to R411 billion in fiscal 2016/17, or 6,3 per cent more than in fiscal 2015/16. 12 These are adjustments to total expenditure arising from timing differences between the recording of transactions and bank clearances, along with late departmental requests for funds. 12 After taking into account cash-flow adjustments , the cash-flow expenditure of national government amounted to R278 billion in the first quarter of fiscal 2016/17 – some 6,0 per cent more than in the first quarter of fiscal 2015/16. 13 13 In the calculation by the Bank, revenue excludes premiums on debt portfolio restructuring and loan transactions. These came to R1,6 billion in April–June 2016. National government revenue amounted to R261 billion in the first three months of fiscal 2016/17, or 11,7 per cent more than in the first three months of fiscal 2015/16. The solid increase recorded in national government revenue could mainly be attributed to strong collections of import duties and personal income tax. A significant increase in non-tax revenue also contributed to the growth in national government revenue. National government revenue as a percentage of GDP amounted to 24,5 per cent in April–June 2016, higher than the revenue ratio of 23,5 per cent recorded in April–June 2015. The Budget Review 2016 projected that national government revenue would reach R1 162 billion in fiscal 2016/17, representing a year-on-year rate of increase of 8,7 per cent. Taxes on income, profits and capital gains increased at a solid year-on-year rate of 7,6 per cent in the first quarter of fiscal 2016/17. According to the Budget Review 2016, this tax category is expected to record a year-on-year rate of increase of 10,1 per cent for the whole of fiscal 2016/17. Personal income tax collections defied the difficult economic conditions and rose at a double-digit rate but were nevertheless lower than expected, owing to lower-than-anticipated pay-as-you-earn receipts from the electricity, gas and water sector. Consistent with the subdued economy, corporate income tax collections increased at a sluggish year-on-year rate of 3,3 per cent, almost in line with budgeted projections. Brisk collections of taxes on property stemmed from a significant year-on-year rate of increase of 17,2 per cent in transfer duties in the first quarter of fiscal 2016/17. However, the strong increase in transfer duties was partly offset by a decrease in securities transfer tax collected. 70 Quarterly Bulletin September 2016

Composition of gross tax revenue in April–June 2016 Taxes on goods and services recorded a marginal year-on-year rate of increase in the first three months of fiscal 2016/17. Net value-added tax receipts remained broadly unchanged in the first quarter of fiscal 2016/17 as both domestic VAT and import VAT collections registered year-on-year increases, but was countered by a solid rise in VAT refunds mainly to the finance, wholesale and retail trade and the manufacturing sectors. Proceeds from the general fuel levy and excise duties, which also contribute to taxes on goods and services, recorded year-on-year rates of increase. These increases were, however, weighed down by the sluggish VAT collections. National government revenue in fiscal 2016/17 Originally budgeted Full 2016/17 Actual Apr–Jun 2016 Revenue source Percentage change* Percentage change* R billions R billions Taxes on income, profits and capital gains ......... Income tax on individuals ................................ Income tax on companies ............................... Payroll taxes ....................................................... Taxes on property............................................... Taxes on goods and services ............................. Value-added tax (VAT) ..................................... Domestic VAT ............................................. Import VAT.................................................. Taxes on international trade and transactions ..... Import duties.................................................. Other revenue** .................................................. Less: SACU*** payments.................................... Total revenue ..................................................... 668,4 442,2 200,9 17,6 15,5 418,6 301,3 322,4 164,0 54,5 54,0 26,8 39,4 1 162,0 10,1 13,6 3,9 15,9 2,7 8,5 7,2 8,4 8,8 16,0 16,8 -46,0 -22,7 8,7 156,5 97,3 51,9 3,4 3,8 83,6 57,2 75,7 29,4 9,6 9,4 14,4 9,9 261,4 7,6 10,2 3,3 -5,6 7,2 1,6 0,0 6,2 8,5 19,2 19,5 230,2 -25,7 11,7 * ** Fiscal 2015/16 to fiscal 2016/17 Including extraordinary receipts, but excluding premiums on debt portfolio restructuring and loan transactions that amounted to R1,6 billion in April-June 2016 Southern African Customs Union *** 71 Quarterly Bulletin September 2016 3,7% 3,7% 5,3% 5,6% 37,9% 22,3% 20,2% 1,5% Personal income taxCorporate income tax Taxes on property Value-added tax General fuel levySpecific excise duties Import dutiesOther tax revenue

Taxes on international trade and transactions outstripped domestic inflation by a large margin and increased at a year-on-year rate of 19,2 per cent in April–June 2016, owing to robust collections of import duties. The increase in import duties during the period under review emanated mainly from a relatively weaker domestic currency. During the period under review, non-tax revenue increased substantially on a year-on-year basis due to higher national revenue receipts arising from revaluation profits on foreign currency investment transactions. Cumulative deficit of national goverment R billions 200 160 120 80 40 0 A M J J A S O N D J F M Fiscal years Originally budgeted deficit of R156 billion for fiscal 2016/17 The net result of national government revenue and expenditure was a cash-book deficit of R29,7 billion in the first quarter of fiscal 2016/17 – some R7,0 billion less than the cash-book deficit recorded in the first quarter of fiscal 2015/16. The narrowing of the cash-book deficit was driven by the higher rate of increase recorded in national government revenue, which exceeded the increase recorded in expenditure. As a ratio of GDP, the cash-book deficit amounted to 2,8 per cent in April–June 2016 – lower than the deficit of 3,7 per cent recorded in April–June 2015. In the Budget Review 2016, the cash-book deficit for the full fiscal 2016/17 is projected to amount to R156 billion, or 3,6 per cent of GDP. The primary balance (the deficit/surplus calculated by excluding interest payments from total expenditure) recorded a deficit of R7,4 billion in the first quarter of fiscal 2016/17 or 0,7 per cent of GDP. This demonstrates progress in the process of fiscal consolidation when compared with the primary deficit of R17,2 billion, or 1,7 per cent as a ratio of GDP, registered in the first quarter of fiscal 2015/16. The cash-flow deficit of national government amounted to R15,5 billion in April–June 2016, some R7,8 billion lower than the amount recorded in the same period of the previous fiscal year. The lower cash-flow deficit resulted from a year-on-year increase in cash-flow revenue which outpaced that of expenditure. After accounting for the cost on revaluation of foreign debt at maturity, the net borrowing requirement of national government amounted to R23,2 billion in April–June 2016, marginally lower than the amount recorded during the same period of 2015. 72 Quarterly Bulletin September 2016 Actual 2015/16 Actual 2016/17

