| Fair Values of Financial Instruments. | Note 16. Fair Values of Financial Instruments.
The fair value amounts recorded on the Statements of Condition or presented in the note disclosures have been determined by the FHLBNY using available market information and best judgment of appropriate valuation methods.
Estimated Fair Values — Summary Tables
The carrying values, estimated fair values and the levels within the fair value hierarchy were as follows (in thousands):
June 30, 2015
Estimated Fair Value
Financial Instruments
Carrying Value
Total
Level 1
Level 2
Level 3 (a)
Netting Adjustment and Cash Collateral
Assets
Cash and due from banks
$
$
$
$
—
$
—
$
—
Securities purchased under agreements to resell
—
—
—
Federal funds sold
—
—
—
Available-for-sale securities
—
—
Held-to-maturity securities
—
—
Advances
—
—
—
Mortgage loans held-for-portfolio, net
—
—
—
Accrued interest receivable
—
—
—
Derivative assets
—
—
)
Other financial assets
—
—
—
Liabilities
Deposits
—
—
—
Consolidated obligations
Bonds
—
—
—
Discount notes
—
—
—
Mandatorily redeemable capital stock
—
—
—
Accrued interest payable
—
—
—
Derivative liabilities
—
—
)
Other financial liabilities
—
—
—
December 31, 2014
Estimated Fair Value
Financial Instruments
Carrying Value
Total
Level 1
Level 2
Level 3 (a)
Netting Adjustment and Cash Collateral
Assets
Cash and due from banks
$
$
$
$
—
$
—
$
—
Securities purchased under agreements to resell
—
—
—
Federal funds sold
—
—
—
Available-for-sale securities
—
—
Held-to-maturity securities
—
—
Advances
—
—
—
Mortgage loans held-for-portfolio, net
—
—
—
Accrued interest receivable
—
—
—
Derivative assets
—
—
)
Other financial assets
—
—
—
Liabilities
Deposits
—
—
—
Consolidated obligations
Bonds
—
—
—
Discount notes
—
—
—
Mandatorily redeemable capital stock
—
—
—
Accrued interest payable
—
—
—
Derivative liabilities
—
—
)
Other financial liabilities
—
—
—
(a)
Level 3 Instruments — The fair values of non-Agency private-label MBS and housing finance agency bonds were estimated by management based on pricing services. Valuations may have required pricing services to use significant inputs that were subjective because of the current lack of significant market activity so that the inputs may not be market based and observable.
Fair Value Hierarchy
The FHLBNY records available-for-sale securities, derivative assets, derivative liabilities, and consolidated obligations and advances elected under the FVO at fair value on a recurring basis. On a non-recurring basis, held-to-maturity securities determined to be OTTI and written down to their fair values, and mortgage loans held-for-portfolio that are written down to their fair values or are foreclosed as Other real estate owned (“REO”) and written down to their fair values are reported on a non-recurring basis in the period the fair values are recorded in the Statements of Condition.
The accounting standards under Fair Value Measurement defines fair value, establishes a consistent framework for measuring fair value and requires disclosures about fair value measurements. Fair value is defined as the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date. Among other things, the standard requires the FHLBNY to maximize the use of observable inputs and minimize the use of unobservable inputs when measuring fair value. The standard specifies a hierarchy of inputs based on whether the inputs are observable or unobservable. Observable inputs reflect market data obtained from independent sources, while unobservable inputs reflect the FHLBNY’s market assumptions.
These two types of inputs have created the following fair value hierarchy, and an entity must disclose the level within the fair value hierarchy in which the measurements are classified for all assets and liabilities measured on a recurring or non-recurring basis:
·
Level 1 Inputs — Quoted prices (unadjusted) for identical assets or liabilities in an active market that the reporting entity can access on the measurement date.
·
Level 2 Inputs — Inputs other than quoted prices within Level 1 that are observable inputs for the asset or liability, either directly or indirectly. If the asset or liability has a specified (contractual) term, a Level 2 input must be observable for substantially the full term of the asset or liability. Level 2 inputs include the following: (1) quoted prices for similar assets or liabilities in active markets; (2) quoted prices for identical or similar assets or liabilities in markets that are not active; (3) inputs other than quoted prices that are observable for the asset or liability (e.g., interest rates and yield curves that are observable at commonly quoted intervals, and volatilities).
