 | One Astoria Federal Plaza Lake Success, NY 11042-1085 (516) 327-3000 |
October 18, 2007
Mr. Kevin W. Vaughn
Branch Chief
Division of Corporation Finance
United States Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549
Mail Stop 4561
| Re: | Astoria Financial Corporation |
| Form 10-K for Fiscal Year Ended December 31, 2006 | |
| Form 10-Q for Fiscal Quarter Ended March 31, 2007 | |
| Form 10-Q for Fiscal Quarter Ended June 30, 2007 | |
| File No. 001-11967 |
Dear Mr. Vaughn:
We wish to acknowledge receipt of your letter dated September 19, 2007 concerning the Securities and Exchange Commission’s review of the financial statements and related disclosures contained in Astoria Financial Corporation’s (the “Company”) Form 10-K for the Fiscal Year Ended December 31, 2006, Form 10-Q for the Fiscal Quarter Ended March 31, 2007 and Form 10-Q for the Fiscal Quarter Ended June 30, 2007. The following sets forth the Company’s responses to your comments.
Form 10-K for the year ended December 31, 2006
Non-performing Assets, page 10
1. | In your discussion here on page 10, you state that your non-performing loans continue to remain at low levels in relation to the size of your portfolio. Further, you disclose an overall decline in non-performing loans and assets. However, you also disclose here that you sold $10.1 million in non-performing loans in the third quarter of 2006. Please revise your discussion here, as well as elsewhere in your filing where you refer to decreases in non-performing loans, to more clearly address the impact of that sale on the level of and trends within non-performing assets. Clarify how much of the change in the level of non- |
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 2
performing loans was due to actual improvements in credit quality and how much was due to the sale of non-performing loans. If true, revise to specifically disclose that without the sale of these non-performing loans during the third quarter, the level of your non-performing loans at year end would have been higher than the previous year.
Of the $10.1 million in non-performing loans sold during the third quarter of 2006, $5.5 million were non-performing as of December 31, 2005. The remainder became non-performing during 2006. Since these loans were sold in the third quarter of 2006, we are unable to determine with any degree of certainty whether some or all of these loans would have remained non-performing as of December 31, 2006, particularly in light of our aggressive collection efforts and prior experience with other borrowers. However, assuming the $10.1 million of non-performing loans sold were not sold and were both outstanding and non-performing at December 31, 2006, our non-performing loans would have totaled $69.6 million, or an increase of $4.5 million from December 31, 2005, and our non-performing assets would have totaled $70.2 million, or an increase of $4.1 million from December 31, 2005. Additionally, at December 31, 2006, our ratio of non-performing loans to total loans would have increased to 0.46%, our ratio of non-performing assets to total assets would have increased to 0.33% and the allowance for loan losses as a percentage of total non-performing loans would have decreased to 114.92%.
We do not believe that the above information would materially enhance the disclosures provided in our December 31, 2006 Form 10-K. Additionally, we believe that information based on assumptions about the performing or non-performing status of loans subsequent to sale would require speculation on our part as to future events, and therefore, would not be meaningful or decision-useful to the reader. Therefore, we believe that a revision to the December 31, 2006 Form 10-K, the March 31, 2007 Form 10-Q or the June 30, 2007 Form 10-Q, or inclusion of such information in future filings, is not necessary.
Critical Accounting Policies
Allowance for Loan Losses, page 42
2. | If true, please revise the first complete paragraph on page 44 to confirm that the balance of your allowance for loan losses represents your best estimate of the probable inherent losses in your loan portfolio as of the balance sheet dates presented. Similarly revise the disclosures related to your table at the bottom of page 67 to include similar disclosure as of each of the balance sheet dates presented. |
We believe the disclosure in the last sentence of the first complete paragraph on page 44 in our December 31, 2006 Form 10-K conveys substantially the same information as the Staff’s requested text. As indicated in our discussion in the fourth paragraph on page 43 for the “Allowance for Loan Losses,” our determination of the allowance for loan losses includes a review of a range of historical losses (charge-offs) and coverage percentages, which are an integral part of our judgment in developing estimated loss percentages for our portfolio. In
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 3
conjunction with a variety of other factors, which are also discussed in the fourth paragraph on page 43, we determine a range of appropriate allowance coverage percentages for each of our portfolio segments. Additionally, although we do not have a comparable statement regarding our allowance for loan losses on page 67, we do have such a statement in the last sentence of the second complete paragraph on page 63. Since our December 31, 2006 Form 10-K included disclosures that we believe are comparable to the Staff’s requested text, we believe that a revision to the aforementioned report is not necessary.
In the March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q, we have disclosures which also convey substantially the same information as the text requested by the Staff. In the March 31, 2007 Form 10-Q, such text appears in the last sentence of the third full paragraph on page 16 and in the third sentence of the first paragraph on page 30. In the June 30, 2007 Form 10-Q, such text appears in the last sentence of the second full paragraph on page 17 and in the third sentence of the last paragraph on page 33. Since our March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q included disclosures that we believe are comparable to the Staff’s requested text, we believe that revisions to the aforementioned reports are not necessary.
We will, in future filings beginning with our September 30, 2007 Form 10-Q, include a statement that the balance of our allowance for loan losses represents management’s estimate of the probable inherent losses in our loan portfolio as of each balance sheet date presented.