National government financing in fiscal 2016/17 R billions Originally budgeted Actual Apr–Jun 2015 Actual Apr–Jun 2016 Item or instrument 1 Full 2016/17 3 23,3 4 156,3 3 15,5 Deficit2 ...................................................................... Plus: Cost/profit on revaluation of foreign debt at redemption5 ............................................................... Net borrowing requirement ...................................... 0,4 8,1 7,7 23,7 164,5 23,2 Treasury bills.............................................................. 1,7 25,0 12,1 Domestic government bonds .................................... 37,0 116,2 34,3 Foreign bonds and loans ........................................... -0,9 15,9 11,6 Change in available cash balances6 ........................... Total net financing7 ................................................... -14,1 23,7 7,3 164,5 -34,9 23,2 1 2 3 4 5 6 7 Budget Review 2016 Including extraordinary receipts and payments Cash-flow deficit which differs from the cash-book deficit Cash-book deficit Cost + profit - Increase – decrease + Components may not add up to totals due to rounding The borrowing requirement was predominantly financed domestically through the issuance of rand-denominated bonds. Government’s net issues of such bonds amounted to R34,3 billion in the first quarter of fiscal 2016/17. Of this amount, inflation-linked bonds accounted for R21,8 billion, bringing the outstanding balance of these instruments to R431 billion at the end of June 2016. These bonds represent 26,4 per cent of total domestic marketable bonds in issue. Net issues of Treasury bills and short-term loans from the CPD amounted to R12,1 billion in the first three months of fiscal 2016/17. National government plans to raise some R116 billion through domestic long-term government bonds and R25 billion through Treasury bills in fiscal 2016/17. Treasury bills were issued at an average interest rate of 6,7 per cent in the first quarter of fiscal 2016/17, while the average yield on funds raised through conventional bonds amounted to 8,7 per cent and the average real yield on inflation-linked bonds to 1,8 per cent per annum. The average unexpired maturity of national government’s domestic marketable bonds increased slightly from 179 months as at 31 March 2016 to 181 months as at 30 June 2016. Net issuance of foreign bonds and loans amounted to R11,6 billion in the first quarter of fiscal 2016/17. Of this amount, R17,9 billion or US$1,25 billion were proceeds resulting from the 10-year USA bond issued with a coupon rate of 4,875 per cent on 14 April 2016 and the difference constituted redemptions of bonds that had been issued previously. The average unexpired maturity of the foreign marketable bonds of national government increased from 102 months as at 31 March 2016 to 107 months as at the end of June 2016, pushed higher by the newly issued bond that initially had a maturity of 120 months. National government’s available cash balances increased by R34,9 billion in the first quarter of fiscal 2016/17, bringing these balances to R241 billion at the end of June 2016. Of this total, deposits with the Bank accounted for 72,2 per cent or R174 billion, while the remaining balances were held with commercial banks. National government’s domestic debt, which accounts for about 90 per cent of total gross loan debt, increased from R1 819 billion to R1 894 billion from the end of March 2016 to the end of June. The increase in national government debt during this period could mainly be attributed to a surge in net-issuance of domestic marketable bonds and Treasury bills. As a ratio of GDP, domestic debt increased from 44,5 per cent at the end of March 2016 to 45,5 per cent at the end of June 2016. The Budget Review 2016 estimated this ratio at 45,7 per cent for fiscal 2016/17. 73 Quarterly Bulletin September 2016

Foreign debt constitutes approximately 10 per cent of the government’s total gross loan debt, and increased from R200 billion at the end of March 2016 to R205 billion at the end of June 2016 on account of the combined impact of revaluations and net issuances. As a percentage of GDP, foreign debt remained broadly unchanged at 4,9 per cent between March and June 2016. The Budget Review 2016 estimated this ratio at 5,2 per cent for fiscal 2016/17. Total gross loan debt of national government R billions Per cent 60 2 500 50 2 000 40 1 500 30 1 000 20 500 10 0 0 2006/07 2008/09 2010/11 2012/13 2014/15 2016/17 As at 31 March of each fiscal year * As at 30 June 2016 Originally budgeted Actual Percentage of gross domestic product (right-hand scale) Total gross loan debt of national government, consisting of all domestic and foreign debt instruments, amounted to R2 099 billion at the end of June 2016, relative to R2 019 billion recorded at the end of March 2016. As a ratio of GDP, total gross loan debt rose from 49,4 per cent to 50,5 per cent between March and June 2016. The Budget Review 2016 projected this ratio at 50,9 per cent for fiscal 2016/17 as a whole. Subtracting the government’s deposits with the Bank and the private banking sector, national government’s net loan debt rose by R45 billion over the course of the first quarter of fiscal 2016/17 and amounted to 44,7 per cent of GDP at the end of June 2016. 74 Quarterly Bulletin September 2016 *

