[Logo]
BCSB BANCORP, INC.
Mailing Address: P.O. Box 397, Perry Hall, MD 21128
Office Location: 4111 E. Joppa Road, Baltimore, MD 21236
410-256-5000
www.baltcosavings.com
E-Mail: info@bcsb.net
_____________________________________________________________________________________________________________________________________________
March 12, 2010
Securities and Exchange Commission
100 F Street, NE
Washington, DC 20549
Re: BCSB Bancorp, Inc.
Form 10-K for Fiscal Year Ended September 30, 2009
Form 10-Q for Fiscal Quarter Ended December 31, 2009
File No. 0-53163
Ladies and Gentlemen:
On behalf of BCSB Bancorp, Inc. (the “Company”), this letter is provided in response to comments provided by Securities and Exchange Commission (the “Commission”) staff, by letter dated March 2, 2010, with respect to the Company’s Form 10-K for the year ended September 30, 2009 (the “Form 10-K”) and the Company’s Form 10-Q for fiscal quarter ended December 31, 2009. The staff’s comment and the Company’s response is set forth below:
Form 10-Q as of December 31, 2009
Financial Statements, page 3
Note 4 – Mortgage-Backed Securities, page 10
Comment No. 1. We note the significant unrealized losses related to your collateralized mortgage obligation securities at September 30, 2009 and December 31, 2009. Please provide us a detailed description of the other-than-temporary impairment analysis you performed on the eight securities in unrealized loss positions as of December 31, 2009. Please identify all of the evidence you considered, explain the relative significance of each piece of evidence, and identify the primary evidence on which you relied to support a realizable value equal to or greater than the carrying value of the investment. |
Response to Comment 1.
As noted in the comment, significant unrealized losses existed within our securities portfolio at September 30, 2009 and December 31, 2009. At both dates, more than 99% of the unrealized losses were associated with four private label collateralized mortgage obligations (“CMOs”). The other unrealized losses relate to FNMA mortgage backed securities, for which |
Securities and Exchange Commission
receipt of principal is fully guaranteed. Accordingly, this response is directed toward our other-than-temporary impairment (OTTI) analysis on the CMOs. It should be noted that, as a result of our analysis, an impairment charge of $500,000 was recorded during the three months ended June 30, 2009 on CMO securities available for sale. Updated reviews since that period have resulted in no expectation of material additional impairment related to credit loss. As a result, no subsequent impairment charges have been recorded. |
The following addresses your request for a detailed description of our analysis. We engaged the services of independent third party professionals to analyze the OTTI status of the CMOs. This firm was in no way involved in the original purchase of the securities. The CMO OTTI methodology is formulated within The FASB Accounting Standards Codification. The goal of the process is to determine the expected cash flows of the instrument in question as well as the expected “credit loss component”. In order to derive these two values, the model addresses each component of the net present value calculation. The model utilized is Intex, the industry standard credit model for structured product. The validity of the model can be assessed by the rationality of the assumptions applied, which are inferred fr om as much available market information as possible. As prescribed by The FASB Accounting Standards Codification, the following information is useful in this process:
| 1. | Quoted prices that are not current for the same or similar security |
| 2. | Relevant reports issued by analysts and rating agencies |
| 3. | The current level of interest rates and any directional movements in relevant indices |
| 4. | Information about the performance of the underlying mortgage loans, such as delinquency and foreclosure rates, loss experience, and prepayment rates |
| 5. | Other observable inputs |
NPV Model Decomposed
Delving into the foundation of the assumptions, the net present value model components are discussed below. The model results hinge on the following three assumptions:
| 1. | Loss Severity assumption |
| 2. | CDR assumption (likelihood of default) |
Loss Severity assumption – To determine the loss severity assumption, the following was performed:
The loss severity assumption is simplified by access to Intex’s historical database. Within this data, the most recent monthly realized loss severity rates by issuer and origination
Securities and Exchange Commision
March 12, 2010
Page 3
year are surveyed (or if the data is security specific). Therefore, the most recent severity results are presented in order to build an assumption.
CDR assumption – The creditworthiness of a borrower is most often determined by factors such as debt service coverage ratio, LTV, FICO, or Asset Documentation (Full, Low, or No). However, the most defining characteristic of a loan’s expected performance is whether the loan is current or delinquent. Therefore, all of the loans within the pool are segregated by their status category.
Next, information is incorporated from access to Loan Performance Data, a service provider that has compiled a database of loan performance statistics for the last few decades. Default rates are then obtained by state and by aging bucket (30, 60, 90 days past due) for applicable periods of the CMO’s under review.
Probabilities are derived called “transition probabilities” or “migration probabilities” for the specified dataset. In essence, these values represent the historical likelihood of a 30 Day Delinquent loan moving through the credit pipeline to default. These values are also derived for 60 Day and 90 Day Delinquent loans. As a result, these probabilities are used within the CDR assumptions for all loans falling into these categories.
The current loan category is stratified by FICO scores and Asset Doc (Full, Low, No Doc). Once again, the Loan Performance Database is used for historical probabilities of loans meeting these characteristics moving into default. The average default rate is then calculated as well as its standard deviation. Two standard deviations are then added to this mean for current loan CDR assumptions, representing an assumed stress scenario.
Prepayment assumption – This component is based on historical activity. As most of these instruments have gravitated towards a baseline voluntary prepayment stream over the previous year, recent history gives a good (and conservative) estimate of future voluntary prepayment results.
Summarized
These assumptions return both a “credit loss component” as well as expected cash flows that should be viewed as conservative, yet rational. Finally, several scenarios are created using these credit assumptions but varying the prepayment speed to reflect the sensitivity of the security.
* * *
Securities and Exchange Commission
March 12, 2010
Page 4
The Company hereby acknowledges that:
| · | 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. |
Please direct any further comments or questions to the undersigned at (410) 248-1164.
Very truly yours,
/s/ Anthony R. Cole
Anthony R. Cole
Chief Financial Officer
cc: Mr. Amit Pande
Ms. Babette Cooper
Securities and Exchange Commission
Gary R. Bronstein, Esq.
Joel E. Bronstein, Esq.
Kilpatrick Stockton LLP
Mr. Conor Quinn
Stegman & Company Professional Association
Stegman & Company Professional Association
