CORRESP 1 filename1.htm SFCALLLSECLetter4redacted


FOIA Confidential Treatment Requested Pursuant to Rule 83

November 1, 2013

Mr. Amit Pande
Accounting Branch Chief
United States Securities and Exchange Commission
100 F Street, NE
Washington, DC 20549

Re:         Sterling Financial Corporation (“Sterling”)
ALLL Determinations in Form 10-K for the Fiscal Year Ended
December 31, 2012
Filed February 27, 2013
File No. 000-34696

Dear Mr. Pande:

This letter attempts to summarize, in one consolidated response, recent discussions and correspondence regarding Sterling’s Allowance for Loans and Lease Losses (“ALLL”) determinations in the Form 10-K for the fiscal year ended December 31, 2012, filed February 27, and Sterling’s proposed enhanced ALLL disclosure for future public filings. These topics were also discussed during our phone conference on October 10, 2013, which further supplemented related points from SEC comments, and Sterling responses, dated August, 13, 2013, July 25, 2013, June 18, 2013, May 17, 2013 and May 6, 2013. We hope you find this summary to be helpful in understanding Sterling’s ALLL methodology.

Because of the commercially sensitive nature of certain information contained herein, this submission is accompanied by a request for confidential treatment for selected portions of text on page SFC-004 (the “Confidential Material”). We have provided a copy of our request for confidential treatment to the Office of Freedom of Information and Privacy Act pursuant to Rule 83 of the Securities and Exchange Commission's Rules on Information and Requests (17 C.F.R. § 200.83). For the Staff’s reference, we have marked this letter to show where the Confidential Material has been redacted from the version filed via EDGAR and for which Sterling is requesting confidential treatment.

In accordance with Rule 83, Sterling requests confidential treatment of the Confidential Material marked in this response letter. Please promptly inform the undersigned of any request for disclosure of the Confidential Material made pursuant to the Freedom of Information and Privacy Act or otherwise so that the undersigned may substantiate the foregoing request for confidential treatment in accordance with Rule 83.

In accordance with Rule 83, this Letter has also been clearly marked with the legend “Confidential Treatment Requested by Sterling Financial Corporation” and each page is marked for the record with the identifying numbers and code “SFC-001” through “SFC-0011.”




SFC-001
    

Confidential Treatment Requested by Sterling Financial Corporation

1.
Description of ALLL Estimated Loss Calculation under the ALLL Base Model

Sterling’s ALLL calculation uses the probability of default and the loss given default to determine an estimate of the losses inherent in the loan portfolio. The probability of default is determined using historical annual migration to loss data, which is then adjusted with an estimate for the various relevant qualitative factors, and an estimated loss emergence period. The loss given default is calculated using the actual/historical life-to-date losses net of recoveries on loans in the portfolio, the last twelve months of losses on OREO (including losses on foreclosure, write-downs, and losses on sale) and discounted note sales. In addition to these quantitative historical calculations, a qualitative assessment is completed for every loan category, which includes an assessment of low, medium, or high risk for each of the following eight qualitative factors:
• Trends in delinquencies, classified & non-accrual;
• Changes in nature/volume of the portfolio;
• Changes in credit concentrations;
• Changes in economic conditions;
• Changes in charge-offs and collateral values;
• Quality of management and lending personnel;
• Quality and effectiveness in lending policies; and
• Competitive, legal & regulatory environment.
These factors closely track to the factors in SEC Staff Accounting Bulletin No. 102. The impact of the eight qualitative factors is made at the loan category level based on a simple average of the low, medium or high assessments attributed to each loan category.1 For example, a loan category with 6 low assessments, 1 medium assessment and 1 high assessment in the various factors would result in a low qualitative risk loan category. The overall qualitative rating of low, medium or high is then assigned a quantitative value (a typical value of 10%, 25% and 50% for low, medium, and high, respectively) that is then applied to increase the calculated probability of default. This calculation of estimated losses using the quantitative probability of default and the loss given default calculations, along with the qualitative assessment described above, is referred to as the “ALLL Base Model.” Sterling's ALLL Committee (the “Committee”) believes that the limited look back period is the main limitation of the ALLL Base Model. As mentioned above, the historical losses incorporated in the model are based on loans that are still in the portfolio as well as the last twelve months of OREO and note sale activity. This limited look back period could exclude loans with significant losses that were realized prior to the last twelve months. Another model limitation is that the initial qualitative factors are incorporated by multiplying the assigned quantitative value, as described above, by the probability of default. As such, decreases in the migration to loss and the loss given default also decrease the resulting qualitative reserve even in periods when the Committee has concluded that no changes to the qualitative factors are necessary. This limitation further amplifies the impact of the limited look back period in economic environments that include significant increases or decreases in charge-offs.
1See Selected Loan Loss Allowance Methodology and Documentation Issues, SEC Staff Accounting Bulletin: No. 102, July 6, 2001, available at http://www.sec.gov/interps/account/sab102.htm.
 