Statement of the Monetary Policy Committee 21 July 2016 Issued by Lesetja Kganyago, Governor of the South African Reserve Bank, at a meeting of the Monetary Policy Committee in Pretoria The United Kingdom (UK) vote to leave the European Union has dominated the global landscape in the past month. The outlook for the global economy has become more uncertain as the potential consequences of this development are being assessed. The lack of clarity regarding the process going forward has had significant implications for global growth and interest rates as the prospects for further near-term monetary policy tightening by the United States (US) Federal Reserve (Fed) recede. While there have been spillovers to the South African financial markets, particularly the exchange rate, the direct short-term impact on South African growth and trade is likely to be fairly limited. Domestic growth has surprised further on the downside, and the outlook remains constrained. At the same time, domestic inflation outcomes have surprised marginally on the downside, but an extended breach of the target is still expected. The year-on-year inflation rate as measured by the consumer price index (CPI) for all urban areas moderated to 6,1 per cent in May, before rising to 6,3 per cent in June. Food price inflation measured 11,0 per cent in June, below the recent peak of 11,3 per cent in April, and was marginally negative on a month-on-month basis. Goods price inflation measured 6,7 per cent, up from 6,6 per cent in May, while services increased by 5,8 per cent, up from 5,7 per cent in May. The Bank’s measure of core inflation, which excludes food, fuel and electricity, measured 5,6 per cent, up from 5,5 per cent previously. Producer price inflation for final manufactured goods continued its downward trend, declining from 7,0 per cent in April to 6,5 per cent in May, mainly due to a further moderation in price inflation of food, beverages and tobacco products. This was below the consensus forecast of 6,9 per cent. The impact of the drought is still evident in the increase in prices of agricultural products, particularly cereals and other crops as well as live animals and animal products. The latest inflation forecast of the Bank shows a marginal improvement compared to the previous forecast. Nevertheless, inflation is still expected to accelerate further this year and is only expected to return to within the target range of 3–6 per cent during the third quarter of 2017. Inflation is now expected to average 6,6 per cent in 2016 and 6,0 per cent in 2017, compared with 6,7 per cent and 6,2 per cent previously. In 2018 inflation is expected to average 5,5 per cent, marginally up from the previous forecast. The expected peak, at 7,1 per cent in the fourth quarter of 2016, is down from 7,3 per cent. The downward revisions are partly due to lower administered price inflation (mainly petrol), despite a small upward adjustment in the international oil price assumption. The forecast for core inflation is lower than the previous forecast in the near term, mainly due to the lower starting point, but higher in the outer period. Core inflation is expected to moderate from an average of 5,8 per cent in 2016 to 5,3 per cent by 2018. Whereas previously core inflation was expected to breach the upper end of the target range in the third quarter of 2016 for four consecutive quarters, a one-quarter breach, at 6,1 per cent, is now expected in the fourth quarter of this year. Inflation expectations as reflected in the survey conducted by the Bureau for Economic Research (BER) have remained relatively anchored at the upper end of the inflation target range. In the second quarter, average expectations for 2016 were 6,3 per cent, up from 6,2 per cent. Average expectations for 2017 were unchanged at 6,2 per cent and were marginally down to 5,9 per cent for 2018. Expectations of business people were more or less unchanged compared to the previous quarter. Both market analysts and trade union officials revised their near-term forecasts upward, while forecasts for 2018 were revised lower. Average five-year inflation expectations declined from 6,1 per cent to 5,9 per cent in the second quarter, with downward revisions by all groups. 75 Quarterly Bulletin September 2016

The median expectations of market analysts, as reflected in the Reuters Econometer survey conducted in July, show a moderate decline since May. Inflation expectations declined from 6,7 per cent and 6,4 per cent in 2016 and 2017, to 6,6 per cent and 6,1 per cent respectively. The yield differential between inflation-linked bonds and conventional government bonds (break-even inflation expectations) declined across all maturities, but remain elevated. The global economic outlook has been influenced by the outcome of the UK referendum. The financial markets displayed a high degree of volatility in the immediate aftermath of the outcome, but have since stabilised to some extent. Nevertheless, the longer-term real impacts are expected to be negative for global growth, particularly for the UK and Europe, as investment decisions are put on hold during the transition period. A high degree of uncertainty is expected to persist for some time as the magnitude of this slowdown is still unclear and dependent on the nature and speed of the UK disengagement. Since the referendum, global growth forecasts have generally been revised downwards. The outlook for emerging markets has remained relatively subdued, with further downside risks in Turkey following the recent coup attempt and terrorist attacks. Recent outcomes in China indicate some improvement in the growth prospects in response to government stimulus packages, and this has underpinned a stabilisation and moderate upward trend in commodity prices. The International Monetary Fund (IMF) growth forecasts released earlier this week have revised sub-Saharan African growth downwards. The region has been negatively impacted by lower commodity prices and severe drought in some parts. Inflation in the advanced economies remains low and still generally below target. The Brexit vote has also modified expectations regarding monetary policy in these economies. Although the Bank of England kept policy rates on hold at its most recent meeting, expectations are for a reduction in the policy rate in the near future, with the prospect of a resumption of quantitative easing. Markets are now expecting the US policy rate to remain unchanged for some time. Simultaneously, the highly accommodative monetary policy stances of the European Central Bank (ECB) and Bank of Japan are expected to persist, with the possibility of additional stimulus. These developments have reinforced the global search for yield, as yield curves flattened across major advanced economy markets. The number of government bonds trading with negative yields increased further. This has provided additional impetus to renewed capital flows to emerging markets observed since earlier this year. While these flows are expected to persist, they remain sensitive to changes in expectations regarding US monetary policy and general global risk perceptions. The recent volatility experienced by the rand exchange rate has been driven mainly by external factors and changes in global risk perceptions. Although the rand depreciated sharply in the immediate aftermath of the Brexit vote, it has reversed these losses, and more so against the British pound. Since the previous meeting of the Monetary Policy Committee (MPC), the rand has traded in a range of R15,80 and R14,22 against the US dollar, and has appreciated by 11,0 per cent against the US dollar, by 12,9 per cent against the euro and by 23,0 per cent against the British pound. On a trade-weighted basis, the rand has appreciated by 12,2 per cent. The rand has been supported by the global search for yield. There was a sharp increase in non-resident inflows to the domestic bond and equity markets in June and July. Since the beginning of June, net purchases of R107,3 billion by non-residents have been recorded. The rand also responded positively to the improvement in commodity prices and the unexpectedly large trade surplus recorded in May, which followed a small surplus in April. Despite this recent strength, the rand remains vulnerable to possible ‘risk-off’ global scenarios; changes in US monetary policy expectations; and domestic concerns, including the possibility of ratings downgrades later in the year. The domestic economic growth outlook remains extremely challenging, following the contraction in gross domestic product (GDP) in the first quarter of this year. Although this is anticipated to have been the low point of the cycle, the recovery is expected to be weak. The Bank’s latest forecast is for zero per cent growth in 2016, compared with 0,6 per cent previously. Growth 76 Quarterly Bulletin September 2016

