First Midwest | First Midwest Bancorp, Inc. |
November 13, 2006
VIA EDGAR
Mr. John P. Nolan
Accounting Branch Chief
Division of Corporate Finance
Securities and Exchange Commission
100 F Street, N.E., Mail Stop-4561
Washington, D.C. 20549-4561
Re: | First Midwest Bancorp, Inc. (the "Company") | ||||||
Form 10-K for Fiscal Year Ended December 31, 2005 Filed March 2, 2006 | |||||||
Commission File No. 0-10967 | |||||||
Dear Mr. Nolan:
On behalf of the Company, set forth below is the response to the comments raised in the letter of the Staff dated October 31, 2006 with respect to the Company's Annual Report on Form 10-K for the year ended December 31, 2005. For your convenience, the Staff's comments are repeated in italics below, each followed by the Company's response.
1. | We note your response to comment three from our letter dated May 10, 2006. It appears from the documentation you provided that for cash flow hedges of brokered repurchase agreements and cash flow hedges of prime-based loans, you use both method one and method two of SFAS 133 Implementation Issue No. G-7 (DIG G-7) to measure hedge ineffectiveness. For both cash flow hedges of brokered repurchase agreements and prime-based loans, please tell us: | |
|
o |
how you determine and document at inception and on an on-going basis whether you use method one or method two of DIG G-7 to measure ineffectiveness for each specific hedging transaction; |
o | what facts and circumstances would cause you to use different methods for hedges of the same types of assets or liabilities (for example, to use method one for some hedges of |
Mr. John P. Nolan
November 13, 2006
Page 2
o | how you considered the guidance in paragraph 62 of SFAS 133 that states that ordinarily an entity should assess effectiveness for similar hedges in a similar manner and that the use of different methods for different hedges should be justified. In this regard, tell us whether you believe that a different method is justified simply because there are basis differences; | |
o | how you concluded that your methodology does not result in a two-step test. In this regard, it appears that you may have documented that you assess effectiveness using method one of DIG G-7, but then if during the process of evaluating whether there are any basis differences or whether the balance of the broker repurchase agreements falls below the notional amount of the LIBOR based swaps, you determined either was the case, you would use method two of DIG G-7 to determine the amount of ineffectiveness. Tell us why you wouldn't continue to use method one of DIG G-7 to determine the amount of ineffectiveness in these circumstances. Additionally, tell us whether you believe you would get a different result by using method one versus using method two; and | |
o | how you determined that your hedge documentation specifically identifies the method that will be used to measure hedge ineffectiveness as required by EITF Topic No. D-102. |
We would like to clarify our response to you dated May 23, 2006 and further explain the related hedge documentation provided in support of our prior response.
It is our interpretation that the assessment of hedge effectiveness and the measurement of ineffectiveness are two different concepts under SFAS 133. For example, DIG E-7 indicates that an entity can use regression analysis to assess hedge effectiveness and use the dollar-offset method to measure ineffectiveness. Further, FAS 133 and EITF Topic D-102 clarify that the method of assessing effectiveness and the method to measure ineffectiveness are different concepts, both of which must be addressed. However, in a hedge relationship such as ours where there are no sources of ineffectiveness, other than the timing delay of some of our cash flows versus the swap (even though the reset dates match precisely) which your Comment #2 addresses, the separate concepts of "assessing effectiveness" and "measuring ineffectiveness" tend to blur, which is what we believe you are reacting to in our documentation. We would like to again emphasize that we are not using more than one method to assess effe ctiveness, noting the staff's objection to such a practice in the John James speech at the 2003 AICPA Conference on SEC and PCAOB Developments.
As discussed in DIG G-7, the method of measuring ineffectiveness in a cash flow hedge must be clearly identified as Method 1 - Change in Variable Cash Flows Method, Method 2 - Hypothetical Derivative Method, or Method 3 - Change in Fair Value Method. We believe we have directly complied with that requirement by clearly indicating that Method 2 is our choice.
With this in mind, the Company uses the critical matched terms method under DIG G-9 to assess the effectiveness of its cash flow hedges of broker repurchase agreements and cash flow hedges of prime-based loans. As indicated in the Company's documentation, if the critical terms of the hedged instrument and the hedged item match, no ineffectiveness is assumed related to these cash flow hedges. Reference is made in our hedge documentation to method one of DIG G-7 as it
Mr. John P. Nolan
November 13, 2006
Page 3
provides useful background information on how to determine whether the critical terms match for an interest rate swap and thus aids in interpreting the guidance in DIG G-9 on assuming no ineffectiveness when critical terms of the hedging instrument and the hedged transaction match. However, if at any point in time it is determined the critical terms do not match, the Company would use method 2 (the Hypothetical Derivative method) to measure ineffectiveness.