Provision for Loan Losses, page 56
3. | Please revise your disclosure here to discuss the recent trends in the loan-to-value ratios of your non-performing loans. For example, on page 50 you disclose that the real estate market has softened, and on page 31 you disclose that property values are, in some cases, stagnant or declining. As it appears that you rely heavily on loan to value ratios, among other things, in both your underwriting practices and in your determination of the provision for loan losses, please revise your disclosure to discuss the impact of this trend on management’s determination of the provision for loan losses. Please address these trends in your discussion for both your determination of the specific allowance and the general allowance. |
Please see our response to comment number 5 which incorporates the response to this comment.
4. | You disclose here on page 56 that the composition of your loan portfolio has remained consistent over the last several years. Please revise this section to discuss the significant increases in interest-only loans and no- or low-documentation loans within your portfolios during the last five years. Revise to specifically discuss how you considered these trends and the nature of these loans combined with the softening of the real estate market and declining property values as noted above in determining your allowance for loan losses. |
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 4
Please see our response to comment number 5 which incorporates the response to this comment.
5. | You state at the top of page 57 that “Based on the evaluation of the foregoing factors, (y)our 2006 analyses did not indicate that a change in (y)our allowance for loan losses at December 31, 2006 was warranted.” Please revise this paragraph to specifically disclose how you determined to decrease the amount of your allowance for loan losses during each of the last several years through charge-offs with no provision expense. |
The following incorporates our responses to comment numbers 3, 4, 5, 6, 13 and 14. We believe that a combined response, addressing the various inter-related factors affecting our allowance for loan losses and provision for loan losses, will facilitate the Staff’s review. Our response first addresses the Staff’s comments related to the December 31, 2006 Form 10-K and then addresses the Staff’s comments related to the March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q. Within our responses for each period, we further identify our comments as being related to our portfolio composition, non-performing loans, historical loss experience and current market conditions.
Please note that on May 3, 2006 we responded to a comment in a Staff letter dated April 19, 2006 which addressed the allowance for loan losses and the provision for loan losses at and for the three years ended December 31, 2005. Our response was filed with the Securities and Exchange Commission, as EDGAR Correspondence, on May 3, 2006. We therefore have focused our response provided herein on the allowance for loan losses and the provision for loan losses for periods subsequent to December 31, 2005, although data from earlier periods has been incorporated to the extent such data is relevant to our response.
December 31, 2006 Form 10-K
Our Form 10-K for the fiscal year ended December 31, 2006 discusses in detail the various qualitative and quantitative factors we consider in determining our provision for loan losses and our allowance for loan losses. We discuss our methodology for determining our provision for loan losses and the impact that these factors and trends have on our evaluation. As permitted by the instructions to Form 10-K, we provide cross-references, when applicable, to further enhance those discussions. The sections that we cross reference are “Critical Accounting Policies” and “Asset Quality.” The “Critical Accounting Policies – Allowance for Loan Losses” section of “Management’s Discussion and Analysis of Financial Condition and Results of Operations” (the “MD&A”) on pages 42 through 44 of the December 31, 2006 Form 10-K discusses the analyses we perform and the factors we consider in determining our allowance for loan losses. The “Asset Quality” section of the MD&A on pages 65 through 68 of the December 31, 2006 Form 10-K includes tables of our non-performing loans and assets, classified assets, provision for loan losses and allowance for loan losses.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 5
As discussed on pages 42 through 44 in “Critical Accounting Policies – Allowance for Loan Losses” in the December 31, 2006 Form 10-K, our determination of the allowance for loan losses includes a review of our historical loss experience and delinquency levels, by loan type and segment, which are an integral part of our judgment in developing estimated loss percentages for our portfolio. In conjunction with a variety of other factors, we determine a range of estimated losses and appropriate allowance coverage percentages for each of our portfolio segments. Other factors include, but are not limited to the size, composition, risk profile, delinquency levels and cure rates of our portfolio as well as our credit administration and asset management philosophies and procedures. In addition, we evaluate and consider the impact that existing and projected economic and market conditions may have on the portfolio, as well as known and inherent risks in the portfolio. We also evaluate and consider the allowance ratios and coverage percentages set forth in both peer group and regulatory agency data and any comments from the OTS resulting from their review of our general valuation allowance methodology during regulatory examinations. Our focus, however, is primarily on our historical loss experience and the impact of current economic conditions. After evaluating these variables, we determine appropriate allowance coverage percentages for each of our portfolio segments and the appropriate level of our allowance for loan losses. Additionally, within “Asset Quality,” our historical loss experience is disclosed in the second table on page 67 of the December 31, 2006 Form 10-K and our non-performing assets and loans are disclosed in the tables on pages 65 and 66 of the December 31, 2006 Form 10-K. We believe the allowance for loan losses has been established and maintained at levels that reflect our estimate of probable loss inherent in our loan portfolio, giving consideration to the composition and size of our loan portfolio, our charge-off experience and our non-accrual and non-performing loans, as well as the changes in loan-to-value ratios of loans in our portfolio and the changes in market conditions.