rates of 1,1 per cent and 1,5 per cent are forecast for the next two years, down from 1,3 per cent and 1,7 per cent previously. The Bank’s estimate of potential output has been revised down marginally to 1,4 per cent in 2016, rising to 1,7 per cent in 2018. This growth outlook is corroborated by the persistent negative trend in the Bank’s leading indicator of economic activity. Business confidence remains low with the Rand Merchant Bank (RMB)/BER business confidence indicator falling to its lowest level since 2009 in the second quarter of this year. Recent economic data suggest that positive growth was recorded in the second quarter, with the mining and manufacturing sectors expected to add positively to growth. The more positive trend in manufacturing is consistent with the Barclays Purchasing Managers’ Index (PMI) which has been above the neutral 50-point level since March. The BER manufacturing confidence index also improved, but still remains at low levels. The real value of building plans passed is indicative of some improvement in the sector, particularly with respect to residential construction. However, this is not reflected in the First National Bank (FNB)/BER building confidence index which declined further in the second quarter. Underlying the negative performance of the economy during the first quarter was the sharp contraction in growth in gross fixed capital formation for the second consecutive quarter by both the private sector and general government. These trends have contributed to the persistence of high rates of unemployment in the economy. Although formal non-agricultural employment increased in the first quarter of 2016, this was largely due to temporary employment opportunities created by the Independent Electoral Commission in preparation for the local government elections. By contrast, private-sector employment contracted during the first quarter. Growth in consumption expenditure by households also contracted in the first quarter, with a sharp slowdown in expenditure on durable goods in particular. The FNB/BER consumer confidence index declined in the second quarter to low levels, following a moderate improvement in the previous quarter. The BER retail confidence index also declined sharply in the second quarter. While new vehicle sales remained weak, vehicle exports recorded strong growth in the second quarter. Despite the surprise increase in retail and wholesale trade sales in May, consumption expenditure by households is expected to remain subdued given the low consumer confidence, high debt levels, rising costs of debt servicing and slow employment growth. Consumption expenditure has been further constrained by the absence of significant wealth effects owing to the weak performance of asset markets, particularly the housing market. Credit extension to households also remains weak, with negative real rates of growth. Average wage growth has remained relatively stable, but there are risks of increases in excess of inflation and productivity gains. There are still no clear signs of a recovery in the agricultural sector, and food price inflation is expected to remain elevated for some time. The Bank expects food price inflation to peak at 12,6 per cent during the final quarter of this year. There are some encouraging signs of moderation in some food categories at both the producer and consumer price levels, and futures prices of grains have declined markedly. Should food prices stabilise or decline in the coming months, a potential exists for significant downside base effects next year. Exchange rate developments will also be critical in this respect as will global food prices which have reversed their recent negative trend. Despite supply disruptions and curtailments by a number of producers, the price of Brent crude oil has mostly traded within a band of US$45 and US$50 since the previous meeting of the MPC. The Bank’s model assumes a very moderate upward trend in oil prices over the forecast period, but there may be a degree of downside risk in the short term, with some upside risk in the outer period as global demand recovers. The domestic petrol price increased by a cumulative 60 cents per litre in the past two months due to both the exchange rate and international prices. However, the recent appreciation of the rand coupled with a lower average oil price has resulted in a substantial recovery in the petrol price, and a sizeable reduction in retail prices is expected in August. 77 Quarterly Bulletin September 2016

The MPC remains concerned about the weak economic growth outlook and the medium-term inflation trajectory which will remain outside the target range of 3–6 per cent until the second half of next year. Nevertheless, there have been some improvements in the near-term inflation prospects following successive downside surprises. This is also the case for core inflation, where the expected breach of the upper end of the target range is now less protracted. While the risks to the inflation forecast are assessed to remain on the upside, these risks have moderated somewhat. Global uncertainties appear to have delayed monetary policy tightening in advanced economies. The impact of the more appreciated rand exchange rate on the inflation outlook will depend to a large extent on whether the exchange rate is sustained at these stronger levels. Current exchange rate levels are stronger than those implicit in the forecast, providing some buffer to the projections. The rand remains sensitive to both domestic and external developments, and the recent trends can be quickly reversed. The outlook is clouded by the uncertainty surrounding the longer-term market and global growth implications of Brexit. The implications for the rand and domestic growth, and ultimately inflation, could vary quite significantly depending on which scenario plays out. A weaker global growth scenario could also imply that there may be a degree of downside risk to the international oil price assumption, which was adjusted upward. A combination of a stronger exchange rate and subdued international oil prices would have a favourable impact on domestic petrol prices in the coming months. However, the MPC assesses the longer-term risk to this assumption to be on the upside as global growth and oil demand are expected to recover. Food prices remain an upside risk in the near term. However, there could be a sharp decline in agricultural prices next year, should favourable weather patterns transpire as forecast. The absence of demand pressures and weak consumption expenditure growth may also contribute to downside risks. While the MPC remains concerned about the overall inflation trajectory, the assessment of the balance of risks to the inflation outlook and the weak domestic economy has provided some room to delay further tightening of the monetary policy stance for now. Accordingly, the MPC has unanimously decided to keep the repurchase rate unchanged at 7,0 per cent per annum. The MPC is aware that some of the favourable factors that contributed to this decision could reverse quickly and it remains ready to react appropriately to any significant change in the inflation outlook. 78 Quarterly Bulletin September 2016

