VIA EDGAR
January 7, 2013
Bryan J. Pitko
Division of Corporation Finance
Securities and Exchange Commission
100 F Street, N.E.
Washington, DC 20549
| Re: | Canopius Holdings Bermuda Limited |
| Registration Statement on Form S-4 |
| Filed November 19, 2012 |
| File No. 333-183661 |
Dear Mr. Pitko:
We are responding to Comment 10 included in your letter dated December 3, 2012 to Susan Patschak, Chief Executive Officer of Canopius Holdings Bermuda Limited (“Canopius Bermuda”), with respect to the preliminary joint proxy statement/prospectus of Canopius Bermuda and Tower Group, Inc. (“Tower”) included in the above-referenced registration statement, as well as subsequent discussions between Tower and the Staff of the Division (“Staff”) of Corporate Finance of the Securities and Exchange Commission relating thereto.
For your convenience, we have repeated the Staff’s comment to which this letter is responding in italicized typeface before the response by Tower. In addition, we note our previous response to the Staff’s comment in a letter dated December 4, 2012 from Joseph L. Seiler, III of Drinker, Biddle & Reath LLP to the Staff.
Tower Group’s Form 10-Q for the Quarterly Period Ended March 31, 2012
Notes to Consolidated Financial Statements
Note 2—Accounting Policies and Basis of Presentation
Reclassification and Adjustments, page 6
| 10. | We are still evaluating your response to prior comment 15. Please address the following: |
| • | Revise the Net Income Attributable to Tower heading to Consolidated Net Income in the second table on page 10 of your response; |
| • | Tell us how you amortized a deferred tax asset that you never recorded as a fair value purchase adjustment in connection with the acquisition of CastlePoint; |
| • | Clarify why the reversal of the deferred tax valuation allowance would have increased consolidated net income but would have no effect on Net Income Attributable to Tower; and |
120 Broadway, 31st Floor • New York, NY 10271-3199
MAIN: 212.655.2000 FAX: 212.655.2070
| • | Tell us why you analyzed the errors identified prior to the out-of-period adjustments, during the out-of-period adjustments and in total as opposed to on a consolidated basis. |
Response
In response to your comments above and subsequent discussions with the Staff, Tower has discussed the attached SAB 99 assessment (see Appendix) with the Audit Committee of its Board of Directors and believes that the revisions proposed are not material to the financial statements, and that such previously issued financial statements can continue to be relied on. For purposes of reaching Tower’s conclusion, we noted that materiality under GAAP is defined as follows: “The omission or misstatement of an item in a financial report is material if, in the light of surrounding circumstances, the magnitude of the item is such that it is probable that the judgment of a reasonable person relying upon the report would have been changed or influenced by the inclusion or correction of the item.” In its analysis, Tower management and our Audit Committee considered consolidated net income as a measure along with other qualitative measures, but weighted its analysis more heavily toward the Tower stock company results, as we believe this is the measure that is most material to investors. As a result, Tower concluded that the misstatement of the $4.5 million pertaining to the Reciprocal Exchanges was less significant under this definition because the results of the Reciprocal Exchanges would not be incorporated into the analysis of a reasonable investor or reader relying upon Tower’s publicly available Form 10-Q or Form 10-K. Additionally, Tower’s management and Audit Committee believe that the disclosure with respect to the $2.6 million adjustment in the first quarter was adequate to an investor or reader to conclude that such an amount would be a non-recurring benefit to Tower, and that the subsequent revision of this adjustment as a reduction to goodwill rather than as an adjustment to earnings would not have a material effect on book value. Tower revisited this assessment at both the second and third quarters of 2012 and, despite the changing earnings pattern in the second and third quarter, did not change its conclusion.
