VIA EDGAR
January 20, 2012
Division of Corporation Finance
Securities and Exchange Commission
100 F Street N.E.
Washington, D.C. 20549
Attention: John P. Nolan
Re: | S&T Bancorp, Inc. |
Form 10-K
For Fiscal Year Ended December 31, 2010
Filed March 16, 2011
Form 10-Q for Fiscal Quarter Ended September 30, 2011
Filed November 4, 2011
File No. 000-12508
Dear Mr. Nolan:
We are in receipt of, and have reviewed, your supplemental comment letter addressed to Mr. Todd D. Brice, President and Chief Executive Officer of S&T Bancorp, Inc. (S&T or the Company or we), dated January 12, 2012 (the Comment Letter) in connection with the Companys November 17, 2011 response to your correspondence dated November 3, 2011. We are hereby providing to the Commission staff (the Staff) the Companys responses to the Comment Letter. For reference purposes, the text of the Comment Letter has been reproduced herein with responses below each numbered comment.
Form 10-K for the Year Ending December 31, 2010
MD&A, page 25
Allowance for Loan Losses, page 46
1. | We have reviewed your response to comment five of our letter dated November 3, 2011 regarding changes and adjustments to your allowance for loan loss methodology. Please explain and revise future filings to clearly indicate how the absolute high method and average method are computed, what these methods represent and how they are considered and numerically weighted into the allowance for loan loss calculations. Please also include more granular and transparent |
disclosure regarding how and why you expanded the rating scale for qualitative adjustments, including the specific qualitative adjustments you included, any changes to previously included adjustments and why you believe you needed to make these changes. Further, please revise future filings to fully disclose any changes to your allowance for loan loss methodology, why they were made, and how they impacted the financial statements. |
Response 1:
Our allowance for loan losses (ALL) includes a component for historical loss considerations. The historical loss rate is the base factor we adjust with qualitative factors to arrive at portfolio segment level general loss allocation percentages. Prior to the second quarter of 2009, the historical loss rate was determined using the absolute high method, which was the highest annualized loss rate experienced over the prior seven years by loan category. This method had been in use for many years at S&T. It was established over a period during the late 1990s and 2000s characterized by general economic strength and low loan losses. For example, in commercial real estate, the largest portfolio segment, the historical loss rate utilized in the fourth quarter of 2008 was .34%, which was the highest annualized loss rate experienced over the period 2002 2008, with a range from .34% to -.04% of loans. As the impact of the recent financial crisis manifested itself in our loan portfolio, we experienced $34.2 million of net charge offs in the first half of 2009, including losses on one relationship that exceeded $26.5 million. The losses in the first half of 2009 were more than had been experienced in the prior 6 years combined. This method, by annualizing unusually high losses that were skewed by one relationship, when combined with qualitative adjustments, resulted in an unreasonably high required level of ALL that was in excess of $150 million. Management concluded that the calculated level of ALL, especially given the impact of one relationship, was not reflective of losses inherent in the total portfolio. This conclusion is supported by the actual level of losses incurred during 2010 and 2011 which were far less than $150 million. Management elected to enhance the calculation of the base historical loss rate, changing from the absolute high method to the average method. The average method utilized, starting with the second quarter of 2009, computed the base historical loss rate using the annualized average loss rates over the prior five years by loan category. This simple average was considered to be more reflective of our stable underwriting standards while giving consideration to outlying credit events that we experienced in 2009. Our backtesting indicated results comparable to the pre-2009 period.
Management further enhanced the average method starting in the fourth quarter of 2010 to better align the ALL model with regulatory guidance. The calculation of the base historical loss utilizing the average method was shortened to include a 1-2 year loss emergence period depending on loan category over a 4 quarter average. Loss emergence refers to the length of time between a loss event and the charge-off. With the volatility in the current credit cycle, the shorter time horizon is more responsive to the loss emergence periods we are currently experiencing in our portfolio. This change was also backtested indicating comparable results. Management believes this shorter
time horizon provides a better indication of inherent losses in the loan portfolio given the recent economic downturn.