SFC-002
    

Confidential Treatment Requested by Sterling Financial Corporation

Quarterly, the Committee evaluates the estimated losses calculated by the ALLL Base Model for a given period in an effort to increase the likelihood that the ALLL Base Model will result in estimated losses that are most representative of the current inherent losses in the loan portfolio. This includes monitoring the estimated losses produced by the ALLL Base model against management’s expectations based on a review of the significant deviations from historical loss rates and trends in estimated loss rates for recent periods (as described below under “The Validation Process—Application of Management Judgment to the ALLL Base Model”).
2.
The Validation Process—Application of Management Judgment to the ALLL Base Model
As part of the overall ALLL methodology, the Committee has consistently compared the ALLL Base Model estimated loss rates to the actual one-year and three-year (annualized) losses and the trend in estimated loss rates for recent periods, and management's expectations of the inherent losses based on the quantitative and qualitative data. The one-year and three-year (annualized) loss rates are calculated by dividing the actual charge-offs net of recoveries by the average loan balances for the associated measurement period. If the estimated loss rate calculated under the ALLL Base Model is inconsistent with the one-year loss rate, three-year loss rates or the trend in estimated loss rates, then a review of the quantitative and qualitative data is completed to determine whether the historical loss rate more accurately represents the estimated losses inherent in a particular loan category. The determination that the ALLL Base Model estimated loss rate is inconsistent with losses inherent in the portfolio occurs when there is a significant change in the calculated estimate loss rate from the prior periods without a corresponding significant improvement in credit trends, historical loss rates, and/or qualitative factors. The Committee would expect any change in the ALLL Base Model estimated loss rate to be directionally consistent with and proportionate to changes in the historical loss rates, trends in credit quality and/or qualitative factors. A larger than expected deviation in the ALLL Base Model estimated loss rates without a corresponding improvement in the quantitative and qualitative factors is due to the limitations in the ALLL Base Model discussed above. The historical loss rate is only used when it is determined that it is a better estimate of the inherent losses for a particular loan category.

For example, at December 31, 2012, when the Committee evaluated the ALLL Base Model results through the validation process, as described above, it determined that for some loan categories the model was not producing results consistent with the trends in prior period estimated loss rates, historical losses, loan portfolio quarterly credit trends, economic conditions or the Committee’s expectations of the inherent losses. For some loan categories, the ALLL Base Model estimated loss rates were less than the three-year historical loss rates. The Committee did not think these estimated loss rates was reasonable due to the limited improvement in the quantitative and qualitative factors. In addition, the significant declines in these estimated loss rates were not supported by the Committee’s assessment of the remaining credit risk inherent in these loan categories. The Committee determined the most reasonable approach for determining the appropriate allowance for each category, as well as the overall ALLL, was to select the loss rate that the Committee, based on an analysis of all pertinent quantitative and qualitative data, determined to provide a reasonable estimate of the losses inherent in each category and the overall loan portfolio. In addition to using the three-year historical loss rates, the Committee considered other approaches including holding the estimated loss rates at levels consistent with prior period