Portfolio Composition
As evidenced by the table on the top of page 8 of the December 31, 2006 Form 10-K, the composition of our loan portfolio, by property type, has remained consistent over the last several years. At December 31, 2006, our loan portfolio was comprised of 69% one-to-four family mortgage loans, 20% multi-family mortgage loans, 7% commercial real estate loans and 4% other loan categories. Within our one-to-four family portfolio, interest-only loans totaled $6.17 billion, or 60% of our one-to-four-family loans at December 31, 2006 (as disclosed in the second paragraph on page 4 of the December 31, 2006 Form 10-K); $4.97 billion, or 51% at December 31, 2005; $3.13 billion, or 35% at December 31, 2004; and $1.67 billion, or 19% at December 31, 2003. Interest-only loan data for one-to-four family loans is unavailable prior to 2003. We do not originate negative amortization loans, payment option loans, or other loans with short-term interest-only periods. During the second quarter of 2006, we began underwriting our interest-only adjustable rate mortgage (“ARM”) loans based on a fully amortizing thirty year loan. Additionally, effective in 2007, in accordance with federal banking regulatory requirements, we began underwriting our interest-only ARM loans at the fully indexed rate. Based on our underwriting standards and cumulative experience with our interest-only loans, these loans have performed as well as our fully
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 6
amortizing loan products (see discussion below regarding non-performing loans). The respective allowance coverage factors utilized for interest-only and amortizing loans give appropriate recognition to the potential for increased risk of default (and risk of loss) attributable to payment increases on interest-only loans once principal amortization begins.
As disclosed in the third paragraph on page 4 of the December 31, 2006 Form 10-K, within our one-to-four family portfolio, we also have reduced documentation loans which totaled $2.90 billion, or 28% of our one-to-four family loans at December 31, 2006. Reduced documentation loans which are also interest-only loans totaled $2.15 billion, or 21% of our total one-to-four family loans at December 31, 2006. Reduced documentation loans are comprised primarily of SIFA (stated income, full asset) loans which require a potential borrower to complete a standard mortgage loan application and require the verification of a potential borrower’s asset information on the loan application, but not the income information provided. In addition, SIFA loans require the receipt of an appraisal of the real estate used as collateral for the mortgage loan and a credit report on the prospective borrower. SIFA loans comprised approximately 30% of our one-to-four family originations and purchases in 2006 and 2005. Reduced documentation loan data is not available for periods prior to 2005.
Multi-family and commercial real estate interest-only loans totaled $542.6 million, or 13% of our multi-family and commercial real estate loans at December 31, 2006 (as disclosed in the last paragraph on page 5 of the December 31, 2006 Form 10-K) and $372.2 million, or 10% of our multi-family and commercial real estate loans at December 31, 2005. We began originating interest-only multi-family and commercial real estate loans in 2005. These loans do not represent a material component of our loan portfolio.
Our loan-to-value ratios upon origination are low overall and have been consistent over the past several years. The average loan-to-value ratios, based on current principal balance and original appraised value, of total one-to-four family loans outstanding as of December 31, 2006, by year of origination, were 67% for 2006, 69% for 2005, 69% for 2004, 68% for 2003 and 55% for pre-2003 originations. As of December 31, 2006, average loan-to-value ratios, based on current principal balance and original appraised value, of total multi-family and commercial real estate loans outstanding, by year of origination, were 62% for 2006, 66% for 2005, 65% for 2004, 62% for 2003 and 57% for pre-2003 originations.
Non-performing Loans
As discussed in the second full paragraph on page 10 of the December 31, 2006 Form 10-K, our non-performing loans continue to remain at low levels relative to the size of our loan portfolio. Our non-performing loans, which are comprised primarily of mortgage loans, decreased $5.6 million to $59.4 million, or 0.40% of total loans, at December 31, 2006, from $65.0 million, or 0.45% of total loans, at December 31, 2005. Our non-performing mortgage loans totaled $58.6 million at December 31, 2006 and had an average loan-to-value ratio of 71%, based on current principal balance and original appraised value, and totaled $64.5
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 7
million at December 31, 2005 and had an average loan-to-value ratio of 69%, based on current principal balance and original appraised value. Included in non-performing loans at December 31, 2006 were non-performing interest-only one-to-four family loans totaling $14.5 million, or 35% of total non-performing one-to-four family loans, which had an average loan-to-value ratio of 76%, based on current principal balance and original appraised value. There were no non-performing interest-only multi-family and commercial real estate loans at December 31, 2006. Also included in non-performing loans at December 31, 2006 were non-performing reduced documentation one-to-four family loans totaling $16.2 million, or 39% of total non-performing one-to-four family loans, which had an average loan-to-value ratio of 70%, based on current principal balance and original appraised value. The average age of our non-performing mortgage loans since origination was 3.7 years at December 31, 2006. Although the loan-to-value ratios of our non-performing loans, based on current principal balance and original appraised value, are somewhat higher than the loan-to-value ratios of our entire portfolio, they are still, on average, within our current guidelines for loan-to-value ratios at origination.
Historical Loss Experience
Net loan charge-offs are a strong indicator of the impact that loan-to-value ratios have on a portfolio’s overall performance and inherent loss levels over time. Our net charge-offs have remained stable, despite the recent softening of the real estate market, and the changes in the composition of our one-to-four family loan portfolio, and in recent years have been primarily attributable to a small number of loans. Both our net loan charge-offs as a percentage of average loans outstanding and our net mortgage loan charge-offs as a percentage of our average mortgage loans outstanding have remained at or below one basis point for each of the past five years. As disclosed in the second table on page 67 of the December 31, 2006 Form 10-K, our net loan charge-offs totaled $1.2 million for the year ended December 31, 2006, $1.6 million for the year ended December 31, 2005, $363,000 for the year ended December 31, 2004, $425,000 for the year ended December 31, 2003 and $1.0 million for the year ended December 31, 2002. Included in the net charge-offs for the year ended December 31, 2006 was a single multi-family loan charge-off totaling $967,000. Included in the net charge-offs for the year ended December 31, 2005 was a single commercial real estate loan charge-off totaling $650,000. The majority of the net charge-offs for years prior to 2005 were related to consumer loans and lines of credit which we assumed in an acquisition of another financial institution. In reviewing our charge-off experience, particularly for the years ended December 31, 2006 and 2005, we determined that the single events noted above represented unique loans and/or circumstances and were not indicative of a trend of increased charge-offs. We incorporated that determination in our review of the adequacy of our allowance for loan losses and determined that the total balance of our allowances, at the respective year ends remained within the acceptable range of probable losses within our portfolio and that no additional provision for loan losses was required.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 8
Current Market Conditions
In determining our allowance for loan losses, we evaluate and consider the impact that existing and projected economic and market conditions may have on our loan portfolio. The disclosures we made in the December 31, 2006 Form 10-K in the last full paragraph on page 50, regarding the softening of the real estate market, and in the third paragraph under the first risk factor on page 31, regarding stagnant to declining property values, were included to explain the decrease in mortgage loan repayment and refinance activity, which enabled us to grow the mortgage loan portfolio, despite a reduction in mortgage loan originations. These market conditions were not discussed within our disclosure regarding the allowance for loan losses and related provision, as they had not had a discernable negative impact on our trends of non-performing loans, loan-to-value ratios or loan loss experience on our loan portfolio during 2006 as previously discussed.