Statement issued by the South African Reserve Bank 25 July 2016 Incorrect data for the trading of equities by non-residents The South African Reserve Bank (the Bank) has noted the announcement yesterday (24 July 2016) by the JSE Limited (JSE) that a programming error resulted in incorrect data for the trading of equities by non-residents for the period 31 May to 20 July 2016 being reported. The error has since been corrected and the JSE has reassured the Bank that steps are being taken to prevent a reoccurrence. The JSE’s announcement means that incorrect data (combined bond and equity net purchases by non-residents) was quoted in the July 21 Monetary Policy Committee (MPC) statement. The incorrect data did not have any bearing on the Committee’s decision. The revised data for equity transactions by non-residents as published by the JSE yesterday is: May 2016 (net sales of R16,1 billion); June 2016 (net sales of R20,3 billion); July 2016 to date (net purchases of R0,05 billion). The incorrect data recorded net purchases by non-residents as R6,4 billion (May 2016); R63,8 billion (June); and R27,9 billion (July to date). Data for the net purchases of South African bonds by non-residents was not affected by the programming error. Net purchases totalled R7,0 billion in June and for July to date have amounted to R8,6 billion. 79 Quarterly Bulletin September 2016

Summary of assumptions: Monetary Policy Committee meeting on 21 July 2016* 1. Foreign-sector assumptions 2. Domestic-sector assumptions The figures in brackets represent the previous assumptions of the Monetary Policy Committee. * For an explanation of foreign-sector assumptions and domestic-sector assumptions, see pages 82 and 83. 80 Quarterly Bulletin September 2016 Percentage changes (unless otherwise indicated) 1. Real government consumption expenditure .................................... 2. Administered prices........................................................................... – Petrol price .................................................................................... – Electricity price.............................................................................. 3. Potential growth ................................................................................. 4. Repurchase rate (per cent) ................................................................ Actual Forecast 2013 2014 2015 2016 2017 2018 3,8% 1,8% 0,2% 8,7% 6,7% 1,7% 11,8% 7,2% -10,7% 8,7% 7,2% 9,4% 2,0% 1,7% 1,5% 5,00 5,57 5,89 1,5% 1,0% 1,0% (1,5%)(1,0%)(1,0%) 5,7% 7,4% 7,9% (6,1%)(8,5%)(8,1%) 2,2% 7,6% 7,9% (3,8%) (11,5%)(8,7%) 9,6% 8,5% 9,0% (9,6%)(8,5%)(9,0%) 1,4% 1,5% 1,7% (1,5%)(1,6%)(1,8%) 6,91 7,00 7,00 (6,91)(7,00)(7,00) Percentage changes (unless otherwise indicated) 1. Real GDP growth in South Africa’s major trading-partner countries... 2. International commodity prices in US$ (excluding oil).......................... 3. Brent crude (US$/barrel) ....................................................................... 4. World food prices (US$) ........................................................................ 5. International wholesale prices ............................................................... 6. Real effective exchange rate of the rand (index 2010 = 100) ............... 7. Real effective exchange rate of the rand............................................... Actual Forecast 2013 2014 2015 2016 2017 2018 3,0% 3,1% 2,8% -6,4% -9,8% -19,3% 108,8 99,2 52,5 -1,6% -3,8% -18,7% 0,3% -0,1% -3,5% 81,91 79,17 80,08 -10,1% -3,3% 1,1% 2,7% 3,0% 3,3% (2,7%)(3,1%)(3,4%) -8,5% 1,0% 1,5% (-9,0%)(1,0%)(1,5%) 44,6 53,5 57,5 (42,0)(52,0)(57,5) -6,0% 2,0% 3,0% (-7,4%)(2,0%)(3,0%) -1,3% 1,0% 1,2% (-1,2%)(0,9%)(1,1%) 74,01 75,00 75,00 (74,00) (75,00) (75,00) -7,6% 1,3% 0,0% (-7,6%)(1,4%)(0,0%)

Selected forecast results: Monetary Policy Committee meeting on 21 July 2016 Selected forecast results (quarterly) The figures in brackets represent the previous forecasts of the Monetary Policy Committee. Selected forecast results (annual) The figures in brackets represent the previous forecasts of the Monetary Policy Committee. 81 Quarterly Bulletin September 2016 Per cent Actual Forecast 1. Real gross domestic product (GDP) growth ............................ 2. Current account as a ratio to nominal GDP.............................. 2013 2014 2015 2016 2017 2018 2,3% 1,6% 1,3% (2,2%)(1,5%)(1,3%) -5,9 -5,3 -4,3 (-5,8)(-5,4)(-4,4) 0,0% 1,1% 1,5% (0,6%)(1,3%)(1,7%) -4,2 -4,4 -4,7 (-4,6)(-4,7)(-4,9) Year-on-year percentage change Actual Forecast 1 2 3 4 2015 1 2 3 4 2016 1 2 3 4 2017 1 2 3 4 2018 1. Headline inflation............................. 4,2 4,6 4,7 4,9 4,6 2. Core inflation ................................... 5,7 5,6 5,3 5,2 5,5 6,5 6,2 6,5 7,1 6,6 (6,5)(6,4)(6,8)(7,3)(6,7) 5,5 5,5 5,9 6,1 5,8 (5,5)(5,8)(6,2)(6,2)(5,9) 6,6 6,1 5,8 5,5 6,0 (7,0)(6,4)(5,8)(5,5)(6,2) 6,0 5,9 5,4 5,2 5,7 (6,2)(6,0)(5,3)(5,1)(5,7) 5,4 5,5 5,5 5,6 5,5 (5,4)(5,4)(5,4)(5,5)(5,4) 5,2 5,2 5,3 5,4 5,3 (5,1)(5,1)(5,2)(5,3)(5,2)