In the first quarter of 2009, management made two enhancements to the qualitative factors in the ALL model. These enhancements occurred prior to higher actual losses that had a significant impact on the base historical loss rate as detailed above. Management became aware of deteriorating credit quality in the first quarter of 2009 and did not believe that the then structure of the qualitative adjustments existing in our methodology would be sufficient to increase the ALL enough in the future to reflect the risk inherent in the loan portfolio. The first enhancement added a qualitative factor for the evaluation of out-of-state loans, which at the time was becoming a more significant source of credit problems for the Company. This factor was added to the commercial real estate loan category to allow management to increase the ALL for exposure to the higher risk inherent in these loans. The second enhancement was the expansion of the rating scale for qualitative adjustments. Prior to this enhancement, the various qualitative factors for each loan category were assigned a rating between 1 (low) and high (3) and then averaged. This average value was then multiplied by the base historical loss rate to arrive at the reserve required for each loan category. The enhancement expanded the rating scale to between 1 (low) and 4 (high), thus allowing management to increase the impact of the qualitative adjustments during a time of deteriorating credit quality. In the first quarter of 2009, the combined impact of the two enhancements was to increase the allowance in the first quarter of 2009 by approximately $2 million, relative to what it would have been without the enhancements
Management further refined the qualitative factors starting in the fourth quarter of 2010 to better align the ALL model with regulatory guidance. The improved base historical loss rate calculation described above as implemented did not require the multiplier effect provided by the 1 to 4 scale described above in order to have an ALL that reflected inherent loan losses. The qualitative factors were reorganized consistent with regulatory guidance and became basis point adjustments to, instead of a multiplier on, the historical base loss. The change from a multiplier to a basis point adjustment mechanism provides a greater level of precision to management in applying their qualitative decisions. The combination of the enhancements to the average method and the qualitative factors in the fourth quarter of 2010 did not materially change the ALL at December 31, 2010, resulting in an increase in the allowance of less than $0.5 million.
Since December 31, 2010 there have been no further enhancements to the ALL methodology. We will disclose in future filings any enhancements along with the reasons why they were made and the impact on the financial statements.
The following will be included in our 2011 Form 10-K within the Critical Accounting Policies and Judgments section of Managements Discussion and Analysis of Financial Condition and Results of Operations:
Allowance for Loan Losses
Our loan portfolio is our largest category of assets on our Consolidated Balance Sheet. Accordingly, we have designed a systematic ALL methodology used to determine our provision for loan losses and ALL on a quarterly basis. The ALL represents managements estimate of probable losses inherent in the loan portfolio at the balance sheet date and is presented as a reserve against loans in the Consolidated Balance Sheets. The ALL is increased by a provision charged to expense and reduced by charge-offs, net of recoveries. Determination of an adequate ALL is inherently subjective, as it requires estimations of the occurrence of future events, as well as the timing of such events. The ALL may be subject to significant changes from period to period.
The ALL methodology includes two main components: the evaluation of individually impaired loans and the evaluation of groups of homogeneous loans with similar risk characteristics.
S&T individually evaluates for impairment all substandard and nonaccrual commercial loans greater than $0.5 million. A loan is impaired when, based on current information and events, it is probable that a creditor will be unable to collect all amounts due according to the contractual terms of the loan agreement. For all TDRs, regardless of size, as well as all other impaired loans, we conduct further analysis to determine the probable loss and assign a specific reserve to the loan if deemed appropriate. Specific reserves are established based upon the following three impairment methods: 1) the present value of expected future cash flows discounted at the loans effective interest rate, 2) the loans observable market price or 3) the estimated fair value of the collateral if the loan is collateral dependent. These analyses involve a high degree of judgment in estimating the amount of loss associated with specific impaired loans, including estimating the amount and timing of future cash flows, current estimated fair value of the loan and collateral values. S&T primarily uses the estimated fair value of the collateral method as required for collateral dependent loans.