·
Level 3 Inputs — Inputs that are unobservable and significant to the valuation of the asset or liability.
The inputs are evaluated on an overall level for the fair value measurement to be determined. This overall level is an indication of market observability of the fair value measurement for the asset or liability. Changes in the observability of the valuation inputs may result in a reclassification of certain assets or liabilities. These reclassifications are reported as transfers in/out as of the beginning of the quarter in which the changes occur. There were no such transfers in any periods in this report.
The availability of observable inputs can vary from product to product and is affected by a wide variety of factors including, for example, the characteristics peculiar to the transaction. To the extent that valuation is based on models or inputs that are less observable or unobservable in the market, the determination of fair value requires more judgment. Accordingly, the degree of judgment exercised by the FHLBNY in determining fair value is greatest for instruments categorized as Level 3. In certain cases, the inputs used to measure fair value may fall into different levels of the fair value hierarchy. In such cases, for disclosure purposes the level in the fair value hierarchy within which the fair value measurement falls is determined based on the lowest level input that is significant to the fair value measurement in its entirety.
Summary of Valuation Techniques and Primary Inputs
The fair value of a financial instrument that is an asset is defined as the price the FHLBNY would receive to sell the asset in an orderly transaction with market participants. A financial liability’s fair value is defined as the amount that would be paid to transfer the liability to a new obligor, not the amount that would be paid to settle the liability with the creditor. Where available, fair values are based on observable market prices or parameters, or derived from such prices or parameters. Where observable prices are not available, valuation models and inputs are utilized. These valuation techniques involve some level of management estimation and judgment, the degree of which is dependent on the price transparency for the instruments or markets and the instruments’ complexity.
Because an active secondary market does not exist for a portion of the FHLBNY’s financial instruments, in certain cases, fair values are not subject to precise quantification or verification and may change as economic and market factors and evaluation of those factors change. The fair values of financial assets and liabilities reported in the tables above are discussed below:
Cash and Due from Banks — The estimated fair values approximate the recorded book balances, and are considered to be within the Level 1 of the fair value hierarchy.
Interest-Bearing Deposits, Federal Funds Sold and Securities Purchased under Agreements to Resell — The FHLBNY determines estimated fair values of short-term investments by calculating the present value of expected future cash flows of the investments, a methodology also referred to as the Income approach under the Fair Value Measurement standards. The discount rates used in these calculations are the current coupons of investments with similar terms. Inputs into the cash flow models are the yields on the instruments, which are market based and observable and are considered to be within Level 2 of the fair value hierarchy.
Investment Securities — The fair value of investment securities is estimated by the FHLBNY using pricing primarily from pricing services. The pricing vendors typically use market multiples derived from a set of comparables, including matrix pricing, and other techniques.
Mortgage-backed securities — The FHLBNY’s valuation technique incorporates prices from up to four designated third-party pricing services, when available. The FHLBNY’s base investment pricing methodology establishes a median price for each security using a formula that is based on the number of prices received. If four prices are received from the four pricing vendors, the average of the two middle prices is used; if three prices are received, the middle price is used; if two prices are received, the average of the two prices is used; and if one price is received, it is used subject to further validation. Vendor prices that are outside of a defined tolerance threshold of the median price are identified as outliers and subject to additional review, including, but not limited to, comparison to prices provided by an additional third-party valuation service, prices for similar securities, and/or non-binding dealer estimates, or use of internal model prices, which are deemed to be reflective of all relevant facts and circumstances that a market participant would consider. Such analysis is also applied in those limited instances where no third-party vendor price or only one third-party vendor price is available in order to arrive at an estimated fair value. In its analysis, the FHLBNY employs the concept of cluster pricing and cluster tolerances. Once the median prices are computed from the four pricing vendors, the second step is to determine which of the sourced prices fall within the required tolerance level interval to the median price, which forms the “cluster” of prices to be averaged. This average will determine a “default” price for the security. The cluster tolerance guidelines shall be reviewed annually and may be revised as necessary. To be included among the cluster, at June 30, 2015, each price must fall within 7 points (7 points at December 31, 2014) of the median price for residential PLMBS and within 2 points of the median price for GSE-issued MBS. The final step is to determine the final price of the security based on the cluster average and an evaluation of any outlier prices. If the analysis confirms that an outlier is not representative of fair value and that the average of the vendor prices within the tolerance threshold of the median price is the best estimate, then the average of the vendor prices within the tolerance threshold of the median price is used as the final price. If, on the other hand, an outlier (or some other price identified in the analysis) is determined to be a better estimate of fair value, then the outlier (or the other price as appropriate) is used as the final price. In all cases, the final price is used to determine the fair value of the security.