Foreign-sector assumptions 1. Trading-partner gross domestic product (GDP) growth is determined broadly using the Global Projection Model (GPM) of the International Monetary Fund (IMF), which is then adjusted to aggregate the GDP growth rates of South Africa’s major trading partners on a trade-weighted basis. Individual projections are done for the four largest trading partners: the euro area, China, the United States (US) and Japan. The remaining trading partners are grouped into three regions: emerging Asia (excluding China), Latin America, and the Rest of Countries bloc. The assumption takes account of country-specific ‘consensus’ forecasts as well as IMF regional growth prospects. 2. The commodity price index is a weighted aggregate price index of the major South African export commodities based on 2010 prices. The composite index represents the total of the individual commodity prices multiplied by their smoothed export weights. Commodity price prospects generally remain commensurate with global liquidity as well as commodity supply/ demand pressures as reflected by the pace of growth in the trading-partner countries. 3. The Brent crude oil price is expressed in US dollar per barrel. The assumption incorporates an analysis of the factors of supply, demand (using global growth expectations) and inventories of oil (of all grades) as well as the expectations of the US Energy Information Administration (EIA), the Organization of the Petroleum Exporting Countries (OPEC) and Reuters. 4. World food prices uses the composite food price index of the Food and Agriculture Organization of the United Nations (FAO) in US dollar. The index is weighted using average export shares and represents the monthly change in the international prices of a basket of five food commodity price indices (cereals, vegetable oil, dairy, meat and sugar). World food price prospects incorporate selected global institution forecasts for food prices as well as imbalances from the anticipated trend in international food supplies relative to expected food demand pressures. 5. International wholesale prices refers to a weighted aggregate of the producer price indices (PPIs) of South Africa’s major trading partners, as per the official real effective exchange rate calculation of the South African Reserve Bank (the Bank). Although individual country consumer price index (CPI) inflation forecasts provide a good indication for international wholesale price pressures, the key drivers of the assumed trend in global wholesale inflation are oil and food prices as well as expected demand pressures emanating from the trends in the output gaps of the major trading-partner countries. Other institutional forecasts for international wholesale prices are also considered. 6. The real effective exchange rate is the nominal effective exchange rate of the rand deflated by the producer price differential between South Africa and an aggregate of its trading-partner countries (as reflected in the Quarterly Bulletin published by the Bank). Although the nominal rate is a weighted average of South Africa’s 20 largest trading partners, particular focus is placed on the rand outlook against the US dollar, the euro, the Chinese yuan, the British pound and the Japanese yen. The assumed trend in the real effective exchange rate remains constant from the latest available quarterly average over the projection period. However, due to the time delay in the calculation of the real effective exchange rate, the most recent trend in the nominal effective exchange rate is adjusted with the assumed trend for the domestic and foreign price differential for the current quarter. This may result in a technical annual adjustment over the current and next forecast year that differs from zero. 82 Quarterly Bulletin September 2016

Domestic-sector assumptions 1. Government consumption expenditure (real) is broadly based on the most recent National Treasury budget projections. However, since these projections take place twice a year, the most recent actual data points also play a significant role in the assumptions process. 2. Administered prices represent the total of regulated and non-regulated administered prices as reflected by Statistics South Africa (Stats SA). Their weight in the consumer price index basket is 18,48 per cent and the assumed trend over the forecast period is largely determined by the expected pace of growth in petrol prices, electricity tariffs, school fees as well as water and other municipal assessment rates. The petrol price is an administered price and comprises 5,68 per cent of the overall basket. The basic fuel price (which currently accounts for roughly half of the petrol price) is determined by the exchange rate and the price of petrol quoted in US dollars at refined petroleum centres in the Mediterranean, the Arab Gulf and Singapore. The remainder of the petrol price is made up of wholesale and retail margins as well as the fuel levy and contributions to the Road Accident Fund (RAF). Since most taxes and retail margins are changed once a year, the assumed trajectory of the petrol price largely reflects the anticipated trend in oil prices and the exchange rate. The electricity price is an administered price measured at the municipal level with a weight of 4,13 per cent in the CPI basket. Electricity price adjustments generally take place in July and August of each year, and the assumed pace of increase over the forecast period reflects the multi-year price determination (MYPD) by the National Energy Regulator in respect of Eskom, with a slight adjustment for measurement at municipal level. 3. The pace of potential growth is derived from the Bank’s semi-structural potential output model. The measurement accounts for the impact of the financial cycle on real economic activity and introduces economic structure via the relationship between potential output and capacity utilisation in the manufacturing sector (see South African Reserve Bank Working Paper Series WP/14/08). 4. The repurchase rate (commonly called the ‘repo rate’) is the official monetary policy instrument and represents the interest rate at which banks borrow money from the Bank. Although the rate is held constant over the forecast period, this assumption is relaxed in alternative scenarios where, for instance, the policy rate responds to deviations of output from its potential and the gap between future inflation and the inflation target, in other words, via a stylised ‘Taylor rule’, one that is based on market expectations of the future path of the policy rate, and other paths (as requested). 83 Quarterly Bulletin September 2016