The factors that Tower management and its Audit Committee relied upon in making this judgment included the following:
| 1) | The out of period adjustment that pertained to the Reciprocal Exchange Business of $4.5 million affected Net Income, but did not affect Earnings per Share, Net Income Available to Tower Common Shareholders (the bottom line amount of Tower’s statement of operations) or Book Value per Share. This is because the results of the Reciprocal Exchanges are not for the benefit of investors, but are rather for the benefit of policyholders of the Reciprocal Exchanges. This has been disclosed in Tower’s financial reports since the consolidation of the Reciprocal Exchanges in 2010, and is well known and understood by the analysts that track Tower’s performance for investors. All of the analyst models exclude the operating results of the Reciprocal Exchanges in their review of Tower’s operations. Moreover, other users of the financial statements (regulators, policyholders and rating agencies) that are interested in the results of the Reciprocal Exchanges use the statutory financial statements of the Reciprocal Exchanges, which are filed on an individual entity basis and which were not affected by the adjustment. |
| 2) | The $2.6 million adjustment and its subsequent correction to a reduction of goodwill as opposed to a reduction of tax expense was very clearly disclosed as a non-recurring item and was discounted by analysts and investors as a component of Tower’s results. Tower was very transparent about this adjustment in its disclosure with respect to the first quarter of 2012, providing investors and analysts with a reconciliation of the recorded tax rate to the tax rate without the impact of the adjustment. All of the analyst reports on Tower’s first quarter results noted the impact of the adjustment and none of the models considered any further tax benefits in their estimation of Tower’s future operating results. Moreover, the impact of this adjustment will not change Tower’s tangible book value per share (it will reduce Tower’s book value per share by 8 cents, or .3%). Tangible book value, while a non-GAAP measure, is a measure that is a focal point of most sell side analysts that cover the insurance industry. Dowling and Partners, a sell side analyst firm, noted “We believe tangible book value growth per share remains the only metric that reflects true “value creation”. Over the long run, disparate results in TBV growth/share will be directly reflected in disparate moves in stock prices.” Additional commentary from other sell-side analysts is included in the attached SAB 99 materiality assessment. |
| 3) | The disclosed subsequent event in Tower’s 3rd quarter earnings release pertaining to Superstorm Sandy effectively created a material change in Tower’s expected trend for the year, and such material change would not be affected in any material way by the adjustment noted. Tower disclosed in its earnings release and in its September 30, 2012 Form 10-Q its preliminary after-tax charge associated with Superstorm Sandy. This charge, which was estimated at $55 million to $68 million after-tax, or $1.45 to $1.78 per share, has effectively reset analysts’ expectations for Tower’s fourth quarter and year to date 2012 performance, and against this significant change, the adjustment is clearly not material. |
While there are many other qualitative analyses included in the SAB 99 materiality assessment, these were the most substantive qualitative factors that led Tower management and its Audit Committee to conclude that the filing of a Form 8-K under Item 4.02 was not necessary and reliance can continue to be placed on our previously issued annual and quarterly financial statements.
Management has elected to revise its previously issued financial statements to record the misstatements in the periods in which they originated. Tower will revise its consolidated balance sheet as of December 31, 2011 and 2010 and the statements of income and comprehensive income, cash flows and changes in shareholders’ equity for each of the three years in the period ended December 31, 2011. These financial statements will be revised in its 2011 Form 10-K/A. In addition, the Company will also revise its 2012 and 2011 quarterly financial statements presented in the March 31, 2012, June 30, 2012 and September 30, 2012 Form 10-Qs.
In the revised Form 10-K/A and Form 10-Q/As, Tower’s disclosure will comply with ASC 250-10-50-7. This includes disclosing in tabular format the effect of the correction on each financial statement line item and earnings per-share amounts for each period presented and the cumulative effect on retained earnings or other appropriate components of equity in the balance sheet as of the beginning of the earliest period presented. In addition, Tower will add discussion on the more significant adjustments, including the $4.5 million and $2.6 million income tax items. This disclosure will be incorporated in our filed amendment.
The attached SAB 99 analysis has not given consideration to a deferred tax adjustment currently being evaluated by the Company pertaining to the opening acquisition balance sheet of the Reciprocal Businesses. This adjustment would not have any effect on consolidated or Tower Group Inc. revenues, income or operating cash flows in any of the periods presented nor would it have any effect on Tower Group Inc.’s Stockholder’s Equity.
******
Should you have any questions concerning this letter please call William (Bill) Hitselberger at (212) 655-2067.
Very truly yours,
/s/ William E. Hitselberger
William E. Hitselberger
To: Tower Accounting Files
From: Steve Ciardiello
cc: Bill Hitselberger
Date: December 12, 2012
Re: 2012 SAB 99 Assessment
The SEC released Staff Accounting Bulletin 99 (SAB 99) in August 1999. SAB 99 reiterates the exclusive reliance on certain quantitative benchmarks to assess materiality in preparing financial statements is inappropriate and other qualitative factors must also be considered. In addition, Staff Accounting Bulletin 108 (SAB 108) was issued in September 2006 requiring errors to be evaluated on each of the Company’s financial statements. This evaluation must employ the “dual approach” which requires quantification under the “rollover” and “iron curtain” methods.
The purpose of this memo is to document management’s assessment of the unadjusted misstatements identified in the current and previously issued consolidated financial statements as directed by SAB 99 and SAB 108. Our assessment will evaluate the quantitative and qualitative effects of unadjusted misstatements.