SFC-003
    

Confidential Treatment Requested by Sterling Financial Corporation

loss rates, and adjusting the quantitative amounts assigned to the qualitative factors in order to get the ALLL Base Model to provide calculated estimated loss rates consistent with the Committee's expectations of the inherent losses.
As this deviation first emerged in the fourth quarter of 2012, the Committee determined that rather than making wholesale changes to the ALLL Base Model, it was a better approach to use the three-year historical loss rates for certain loan categories after a thorough review of both the quantitative and qualitative factors. The key to any ALLL model is management’s oversight and review of the overall process in determining that the end result (the cumulative ALLL for all loan categories) is reasonable when considering the loan portfolio as a whole, as well as when compared to prior periods and to peer financial institutions. Ultimately, the Committee evaluates all pertinent quantitative and qualitative data in the context of the significant collective years of banking industry experience of its members, specifically their experience through the recent credit cycle and their direct knowledge of Sterling’s loan portfolio, in selecting loss rates that are most representative of the inherent losses in the loan portfolio. Sterling’s ALLL Base Model is imprecise by nature and also is a significant management estimate designed to provide a reasonable estimate of the inherent losses in the loan portfolio. The Committee believes that this process follows the guidance provided for in the 1996 Joint Interagency Guidance, including the statement that “[a]rriving at an appropriate allowance involves a high degree of management judgment and results in a range of estimated losses” and “[d]etermining the allowance for loan losses is inevitably imprecise, and an appropriate allowance falls within a range of estimated losses.”2 
The Committee is comfortable with the process undertaken to select the individual inputs and, more importantly, the cumulative ALLL for all loan categories at December 31, 2012 was compliant with GAAP and the SEC Staff Accounting Bulletin: No. 102, which states that “registrants [may] employ numerous procedures when validating the reasonableness of their loan loss allowance methodology and determining whether there may be deficiencies in their overall methodology or loan grading process…[including]…[a] review of trends in loan volume, delinquencies, restructurings, and concentrations [and] [a] review of previous charge-off and recovery history, including an evaluation of the timeliness of the entries to record both the charge-offs and the recoveries.3

[***]

[TEXT FROM PAGE SFC-004 OF THIS LETTER HAS BEEN OMITTED AND PROVIDED SEPARATELY TO THE SECURITIES AND EXCHANGE COMMISSION PURSUANT TO A REQUEST FOR CONFIDENTIAL TREATMENT UNDER RULE 83.]
2 See Joint Interagency Letter to Financial Institutions, July 12, 1999, jointly issued by the Securities and Exchange Commission, Federal Deposit Insurance Corporation, Federal Reserve Board, Office of the Comptroller of the Currency and Office of Thrift Supervision, available at http://www.federalreserve.gov/boarddocs/srletters/1999/sr9922a1.pdf.

 
 
 
 
 
3 See Selected Loan Loss Allowance Methodology and Documentation Issues, SEC Staff Accounting Bulletin: No. 102, July 6, 2001, available at http://www.sec.gov/interps/account/sab102.htm.

 
 
 
 
 


SFC-004
    

Confidential Treatment Requested by Sterling Financial Corporation

3.
Actual Historical Loss Rates and Calculated Estimated Loss Rates for the Year Ended December 31, 2012
For the year ended, December 31, 2012, the following table sets forth the actual historical loss rates and ALLL Base Model calculated loss rates for the six loan categories for which the actual three-year (annualized) loss rates exceeded the calculated estimated loss rates.

Table 1
  
(A)
One Year Actual Loss Rates
(B)
Three Year Annualized Actual Loss Rates
(C)
12/31/2012 ALLL Base Model Estimated Loss Rates
B-C Difference
 
 
 
 
 
Residential
0.78
%
2.44
%
1.29
%
1.63
%
Investor CRE
0.94
%
1.91
%
1.60
%
0.31
%
Multifamily Perm
0.04
%
0.92
%
0.73
%
0.19
%
CRE Owner
1.09
%
1.52
%
1.49
%
0.02
%
Consumer RE Secured 1st Lien
0.79
%
1.02
%
0.97
%
0.05
%