Our discussion on page 31 of the December 31, 2006 Form 10-K suggests stagnant to declining property values beginning in 2006. This is further supported by the Federal Housing Finance Board’s Monthly Interest Rate Survey, which is used by the Office of Federal Housing Enterprise Oversight in establishing conforming loan limits, which indicated a rise in property values through October 2005 and only a modest decrease of 0.16% from October 2005 to October 2006. As a result, the vast majority of loans in our portfolio originated prior to 2006 would most likely have current loan-to-value ratios lower than those at the origination date, and loans originated in 2006 may have only slightly higher loan-to-value ratios than at origination. Our policy regarding obtaining updated loan values is discussed further below. These loan-to-value ratio trends apply to both our fully amortizing and interest-only loan portfolios. We continue to monitor these market trends and their impact, if any, on our loan portfolio and consider their effect in determining our allowance for loan losses.
Specific valuation allowances are applied to loans where the current estimated value of the loan (using the market price of the loan, if one exists; the estimated fair value of the collateral, for collateral dependent loans; or the present value of expected future cash flows), less estimated selling costs, is lower than our recorded investment in the loan. Generally, loans which are classified by the Asset Classification Committee have a component of our general valuation allowance allocated to them. These allocations are considered in the determination of our total allowance for loan losses. Should the collateral value, less estimated selling costs, of a classified loan fall below the recorded investment of that loan, a specific valuation allowance is established for this difference through the reclassification of the general valuation allowance. Updated estimates of value are obtained when loans are classified by our Asset Classification Committee as either substandard or doubtful, as well as special mention and watch list loans in excess of $2.5 million. For loans meeting these criteria, we update our estimate of the value of the property securing the loan through either an updated appraisal, a visual inspection of the property or a market analysis of comparable homes in the area. For multi-family and commercial real estate loans, we may also perform a cash flow analysis for the property based on current borrower financial statements. The
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 9
impact of current market conditions on the value of a particular property is reflected in the results of our updated valuation analysis.
In summary, our non-performing loans and charge-offs continue to remain stable and at low levels relative to the size of our loan portfolio. Additionally, our loan-to-value ratios upon origination are low overall and have been consistent over the past several years. In evaluating our allowance for loan losses for the year ended December 31, 2006, we continued to look at market conditions, including real estate values. However, those market observations had not had a discernable negative impact on our trends of non-performing loans, loan-to-value ratios or loan loss experience as of and for the year ended December 31, 2006, and did not indicate that we would experience a significant change in those trends. As such, considering that we believe each of the foregoing items was adequately disclosed in our filing, we believe that revisions to the December 31, 2006 Form 10-K are not necessary.
March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q
Similar to the disclosures in the December 31, 2006 Form 10-K, our March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q discuss in detail the various qualitative and quantitative factors we consider in determining our provision for loan losses and our allowance for loan losses. The “Critical Accounting Policies – Allowance for Loan Losses” section of the MD&A on pages 15 and 16 of the March 31, 2007 Form 10-Q and pages 16 and 17 of the June 30, 2007 Form 10-Q discuss the analyses we perform and the factors we consider in determining our allowance for loan losses. The “Asset Quality” section of the MD&A on pages 31 through 33 of the March 31, 2007 Form 10-Q and pages 36 through 39 of the June 30, 2007 Form 10-Q include tables of our non-performing loans and assets, provision for loan losses and allowance for loan losses.
Portfolio Composition
There has not been a significant change in our portfolio composition from December 31, 2006. One-to-four family interest-only loans totaled $6.42 billion, or 62% of our one-to-four-family loans at March 31, 2007 and $7.01 billion, or 64% of our one-to-four family loans at June 30, 2007. Multi-family and commercial real estate interest-only loans totaled $578.2 million, or 14% of our multi-family and commercial real estate loans at March 31, 2007 and $602.3 million, or 15% of our multi-family and commercial real estate loans at June 30, 2007. Reduced documentation loans totaled $2.89 billion, or 28% of our one-to-four family loans at March 31, 2007 and $2.95 billion, or 27% of our one-to-four family loans at June 30, 2007.