The FHLBNY has also established that the pricing vendors use methods that generally employ, but are not limited to benchmark yields, recent trades, dealer estimates, valuation models, benchmarking of like securities, sector groupings, and/or matrix pricing. To validate vendor prices of PLMBS, the FHLBNY has also adopted a formal process to examine yields as an additional validation method. The FHLBNY calculates an implied yield for each of its PLMBS using estimated fair values derived from cash flows on a bond-by-bond basis. This yield is then compared to the implied yield for comparable securities according to price information from third-party MBS “market surveillance reports”. Significant variances or inconsistencies are evaluated in conjunction with all of the other available pricing information. The objective is to determine whether an adjustment to the fair value estimate is appropriate.
Based on the FHLBNY’s review processes, management has concluded that inputs into the pricing models employed by pricing services for the FHLBNY’s investments in GSE securities are market based and observable and are considered to be within Level 2 of the fair value hierarchy. The valuation of the private-label securities, all designated as held-to-maturity, may require pricing services to use significant inputs that are subjective and are considered to be within Level 3 of the fair value hierarchy. This determination was made based on management’s view that the private-label instruments may not have an active market because of the specific vintage of the securities as well as inherent conditions surrounding the trading of private-label MBS, so that the inputs may not be market based and observable. At June 30, 2015, one held-to-maturity private -label MBS was deemed OTTI and was written down to its fair values, so that its carrying values recorded in the balance sheet equaled its fair values, which was classified on a nonrecurring basis as Level 3 financial instruments under the valuation hierarchy. No held-to-maturity securities were recorded at fair values on a non-recurring basis at December 31, 2014, as no MBS were determined to be OTTI.
Housing finance agency bonds — The fair value of housing finance agency bonds is estimated by management using information primarily from pricing services. Because of the current lack of significant market activity, their fair values were categorized within Level 3 of the fair value hierarchy as inputs into vendor pricing models may not be market based and observable.
Advances — The fair values of advances are computed using standard option valuation models. The most significant inputs to the valuation model are (1) consolidated obligation debt curve (“CO Curve”), published by the Office of Finance and available to the public, and (2) LIBOR swap curves and volatilities. Both these inputs are considered to be market based and observable as they can be directly corroborated by market participants.
The FHLBNY determines the fair values of its advances by calculating the present value of expected future cash flows from the advances, a methodology also referred to as the Income approach under the Fair Value Measurement standards. The discount rates used in these calculations are equivalent to the replacement advance rates for advances with similar terms. In accordance with the Finance Agency’s advances regulations, an advance with a maturity or repricing period greater than six months requires a prepayment fee sufficient to make a FHLBank financially indifferent to the borrower’s decision to prepay the advance. Therefore, the fair value of an advance does not assume prepayment risk.
The inputs used to determine fair value of advances are as follows:
·
CO Curve. The FHLBNY uses the CO Curve, which represents its cost of funds, as an input to estimate the fair value of advances, and to determine current advance rates. This input is considered market observable and therefore a Level 2 input.
·
Volatility assumption. To estimate the fair value of advances with optionality, the FHLBNY uses market-based expectations of future interest rate volatility implied from current market prices for similar options. This input is considered a Level 2 input as it is market based and market observable.
·
Spread adjustment. Adjustments represent the FHLBNY’s mark-up based on its pricing strategy. The input is considered as unobservable, and is classified as a Level 3 input. The spread adjustment is not a significant input to the overall fair value of an advance.
The FHLBNY creates an internal curve, which is interpolated from its advance rates. Advance rates are calculated by applying a spread to an underlying “base curve” derived from the FHLBNY’s cost of funds, which is based on the CO Curve, inputs to which have been determined to be market observable and classified as Level 2. The spreads applied to the base advance pricing curve typically represent the FHLBNY’s mark-ups over the FHLBNY’s cost of funds, and are not market observable inputs, but are based on the FHLBNY’s advance pricing strategy. Such inputs have been classified as a Level 3 input. For the FHLBNY, Level 3 inputs were considered as not significant.