The ALL methodology for groups of homogeneous loans, known as the general reserve, is comprised of both a quantitative and qualitative analysis. We first apply historical loss rates to pools of loans with similar risk characteristics. The historical loss rates are calculated by historical charge-offs that have occurred within each pool of loans over the loss emergence period. Management has estimated the loss emergence period to be two years for commercial real estate loans and one year for all other commercial loan portfolio segments. Historical loss rates for the consumer loan portfolio segments are determined using a one year loss emergence period. After consideration of the loss calculations, management applies additional qualitative adjustments so that the ALL is reflective of the inherent losses that exist in the loan portfolio at the balance sheet date. Qualitative adjustments are made based upon changes in economic conditions, loan portfolio and asset quality data and credit process changes, such as credit policies or underwriting standards. The evaluation of the various components of the ALL requires considerable judgment in order to estimate inherent loss exposures.
At December 31, 2011, approximately 92 percent of the ALL related to the commercial loan portfolio. Commercial loans comprise approximately 73 percent of our loan portfolio. Commercial loans have been more impacted by the economic slowdown in our markets. The ability of customers to repay commercial loans is more dependent upon the success of their business, continuing income and general economic conditions. Accordingly, the risk of loss is higher on such loans compared to consumer loans, which have incurred lower losses in our market.
There are many factors affecting the ALL; some are quantitative, while others require qualitative judgment. Although management believes its process for determining the ALL adequately considers all of the factors that could potentially result in credit losses, the process includes subjective elements and may be susceptible to significant change. To the extent actual outcomes differ from management estimates, additional provisions for loan losses could be required and could adversely affect S&Ts earnings or financial position in future periods.
Form 10-Q for the Nine Month Period Ending September 30, 2011
Financial Statements
Note 7 Allowance for Loan Losses, page 20
2. | In this section, we note the significant increase in commercial real estate impaired loans that did not require a related allowance, in the three month period ending September 30, 2011. We further note your disclosure on page 32 stating that in the three months ending September 30, 2011, no new significant impaired loans requiring a specific reserve were identified. Please tell us and expand upon this disclosure in future filings to discuss the nature of the commercial real estate impaired loans. Please also discuss the underlying collateral supporting the loan(s), the amount of any charge-offs recognized for the loan(s), as applicable, the date and amount of the last appraisal obtained for the underlying collateral, and the estimated fair value of the underlying collateral as of the period presented. |
Prior to September 30, 2011, accruing troubled debt restructurings (TDRs) were included in the TDR disclosures on page 23 of the June 30, 2011 Form 10-Q, where accruing and non-accruing TDRs are disclosed. However, accruing TDRs were not included in the impaired loan disclosure. Non-accruing TDRs were included in the impaired loan disclosures and the nonaccrual loan disclosures.
The significant increase from June 30, 2011 of $18.7 million of commercial real estate impaired loans without a related allowance recorded is a result of reporting all accruing TDRs as impaired loans. During the period ended September 30, 2011, management obtained clarification around the reporting of accruing TDRs through our regulators and external accountants. As a result, S&T began reporting both accruing and non-accruing
TDRs as impaired loans for the life of the loan. Including accruing TDRs in the impaired disclosure category had an immaterial impact on the allowance for loan losses and no impact on non-performing loan totals or reported TDRs in current or prior periods.
The following represents additional information as of September 30, 2011 relating to our commercial real estate impaired loans:
As of September 30, 2011, commercial real estate loans of $50.6 million comprised 73% of the total impaired loans of $69.0 million. These impaired loans are collateralized primarily by commercial real estate properties such as retail or strip malls, office buildings, hotels and various other types of commercial purpose properties. These loans are generally considered collateral dependent and charge-offs are recorded when a confirmed loss exists. Approximately $7.3 million of charge-offs have been recorded relating to these commercial real estate loans over the life of these loans. It is S&Ts policy to order appraisals on an annual basis on impaired loans or sooner if facts and circumstances warrant otherwise. As of September 30, 2011, updated appraisals completed within the last 16 months exist for these loans and aggregate an estimated fair value of approximately $57.3 million.
We will include disclosure of this type in future filings.
If you should have any questions concerning the enclosed matters, please contact the undersigned at 724-427-2347.
Sincerely, | ||
/s/ Melanie Hubler | ||
Name: | Melanie Hubler | |
Title: | Senior Vice President, Controller |
cc: | Todd D. Brice, Chief Executive Officer, S&T Bancorp, Inc. |
Mark Kochvar, Chief Financial Officer, S&T Bancorp, Inc.
Paul D. Freshour, Arnold & Porter LLP