Our average loan-to-value ratios upon origination continue to be low overall and have remained consistent with the loan-to-value ratios upon origination at December 31, 2006. As of June 30, 2007, average loan-to-value ratios, based on current principal balance and original appraised value, of total one-to-four family loans outstanding, by year of origination, were 66% for the six months ended June 30, 2007, 67% for 2006, 69% for 2005, 68% for
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 10
2004, 61% for 2003 and 53% for pre-2003 originations. As of June 30, 2007, average loan-to-value ratios, based on current principal balance and original appraised value, of total multi-family and commercial real estate loans outstanding, by year of origination, were 64% for the six months ended June 30, 2007, 67% for 2006, 67% for 2005, 64% for 2004, 60% for 2003 and 55% for pre-2003 originations.
Non-performing Loans
The balances of our non-performing loans and their related average loan-to-value ratios at March 31, 2007 and June 30, 2007 were consistent with the balances and ratios at December 31, 2006. Our non-performing loans totaled $67.9 million, or 0.45% of total loans, at March 31, 2007 (as disclosed in the first paragraph on page 30 of the March 31, 2007 Form 10-Q) and $64.0 million, or 0.41% of total loans, at June 30, 2007 (as disclosed in the first paragraph on page 34 of the June 30, 2007 Form 10-Q). Our non-performing mortgage loans totaled $66.8 million at March 31, 2007 and had an average loan-to-value ratio of 71%, based on current principal balance and original appraised value, and totaled $63.0 million at June 30, 2007 and had an average loan-to-value ratio of 72%, based on current principal balance and original appraised value. The average age of our non-performing loans since origination was 3.8 years at both March 31, 2007 and June 30, 2007.
Historical Loss Experience
Our annualized net mortgage loan charge-offs remained at or below one basis point of our average mortgage loans outstanding for the three months ended March 31, 2007 and the three and six months ended June 30, 2007. Our annualized net loan charge-offs as a percentage of average loans outstanding was less than one basis point for the three months ended March 31, 2007 and two and one basis points, respectively, for the three and six months ended June 30, 2007. We had net recoveries of $155,000 for the quarter ended March 31, 2007 and net charge-offs of $698,000 for the quarter ended June 30, 2007. Similar to our response for the year ended December 31, 2006, our evaluation of the adequacy of our allowance for loan losses resulted in our determination that the total balances of the allowances were within our range of probable losses and no additional provision was warranted.
Current Market Conditions
In evaluating our allowance for loan losses for the quarters ended March 31, 2007 and June 30, 2007, we continued to look at market conditions, including real estate values. However, the changes in the economy and real estate market for the quarters ended March 31, 2007 and June 30, 2007 did not have a discernable negative impact on our trends of non-performing loans, loan-to-value ratios or loan loss experience for those periods and did not indicate that we would experience a significant change in those trends, as evidenced by the trend data included in our March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 11
In summary, there were no significant changes in our portfolio composition, loan-to-value ratios, non-performing loans and loan loss experience, or any discernible negative impact of current market observations on our portfolio for the quarters ended March 31, 2007 and June 30, 2007 compared to the year ended December 31, 2006. As such, we believe that revisions to the March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q are not necessary.
We will, in future quarterly and annual filings beginning with the September 30, 2007 Form 10-Q, enhance our disclosures to address the balances of interest-only and reduced documentation loans, and the impact, if any, of current market observations and trends in loan-to-value ratios of both total loans and non-performing loans on our determination of the allowance for loan losses.
Asset Quality, page 65
6. | Please revise your disclosure to discuss how the average original loan-to-value ratios upon origination have changed given recent trends in the real estate market. Disclose the loan-to-value ratios based on the most recent data obtained and discuss how changes in these trends impacted your determination of the provision for loan losses. Disclose specific steps you have taken to mitigate the additional credit risk apparently evidenced by these trends. Specifically disclose the extent to which you have obtained updated appraisal data for collateral supporting the portfolio and in particular your interest-only loans. |
Please see our response to comment number 5 which incorporates the response to this comment.
7. | Please tell us the amount of loans that were restructured that qualified as “troubled debt restructurings” under SFAS 15 during each of the last five years. |
There were no loans that were restructured that qualified as “troubled debt restructurings” under SFAS 15 during any of the five years ended December 31, 2006.
Consolidated Financial Statements
Summary of Significant Accounting Policies, page 86
General
8. | Please address the following regarding your recent sales of non-performing loans: |
· | Tell us how you accounted for the transfer of the non-performing loans upon your decision to sell them; |
· | Tell us the amount of the gain or loss on sale of non-performing loans and where it was recorded in your financial statements; |
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 12
· | Please tell us whether there were any portions of the allowance allocated to these loans on the date of transfer, and tell us the amount of any charge-offs against the allowance for any of these loans sold; and, |
· | Tell us the nature and extent of any recourse provisions associated with the sale of these loans, as well as the amount of any recourse reserves, payments, and repurchases associated with these loans. |
With regard to our sales of non-performing loans during 2006:
| - | Upon our decision to sell certain non-performing loans held in portfolio, we reclassified them to held-for-sale at the lower of cost or fair value, less estimated selling costs. Any reductions in the loans’ values were reflected as a write-down of the recorded investment in the loans resulting in a new cost basis, with a corresponding reduction in the allowance for loan losses. |
| - | We recognized a $43,000 loss on sales of non-performing loans which was included in net gain on sales of loans, which is a component of mortgage banking income, net, in our consolidated statement of income. |
| - | On the date of transfer, $1.8 million of the allowance for loan losses was allocated to these non-performing loans. Charge-offs against the allowance for loan losses for these non-performing loans totaled $967,000, substantially all of which related to one non-performing multi-family loan. |
| - | Our non-performing loans are sold without recourse. One non-performing loan sold in 2006 totaling $176,000 was repurchased in 2007, at the Company’s request to the purchaser, as a result of litigation between the Company and the borrower. We determined it would be more efficient to respond to the litigation if the Company owned the loan. |
For the quarter ended March 31, 2007, we recognized a $9,000 loss on sales of $2.3 million in non-performing loans. On the date of transfer, $233,000 of the allowance for loan losses was allocated to these non-performing loans. There were no charge-offs against the allowance for loan losses for these non-performing loans. These loans were sold without recourse and there have been no repurchases of such loans.