To determine the appropriate classification of the overall measurement in the fair value hierarchy of an advance, an analysis of the inputs to the entire fair value measurement was performed at June 30, 2015 and December 31, 2014.
If the unobservable spread to the FHLBNY’s cost of funds was not significant to the overall fair value, then the measurement was classified as Level 2. Conversely, if the unobservable spread was significant to the overall fair value, then the measurement would be classified as Level 3. The impact of the unobservable input was calculated as the difference in the value determined by discounting an advance’s cash flows using the FHLBNY’s advance curve and the value determined by discounting an advance’s cash flows using the FHLBNY’s cost of funds curve. Given the relatively small mark-ups over the FHLBNY’s cost of funds, the results of the FHLBNY’s quantitative analysis confirmed the FHLBNY’s expectations that the measurement of the FHLBNY’s advances was Level 2. The unobservable mark-up spreads were not significant to the overall fair value of the instrument. A quantitative threshold for significance factor was established at 10%, with additional qualitative factors to be considered if the ratio exceeded the threshold.
The FHLBNY has elected the FVO designation for certain advances and recorded their fair values in the Statements of Condition for such advances. The CO Curve was the primary input, which is market based and observable. Inputs to apply spreads, which are FHLBNY specific, were not material. Fair values were classified within Level 2 of the valuation hierarchy.
Accrued Interest Receivable and Other Assets — The estimated fair values approximate the recorded book value because of the relatively short period of time between their origination and expected realization.
Mortgage Loans (MPF Loans)
A. Principal and/or Most Advantageous Market and Market Participants — MPF Loans
The FHLBNY may sell mortgage loans to another FHLBank or in the secondary mortgage market. Because transactions between FHLBanks occur infrequently, the FHLBNY has identified the secondary mortgage market as the principal market for mortgage loans under the MPF programs. Also, based on the nature of the supporting collateral to the MPF loans held by FHLBNY, the presentation of a single class for all products within the MPF product types is considered appropriate. As described below, the FHLBNY believes that the market participants within the secondary mortgage market for the MPF portfolio would differ primarily whether qualifying or non- qualifying loans are being sold.
Qualifying Loans (Unimpaired mortgage loans) — The FHLBNY believes that a market participant is an entity that would buy qualifying mortgage loans for the purpose of securitization and subsequent resale as a security. Other government-sponsored enterprises (“GSEs”), specifically Federal National Mortgage Association (“Fannie Mae”) and Federal Home Loan Mortgage Corporation (“Freddie Mac”), conduct the majority of such activity in the United States, but there are other commercial banks and financial institutions that periodically conduct business in this market. Therefore, the FHLBNY has identified market participants for qualifying loans to include (1) all GSEs, and (2) other commercial banks and financial institutions that are independent of the FHLBank System.
Non-qualifying Loans (Impaired mortgage loans) — For the FHLBNY, non-qualifying loans are primarily impaired loans. The FHLBNY believes that it is unlikely the GSE market participants would willingly buy loans that did not meet their normal criteria or underwriting standards. However, a market exists with commercial banks and financial institutions other than GSEs where such market participants buy non-qualifying loans in order to securitize them as they become current, resell them in the secondary market, or hold them in their portfolios. Therefore, the FHLBNY has identified the market participants for non-qualifying loans to include other commercial banks and financial institutions that are independent of the FHLBank System.
B. Fair Value at Initial Recognition — MPF Loans
The FHLBNY believes that the transaction price (entry price) may differ from the fair value (exit price) at initial recognition because it is determined using a different method than subsequent fair value measurements. However, because mortgage loans are not measured at fair value in the balance sheet, day one gains and losses would not be applicable. Additionally, all mortgage loans were performing at the time of origination.
The FHLBNY receives an entry price from the FHLBank of Chicago, the MPF Provider, at the time of acquisition. This entry price is based on the TBA rates, as well as exit prices received from market participants, such as Fannie Mae and Freddie Mac. The price is adjusted for specific MPF program characteristics and may be further adjusted by the FHLBNY to accommodate changing market conditions. Because of the adjustments, in many cases the entry price would not equal the exit price at the time of acquisition.