Quantitative and Qualitative Assessment:
The iron curtain method represents the correcting of misstatements to ensure the balance sheet is properly stated as of the reporting date and the effects of those corrections on reported Net Income (“Balance Sheet” approach). The rollover method evaluates the out-of-period errors recorded in the current period together with the current period identified errors (“Income Statement” approach). As a rule of thumb, individual and aggregated misstatements exceeding 5% of annual Net Income are not deemed to be material by the Company without a qualitative analysis. This analysis will include the impact to Net Income Attributable to Tower Group, situations when Tower is close to a “break-even” point, and the effect of the errors to investors, analysts and other users of the financial statements. However, this rule of thumb is less apparent when considering misstatements on quarterly financial statements.
Tower management believes that while a 5% materiality threshold for quarterly financial reporting should be used as an indicator as to whether the sum of aggregated misstatements are material to the quarterly consolidated financial statements, it should not be considered a rigid threshold. Other factors should also be considered in addition to the 5% threshold, including the effects of the misstatements on annual consolidated financial statements.
In addition, out-of-period adjustments (“OOPAs”) will be recorded in the period in which items were identified. In evaluating OOPAs for materiality, management will follow Accounting Standard Codification 270-10-45-16 and assess the OOPA with respect to estimated net income for the full fiscal year and to the trend in earnings.
1 | Page
Assessment of recorded and unrecorded SUDs on 2012 consolidated financial statements:
The following schedule quantifies the misstatements (both “originating” and OOPAs) against pre-tax income, income tax expense, and net income. For originating misstatement purposes, the Company quantified the amounts against quarterly results. Management used projected annual net income for the OOPAs. However, due to our unique situation where we have both originating and OOPAs, the Company also quantified each of these against the quarterly and annual results.
The schedule below presents the effect of the misstatements using the Roll-Over method. In addition, Tower also performed various qualitative analyses further in this memo:
| (in thousands) | 2012 Quarter-to-Date | 2012 Year-to-Date | ||||||||||||||||||||||
| 1Q | 2Q | 3Q | 1Q | 2Q (c) | 3Q | |||||||||||||||||||
Pre-tax income (loss)—Reported | 31,491 | (25,546 | ) | 34,961 | 31,491 | 5,945 | 40,906 | |||||||||||||||||
Misstatement | (90 | ) | 255 | — | (90 | ) | 165 | 165 | ||||||||||||||||
%—total misstatements | -0.3 | % | -1.0 | % | 0.0 | % | -0.3 | % | 2.8 | % | 0.4 | % | ||||||||||||
Income tax exp (benefit)—Reported | 1,850 | (7,239 | ) | 9,500 | 1,850 | (5,389 | ) | 4,111 | ||||||||||||||||
Misstatement—originating items | (3,743 | ) | — | — | (3,743 | ) | (3,207 | ) | (3,207 | ) | ||||||||||||||
Misstatement—OOPA (a) | (3,989 | ) | 536 | — | (3,989 | ) | (3,989 | ) | (3,989 | ) | ||||||||||||||
%—total misstatements | -417.9 | % | -7.4 | % | 0.0 | % | -417.9 | % | 133.5 | % | -175.0 | % | ||||||||||||
Net income (loss)—Reported | 29,641 | (18,307 | ) | 25,461 | 29,641 | 11,334 | 36,795 | |||||||||||||||||
Misstatement—originating items | (3,940 | ) | — | — | (3,940 | ) | (3,149 | ) | (3,149 | ) | ||||||||||||||
Misstatement—OOPA (a) | (3,882 | ) | 791 | — | (3,882 | ) | (3,882 | ) | (3,882 | ) | ||||||||||||||
%—originating items | -13.3 | % | 0.0 | % | 0.0 | % | -13.3 | % | -27.8 | % | -8.6 | % | ||||||||||||
%—originating items and OOPA | -26.4 | % | -4.3 | % | 0.0 | % | -26.4 | % | -62.0 | % | -19.1 | % | ||||||||||||
Projected Annual Net income (b) | 110,417 | 63,807 | (6,396 | ) | 110,417 | 63,807 | (6,396 | ) | ||||||||||||||||
Misstatement—originating items | (3,940 | ) | — | — | (3,940 | ) | (3,149 | ) | (3,149 | ) | ||||||||||||||
Misstatement—OOPA (a) | (3,882 | ) | 166 | — | (3,882 | ) | (3,882 | ) | (3,882 | ) | ||||||||||||||
%—originating items | -3.6 | % | 0.0 | % | 0.0 | % | -3.6 | % | -4.9 | % | 49.2 | % | ||||||||||||
%—OOPAs | -3.5 | % | 0.3 | % | 0.0 | % | -3.5 | % | -6.1 | % | 60.7 | % | ||||||||||||
| (a) | Out-of-period adjustment (“OOPA”) represents, primarily, the reversal of a deferred tax valuation allowance and a reduction to certain deferred tax liabilities. Tower recorded this in 1Q 2012; the reversal should have been recorded in 3Q 2010 and 2009. |
| (b) | Projected Annual Net Income in 1Q, 2Q and 3Q represents Tower’s estimated annual net income at each quarter end. Change in projected annual income is primarily attributable to significant reserve strengthening taken in 2Q 2012 and estimates for Superstorm Sandy, which Tower estimated prior to filing of the Form 10-Q for September 30, 2012. |
| (c) | The change in year-to-date misstatements-originating relate to 1Q 2012 misstatements that were corrected in 2Q 2012. |
2 | Page