For the quarter ended June 30, 2007, we recognized no gain or loss on sales of $3.3 million in non-performing loans. On the date of transfer, $244,000 of the allowance for loan losses was allocated to these non-performing loans. Charge-offs against the allowance for loan losses for these non-performing loans totaled $315,000. These loans were sold without recourse and there have been no repurchases of such loans.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 13
Exhibit 31
9. | Please revise this exhibit to include the language related to internal controls exactly as stated in Item 601(b)(31)(4) of Regulation S-K. |
Exhibits 31.1 and 31.2 of the December 31, 2006 Form 10-K, will be revised to include the language related to internal controls exactly as stated in Item 601(b)(31)(4) of Regulation S-K. The introduction to Item 4. will be revised to read as follows:
| 4. | The registrant’s other certifying officer and I are responsible for establishing and maintaining disclosure controls and procedures (as defined in Exchange Act Rules 13a – 15(e) and 15d – 15(e)) and internal control over financial reporting (as defined in Exchange Act Rules 13a – 15(f) and 15d – 15(f)) for the registrant and have: |
The remaining language in Exhibits 31.1 and 31.2 was in agreement with the language stated in Item 601(b)(31)(4) of Regulation S-K.
The above revision will also be made to Exhibits 31.1 and 31.2 of the March 31, 2007 Form 10-Q and the June 30, 2007 Form 10-Q. Consistent with our discussion with Ms. Rebekah Moore on October 16, 2007, we will withhold the filing of our amended December 31, 2006 Form 10-K, March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q until the Staff’s review is complete.
Form 10-Q for the Period Ended March 31, 2007
10. | Please revise your disclosure to address the above comments as applicable, including but not limited to the comment related to your certification language. |
Based on our responses to comment numbers 1 through 9, other than the revisions to Exhibits 31.1 and 31.2, we believe that additional revisions to the March 31, 2007 Form 10-Q are not necessary.
Form 10-Q for the Period Ended June 30, 2007
11. | Please revise your disclosure to address the above comments as applicable, including but not limited to the comment related to your certification language. |
Based on our responses to comment numbers 1 through 9, other than the revisions to Exhibits 31.1 and 31.2, we believe that additional revisions to the June 30, 2007 Form 10-Q are not necessary.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 14
Notes to Consolidated Financial Statements
Note 7, Impact of Accounting Standards and Interpretations, page 11
12. | In your disclosure you indicate that you do not expect the adoption of SFAS 157 will have a material impact on your financial condition or results of operations. We note in your disclosures on page 23 that you hold non-GSE issuance REMICs and CMOs in your available for sale portfolio. Further, we note your disclosure in your 2006 Form 10-K on page 117 that the fair value of your available for sale and held to maturity securities are based on published or securities dealers’ estimated market values. Please revise your disclosures beginning in your next Form 10-Q to address the following bullets: |
· | Revise your future filings to disclose the potential impact of current market trends for these types of securities, if material, and tell us what you consider in making the determination; |
· | Please revise your future filings to clarify the extent to which you use actual published prices of the same securities, and the extent to which you use dealers’ estimated market values to determine the fair value of these securities; |
· | Please revise your future filings to disclose the models and assumptions used in the dealers’ estimated market values, and how the current market trends may affect the availability of quoted market prices for similar instruments. |
· | Please provide us with a copy of your proposed disclosures. |
Our investment portfolio is comprised primarily of fixed rate mortgage-backed securities guaranteed by a government sponsored enterprise (“GSE”), as issuer. GSE issuance mortgage-backed securities comprised 92% of our securities portfolio at June 30, 2007. Non-GSE issuance mortgage-backed securities comprised 6% of our securities portfolio at June 30, 2007 and had an amortized cost of $297.9 million, 15% of which are classified as available-for-sale and 85% of which are classified as held-to-maturity. Based on the disclosure documents for our non-GSE issuance securities, none were backed by pools consisting primarily of subprime mortgage loans. Our non-GSE issuance securities have either a AAA credit rating or an insurance wrap and they perform similarly to our GSE issuance securities. While the recent mortgage market conditions reflecting credit quality concerns might significantly impact lower grade securities, the impact on our non-GSE securities has not been significant. Our investment portfolio is primarily impacted by changes in interest rates, as we disclosed on page 18 in the “Securities Impairment” section of “Critical Accounting Policies” in Item 2. MD&A of the June 30, 2007 Form 10-Q.
The fair values for substantially all of our securities are obtained from an independent nationally recognized pricing service. We use third party brokers to obtain prices for a small portion of the portfolio that we are not able to price using our third party pricing service.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 15
Our independent pricing service provides us with prices which are most appropriately categorized as Level 2 inputs under SFAS No. 157, as quoted prices in active markets for identical assets are generally not available for the majority of securities in our portfolio. The modeling techniques they use to determine pricing for our mortgage-backed securities, which represent 98% of our securities portfolio, include an option-adjusted spread model and a volatility-driven, multi-dimensional single cash flow stream model, depending on the characteristics of a given security. The inputs to these models include benchmark yields, reported trades, broker/dealer quotes, issuer spreads, two-sided markets, benchmark securities, bids, offers and reference data. Additionally, new issue data and monthly payment information is used and market indicators, industry and economic events are monitored. As indicated above, based on the high quality of our investment portfolio, we do not believe that recent mortgage market conditions related to credit quality concerns will significantly impact the pricing of our portfolio or our ability to obtain reliable prices.