As discussed in our prior response, after considering quantitative and qualitative data at December 31, 2012, Sterling used the historical three-year loss rates for these six loan categories. Due to the continuing economic uncertainty, high levels of historic losses experienced in these loan categories by Sterling, and the Committee’s understanding of the remaining credit risk inherent in these loan categories, the Committee concluded that the selected historical losses were more reflective of the inherent losses than the ALLL Base Model calculated estimated loss rates. As discussed above, the ALLL Base Model results for these loan categories were artificially low due to the methodology for incorporating the qualitative factors and the limited loss look back period used in the ALLL Base Model. The limited look back period resulted in some significant losses beyond the last twelve months not being taken in to consideration, whereas the three-year historical loss rate includes all losses for the last three years. The Committee believes it is appropriate to place greater reliance upon specific quantitative data, such as the three year loss history, as loan loss periods are typically extended after exiting a credit cycle. The periods of higher losses are beginning to have less impact on the calculated estimated loss rates for many of the loan categories as a greater number of periods with lower losses are being included in the calculation and each period is weighted equally in the loss migration data analysis.
The net aggregate impact to the ALLL of using the annualized three-year loss rates for the six loan categories in the table above was $16.4 million at December 31, 2012. The two loan categories impacted the most by this adjustment were Residential (which includes Non Owner and Owner) and Investor CRE loans, which had $9.7 million and $3.3 million impact, respectively. The Committee deemed the impact for the other loan categories to be immaterial to the overall ALLL calculation.


SFC-005
    

Confidential Treatment Requested by Sterling Financial Corporation

a)
Residential and Investor CRE Categories
The following is provided as an example of the process the Committee goes through in validating the ALLL Base Model results. This example shows the specific items considered in validating the ALLL Base Model results for Residential real estate and the Investor CRE loans. The following table provides data points that were considered by the Committee in validating the ALLL Base Model results for Residential real estate and Investor CRE loans.
Table 2
 
One Year Actual Loss Rates
Three Year Annualized Actual Loss Rates
12/31/2012 Estimated Loss Rates as Adjusted
12/31/2012 Estimated Loss Rates as Calculated by the ALL Base Model
12/31/11 Estimated Loss Rates as Calculated by the ALL Base Model
12/31/10 Estimate Loss Rates as Calculated by the ALL Base Model
Residential
0.78%
2.44%
2.44%
1.29%
2.14%
2.35%
Investor CRE
0.94%
1.91%
1.91%
1.60%
4.16%
3.65%

b)
Residential
For the Residential category, the reduction in the ALLL Base Model calculated loss rate from 2.14% as of December, 31 2011 to 1.29% was determined to be inconsistent with the Committee’s expectations of the inherent losses based on the qualitative factors for this loan category in the region, as well as the Committee’s experience and knowledge of this loan category for Sterling. The large decline in the calculated estimate loss rate when compared to the prior year estimated loss rate and the historical three-year loss rate was being caused by recent periods of lower probability of default which is a reflection of the continued low interest rate environment that has resulted in lower payments for many borrowers. Where the actual three-year historical loss rate calculation weights all losses within the period equally, the ALLL Base Model, as discussed as a limitation above, relies on more certain periods being weighted more than others. Given that the regional economy has only partially recovered, the Committee did not think a decline in the estimated loss rates on this loan category was supportable. Specific economic factors considered in determining that the reduction in the calculated estimated loss rates was not supportable were the following:
•
Specific comments regarding Residential Real Estate loans included in the Federal Reserve Bank's 12th District, where Sterling’s lending is concentrated, report dated November 21, 2012 included the following:
◦
large portion of homeowners in the District remain underwater on their mortgages;
◦
foreclosure rates remain elevated; and
◦
consumer confidence may remain weak until home prices show sustained support and increases.