The schedule above highlights certain “Misstatement / Net Income” ratios that exceed the 5% threshold. In particular in 1st quarter 2012, the ratio was 26.4% for all misstatements (and 13.3% for the originating errors, excluding the OOPAs). In evaluating these misstatements from a qualitative perspective, while management recognizes the 13.3% error to net income exceeds the 5% rule of thumb, as discussed above, the 5% rule of thumb is only one metric when looking at the originating misstatements for quarterly results. We also view the effect of the error on projected annual net income to be important as well, and in this situation, that misstatement % (for originating entries), was reduced to 3.6% (and for all misstatements, including the OOPA this was 7.1%)
In addition, we also looked to the misstatements in combination with the other qualitative factors considered below, which include the impact to Net Income Attributable to Tower and the effect on metrics used by Tower’s analysts, and concluded the amount did not cause the financial statements to be materially incorrect and would not influence a user of the financial statements.
Management also considered the misstatements relative to the 2Q 2012 year-to-date and 3Q 2012 year-to-date figures, and also considered the impact against projected net income. However, since the Company’s projected net income declined due to a significant 2nd quarter reserve strengthening and 4th quarter Superstorm Sandy, management also considered the magnitude of the error against the Company’s three-year net income average.
After consideration of all the quantitative and qualitative matters discussed within this document, management believes the misstatements for each of the quarters in 2012 are not material to users of the financial statements.
See the following qualitative analysis below:
Qualitative Assessment:
The following items were considered by management. In connection with this analysis, it is important to note that all unadjusted misstatements were identified by Tower during its reconciliation, account analysis, and consolidated financial statement preparation process and disclosed to our independent accountants, PricewaterhouseCoopers and to the Tower Audit Committee.
| 1. | Do the misstatements arise from an item capable of precise measurement or whether it’s an estimate and the degree of imprecision inherent in the estimate? |
The tax misstatements identified above are capable of precise measurement. The tax items originated from accounting for income taxes associated with business combinations.
The non-tax related misstatements are nearly all capable of precise measurement.
3 | Page
| 2. | Do the misstatements mask a change in earnings or other trends or change a loss into income or vice versa? |
No. We believe the misstatements do not mask a change in earnings or other trends or change a loss into income or vice-versa.
| • | Pre-tax income: The misstatements have very little effect on pre-tax income. |
| • | Income tax expense: The tax misstatements reduce the income tax expense caption in the quarter-to-date and year-to-date income tax expense line and do not cause an expense to be a benefit in any period except as follows: in 2nd quarter year-to-date tax expense, Tower reported a benefit of $5.4 million that otherwise would have been an expense of $2.2 million. This is due in large part to the near breakeven year-to-date results as of June 30, which resulted in a low amount of pre-tax income and tax expense, and accordingly, the tax credit recorded resulted in tax expense to income which would not have occurred at a more “normal” level of pre-tax income. The impact of these tax adjustments are disclosed in both Tower’s earnings statement (it was noted on the first page of the earnings release) as well as in the notes to the financial statements. This did not change any trend in 2Q year-to-date net income. The Company also notes that in a review of analyst reports (discussed further below), the analysts focus on QTD results as opposed to year-to-date. Accordingly, the Company believes this item does not result in the 2Q YTD financials being materially misstated taken in connection with the other analysis presented herein. See item 11 below for discussion on analysts. |
| • | Net income:The misstatements do NOT alter trends in net income and do not cause a net loss to be reported as net income or vice-versa. |
| 3. | Does the misstatement hide a failure to meet analysts’ consensus expectations? |
No. The misstatements do not alter any trends, except for the change in 2Q YTD income tax expense, which was discussed in item 2 above. Also, the misstatements do not hide a failure to meet analysts’ consensus expectations. The tax misstatement and OOPA were disclosed in Tower’s Form 10-Q’s and Tower’s quarterly earnings releases, and in the earnings release, Tower disclosed the tax rate excluding the effect of the misstatement. In addition, Tower noted the analysts explicitly identified the tax misstatement as a non-recurring item in their quarterly reports, and did not use the reported rate in their trend analysis.
