November 18, 2011
Ms. Stephanie L. Hunsaker
Senior Assistant Chief Accountant
United States Securities and Exchange Commission
Division of Corporation Finance
100 F Street, N.E.
Washington, D.C. 20549
Re: File No. 1-9861
Dear Ms. Hunsaker:
This letter serves as the response of M&T Bank Corporation (“M&T” or “the Company”) to your letter dated November 3, 2011. Your comments and our responses to those comments follow.
Form 10-K for Fiscal Year Ended December 31, 2010
Item 7. Management’s Discussion and Analysis
Selected Residential Real Estate-Related Loan Data, page 62
| 1. | Refer to your response to prior comment four. We note that home equity lines and loans represent 11.7% and 12.6% of your loan portfolio as of June 30, 2011 and December 31, 2010, respectively and 63% and 64% of those loans are junior liens, respectively. Please revise your disclosure in future filings to address the following: |
| • | State that only approximately 13% of your junior lien home equity portfolio was behind a first lien mortgage loan held or serviced by you, and if true, state that for the remainder of junior lien home equity lines and loans that are behind a first lien held or serviced by another financial institution, you are not able to and do not track whether the first lien is in default; |
| • | State how that lack of information on the first lien performance is taken into account in developing your allowance for loan losses on these junior liens, and state what leading indicators you track in order to assess the credit quality of your junior liens (e.g., FICO scores, credit bureau statistics); |
November 18, 2011
Page 2
| • | State that home equity line of credit terms generally have an open draw period of ten years followed by an amortization period of up to twenty years, and the weighted-average remaining draw periods for home equity lines of credit were approximately six years at December 31, 2010, and approximately 98% of all outstanding balances of home equity lines of credit related to lines that were still in the draw period; and |
| • | State that you estimate approximately one-fifth of outstanding balances are associated with home equity line of credit borrowers that while in the draw period are making contractually allowed payments that do not include any repayment of principal. |
Response: We will revise our future Form 10-K filings to include the requested information. An example of such disclosure follows:
At December 31, 2011 approximately X% of M&T’s home equity portfolio consisted of junior lien loans that were behind a first lien mortgage loan that was not owned or serviced by M&T. For the junior lien loans where an entity other than the Company held a first lien mortgage, the Company cannot precisely determine whether there is a delinquency on such first lien mortgage. In monitoring the credit quality of its home equity portfolio for purposes of determining the allowance for credit losses the Company reviews delinquency and nonaccrual information and considers recent charge-off experience as presented in Tables X and X. Additionally, the Company generally evaluates home equity loans and lines of credit that are more than 150 days past due for collectability on a loan-by-loan basis and the excess of the loan balance over the net realizable value of the property collateralizing the loan is charged-off at that time. Home equity line of credit terms vary but such lines are generally originated with an open draw period of ten years followed by an amortization period of up to twenty years. At December 31, 2011, approximately X% of all outstanding balances of home equity lines of credit related to lines that were still in the draw period, the weighted-average remaining draw periods were approximately X years, and approximately X% were making contractually allowed payments that do not include any repayment of principal.
Other Income, page 68
| 2. | We note your response to prior comment six. Please revise your disclosure in future filings to address the following related to your BLG impairment analysis: |
| • | The period in which you estimate BLG will return to profitability; |
| • | Clarification of how the discount rates of between 8-17% were determined as of June 30, 2011 and disclose the weighted average rate used; and |
| • | A description of reasonably likely events and/or changes in circumstances that could be expected to negatively affect the key assumptions (e.g., reasonably likely scenarios of BLG not returning to profitability in the period estimated and the related impact on your results of operations). |
November 18, 2011
Page 3
Response: The operating losses of BLG are a result of disruptions in the residential and commercial real estate markets and reflected provisions for losses associated with securitized loans and other loans held by BLG and other administrative costs. The provision for loan losses represented the significant majority of BLG’s net loss for the year. Such provision primarily pertained to loans in non-recourse securitization trusts that BLG consolidates in its financial statements. Under generally accepted accounting principles (“GAAP”), such losses are required to be recognized by BLG despite the fact that many of the securitized loan losses will ultimately be borne by the underlying third party bond-holders. As these loan losses are realized through later foreclosure and still later sale of real estate collateral, the underlying bonds will be charged-down resulting in BLG’s future recognition of debt extinguishment gains. The timing of such debt extinguishment is largely dependent on the timing of loan workouts and collateral liquidations and given ongoing loan loss provisioning it is difficult to project when BLG will return to profitability. As we have noted in our disclosure, we have projected no further commercial mortgage origination and securitization activities by BLG in assessing M&T’s investment for other-than-temporary impairment (“OTTI”). We also note in our disclosure that as a result of past securitization activities, BLG is still entitled to cash flows from mortgage assets that it owns or that are owned by its affiliates and to participate in the value generation activities of its affiliates that provide asset management and other services through its contractual right to receive a portion of the distributions made by its affiliates if and when such distributions are made. Accordingly, the Company believes that BLG will realize value that could be available for distribution to its owners, including M&T, despite a lack of positive GAAP-earnings from its core mortgage origination and securitization activities. To this point, BLG’s affiliates have reinvested their earnings to generate additional servicing and asset management activities further contributing to the value of those affiliates that inures to the benefit of BLG and, ultimately, M&T. As such, we do not believe that the timing of when BLG will return to a GAAP net income position is a key factor in our OTTI analysis.