We will, upon adoption of SFAS 157 on January 1, 2008, include in the notes to our financial statements the required fair value disclosures relative to input levels, models and assumptions to comply with the standard’s requirements. These disclosures will be provided beginning with our March 31, 2008 Form 10-Q. However, we believe our current disclosure in Note 7. “Impact of New Accounting Standards and Interpretations” included in Item 1. “Financial Statements (Unaudited)” of our June 30, 2007 Form 10-Q, regarding the impact of our adoption of SFAS 157, is accurate and complete.
We will, in future filings within the “Securities Impairment” section of “Critical Accounting Policies” in Item 2. MD&A, beginning with the September 30, 2007 Form 10-Q, expand our discussion regarding how we obtain fair values for our securities portfolio as well as clarify the composition of our securities portfolio and the impact, if any, of current market trends on our securities portfolio. A draft of our proposed disclosures is attached as Appendix A.
Provision for Loan Losses, page 33
13. | Your disclosure on page 37 indicates that 39% of the non-performing 1-4 single family mortgage loans are comprised of reduced documentation loans and another 39% is comprised of interest-only loans. Given the emphasis placed on loan-to-value ratios and borrower credit scores upon origination and in your ongoing asset quality monitoring, please revise to discuss the impact of changes in the economy and the real estate market in your determination of the provision for loan losses for both the specific and general portions of the allowance. |
Please see our responses to comment numbers 5 and 11 which incorporate the response to this comment.
14. | Please revise to specifically disclose how you considered the increases in non-performing loans even after your sales of non-performing loans, as well as the increases in other |
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 16
problem loans when determining to allow your allowance to decrease through charge-offs with no additional provision expense. |
Please see our responses to comment numbers 5 and 11 which incorporate the response to this comment.
Asset Quality, page 36
15. | You disclose here on page 37 that within your portfolio you have loans that have certain attributes found in subprime lending, even though subprime lending is not a market you have actively pursued. Please revise to quantify the amount of loans as of December 31, 2006 and June 30, 2007 that have certain attributes in subprime lending. Revise to describe the attributes you are referring to here. |
The market does not apply a uniform definition of what constitutes “subprime” lending. Our reference to subprime lending relies upon the “Statement on Subprime Mortgage Lending” issued by the Office of Thrift Supervision and the other federal bank regulatory agencies (the “Agencies”) on June 29, 2007, published at 72 Fed. Reg. 37569 (July 10, 2007), which further references the “Expanded Guidance for Subprime Lending Programs” (the “Expanded Guidance”) issued by the Agencies by press release dated January 31, 2001.
In the first paragraph under “Applicability of Guidance – Exclusions” in the Expanded Guidance, the Agencies indicated that “subprime lending does not refer to individual subprime loans originated and managed, in the ordinary course of business, as exceptions to prime risk selection standards. The Agencies recognize that many prime loan portfolios will contain such accounts.” The Agencies also excluded prime loans that develop credit problems after acquisition and community development loans from the subprime arena.
According to the Expanded Guidance, subprime loans are other loans to borrowers which display one or more characteristics of reduced payment capacity. Five specific criteria are set forth including having a FICO score of “660 or below (depending on the product/collateral).” The Expanded Guidance also indicates that the criteria listed are not exhaustive and are not meant to define specific parameters for all subprime borrowers and may not match all markets or institutions’ specific subprime definitions.
Based upon the definition and exclusions described above, we are a prime lender. Our disclosure, on page 37 of the June 30, 2007 Form 10-Q, that states that within our portfolio are loans that have certain attributes of subprime lending refers to our having loans in our portfolio that, at the time of origination, had FICO scores of 660 or below. However, as we are a portfolio lender, we review all data contained in borrower credit reports and do not base our underwriting decisions solely on FICO scores. We believe the aforementioned loans, when made, were amply collateralized and otherwise conformed to our prime lending standards and do not present a greater risk of collectibility or other asset quality risk than loans to other borrowers with higher credit scores.
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 17
Credit scores are not available for all of the one-to-four family loans on our mortgage loan system. However, substantially all of our one-to-four family loans originated since March 2005 do have credit scores available on our mortgage loan system. One-to-four family loans, with credit scores available on our mortgage loan system, which had FICO scores of 660 and below totaled $741.9 million, or 10%, of one-to-four family mortgage loans with credit scores available on our mortgage loan system at June 30, 2007 and $765.8 million, or 11%, at December 31, 2006. At both dates, such loans had an average loan-to-value ratio, based on current principal balance and original appraised value, of 70%.
Since we do not consider the loans in our portfolio with FICO scores of 660 or below to be subprime loans, we believe that quantifying those loans in our disclosure would not provide meaningful and decision-useful information to the reader. Additionally, we do not believe the total amount of such loans is material or that it represents a significant risk concentration for disclosure purposes. As such, we believe that an amendment to the June 30, 2007 Form 10-Q is not necessary. We will, in future filings beginning with our September 30, 2007 Form 10-Q, include the more descriptive language noted above in paragraphs two, three and four of our response to this comment to more clearly disclose that we do not originate or hold subprime loans in our portfolio.