| 4. | Do the misstatements concern a segment or other portion of the business that has been identified as significant to the Company’s operations and profitability? |
No. The tax misstatements do not concern any segment reporting or other portion of the business that has been identified as significant to the Company’s operations and profitability.
4 | Page
| 5. | Do the misstatements affect the Company’s compliance with regulatory requirements? Loan covenants? Other contractual requirements? |
No. The misstatements would have no affect on compliance with regulatory, loan, or other contractual requirements.
| 6. | Do the misstatements affect certain metrics critical to senior management’s compensation plans? |
No.
| 7. | Were the misstatements deemed intentional or conceal an unlawful transaction? |
No. None of the misstatements were deemed intentional or conceal an unlawful transaction.
| 8. | Other Considerations—Net Income vs. Net Income Attributable to Tower Shareholders |
As disclosed in the Company’s consolidated financial statements, certain accounting standards require the Company to consolidate the Reciprocal Exchanges in the balance sheet and income statement; however, these amounts are then “removed” as amounts related to non-controlling interests to report Tower equity and net income and EPS. For this reason, management believes the caption “Net Income Attributable to Tower Shareholders ,” which is the last caption in our income statement is thebottom line figure. Management believes Net Income Attributed to Tower Shareholders needs to be considered the primary consideration when evaluating misstatements for materiality.
This view is supported by the fact that Earnings Per Share is calculated from Net Income Attributed to Tower Shareholders and the analysts that track Tower are solely focused on net income available to Tower Group. These analysts include: William Blair; FBR Capital Markets & Co.; Keefe, Bruyette & Woods; and Sidoti & Company, LLC. Published reports by each of these analysts make reference to Net Income Attributed to Tower, Inc. and all key metrics such as EPS and book value per share are representative of Tower results only (excluding the Reciprocal Exchanges). The following excerpts were taken from analyst reports to support the assertion:
| • | Keefe, Bruyette & Woods analyst report as of 8/6/2012: “While the reciprocal exchanges that Tower Group manages are consolidated, our comparisons and commentary reflect Tower’s stock companies only”. |
| • | FBR Flash as of 8/6/2012: “We forecast fee income separately as we model the personal lines segment excluding reciprocals”. |
| • | William Blair report dated July 31, 2012 and Sidoti & Company report dated August 7, 2012 presented tables with net income and operating income. In determining such income numbers, we noted that the net income attributable to reciprocals was excluded. |
5 | Page
In addition, the Company’s earnings release and investor presentation highlight Tower-only net income, operating income, earnings per share, and book value per share from a common shareholder’s perspective and therefore exclude the Reciprocal Exchanges. Accordingly, management believes that, while accounting standards require the Reciprocal Exchanges be consolidated in the Company’s financial statements, the key users of the financial statements focus solely on the results that are attributed to Tower Group, Inc.
Management also considered that users associated with the Reciprocal Exchanges financial statements (NCI) would not be looking to the Tower financial statements to gain information about the performance of those entities, but would look (at least primarily) to the separate reciprocal financial statements which are produced on a quarterly basis. The Reciprocal Exchange information contained in Tower’s consolidated financial statements comprise two different reciprocal entities (New Jersey Skylands Insurance Association and Adirondack Insurance Exchange), and do not provide sufficient level of detail to disclose the results of each of the reciprocal exchanges separately. As such, Net Income Attributable to Noncontrolling Interest has little relevance to a user of either New Jersey Skylands or Adirondack Insurance Exchange. Such users would review the individual exchange financial reports that are filed with its regulators. These statements are prepared in accordance with statutory accounting rules, which is an other comprehensive basis of accounting rather than US GAAP. Importantly, this OOPA has no impact on these financial statements.
