- FHN Dashboard
- Financials
- Filings
- Holdings
- Transcripts
- ETFs
- Insider
- Institutional
- Shorts
-
CORRESP Filing
First Horizon (FHN) CORRESPCorrespondence with SEC
Filed: 2 Jun 05, 12:00am
June 2, 2005
Mr. Paul Cline
Senior Accountant
Securities and Exchange Commission
Division of Corporate Finance
100 F Street, N.E.
Washington, D.C. 20549
Re: | First Horizon National Corporation Form 10-K Filed March 14, 2005 File No. 001-15185 |
Dear Mr. Cline:
First Horizon National Corporation (the “Company”) is in receipt of the letter from the staff (the “Staff”) of the Securities and Exchange Commission, dated May 5, 2005, regarding the above-referenced filing.
To facilitate the Staff’s review, we have included in this letter the captions and numbered comments from the Staff’s comment letter in bold, italicized text, and have provided our responses immediately following each comment.
In some of our responses, we have agreed to supplement the disclosure in our future filings. We are doing that in the spirit of cooperation with the Staff, and not because we believe our prior filing is materially deficient or inaccurate. Accordingly, any changes implemented in our future filings should not be taken as an admission that our prior disclosures were in any way deficient.
We have also indicated in some of our responses that we believe no change in disclosure is appropriate, and have explained why. We understand the Staff’s comments, even where a disclosure change is requested or suggested, to be based on the Staff’s understanding based on information available to it, which may be less than the information available to us. Accordingly, we understand those Staff comments may be withdrawn or modified based on the additional explanation or information we provide.
In response to your first, third and fourth comments, we have agreed to revise our disclosures in the notes to our audited consolidated financial statements, specifically Note 1 — Summary of Significant Accounting Policies, in our future filings. For your convenience, we have included a sample of the disclosure that we currently intend to include in future filings as Annex I to this letter. Please note that the proposed disclosure includes additional information intended to clarify which of the Company’s hedging relationships are strictly economic in nature, and which are entered into both for economic and hedge accounting benefits. In response to your second comment, we have included the requested supplemental information.
Page 1
1. Summary of Significant Accounting Policies – page 60
Please revise your SFAS No. 133 derivative disclosures as follows:
a. | Clearly describe the specific methodology used to evaluate both prospective and retrospective hedge effectiveness for each hedging relationship (for example, the pipeline and the warehouse, MSR’s, interest-only securities) |
b. | Disclose how often hedge effectiveness tests are performed |
We will revise the Mortgage Banking section of Note 1 — Summary of Significant Accounting Policies in future filings so that it more fully describes both our prospective and retrospective evaluations of hedge effectiveness, including the frequency with which such tests are performed. (See paragraphs 3 and 5 of our proposed Mortgage Banking disclosure.)
To supplementally answer your comment, in accordance with SFAS No. 133, as amended by SFAS No. 149, interest rate lock commitments (IRLC) are considered derivatives and are marked to market. The interest only securities are classified as trading securities and are fair value assets. Neither IRLC nor interest only securities are eligible for hedge accounting pursuant to SFAS No. 133 although both are hedged for economic purposes, as described in our proposed revised disclosure. (See paragraphs 1 and 8 of our proposed Mortgage Banking disclosure.)
c. | Quantify and disclose where on the balance sheet and income statement you classify your SFAS 133 qualifying fair value hedges and related hedged items for each major hedging relationship. |
The only SFAS No. 133 qualifying hedges relate to the warehouse and MSR. The Mortgage Banking section of Note 1 — Summary of Significant Accounting Policies in our future filings will include the dollar amounts and balance sheet classification of assets which are hedged pursuant to SFAS No. 133 and their related hedges. (See paragraphs 4 and 6 of our proposed Mortgage Banking disclosure.) In addition, we are clarifying our ineffectiveness disclosures which are already included in Note 1 to address your request for income statement quantification and location. The balance sheet classification of MSR was not separately disclosed in Note 1 as this asset is shown separately on the face of the balance sheet.
In addition, we will revise the Interest Rate Risk Management section of Note 1 — Summary of Significant Accounting Policies in our future filings to quantify and disclose the balance sheet location of qualifying SFAS No. 133 fair value hedges and related hedged items. (See Interest Rate Risk Management section of our proposed disclosure.)