M&T discounts the various projections used in its OTTI analysis using discount rates that range from 8% to 17%. To determine these discount rates M&T considered the sources of potential future distributions. One such source is certain subordinated mortgage assets related to past securitization activities of BLG and its affiliates. The realization of such assets is highly dependent on the eventual default and loss severity on the underlying mortgage loan collateral. M&T discounted projected cash flows from these mortgage assets at a high rate of 17%, which was indicative of the inherent risk of those assets. The other potential sources consist of established business activities of BLG’s affiliates in servicing mortgage loans and other assets as well as performing asset management services. Through our experience and by reference to another market transaction, we estimate that an investor in this type of long-term investment would require an 8% return and, accordingly, we believe this is reasonably indicative of an
appropriate discount rate. The weighted-average discount rate used in our analyses was 8.3%.
November 18, 2011
Page 4
M&T will expand its future disclosures to include this information about discount rates used in its OTTI projections.
M&T’s analysis at June 30, 2011 estimated that the value attributed to BLG and its relationship with its affiliates that inures to the benefit of M&T exceeded the carrying value of M&T’s investment in BLG by 6%. M&T’s OTTI analysis was dependent on the current and forecasted asset management, loan servicing and other fee generating activities of BLG’s affiliates and the realizability of mortgage assets held by BLG and affiliates. Although management believes its estimates are reasonable, should sufficient value not be generated that contractually inures to the benefit of M&T, an OTTI charge may be required in a future period for some portion of the remaining book value of M&T’s investment in BLG. Reasonably possible events that could be expected to negatively affect the key assumptions in the OTTI analysis include BLG’s affiliates not generating as much value from their asset management, loan servicing and other fee generating activities as forecasted by M&T as a result of slower growth or lower profit margins. Should such actual results fall short of M&T’s projections, the value attributed to BLG that inures to the benefit of M&T could become less than the carrying value of M&T’s investment in BLG and M&T could be required to recognize an OTTI, unless other value generating activities are initiated.
With respect to specific disclosures about BLG and its affiliates, M&T is cognizant of the fact that such entities are closely-held businesses that do not otherwise disclose financial information publicly. Nevertheless, we will expand our disclosures in future filings to provide the additional information as noted above with respect to our OTTI assessment.
Capital, page 84
Privately Issued Mortgage-Backed Securities Classified as Available for Sale, page 85
| 3. | Refer to your response to prior comment seven. We note your disclosure on page 89 of your June 30, 2011 Form 10-Q that indicates bonds in which you have recognized an other-than-temporary impairment charge had a projected weighted-average default percentage of 24% and weighted-average loss severity of 52%, and for bonds without an other-than-temporary impairment charge, the projected weighted-average default percentage was 11% and the weighted-average loss severity was 42%. Thus, it appears that if the projected default percentage and loss severity increase in future periods, it would be reasonably possible for you to record an impairment charge on the privately issued mortgage-backed securities. We also note that the remaining credit enhancement and current payment status of the mortgage-backed securities backed by Prime and Alt-A Hybrid ARMs continued to deteriorate from December 31, 2010 to June 30, 2011. Please provide further revised disclosure in future |
November 18, 2011
Page 5
| filings to clarify the relationship between the weighted-average default percentage and weighted-average loss severity on page 89 and the credit enhancement and current payment status disclosed on page 88. For example, clarify if true, that since your disclosures of the remaining credit enhancements and current payment status are based on unpaid principal balance, that the amounts do not reflect “expected” future credit losses in the disclosed percentages of remaining credit enhancements and current payment status, and that this would be in contrast to your disclosure of the projected default percentage and loss severity, which is based on future expectations. In your disclosure, consider providing a sensitivity analysis in which you disclose a reasonably likely change in the remaining credit enhancement and in the current payment status that would lead to additional impairment charges of a specified amount. |
Response: We believe that the footnotes to the table on page 88 clearly indicate how the credit enhancement and current payment status percentages are calculated and that the information is as of June 30, 2011. Nevertheless, we will clarify in future filings that the assumptions discussed on page 89 are forward-looking so as to measure OTTI resulting from expected performance, while the disclosures in the table on page 88 are point in time measurements as of each reporting date. We will also provide a sensitivity analysis with respect to the potential additional impairment charges that could result from a change in the estimated mortgage loan default and severity assumptions disclosed. An example of such disclosure would be that “a X% increase in the estimated default rate assumption and a X% increase in the severity rate assumption would have increased the OTTI charge by $X million for the year ended December 31, 2011.”