The misstatements had the following impact to Net Income Attributable to Tower:
| (in thousands) | 2012 Quarter-to-Date | 2012 Year-to-Date | ||||||||||||||||||||||
| 1Q | 2Q | 3Q | 1Q | 2Q | 3Q | |||||||||||||||||||
Tower Net income (loss)—Reported | 20,634 | (15,113 | ) | 21,629 | 20,634 | 5,521 | 27,150 | |||||||||||||||||
Misstatement—originating items | 2,154 | 166 | — | 2,154 | 3,149 | 3,149 | ||||||||||||||||||
Misstatement—OOPA | (682 | ) | — | — | (682 | ) | 16 | 16 | ||||||||||||||||
%—originating items | 10.4 | % | -1.1 | % | 0.0 | % | 10.4 | % | 57.0 | % | 11.6 | % | ||||||||||||
%—total misstatements | 7.1 | % | -1.1 | % | 0.0 | % | 7.1 | % | 57.3 | % | 11.7 | % | ||||||||||||
Projected Annual Tower Net income | 102,000 | 58,102 | (8,580 | ) | 102,000 | 58,102 | (8,580 | ) | ||||||||||||||||
Misstatement—originating items | 2,154 | 166 | — | 2,154 | 3,149 | 3,149 | ||||||||||||||||||
Misstatement—OOPA | (682 | ) | — | — | (682 | ) | 16 | 16 | ||||||||||||||||
%—total misstatements | 1.4 | % | 0.3 | % | 0.0 | % | 1.4 | % | 5.4 | % | -36.9 | % | ||||||||||||
Tower recognizes the misstatement to Tower Net Income amount by 7.1% in 1Q. In evaluating this amount, Tower also looks to the impact to projected annual Tower net income (as discussed above, we believe additional metrics should also be considered on a quarterly basis rather than just the 5% metric). Management considered the caption of the misstatements (primarily income tax expense) and also the disclosure in Note 2 to the quarterly financial statements which discloses the nature and amount of the errors. Additionally, the amount was less than 1.5% of Tower’s projected net income. We believe the amount to not be material to the consolidated financial statements.
In evaluating the impact on Tower Net Income, management recognizes the YTD 2Q 2012 misstatement percentage exceeds 50% and for YTD 3Q it exceeds 11%.
6 | Page
In addition, the estimated percentage of misstatements to Projected Annual Tower Net Income is 36.9% as of 3QYTD. While large percentages, management recognizes this is a result of the Company being in a “break-even” or close to “break-even” position resulting from the 2Q 2012 $65 million reserve strengthening (pre-tax) and estimated losses from Super storm Sandy of $95 million (pre-tax) in the fourth quarter 2012.
To assist with this evaluation, the Company evaluated the misstatements against the estimated three-year Tower net income average (2010, 2011 and estimated 2012), or $51.8 million. The $2.4 million misstatement is 4.6% of $51.8 million, which is under the 5% quantitative benchmark.
| 9. | Other Considerations—Effective tax rate |
The table below presents the impact to the effective tax rate for the consolidated balances:
| 2012 QTD | 2012 YTD | |||||||||||||||||||||||
| 1Q | 2Q | 3Q | 1Q | 2Q | 3Q | |||||||||||||||||||
Pre-tax income | 31,491 | (25,546 | ) | 34,961 | 31,491 | 5,945 | 40,906 | |||||||||||||||||
Tax expense—reported | 1,850 | (7,239 | ) | 9,500 | 1,850 | (5,389 | ) | 4,111 | ||||||||||||||||
Tax expense—misstatements | 7,732 | (536 | ) | — | 7,732 | 7,196 | 7,196 | |||||||||||||||||
Tax expense—revised | 9,582 | (7,775 | ) | 9,500 | 9,582 | 1,807 | 11,307 | |||||||||||||||||
ETR—reported | 6 | % | 28 | % | 27 | % | 6 | % | -91 | % | 10 | % | ||||||||||||
ETR—revised | 30 | % | 30 | % | 27 | % | 30 | % | 30 | % | 28 | % | ||||||||||||
1st quarter reported ETR was 6%. Management believes this amount is mitigated by the disclosure included in the 1Q Form-10Q highlighting the tax misstatements and analysts explicitly identified this in their reports (these facts are discussed elsewhere in this document). In addition, the 2Q YTD tax rate of 91% is primarily due to the close to break-even position through June 30, 2012. The Company believes the trend in the ETR for 2Q YTD, while not in trend, does not cause the financial statements to be misleading as the 2Q QTD ETR is as expected and the break-even type results make for the 2Q YTD % amounts to be skewed.