16. | Revise to quantify the amount of interest-only loans in your portfolio as of June 30, 2007 and December 31, 2006. Quantify the amount of low- or no-documentation loans as well. Quantify the amount of interest-only loans that are also low- or no-documentation loans. |
As previously discussed in our response to comment number 5, interest-only loans totaled $7.61 billion at June 30, 2007, of which $7.01 billion were one-to-four family loans and $602.3 million were multi-family and commercial real estate loans and represented 64% of total one-to-four family loans and 15% of total multi-family and commercial real estate loans, respectively. Interest-only loans totaled $6.72 billion at December 31, 2006, of which $6.17 billion were one-to-four family loans and $542.6 million were multi-family and commercial real estate loans and represented 60% of total one-to-four family loans and 13% of total multi-family and commercial real estate loans, respectively. Reduced documentation loans totaled $2.95 billion, or 27% of total one-to-four family loans, at June 30, 2007 and $2.90 billion, or 28% of total one-to-four family loans, at December 31, 2006. Interest-only loans that are also reduced documentation loans totaled $2.26 billion at June 30, 2007 and $2.15 billion at December 31, 2006, or 21% of total one-to-four family loans at both dates indicated. Reduced documentation loans are comprised primarily of SIFA loans.
In the last paragraph on page 45 of the June 30, 2007 Form 10-Q, we disclosed the balance of interest-only loans by property type as of June 30, 2007. Additionally, the changes in the balances of reduced documentation loans and interest-only reduced documentation loans from December 31, 2006 to June 30, 2007 are not material. As such, we believe the disclosures included in our June 30, 2007 Form 10-Q are sufficient and revisions are not necessary. We will, in future filings beginning with the September 30, 2007 Form 10-Q,
Mr. Kevin W. Vaughn
United States Securities and Exchange Commission
Page 18
disclose the dollar amounts of interest-only loans, reduced documentation loans and interest-only reduced documentation loans as of each balance sheet date presented.
| The Company acknowledges the following: |
| ▪ | the Company is responsible for the adequacy and accuracy of the disclosure in the filing; |
| ▪ | Staff comments or changes to disclosure in response to Staff comments do not foreclose the Commission from taking any action with respect to the filing; and |
| ▪ | the Company may not assert staff comments as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States. |
We believe that our letter is responsive to your comments and that other than the revisions to Exhibits 31.1 and 31.2 for the December 31, 2006 Form 10-K, March 31, 2007 Form 10-Q and June 30, 2007 Form 10-Q, no additional revisions to the aforementioned reports are necessary.
Please contact the undersigned at (516) 327-7892 or (516) 327-7820, respectively, if you have any questions.
Sincerely,
| /s/ Monte N. Redman | /s/ Frank E. Fusco | |
| Monte N. Redman | Frank E. Fusco | |
| President and Chief Operating Officer | Executive Vice President, Treasurer | |
| Astoria Financial Corporation | and Chief Financial Officer | |
| Astoria Financial Corporation |
Appendix A
(Additional disclosures appear in italics.)
Critical Accounting Policies
Securities Impairment
Our available-for-sale securities portfolio is carried at estimated fair value, with any unrealized gains and losses, net of taxes, reported as accumulated other comprehensive income/loss in stockholders’ equity. Debt securities which we have the positive intent and ability to hold to maturity are classified as held-to-maturity and are carried at amortized cost. The fair values for substantially all of our securities are obtained from an independent nationally recognized pricing service. We use third party brokers to obtain prices for a small portion of the portfolio that we are not able to price using our third party pricing service.
Our investment portfolio is comprised primarily of fixed rate mortgage-backed securities guaranteed by a government sponsored enterprise, or GSE, as issuer. GSE issuance mortgage-backed securities comprised 9x% of our securities portfolio at September 30, 2007. Non-GSE issuance mortgage-backed securities at September 30, 2007 comprised x% of our securities portfolio and had an amortized cost of $xxx.x million, 1x% of which are classified as available-for-sale and 8x% of which are classified as held-to-maturity. Based on the disclosure documents for our non-GSE issuance securities, none were backed by pools consisting primarily of subprime mortgage loans. Our non-GSE issuance securities have either a AAA credit rating or an insurance wrap and they perform similarly to our GSE issuance securities. While the recent mortgage market conditions reflecting credit quality concerns might significantly impact lower grade securities, the impact on our non-GSE securities has not been significant. Based on the high quality of our investment portfolio, we do not believe that current market conditions will significantly impact the pricing of our portfolio or our ability to obtain reliable prices.
Our investment portfolio is primarily impacted by changes in interest rates. In general, as interest rates rise, the fair value of fixed rate securities will decrease; as interest rates fall, the fair value of fixed rate securities will increase. We conduct a periodic review and evaluation of the securities portfolio to determine if the decline in the fair value of any security below its cost basis is other-than-temporary. We generally view changes in fair value caused by changes in interest rates as temporary, which is consistent with our experience. If we deem such decline to be other-than-temporary, the security is written down to a new cost basis and the resulting loss is charged to earnings as a component of non-interest income. At September 30, 2007, we had xxx securities with an estimated fair value totaling $x.xx billion which had an unrealized loss totaling $xxx.x million, substantially all of which have been in a continuous unrealized loss position for more than twelve months. Substantially all of these securities are guaranteed by a GSE as issuer. At September 30, 2007, the impairments are deemed temporary based on the direct relationship of the decline in fair value to movements in interest rates, the estimated remaining life and high credit quality of the investments and our ability and intent to hold these investments until there is a full recovery of the unrealized loss, which may be until maturity. There were no other-than-temporary impairment write-downs during the nine months ended September 30, 2007.