Note on unit labour cost measurement in South Africa 1 by I Venter and E Botes 1 The views expressed are those of the authors and do not necessarily reflect the views of the South African Reserve Bank (the Bank). 1. Introduction Changes in the cost of labour per unit of real output – unit labour cost or ULC – provide a good indication of price pressures emanating from labour market conditions in an economy. Consequently, ULC is an important macroeconomic indicator for monetary policy purposes and as such serves as a key explanatory variable in a number of equations within the Bank’s suite of macroeconometric models. The Bank’s Economic Research and Statistics Department currently calculates two different nominal ULC measures2: an economy-wide measure compiled by the Macro Models Unit that feeds into the econometric models, and a formal non-agricultural sector measure based on the Quarterly Employment Statistics (QES) survey, compiled by the Business Cycle Analysis Division that is published regularly in the Quarterly Bulletin. 2 All ULC analyses are in nominal terms, unless stated otherwise. 2. Definitions and economic relevance Labour costs are the costs incurred by employers in the employment of labour. Labour costs are of particular interest as they impact on the competitiveness of organisations and also on employers’ willingness to invest and employ workers. Labour costs, i.e. salaries and wages earned by employees, also contribute most to household income, and as such have a significant impact on household consumption expenditure and saving decisions. Thus, understanding changes in wages and labour costs is important in assisting the formulation of macroeconomic policy. 2.1 Unit labour cost and labour productivity Unit labour cost is defined as the total labour cost per unit of output. Nominal unit labour cost (NULC) can be calculated as the ratio of total labour cost to real output (equation 1). NULC = { Total labour cost } (1) Real output 3 Average labour cost per hour worked is calculated as the ratio of total labour cost to hours worked. 3 Alternatively, NULC can be calculated as the ratio of average labour cost per hour worked to real output per hour worked (i.e. labour productivity). Since it is difficult to obtain data on hours worked for some countries (including South Africa), the number of persons employed is usually used as an alternative measure of labour input when calculating labour productivity. The alternative calculation for NULC is expressed in equation 2. { Total labour cost } Number of employees NULC = { Real output } Number of employees (2) where Labour productivity = { Real output Number of employees } (3) As shown in equations 2 and 3, ULC represents a direct link between labour productivity and the cost of labour used in producing output. An increase in ULC implies that growth in average employee compensation exceeds growth in labour productivity, which may create pressure on producer prices. Thus, a rise in ULC represents an increased reward for labour’s contribution to output. However, a rise in labour cost that is higher than the rise in labour productivity may threaten a country’s cost competitiveness, if other costs are not adjusted to compensate. 84 Quarterly Bulletin September 2016

2.2 Building blocks of the Bank’s two unit labour cost measures As mentioned, the Bank currently calculates two different ULC measures; an economy-wide ULC measure utilised in the Bank’s econometric models, as well as a formal non-agricultural ULC measure published regularly in the Quarterly Bulletin. In calculating the economy-wide ULC measure, equation 1 is applied: Economy-wide ULC = { Total compensation of employees } (4) Real gross domestic product Similarly, the formula for calculating the formal non-agricultural ULC measure is: Formal non-agricultural ULC = { Gross earnings in the formal non-agricultural sector } Real non-agricultural gross value added (5) When comparing the two numerators, it should be noted that total compensation of employees is obtained from the national accounts, and thus covers the whole economy. According to the 2008 System of National Accounts (SNA) (page 139), compensation of employees is defined as: “the total remuneration, in cash or in kind, payable by an enterprise to an employee in return for work done by the latter during the accounting period. Compensation of employees has two main components: a. wages and salaries payable in cash or in kind, including enhanced payments and special allowances (e.g. overtime), regular supplementary allowances (e.g. housing), ad hoc bonus payments, as well as commissions, gratuities or tips received by employees; and b. social insurance contributions payable by employers, which include contributions to social security schemes; actual social contributions to other employment-related social insurance schemes and imputed social contributions to other employment-related social insurance schemes.” Gross earnings of employees in the formal non-agricultural sector are obtained from the Statistics South Africa (Stats SA) Quarterly Employment Statistics (QES) survey and are defined in Stats SA’s QES as: “the payments for ordinary-time, standard or agreed hours during the reference period for all permanent, temporary, casual, managerial and executive employees before taxation and other deductions for the reference period. This includes salaries and wages; commission if a retainer, wage or salary was also paid; employer’s contribution to pension, provident, medical aid, sick pay and other funds; allowances; etc., but excludes earnings of sole proprietors or partners of unincorporated businesses; commission where a retainer, wage or salary was not paid; payments to subcontractors and consultants who are not part of the enterprise; and severance, termination and redundancy payments. Gross earnings are the total sum of the earnings, including performance and others bonuses, as well as overtime payments for the three months of the reference quarter.” Although the scope of the compensation of employees and gross earnings measures is fairly similar (both exclude earnings of business owners and the self-employed, while differences relate mainly to certain types of allowances and commissions), the methodology applied to compile each variable differs noticeably. Gross earnings is the sum of earnings measured directly from the payrolls of companies surveyed in Stats SA’s QES survey. Conversely, as a national accounts estimate compensation of employees is very seldom the result of direct measurement but rather estimated using a variety of sources for the various economic sectors. 85 Quarterly Bulletin September 2016