Notes to Financial Statements
5. Allowance for credit losses, page 121
| 4. | Refer to your response to prior comment 13 where you state that you do not utilize weighted-average LTV ratios at a portfolio level in evaluating losses on commercial real estate loans. Please tell us whether you utilize current LTV ratios at a loan level and if you are able to aggregate such information related to a subset of this loan class (e.g., larger balance loans that fall into each of the following buckets: 100% +, 80% – 100%, and below 80%). If you do not have this information for your entire commercial real estate loan portfolio, but you have some subset of this information, please disclose that LTV information and disclose that for the remaining loans in that category the current LTV is not available. Additionally, refer to your response where you state that you do not utilize weighted-average LTV ratios in estimating your allowance for credit losses associated with residential real estate and home equity loans and lines of credit, but you do consider loss severity associated with recent charge-offs which is driven by collateral |
November 18, 2011
Page 6
| values in the market place. Please tell us what metric related to collateral values that you do use in estimating your allowance for residential real estate and home equity loans and lines of credit, and revise your disclosure in future filings to include that collateral value metric for each loan class. If you do not utilize any metrics or credit quality indicators related to collateral values for these loan classes, please revise your disclosure in future filings to state that fact. |
Response: As indicated, we do not utilize weighted-average LTV ratios at a portfolio level in evaluating losses on commercial real estate loans. Instead, several factors including payment status, industry, geographic location, business performance and estimated collateral values on commercial real estate loans are utilized in assigning loan grades on an individual loan basis. For nonaccrual loans, specific allowances are oftentimes computed using the collateral method which compares the book value of the loan to the net realizable value of the collateral. Current LTV ratios on commercial real estate loans are not systemically calculated or tracked across the portfolio.
In estimating the allowance for credit losses on loans secured by residential real estate, including home equity loans and lines of credit, M&T considers recent charge-off history and trends, which are driven by current collateral values in the market place as well as the amount of loan defaults. Loss rates on such loans are determined by reference to recent charge-off history and are evaluated (and adjusted if deemed appropriate) through consideration of other factors including near-term forecasted loss estimates developed by M&T’s Credit Department. In arriving at such forecasts, M&T considers the current estimated fair value of its collateral based on geographical adjustments for home price depreciation/appreciation. Again, weighted-average LTV’s are not utilized in the determination of the allowance, but collateral values and their impact on recent weighted-average charge-off rates do contribute to the analysis.
We believe that we have disclosed the relevant metrics that we consider in establishing the allowance for credit losses for loans secured by real estate, but we will further clarify our procedures in future filings by including the information noted above.