Given our clarification that the Company only uses one method to measure ineffectiveness, we have not addressed the remaining questions in #1 above, as they are not applicable to the Company, because they are based on the presumption that the Company uses two methods to measure ineffectiveness.
2. | We note your response to comment six of our letter dated May 10, 2006 regarding the differences in timing that may exist between the prime-based loans and the swap used to hedge the interest payment on those loans when measuring hedge ineffectiveness. Your response indicates that generally the difference in payment dates of the hedged loans and the swap generally was never more than 10 days and that since the actual difference in payment dates is small, you determined the impact of timing differences in interest payments would be immaterial. As a result, you have not recognized any ineffectiveness associated with this timing difference. We believe that payment-timing differences between the hedged item and the hedging instrument create basis differences, even when the indexes and reset dates are the same. Your response seems to indicate that you believe that while the differences in payment dates would lead to a small amount of ineffectiveness, you could intuitively determine that the amount of ineffectiveness created by these timing differences would be immaterial to the financial statements. However, outside of the shortcut method and the very limited situation of the commodity forward described in paragraph 65 of SFAS 133, we do not believe that SFAS133 permits a company to rely on matching what it believes are the most critical terms and intuition for measuring ineffectiveness, when there are known sources of ineffectiveness. In cases where a company is using some type of critical terms matching in its assessment of effectiveness, we believe in the initial prospective assessment a company must, at a minimum, do some type of quantification, whether stress tests, shocks, regressions, or other statistical analysis to determine the parameters within which the company's evaluation that there wouldn't be a material amount of ineffectiveness are valid. Then, on an ongoing basis, in addition to monitoring for changes in critical terms or counterparty credit changes, the company must monito r that changes in the environment have not occurred such as to be outside the range the company established upfront. |
While you did not do this upfront assessment, you did perform an ongoing assessment of whether there were any changes in the critical terms or the counterparty credit, as well as an evaluation of the impact of the timing difference in interest payments at December 31, 2005 over the remaining life of the three prime-based swaps. As a result, we will not require that the company calculate the impact of not applying hedge accounting for these swaps. However, we believe you need to revise your methodology to comply with this comment on an ongoing basis. Alternative, if you disagree with our conclusions, please tell us why. |
Mr. John P. Nolan
November 13, 2006
Page 4
The Company will revise its methodology for measuring ineffectiveness as requested above to include an upfront test (whether stress tests, shocks, regression or other statistical analysis) to quantify the range in which the impact of timing differences in interest payments would not result in a material amount of ineffectiveness and continue to monitor, on an ongoing basis, that the results are not outside the range established upfront.
3. | We note that you appear to be using a first payments approach for your cash flow hedges of the broker repurchase agreement accounts and for your cash flow hedges of prime-based loans. Please tell us whether you are designating one swap for certain groups of cash flows or whether you are layering on swaps. If you are layering on swaps, please describe the methodology used for doing so.
|
As the Company only uses the first payments approach discussed in DIG G-25 for its cash flow hedges of prime-based loans, this discussion only relates to the Company's cash flow hedges of prime-based loans. The Company acknowledges that it does layer on swaps. When layering on swaps, the first swap with a notional amount of $25 million is designated against the first interest payments received on prime-based loans up to the notional amount of the swap. The second swap with a notional amount of $25 million is designated against the next interest payments received on prime-based loans up to the notional amount of the swap. This same process is repeated for the third swap with a notional amount of $25 million. The designation of the swap against interest payments received is done on a chronological basis, based on the payment receipt date of each individual loan payment. The Company verifies that the balance of the prime-based loans exceeds the notional amount of the swap.
As you requested, the Company acknowledges that:
o | The Company is responsible for the adequacy and accuracy of the disclosure in the filings; | |
o | Staff comments or changes to disclosures in response to staff comments in the filings do not foreclose the Commission from taking any action with respect to the filing; and | |
o | 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. |
If you have further questions or require any additional information, please contact the undersigned at (630) 875-7283 or Barbara E. Briick, Senior Vice President and Corporate Controller, at (630) 875-7459.
Very truly yours,
First Midwest Bancorp, Inc.
By: MICHAEL L. SCUDDER
Michael L. Scudder, Executive Vice President,
Chief Financial Officer
Mr. John P. Nolan
November 13, 2006
Page 5
cc: Lisa Haynes, Staff Accountant, SEC
John M. O'Meara, CEO, First Midwest Bancorp, Inc.
Barbara E. Briick, SVP and Controller, First Midwest Bancorp, Inc
Jonathan A. Koff, Chapman and Cutler LLP
David Doyle, Ernst Young LLP