3. Analysis of correlation with inflation measures Unit labour cost currently feeds into the Bank’s macroeconometric models as explanatory variable for core consumer price inflation as well as for final manufactured producer price inflation. In both instances, the economy-wide ULC measure is used. In this section, both ULC measures are compared with the two dependant variables in the models to see which ULC measure correlates better with each dependant variable. Correlation coefficients were calculated between each dependant variable and each ULC measure, using quarterly data over the period 1980 to 2015. In addition, the two ULC measures were lagged for up to six quarters in order to ascertain the degree to which current ULC changes explain future changes in the inflation measures. 0,85 0,55 Economy-wide ULC 0,70 0,35 The charts above show the results of the correlation analysis between core consumer price inflation and the two ULC measures, and between manufacturing producer price inflation and the two ULC measures respectively. Both ULC measures correlate better with core consumer price inflation than with manufacturing producer price inflation. Over the full sample period, the contemporaneous correlation coefficients of both ULC measures with both inflation measures are the highest, deteriorating slightly as the lag period is increased. It should be noted that this result is dependent on the sample period used. The more relevant finding, however, is that the correlation coefficients of the economy-wide ULC measure with both inflation measures are consistently higher than those of the formal non-agricultural ULC measure. This could largely be due to the fact that the economy-wide ULC measure is much smoother than the formal non-agricultural measure, which would make it more desirable for modelling and forecasting purposes. 86 Quarterly Bulletin September 2016 Correlation analysis: Core CPI vs Correlation analysis: Manufacturing ULC measuresPPI vs ULC measures Correlation coefficientCorrelation coefficient 0,90 0,70 0,65 0,60 0,80 Economy-wide ULC 0,75 0,50 0,45 0,40 0,65 0,60 0,30 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Quarters lagQuarters lag Formal non-agricultural ULC Formal non-agricultural ULC

4. Empirical analysis of the link between unit labour cost, wages and productivity growth As discussed in section 2.1, ULC can be calculated using either equation 1 or 2. Expressing each term in equation 2 in natural logarithms, it can be rewritten as: In ULC = ln { Total labour costs } – In { Real output Number of employees } Number of employees (6) Since natural logarithms can be substituted for growth rates, the formula can be rewritten further as: ULC = { Total labour costs } – { Real output } (7) Number of employees Number of employees where = the percentage change over four quarters. The formula can be simplified further to relate growth in ULC to growth in remuneration per worker and growth in labour productivity: ULC = { Remuneration per worker } – { Labour productivity } (8) The formal non-agricultural ULC measure is calculated using equation 1. However, to confirm whether changes in this ULC measure are consistent with wage and labour productivity growth, it can also be calculated using equation 2. 25 30 30 The chart on the left-hand side above shows year-on-year growth in quarterly gross earnings per worker and in labour productivity from 1980 to 2015, while the chart next to that depicts the difference between these two indicators, representing ULC calculated with equation 2, together with the standard formal non-agricultural ULC measure, calculated with equation 1. These two measures are virtually identical, illustrating the relationship between ULC, gross earnings growth and labour productivity growth. The relationship holds so closely because both gross earnings and employment are obtained from the same source, namely the QES survey, where earnings are surveyed directly from enterprise payrolls and matched to employment. 87 Quarterly Bulletin September 2016 Formal non-agricultural earnings and Growth in gross earnings, labour labour productivity productivity and ULC Percentage change Percentage changes over four quarters over four quarters Percentage points 20 Gross earnings per worker 25 25 15 20 20 10 15 15 5 10 10 0 5 5 Labour productivity -5 0 0 1980 85 90 95 2000 05 10 16 1980 85 90 95 2000 05 10 16 Formal non-agricultural nominal ULC Gross earnings growth per worker minus labour productivity growth (right-hand scale)

A similar calculation was done with the economy-wide ULC measure. The left-hand chart below illustrates year-on-year growth in quarterly compensation of employees per worker and in economy-wide labour productivity from 2009 to 2015. Economy-wide labour productivity was calculated by dividing total GDP by total employment as published in Stats SA’s Quarterly Labour Force Survey (QLFS), which is only available from 2008. Contrary to formal non-agricultural labour productivity, economy-wide labour productivity has mostly contracted since mid-2011 as QLFS employment growth has largely outpaced GDP growth over the period. This was particularly pronounced in 2015, exacerbated by the sharp rise in total employment following the implementation of a new master sample for the QLFS. The right-hand chart below depicts the difference between annual growth in quarterly compensation of employees per worker and in economy-wide labour productivity (representing ULC calculated using equation 2), together with the standard economy-wide ULC measure (calculated using equation 1). Although these two measures correlate fairly well, they are not as close to each other as the two formal non-agricultural measures shown in one of the previous charts. The main reason for the occasional differences lies in the fact that the sources of total employment and that of compensation of employees are completely different, while earnings and employment are matched from the QES payroll data in the formal non-agricultural calculation. 25 12 12 ULC 5. Conclusion The Bank currently calculates two different nominal ULC measures since both are useful, each in its own unique way. The economy-wide ULC measure is important for modelling purposes, since it correlates slightly better with inflation variables, reduces series ‘noise’ and represents a more broad-based measure of ULC. Likewise, the formal non-agricultural ULC measure provides an empirical reflection of the theoretical link between ULC, wage and labour productivity growth and is thus published in the Quarterly Bulletin and used when analysing wage growth and labour productivity developments in the South African economy. 88 Quarterly Bulletin September 2016 Compensation of employees andGrowth in compensation of employees, labour productivity labour productivity and ULC Percentage changes over Percentage changes over four quartersfour quartersPercentage points 20 1010 15 108Economy-wide 8 5 66 0 -5 44 20092011201320152009201120132015 Compensation growth per worker minus economy-wide labour productivity growth (right-hand scale) Compensation per w Economy-wide orker labour productivity

Notes to tables Non-resident and real-estate transactions – Table S–30 While the layout of the tables on page S–30 remains unchanged, it should be noted that the statistics on transactions in domestic bonds by non-residents, published in table KB204 under ‘bonds’, are now sourced from the JSE Limited rather than from Strate Limited. This allows the statistics to incorporate the numerous cancellations of and corrections to transactions effected by the brokers on a daily basis. The consideration statistics as published in the table incorporate cancellations and corrections, but exclude free-of-value trades. 89 Quarterly Bulletin September 2016