SFC-006
    

Confidential Treatment Requested by Sterling Financial Corporation

•
Job loss is still an issue in the 12th District, as the 12th District has recovered 54% of the initial 8.7% job losses versus national which has recovered 67% of the 6.3% job losses nationally. While the unemployment rate is considered in determining the qualitative assessment of the economic conditions, the impact of this qualitative factor was decreased due to the limitations of the ALLL Base Model as a result of the declining probability of default and the Committee determined that the model did not give the necessary weighting of the unemployment rate for this loan category. The unemployment rate is a key variable used by the Federal Reserve Board in the loan stress testing requirements under the Dodd Frank Act.
•
At December 31, 2012, over 35% of Sterling's Residential Real Estate loans had interest rates that were at least 1% greater than market rates, indicating that these borrowers likely are not able to refinance due to collateral values and/or credit issues.
•
At December 31, 2012, approximately 53% of Sterling's Residential Real Estate loans were adjustable-rate mortgages, which currently benefit from the low rate environment. Approximately 45% of these loans are scheduled to have a rate adjustment within the next two years. An increase in interest rates could have a significant impact on the debt servicing ability of many of these borrowers.
•
Residential real estate values have shown some improvement but still remain substantially below peak pre-crisis values. Nationally, values were 33% below peak values as of December 31, 2011 compared to 26% below peak values as of September 30, 2012. As of September 30, 2012 real estate values continue to be more adversely impacted in the 12th district than nationally, with values down 35% from peak for the Pacific region and 32% for the Mountain region.
•
As of December 31, 2012 and 2011, over 20% of Sterling’s Residential Real Estate portfolio was originated in 2006, 2007, and 2008, which was the peak of residential property values. As noted above, residential property values are recovering slowly in the Pacific and Mountain regions.

Based on the foregoing factors, the Committee determined that the actual three-year loss rate provided a better estimate of the inherent losses in the Residential category than the ALLL Base Model calculated loss rate. This determination was based on the Committee’s judgment that the credit quality of this loan category remained consistent with the prior year and the current inherent losses are better reflected by the trend in actual losses experienced over the last three years as reflected in the three-year loss rate than by the unusually low recent losses included in the ALLL Base Model.
c)
Investor CRE Category
For the Investor CRE category, the ALLL Base Model calculated loss rate of 1.60% at December 31, 2012, was a significant decline from the December 31, 2011 estimated loss rate of 4.16%. Using the ALLL Base Model calculated loss rate would have resulted in a 62% reduction in the estimated loss rate in just one year. A 62% decline in this estimated loss rate was not consistent with the Committee's expectations of the inherent losses in the Investor CRE category in Sterling’s market based on the Committee’s knowledge of the market and the local economies. The following items were considered in determining that the ALLL Base model calculated loss rates were not consistent with the Committee's assessment:

SFC-007
    

Confidential Treatment Requested by Sterling Financial Corporation

•
Specific comments regarding CRE loans included in the Federal Reserve Bank's 12th District report dated November 21, 2012 included the following:
◦
core sector CRE property values are down 30% from their peak and some borrowers are left with little or no equity;
◦
concerns that maturing CRE loans have inadequate cash flows and insufficient collateral values; and
◦
current low interest rates have helped keep CRE loan performing, so far, masking credit issues.
•
At December 31, 2012, approximately 66% of Sterling's Investor CRE loans were adjustable-rate loans, as discussed in the Federal Reserve Bank 12th District report these loans benefit from the current low rate environment. Approximately 61% of these loans are scheduled to have an interest rate adjustment within the next two years. An increase in interest rates could have a significant impact on the debt servicing ability of many of these borrowers.
•
Recoveries are distorting Sterling's one year net charge-off rates. Recoveries were 29% of gross charge-offs for Investor CRE loans during 2012 compared to only 11% for three year period ended December 31, 2012. The recoveries experienced during 2012 primarily relate to charge-offs taken prior to 2012. The high level of recoveries during 2012 is causing a temporary distortion in the ALLL Base model results for Investor CRE. The Committee believes this distortion is temporary in nature as recoveries are estimated to normalize in the near future.
•
The delinquency rate for this loan category increased to 3.7% as of December 31, 2012 compared to 3.2% as of December 31, 2011.
•
Close to 50% of Sterling’s Investor CRE portfolio as of December 31, 2012, was originated in 2006, 2007, and 2008, which was the peak of the CRE property values. As noted above, CRE property values are down 30% from their peak values, and there is a concern that the low interest rates have helped keep CRE loans performing.

The Committee did not believe a decrease of this magnitude over one year was supported by the qualitative data. The Committee determined that the trend in the three-year loss rate more closely matched the improvements in the quantitative and qualitative factors during the prior year than the ALLL Base Model calculated loss rate, as such in the Committee's judgment, the three-year loss rate better reflected the losses inherent in this loan category.