| 10. | Other Considerations—Book value per share and Tangible book value per share: |
In addition to Net Income Attributable to Tower, key metrics on which management believes analysts focus include BVPS and TBVP. This is further supported by explicit statements made by Dowling & Partners. In addition, in the proposed merger of Tower into Canopius Bermuda, the Tower common share conversion ratio is dependent upon tangible book value (“TBVPS”). The effect of the misstatements on these metrics for each of the quarters in 2012 is as follows:
| 1Q 2012 | 2Q 2012 | 3Q 2012 | ||||||||||||||||||||||
| As Reported | Revised | As Reported | Revised | As Reported | Revised | |||||||||||||||||||
BVPS | 26.83 | 26.75 | 26.53 | 26.44 | 27.49 | 27.40 | ||||||||||||||||||
TBVPS | 17.86 | 17.88 | 17.31 | 17.32 | 18.32 | 18.34 | ||||||||||||||||||
7 | Page
Management believes the effect of the misstatements on the above metrics is not material. The misstatements would have the effect of increasing TBVPS by 1 to 2 cents.
| 11. | Other Considerations—Users of the Financial Statements |
Tower defines potential users of its financial statements into 5 broad categories:
1) Investors and their information sources (sell side analysts)
2) Policyholders and customer distribution sources
3) Creditors
4) Rating agencies
5) Regulators
In reviewing the reaction of investors and their information sources to our first quarter 2012 results, we believe that the reporting and disclosure around the tax adjustment eliminated risk of misinterpreting trend, or resulted in analysts having a material misunderstanding of our earnings. Affixed below are comments from our significant analysts after our first quarter release.
KBW:
First note:
Tower Group’s beat to our estimate largely reflects a lower-than-expected effective tax rate ($0.05/share benefit).”
“A lower-than-expected tax rate (20.9% versus our 28.6%) benefited EPS by $0.05/share. The lower tax rate reflects a revaluation of tax liabilities and is non-recurring.”
“The most significant difference between actual and our estimate was Tower’s lower-than-expected effective tax rate, which boosted EPS by $0.05/share. A non-recurring revaluation of certain tax liabilities reduced the tax rate by 10 points.”
FBR:
“Absent a benefit from a true-up of tax liabilities, the core result was ~$0.51.”
8 | Page
Sidoti:
“TWGP reported 1Q:12 operating EPS of $0.56, versus our estimate of $0.58 and consensus of $0.55; we note that operating EPS included a roughly $0.07 ($2.6 million) tax benefit.”
“Operating earnings were boosted by a favorable effective tax rate (20.9%, versus 32.4% in 2011) primarily due to a revaluation of tax liabilities resulting in a $2.6 million benefit (roughly 10 percentage points), which added roughly $0.07 to 1Q:12 operating EPS.”
All of the analysts focus solely on operating earnings (GAAP net income adjusted for realized capital gains losses and deal acquisition related costs, net of tax), which is reconciled back to Tower net income in both our 10Q and in our earnings release. Moreover, none of our analysts review the operating results of the reciprocal exchanges, as they understand that these results do not affect EPS, Operating EPS or our disclosed book value or tangible book value per share, all of which exclude operating results and net asset values of the reciprocal exchanges.
In reviewing the reaction of policyholders and customer distribution sources to our earnings release and to our 10Q, it is very common for such constituents to not focus on our GAAP results, and instead to review financial and statistical information prepared by our major rating agency, A.M. Best. This data is uniformly presented on a statutory basis, and GAAP information is not used in our rating model. The adjustment did not have an impact on our statutory financial statements that are used by A.M. Best.
In reviewing from a creditor’s perspective, we note that our credit facility has financial covenants, and that the adjustments would not have an impact on the covenants.
Similarly, the trust preferred securities have covenants; however, these covenants are all based on statutory reporting data.
As mentioned above, our main rating agency A.M Best, formulates its analysis and capital scores on our statutory financial results, as do Fitch and Demotech, the other two rating agencies that cover our company.
As an insurance company, we are required to furnish financial information to our state regulators as well. This information is required to be presented on a legal entity basis, and such data was not affected by the adjustment.
With respect to the adjustment affecting the reciprocal exchange, we believe there are few users of the reciprocal exchange financial information, as the exchanges have no creditors, and have no investors other than the Company (which holds a surplus note due from the reciprocal exchanges). As noted above, the users of the reciprocal financial information are therefore all interested in the statutory financial information of the reciprocal exchanges, which were not affected by the adjustment.