20. Fair value measurements
Assets taken in foreclosure of defaulted loans, page 157
| 5. | Refer to your response to prior comment 17. In addition to the significant writedowns required on foreclosed assets for the years ended December 31, 2010 and 2009, we note your write-downs of $15 million (or 30% of carrying value held as of period end) recorded for the six months ended June 30, 2011 as disclosed on page 43 of your Form 10-Q. Please tell us |
November 18, 2011
Page 7
| whether, and if so how, you consider similar price declines in your allowance for loan loss methodology for loans that are specifically identified as impaired and those that are not specifically identified as impaired, and if so, please revise your disclosure in future filings to address how you consider such price declines. Specifically, address how you consider price declines concentrated in certain regions and/or industries in your allowance methodology such as the write-downs of $19 million on two residential real estate development projects in the Mid-Atlantic region recognized during the second quarter of 2009 as noted in your response. |
Response: After reviewing our disclosure in response to your comment on the declines in fair value of foreclosed assets at June 30, 2011, we discovered that due to a compilation error the $15 million decline on $50 million of such assets was not accurate. The correct disclosure should be a $12 million decline on $31 million of foreclosed assets for the six months ended June 30, 2011. We will correct this disclosure in future filings. We note that this disclosure excludes approximately $140 million of foreclosed assets at June 30, 2011 for which no decreases in fair value were recognized during that same period. Fair value write-downs as a percentage of total foreclosed asset balances at June 30, 2011, therefore, were only 7% of aggregate carrying value and that percentage excludes the impact of any circumstances where fair value may be in excess of carrying value. Because the write-downs that you reference cite only fair value declines on a subset of the real estate holdings they are not necessarily representative of the performance related to all such holdings. For example, M&T recognized $18 million of gains on sales of foreclosed assets during the six months ended June 30, 2011. When coupled with the $12 million of decline in appraised value noted above, it appears that our process of valuing ORE properties is reasonable in all material respects. The net $6 million gain represents 0.4% of the Company’s annualized pre-tax income. Furthermore, included in the $12 million decline in fair value for the six months ended June 30, 2011 were $11 million of decreases on 20 residential builder and developer properties, with the remaining decline coming mainly from residential foreclosed properties. Similar to our response to your prior comment 17, these write-downs were reflective of overall declines in the housing market, the distressed nature of many of the properties, in particular residential real estate development projects that had been foreclosed upon, and the length of time the properties had remained in our inventory. Over the last several years, our disclosures have repeatedly described the downturn in the U.S. economy, significant deterioration in the residential real estate market, declining real estate valuations particularly related to residential real estate and residential real estate development, and our heightened monitoring and collection efforts related to such loans and foreclosed assets.
Similar to assets taken in foreclosure of defaulted loans, as updated appraisals are obtained on collateral supporting defaulted loans, specific reserves on such loans are adjusted accordingly. For loans that are not specifically identified as being impaired, appraisals or other indications of value are utilized in assigning loan grades, which are then utilized in our allowance for credit loss methodology. As disclosed on page 64 of Form 10-K, in those cases where current
November 18, 2011
Page 8
appraisals may not yet be available, prior appraisals are utilized with adjustments, as deemed necessary, for estimates of subsequent declines in value as determined by line of business and/or loan workout personnel in the respective geographic regions. Those adjustments are reviewed for reasonableness by the Company’s loan review department. Estimated values, therefore, give consideration to the geographic region in which the collateral is located. We disclosed on page 75 of M&T’s Form 10-Q at June 30, 2011 that in response to the credit issues experienced by the Company with respect to residential real estate, including loans to builders and developers of residential real estate, the Company has expanded its normal loan review process to conduct detailed reviews of all loans to residential real estate builders and developers that exceeded $2.5 million. Those credit reviews often resulted in commencement of intensified collection efforts, including instances of foreclosure. Such reviews also included an analysis of the value of the supporting collateral. Considering these disclosures, as well as our overall disclosures, we believe that we have addressed how we consider declines in collateral value in our determination of the allowance for credit losses.
Form 10-Q for the Quarterly Period Ended March 31, 2011
Item 2. Management’s Discussion and Analysis
Selected Residential Real Estate Related Loan Data, page 63
| 6. | Refer to your response to prior comment 22. Please revise your disclosure in future filings to explain the reason that your Alt-A junior lien portfolio has lower nonaccrual statistics as compared to Alt-A first mortgages and your junior lien home equity loans and lines have lower nonaccrual statistics than residential first mortgage loans is due in part to your charge-off policy. |
Response: To the extent applicable, we will revise our future filings to disclose if M&T’s charge-off policy with respect to loans secured by residential real estate is impacting the nonaccrual statistics of our second lien portfolios relative to our first lien portfolios.
Form 8-K filed September 12, 2011
Exhibit 99
Key Ratios, page 10
| 7. | We note your presentation of pre-tax, pre-provision earnings is not labeled as non-GAAP. However, you did provide a non-GAAP reconciliation of pre-tax, pre-provision income per share on page 27. It would appear that pre-tax, pre-provision earnings is also a non-GAAP |
November 18, 2011
Page 9
| measure as it represents a measure of profitability, but it excludes one of the most significant costs of the bank, the provision for loan losses. Please revise future filings to label this measure as non-GAAP and provide a reconciliation of the differences between this measure and the most directly comparable GAAP measure. |
Response: We will revise future filings to label the pre-tax, pre-provision earnings as a non-GAAP presentation and also provide a reconciliation of such to GAAP-basis income before taxes.