4.
Proposed Additional Enhanced Disclosure
In light of the numerous correspondences and recent phone conference regarding Sterling’s ALLL methodology, to the extent the facts and circumstances remain the same, we intend to include the following additional enhanced ALLL methodology disclosure as part of Sterling’s Critical Accounting Policies in future public filings. The proposed changes are marked against Sterling’s disclosure in its Form 10-K for the Fiscal Year Ended December 31, 2012.
Allowance for Credit Losses. The allowance for credit losses is comprised of the allowance for loan losses and the reserve for unfunded credit commitments. In general, determining the amount of the allowance requires significant judgment and the use of estimates by management. Sterling maintains an allowance for loan losses to absorb probable losses in the loan portfolio based on a quarterly analysis of the portfolio and expected inherent losses. This analysis is designed to

SFC-008
    

Confidential Treatment Requested by Sterling Financial Corporation

determine an appropriate level and allocation of the allowance for losses among loan classes by considering factors affecting loan losses, including specific and confirmed losses, levels and trends in classified and nonperforming loans, historical loan loss experience, loan migration analysis, current national and local economic conditions, volume, growth and composition of the portfolio, regulatory guidance and other relevant factors. The reserve for unfunded credit commitments includes loss coverage for loan repurchases arising primarily from mortgage banking activities. Management monitors the loan portfolio to evaluate the adequacy of the allowance. The allowance can increase or decrease each quarter based upon the results of management's analysis.
The loan portfolio is grouped into several industry segments for homogeneous loans based on characteristics such as loan type, borrower and collateral. Loan migration to loss data is used to determine the annual probability of default. The annual probability of default is adjusted for the estimated loss emergence period and may be further adjusted based on the assessment of qualitative factors. The qualitative factors are reviewed quarterly to determine a rating for each category of low, medium, and high, that is then quantified into the calculation of the probability of default. The estimated loss emergence period reflects an estimate of the time frame during which losses may be realized. Sterling establishes the expected estimated loss rate on loans using the historical life-to-date losses net of recoveries on charged off and foreclosed loans from remaining in the the most recent 12 months portfolio, the last twelve months of losses on OREO (including losses on foreclosure, write-downs, and losses on sale) and losses on discounted note sales to estimate the amount that would be lost if a default were to occur, which is termed the "loss given default." The probability of default is multiplied by the loss given default to calculate the expected estimated loss rates for each loan class. Once the estimated loss rates for each loan class. The are calculated expected , the results are validated, by comparing this calculated estimated loss is compared for each loan class to the actual one-year and three-year (annualized) losses, net of recoveries. If In addition, the calculated expected inherent loss rate is less than the actual also compared to the prior periods, and the overall trend in loss rates from prior periods. If the calculated loss rates differ significantly from the one-year and three-year actual loss rates and/or is inconsistent with the historical trends, Sterling performs further analyses to identify and evaluate the possible causes of the differences. This includes evaluating the characteristics of the historical loss and recovery data used in calculating the loss rate to determine whether such activity is an appropriate reflection, in management’s judgment, of the current inherent losses in the various loan classes. This analysis takes into consideration the impact of the qualitative factors that increase the probability of default in the calculated loss rates in assessing whether there is additional qualitative risk. Sterling evaluates the results of the model, one and three year loss rates, then the expected loss rate would be set at the greater of the actual one or three year loss rate and the evaluation process and based on their judgment, experience, and specific knowledge of the loan portfolio, determines the most appropriate loss rate for each category based on their evaluation of quantitative data and qualitative factors described above.
Sterling may also maintain an unallocated allowance to provide for other credit losses that may exist in the loan portfolio that are not taken into consideration in establishing estimated loss rates the probability of default and loss given default. The unallocated amount may generally be maintained at higher levels during times of economic uncertainty. The unallocated amount is reviewed at least quarterly based on credit and economic trends.