9 | Page
Assessment of OOPAs on previously issued financial statements:
As of September 30, 2012, the recorded out-of-period adjustments (“OOPAs”) and the impact on previously issued annual consolidated financial statements are as follows:
| OOPAs | ||||
Other operating expenses | $ | (1,235,000 | ) | |
Commission expense | 455,000 | |||
Interest expense | 672,609 | |||
OUE | — | |||
Income tax expense | 3,989,000 | |||
|
| |||
| $ | 3,881,609 | |||
|
| |||
Impact on Prior Period Net Income:
| Net Income— Reported | OOPA | % | ||||||||||
2011 | 72,105 | (780 | ) | -1.1 | % | |||||||
2010 | 99,211 | 4,564 | 4.6 | % | ||||||||
2009 | 96,798 | 98 | 0.1 | % | ||||||||
|
|
|
|
|
| |||||||
| $ | 268,114 | $ | 3,882 | |||||||||
|
|
|
| |||||||||
Based upon the tables above, management has concluded the OOPAs do not have a material impact to previously issued financial statements.
Balance sheet reclasses
Management’s SUD identifies a reclass of approximately $19 million for commission payables that are netted against the Companies premium receivable balance. While management has historically used this presentation, a more accurate accounting would be to reclass the $18.9 million to liabilities in the financial statements. For the 3Q 2012, management has concluded this reclass is not material to the consolidated balance sheet.
In the 3rd quarter 2012, management changed the presentation of cash advances it sends to one of its TPAs. At December 31, 2011, this amount was approximately $4 million, and the Company recorded it as a contra-other liability. As of 3Q 2012, this amount was approximately $18 million. Management concluded this advance should be reported as an asset, and, effective 3Q 2012, amounts is now in other assets. Management considered reclassing the prior period (December 31, 2011) amount from other liability to other assets, and concluded the $4 million was not significant, relative to Tower’s consolidated balance sheet, to warrant this reclass.
10 | Page
Cash Flow Statement
The adjustments discussed herein have no effect on cash flows provided by (used in) operating activities, investing activities or financing activities for each of the quarters in 2012. As discussed in the conclusion below, the Company will be revising its previously issued financial statements (beginning in 2009) to record these adjustments as well as other previously identified immaterial adjustments into the periods in which they originated. The effects of these adjustments will have immaterial impacts to cash flows provided by (used in) operating activities, investing activities or financing activities for some of these previously issued financial statements.
Conclusion
Based on quantitative and qualitative assessments of recorded and unrecorded entries affecting current and prior period financial statements, management has concluded that its previously issued annual and quarterly financial statements are not materially misstated.
However, management has elected to revise its previously issued financial statements to record the misstatements in the periods in which they originated. This includes the tax adjustments disclosed by the Company in its Form 10-Q filings for each of the quarters in 2012 as well as other immaterial misstatements. Tower will revise its consolidated balance sheet as of December 31, 2011 and 2010 and the statements of income and comprehensive income, cash flows and changes in shareholders’ equity for each of the three years in the period ended December 31, 2011. These financial statements will be revised in its 2011 Form 10-K/A. These financial statements will include disclosure as prescribed by ASC 250-10-50-7. In addition, the Company will also revise its 2012 and 2011 quarterly financial statements presented in the March 31, 2012, June 30, 2012 and September 30, 2012 Form 10-Qs
11 | Page
Evaluation of Misstatements for Fraud and Control Deficiencies
Management evaluated each of misstatements listed on the Summary of Unadjusted Differences for any possible effect on the Company’s internal controls over financial reporting. We believe none of the items represent controls weaknesses as each item was identified through management’s reconciliation and detective control process.
In regards to the income tax SUDs, management identified these items in connection with its final “clean-up” of the deferred tax inventory. The Company engaged tax specialists to assist with the preparation of the income tax provision and deferred tax inventory (of which their procedures and output are subject to review and approval by Company management). The tax specialists completed their inventory, and the last items identified were the result of tax matters relating to PGAAP accounting from the Company’s 2009 acquisition of CastlePoint and 2010 acquisition of OneBeacon Personal Lines. The Company now has detailed analysis supporting each item in its deferred inventory; accordingly, management believes no further action is to be taken on the tax matters. In the event the Company has future acquisitions, detailed control procedures will be operationalized to ensure appropriateness of the deferred taxes set-up in any opening balance sheet.
In connection with its testing of the operational effectiveness of controls, management will continue to monitor the quarterly controls in the tax provision area to ensure the Company has the proper design of controls and that they operate effectively through 2012. Management has disclosed these changes to its internal control over financial reporting in its 3Q 2012 Form 10-Q.
Management considered the effect of these control deficiencies on the 2011 consolidated financial statements. The Company had no acquisitions in 2011 and purchase accounting was not required. In addition, the Company has had no such acquisitions in 2012. Management believes the enhancements to the control environment are sufficient to prevent material misstatements in accounting for any future business combination requiring purchase accounting.
12 | Page