Page 2
2. Specifically related to your fair value hedges of MSR, supplementally tell us in detail the following:
a. | The significant components of servicing income in your table on page 10. Specifically tell us the gross amount of hedge gains/losses offset by the change in fair value of the hedged MSR; |
The requested supplemental information with respect to components of servicing income, including hedged MSR gains and losses, is set forth in the following table for the fiscal years ended December 31, 2004, 2003 and 2002:
(thousands of dollars) | |||||||
---|---|---|---|---|---|---|---|
2004 | 2003 | 2002 | |||||
SFAS 133 Basis Adjustment to MSR | (18,340 | ) | 115,673 | (358,492 | ) | ||
SFAS 133 Change in Fair Value MSR Hedge due to Effectiveness | 18,340 | (115,673 | ) | 358,492 | |||
Net Interest Income on Swaps | 79,949 | 111,948 | 108,732 | ||||
Hedge Time Decay | (33,403 | ) | (16,528 | ) | (21,239 | ) | |
Other Hedge Ineffectiveness | 1,373 | 19,715 | 13,255 | ||||
Net G/L MSR Hedges | 47,919 | 115,135 | 100,748 | ||||
Gross Service Fees | 294,315 | 248,375 | 198,828 | ||||
Early Payoff Interest Expense | (21,162 | ) | (34,473 | ) | (20,225 | ) | |
Guarantee Fees | (64,755 | ) | (48,579 | ) | (31,440 | ) | |
Amortization | (152,722 | ) | (132,274 | ) | (117,842 | ) | |
Impairment | (37,062 | ) | (158,324 | ) | (150,162 | ) | |
Ancillary and Other Income | 17,263 | 18,326 | 9,861 | ||||
Total Servicing Income | 83,796 | 8,186 | (10,232 | ) | |||
b. | How you apply paragraph 21 of SFAS 133; |
Per paragraph 21 (subparagraphs b. and c.), MSR are hedgable assets because their value is at risk for changes in interest rates. Consistent with the provisions of paragraphs 20 and 21 (subparagraphs a., b. and f.), the Company hedges the change in the fair value of the MSR asset attributable solely to the change in the benchmark interest rate. The 10-Year LIBOR Swap rate has been selected as the benchmark interest rate, which complies with paragraph 21 (subparagraph f.) and the definition of “benchmark interest rate” as presented in paragraph 540 (as amended by SFAS No. 138). Paragraph 21 (subparagraph a.1.) requires that a hedged portfolio of similar assets must share the risk exposure for which they are hedged. The Company performs a similar asset test on a monthly basis, or more frequently if necessary, to aggregate the MSR assets of its loan servicing portfolio into SFAS No. 133 hedge strata. The similar asset test is performed by measuring price sensitivity over the following rate shock scenarios: up and down 25 bps and 50 bps, respectively.
c. | How you measure both prospective and retrospective hedge effectiveness of this hedging relationship. For instance, how do you calculate the change in the fair value of your MSR related to changes in the benchmark interest rate; |
The Company performs prospective analyses of hedge effectiveness as required by paragraph 20 (subparagraph a.) and consistent with the discussions documented in DIG Issue E7. Prior to acquiring a new hedge instrument, an historical
Page 3
regression analysis is performed between the underlying index of the proposed hedging instrument and the underlying index of the hedged item. The analysis is performed over 6-month, 1-year and 2-year time horizons. High correlation exists when the regression coefficients (r2) for each time horizon are above 80%.
Pursuant to paragraph 20 (subparagraph b.), at the end of each hedge period the Company computes the change in the fair value of the MSR asset due solely to the change in the benchmark interest rate and the change in the fair value of the derivative for each specific risk tranche. The result of this calculation becomes an historical data point. As permitted in DIG Issue E7, the Company performs a regression analysis of all collected data points over a rolling 12-month period to determine retrospective effectiveness. The Company believes that retrospective hedge effectiveness has been established if the regression results meet all of the following criteria:
For all hedge periods up to 12/31/2004
• | Coefficient of correlation (r) > 0.80 |
• | Slope of regression line between 0.80 and 1.25 |
For all hedge periods after 12/31/2004
• | Regression Coefficient (r2) >0.80 |
• | Slope of regression line between 0.80 and 1.25 |
• | Pass F-Test |
The new (post-2004) effectiveness criteria would have had no material impact if applied in 2004, 2003 and 2002. The enhancement was made to conform to evolving industry-best practices.