SFC-009
    

Confidential Treatment Requested by Sterling Financial Corporation

Individual loan reviews are based upon specific quantitative and qualitative criteria, including the size of the loan, loan quality ratings, value of collateral, repayment ability of borrowers and guarantors, as applicable, and historical experience factors. The historical experience factors utilized and allowances for homogeneous loans (such as residential mortgage loans and consumer loans) are collectively evaluated based upon historical loss experience, loan migration analysis, trends in losses and delinquencies, growth of loans in particular markets, and known changes in economic conditions in each particular lending market.
A loan is considered impaired when, based on current information and events, it is probable Sterling will be unable to collect the scheduled payments of principal and interest when due according to the contractual terms of the loan agreement. Factors considered by management in determining impairment include payment status, collateral value and the probability of collecting scheduled principal and interest payments when due. Loans that experience insignificant payment delays and payment shortfalls generally are not classified as impaired. Management determines the significance of payment delays and payment shortfalls on a case-by-case basis, taking into consideration all of the circumstances surrounding the loan and the borrower, including the length of delay, the reasons for the delay, the borrower's prior payment record, the ability and willingness of guarantors to make payments, and the amount of the shortfall in relation to the principal and interest owed. Impairment is measured on a loan-by-loan basis by either the present value of expected estimated future cash flows discounted at the loan's effective interest rate, the loan's obtainable market price or the fair value of collateral if the loan is collateral-dependent.
The fair value of the underlying collateral for real estate loans, which may or may not be collateral-dependent, is determined by using appraisals from qualified external sources. For commercial properties and residential development loans, the external appraisals are reviewed by qualified internal appraisal staff to ensure compliance with appropriate standards and technical accuracy. Appraisals are updated according to regulatory provisions for extensions or restructurings of commercial or residential real estate construction and permanent loans that have not performed within the terms of the original loan. Updated appraisals are also ordered for loans that have not been restructured, but that have stale valuation information, generally defined in the current market as information older than one year, and deteriorating credit quality that warrants classification as substandard.
The timing of obtaining appraisals may vary, depending on the nature and complexity of the property being evaluated and the general breadth of appraisal activity in the marketplace, but generally it is within 30 to 90 days of recognition of substandard status, following determination of collateral dependency, or in connection with a loan's maturity or a negotiation that may result in the restructuring or extension of a real estate secured loan. Delays in timing may occur to comply with actions such as a bankruptcy filing or provisions of an SBA a government guarantee.
Estimates of fair value may be used for substandard collateral-dependent loans at quarter end if external appraisals are not expected to be completed in time for determining quarter end results or to update values between appraisal dates to reflect recent sales activity of comparable inventory or pending property sales of the subject collateral. During periods of declining real estate values, Sterling may record a specific reserve for impaired loans for which an updated valuation analysis has not been completed within the last quarter. Estimates of value are not used to raise a value; however, estimates may be used to recognize deterioration of market values in quarters between

SFC-0010
    

Confidential Treatment Requested by Sterling Financial Corporation

appraisal updates. The judgment with respect to recognition of any provision or related charge-off for a confirmed loss also takes into consideration whether the loan is collateral-dependent or whether it is supported by sources of repayment or cash flow beyond the collateral that is being valued. For loans that are deemed to be collateral-dependent, the amount of charge-offs is determined in relation to the collateral's appraised value. For loans that are not deemed to be collateral-dependent, the amount of charge-offs may differ from the collateral's appraised value because there is additional support for the loan, such as cash flow from other sources.
The reserve for unfunded credit commitments includes loss exposure from Sterling's mortgage banking operations. Loans sold into the secondary market are sold with limited recourse to Sterling, meaning that Sterling may be obligated to repurchase any loans that are not underwritten in accordance with agency guidelines or have borrower misrepresentations.
While management uses available information to provide for loan losses, the ultimate collectability of a substantial portion of the loan portfolio and the need for future additions to the allowance will be influenced by changes in economic conditions and other relevant factors. There can be no assurance that the allowance for credit losses will be adequate to cover all losses, but management believes the allowance for credit losses was appropriate at December 31, 2012 September 30, 2013.
5. Closing Comments

In providing this response, Sterling acknowledges that:

1.
Sterling is responsible for the adequacy and accuracy of the disclosure in the filing;

2.
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

3.
Sterling 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.

Kindly direct any comments or questions to the undersigned at (509) 227-0961 or, in his absence, Andrew J. Schultheis at (509) 838-7239.

Sincerely,

STERLING FINANCIAL CORPORATION


By:    
/s/ Patrick J. Rusnak     
Patrick J. Rusnak
Executive Vice President and Chief Financial Officer



SFC-0011