C. Valuation Technique, Inputs and Hierarchy
The FHLBNY calculates the fair value of the entire mortgage loan portfolio using a valuation technique referred to as the “market approach”. Loans are aggregated into synthetic pass-through securities based on product type, loan origination year, gross coupon and loan term. The fair values are based on TBA rates (or agency commitment rates), as discussed above, adjusted primarily for seasonality. TBA and agency commitment rates are market observable and therefore classified as Level 2 in the fair value hierarchy. However, many of the credit and default risk related inputs involved with the valuation techniques described above may be considered unobservable due to a variety of reasons (e.g., lack of market activity for a particular loan, inherent judgment involved in property estimates). If unobservable inputs are considered significant, the loans would be classified as Level 3 in the fair value hierarchy. At June 30, 2015 and December 31, 2014, fair values were classified within Level 2 of the valuation hierarchy.
The fair values of impaired MPF are generally based on collateral values less estimated selling costs. Collateral values are generally based on broker price opinions, and any significant adjustments to apply a haircut value on the underlying collateral value would be considered to be an unobservable Level 3 input. The FHLBNY may validate the impairment adjustment made to TBA rates by “back-testing” against incurred losses. However, the FHLBNY’s mortgage loan historical loss experience has been insignificant, and expected credit losses are insignificant. Level 3 inputs, if any, are generally insignificant to the total measurement, and therefore the measurement of most loans may be classified as Level 2 in the fair value hierarchy. At June 30, 2015 and December 31, 2014, fair values of impaired loans were classified within Level 2 of the valuation hierarchy as significant inputs to value collateral were considered to be observable. During the six months ended June 30, 2015, loans that were delinquent for 180 days or more were written down to their fair values of $9.1 million on a non-recurring basis as a Level 2 financial instruments as significant inputs to value collateral were considered to be observable.
Consolidated Obligations — The FHLBNY estimates the fair values of consolidated obligations based on the present values of expected future cash flows due on the debt obligations. Calculations are performed by using the FHLBNY’s industry standard option adjusted valuation models. Inputs are based on the cost of raising comparable term debt.
The FHLBNY’s internal valuation models use standard valuation techniques and estimate fair values based on the following inputs:
·
CO Curve and LIBOR Swap Curve. The Office of Finance constructs an internal curve, referred to as the CO Curve, using the U.S. Treasury Curve as a base curve that is then adjusted by adding indicative spreads obtained from market observable sources. These market indications are generally derived from pricing indications from dealers, historical pricing relationships, recent GSE trades and secondary market activity. The FHLBNY considers the inputs as Level 2 inputs as they are market observable.
·
Volatility assumption. To estimate the fair values of consolidated obligations with optionality, the FHLBNY uses market-based expectations of future interest rate volatility implied from current market prices for similar options. These inputs are also considered Level 2 as they are market based and observable.
The FHLBNY has elected the FVO designation for certain consolidated obligation debt and recorded their fair values in the Statements of Condition. The CO Curve and volatility assumptions (for debt with call options) were primary inputs, which are market based and observable. Fair values were classified within Level 2 of the valuation hierarchy.
Derivative Assets and Liabilities — The FHLBNY’s derivatives are executed in the over-the-counter market and are valued using internal valuation techniques as no quoted market prices exist for such instruments. Discounted cash flow analysis is the primary methodology employed by the FHLBNY’s valuation models to measure the fair values of interest rate swaps. The valuation technique is considered as an “Income approach”. Interest rate caps and floors are valued under the “Market approach”. Interest rate swaps and interest rate caps and floors, collectively “derivatives”, were valued in industry-standard option adjusted valuation models, which generated fair values. The valuation models employed multiple market inputs including interest rates, prices and indices to create continuous yield or pricing curves and volatility factors. These multiple market inputs were corroborated by management to independent market data, and to relevant benchmark indices. In addition, derivative valuations were compared by management to counterparty valuations received as part of the collateral exchange process. These derivative positions were classified within Level 2 of the valuation hierarchy at June 30, 2015 and December 31, 2014.
The FHLBNY’s valuation model utilizes a modified Black-Karasinski model that assumes that rates are distributed log normally. The log-normal model precludes interest rates turning negative in the model computations. Significant market based and observable inputs into the valuation model include volatilities and interest rates. The Bank’s valuation model employs industry standard market-observable inputs (inputs that are actively quoted and can be validated to external sources). Inputs by class of derivative were as follows:
Interest-rate related:
·
LIBOR Swap Curve.