Form 8-K filed October 19, 2011
Exhibit 99
Provision for Credit Losses/Asset Quality, page 4
| 8. | We note your disclosure that effective September 30, 2011, you began to separately report other acquired impaired loans, which are defined as loans that ceased performing in accordance with their contractual terms and became impaired subsequent to their acquisition date. We also note your disclosure that the increase of $77 million in such loans from June 30, 2011 was due to loans obtained in the acquisition of Wilmington Trust. Please tell us how you determined that such loans were impaired as of September 30, 2011 and not impaired as of the May 2011 acquisition date and summarize the objective evidence that you utilized to make this determination, including the payment status of the related loans as of each date. Additionally, please clarify whether your disclosure that states that these loans are included in accounting pools that continue to accrue interest means they are in the population of purchased impaired loans accounted for in accordance with ASC 310-30. If so, please clarify whether you would only classify loans that were previously classified as part of purchased impaired loans (as disclosed on page 23 of your June 30, 2011 Form 10-Q) as “other acquired impaired loans” if they became impaired subsequent to their acquisition, or whether you would also use the same classification for acquired loans that were not originally accounted for and classified as purchased impaired since the acquisition date. |
Response: In accordance with our accounting policy, individual acquired loans were identified as purchased impaired if, as of the acquisition date, the loan was 90 days or more past due or full collection of such individually identified loan was not expected. Our disclosure of other acquired impaired loans beginning in the third quarter of 2011 is an attempt to provide transparency on acquired loans that were not individually identified as impaired as of the acquisition date because generally such loans were not delinquent as of that date but subsequently became 90 days or more past due or otherwise ceased performing in accordance
November 18, 2011
Page 10
with their contractual terms. Because such loans were not considered impaired at the acquisition date, they were not in the population of purchased impaired loans accounted for in accordance with ASC 310-30. The other acquired impaired loan category is merely meant to provide information about the performance of individual loans within the population of non-purchased impaired loans. Although such disclosure is not required, we believe that this information may be useful to users of our financial statements. Loans classified as purchased impaired remain in that category and cannot be subsequently disclosed as other acquired impaired loans.
We will clarify this information in future filings.
Non-Interest Income and Expense, page 6
| 9. | We note your discussion of the efficiency ratio on page 7, which is calculated as noninterest operating expenses divided by the sum of taxable-equivalent net interest income and noninterest income (exclusive of gains and losses from bank investment securities and merger-related gains). We also note that you quantify certain of the items that are excluded from your efficiency measure when reconciling some of your other non-GAAP measures, but these amounts are different since you quantify those reconciling items on an after-tax basis but you state on page 11 that when calculating your efficiency ratio, you use pre-tax amounts. Thus, given that your efficiency ratio appears to be a non-GAAP performance measure, please label the measure as non-GAAP and show the calculation of this metric, with quantification of each the reconciling items. |
Response: The numerator in the calculation of the efficiency ratio is “Noninterest Operating Expense” which we have consistently disclosed in previous filings and is disclosed in the Form 8-K on page 18. We will revise future filings to expand our disclosures to present the calculation of the efficiency ratio. The following provides an example of such information that will be included with the reconciliation of quarterly GAAP to non-GAAP measures.
November 18, 2011
Page 11
| Three months ended September 30, |
Nine months ended September 30, |
|||||||||||||||
| Efficiency Ratio |
2011 | 2010 | 2011 | 2010 | ||||||||||||
| (dollars in thousands) | ||||||||||||||||
| Noninterest operating expense |
$ | 618,615 | 466,607 | $ | 1,626,736 | 1,400,729 | ||||||||||
|
|
|
|
|
|
|
|
|
|||||||||
| Taxable-equivalent net interest income |
$ | 623,265 | 575,733 | $ | 1,791,066 | 1,711,322 | ||||||||||
| Other income |
368,382 | 289,899 | 1,184,458 | 821,162 | ||||||||||||
| Less: |
||||||||||||||||
| Gain on bank investment securities |
89 | 1,440 | 150,186 | 1,909 | ||||||||||||
| Net OTTI losses recognized in earnings |
(9,642 | ) | (9,532 | ) | (52,213 | ) | (58,714 | ) | ||||||||
| Gain on acquisition |
— | — | 64,930 | — | ||||||||||||
|
|
|
|
|
|
|
|
|
|||||||||
| Denominator for efficiency ratio |
$ | 1,001,200 | 873,724 | $ | 2,812,621 | 2,589,289 | ||||||||||
|
|
|
|
|
|
|
|
|
|||||||||
| Efficiency ratio |
61.79 | % | 53.40 | % | 57.84 | % | 54.10 | % | ||||||||
|
|
|
|
|
|
|
|
|
|||||||||
* * * *
Should you require additional clarification on any of our responses to your comments please contact me at 716-842-5844.
| Very truly yours, |
| /s/ René F. Jones |
| René F. Jones |
| Executive Vice President and Chief Financial Officer |
| cc: | Michael R. Spychala |