·
Volatility assumption. Market-based expectations of future interest rate volatility implied from current market prices for similar options.
·
Prepayment assumption (if applicable).
·
Federal funds curve (OIS curve).
Mortgage delivery commitments (considered a derivative):
·
TBA security prices are adjusted for differences in coupon, average loan rate and seasoning.
OIS — The FHLBNY incorporates the overnight indexed swap (“OIS”) curves as fair value measurement inputs for the valuation of its derivatives, as the OIS curves reflect the interest rates paid on cash collateral provided against the fair value of these derivatives. The FHLBNY believes using relevant OIS curves as inputs to determine fair value measurements provides a more representative reflection of the fair values of these collateralized interest-rate related derivatives. The OIS curve (federal funds rate curve) is an input to the valuation model. The input for the federal funds curve is obtained from industry standard pricing vendors and the input is available and observable over its entire term structure.
Management considers the federal funds curve to be a Level 2 input. The FHLBNY’s valuation model utilizes industry standard OIS methodology. The model generates forecasted cash flows using the OIS calibrated 3-month LIBOR curve. The model then discounts the cash flows by the OIS curve to generate fair values. Previously, the FHLBNY used the 3-month London Interbank Offered Rate (“LIBOR”) curve as the relevant benchmark curve for its derivatives and as the discounting rate for these collateralized interest-rate related derivatives.
Credit risk and credit valuation adjustments — The FHLBNY is subject to credit risk in derivatives transactions due to the potential non-performance of its derivatives counterparties or a DCO.
To mitigate this risk, the FHLBNY has entered into master netting agreements and credit support agreements with its derivative counterparties for its bilaterally executed derivative contracts that provide for the delivery of collateral at specified levels at least weekly. The computed fair values of the derivatives took into consideration the effects of legally enforceable master netting agreements that allow the FHLBNY to settle positive and negative positions and offset cash collateral with the same counterparty on a net basis.
For derivative transactions executed as a cleared derivative, the transactions are fully collateralized in cash and exchanged daily with the DCO. The FHLBNY has also established the enforceability of offsetting rights incorporated in the agreements for the cleared derivative transactions.
As a result of these practices and agreements and the FHLBNY’s assessment of any change in its own credit spread, the FHLBNY has concluded that the impact of the credit differential between the FHLBNY and its derivative counterparties and DCO was sufficiently mitigated to an immaterial level that no credit adjustments were deemed necessary to the recorded fair value of Derivative assets and Derivative liabilities in the Statements of Condition at June 30, 2015 and December 31, 2014.
Deposits — The FHLBNY determines estimated fair values of deposits by calculating the present value of expected future cash flows from the deposits. The discount rates used in these calculations are the current cost of deposits with similar terms.
Mandatorily Redeemable Capital Stock — The fair value of capital stock subject to mandatory redemption is generally equal to its par value as indicated by contemporaneous member purchases and sales at par value. Fair value also includes an estimated dividend earned at the time of reclassification from equity to liabilities, until such amount is paid, and any subsequently declared dividend. FHLBank stock can only be acquired and redeemed at par value. FHLBank stock is not traded and no market mechanism exists for the exchange of stock outside the FHLBank System’s cooperative structure.
Accrued Interest Payable and Other Liabilities — The estimated fair values approximate the recorded book value because of the relatively short period of time between their origination and expected realization.
Control processes — The FHLBNY employs control processes to validate the fair value of its financial instruments, including those derived from valuation models. These control processes are designed to ensure that the values used for financial reporting are based on observable inputs wherever possible. In the event that observable inputs are not available, the control processes are designed to ensure that the valuation approach utilized is appropriate and consistently applied and that the assumptions are reasonable. These control processes include reviews of the pricing model’s theoretical soundness and appropriateness by specialists with relevant expertise who are independent from the trading desks or personnel who were involved in the design and selection of model inputs. Additionally, groups that are independent from the trading desk, or personnel involved in the design and selection of model inputs participate in the review and validation of the fair values generated from the valuation model. The FHLBNY maintains an ongoing review of its valuation models and has a formal model validation policy in addition to procedures for the approval and control of data inputs. The FHLBNY has concluded that valuation models are performing |