In order to determine the change in the fair value of the MSR asset due to the change in the benchmark interest rate for any given hedge period, the Company compares the value of the MSR portfolio at the beginning of the hedge period with the value at the end of the hedge period using a “Same Tape, Different Assumptions” approach. This approach uses the same loan portfolio (servicing tape) for both valuations. The interest rates used in the valuation model are changed in accordance with the change in the benchmark interest rate. While the interest rates used at the beginning of the period represent actual market rates, the interest rates used at the end of the period represent a parallel shift equal to the change in the benchmark interest rate from the beginning of the period to the end of the period. The interest rate driven assumptions in the MSR valuation model include: mortgage rate, earnings rate, cost of funds rate, and all underlying indices for ARM loans. By parallel shifting only interest rates, the Company can isolate the change in the fair value of the MSR due to the change in the benchmark interest rate, a methodology which complies with the provisions of paragraph 20 (subparagraph b.). This methodology also complies with paragraph 24 of SFAS No. 138 because no contractual cash flows of the MSR are excluded in determining the change in fair value attributable to movements in the benchmark interest rate.
Page 4
d. | We note on Page 41, that you do not specifically hedge the change in fair value of MSR attributed to unanticipated prepayments (representing the difference between actual prepayment experience and estimated prepayments derived from a model). (i) How you calculate the change in fair value related to unanticipated prepayments. |
The Company only hedges the change in fair value attributable to the change in the benchmark interest rate, as permitted by paragraph 21 of SFAS No. 133. Since we use the “Same Tape, Different Assumptions” approach to isolate the change in MSR fair value attributable to the change in the benchmark interest rate, we do not perform a specific calculation to determine the change in fair value due to unanticipated prepayments. However, the impact of all unhedged risks (which includes but is not limited to unanticipated prepayments) is captured by the SFAS No. 140 impairment test. Any changes in the fair value of the MSR asset, whether they are hedged, anticipated or otherwise quantified, are captured when calculating the fair value of the MSR asset as part of the SFAS No. 140 impairment test. Historically, unanticipated prepayments have been the primary driver of impairment expense.
(ii) How you include the effect of unanticipated prepayments on the fair value of MSR in your prospective and retrospective effectiveness calculations. |
Since unanticipated prepayments are not part of the Company’s designated risk, the effect of unanticipated prepayments on the fair value of the MSR is not included in the prospective and retrospective effectiveness calculations.
3. Please revise to disclose how you apply paragraph 27 of SFAS 133 to your MSR impairment tests.
We will revise the Mortgage Banking section of Note 1 — Summary of Significant Accounting Policies in our future filings in order to discuss our tests for impairment and how impairment impacts earnings. (See paragraph 7 of the proposed Mortgage Banking disclosure.)
4. Please revise to disclose if you use different stratification criteria for your MSR impairment tests and for the purposes of grouping similar assets to be designated as a hedged portfolio in your fair value hedges. If you do, tell us how you consider the fair value hedge market value adjustments in your impairment tests.
We will revise the Mortgage Banking section of Note 1 — Summary of Significant Accounting Policies in our future filings in order to disclose that the stratification criteria differ, and to explain that fair value hedge market adjustments are allocated to the loan level which allows for re-stratification for the impairment test. (See paragraph 7 of the proposed Mortgage Banking disclosure.)
Page 5
First Horizon National Corporation acknowledges that it is responsible for the adequacy and accuracy of the disclosure in its Form 10-K, that 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 that 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.
We hope these responses adequately address the questions raised in your letter. Any questions with respect to the foregoing should be directed to the undersigned at (901) 523-5620.
Sincerely,
/s/ Marlin L. Mosby III
Marlin L. Mosby III
Executive Vice President and Chief Financial Officer
Page 6
First Horizon National Corporation Form 10-K
Annex I
Excerpts from Proposed Note 1 Disclosure
(The following are the applicable excerpts from Note 1 — Summary of Significant Accounting Policies as we currently expect them to appear in our future filings)
Derivative financial instruments.
Mortgage Banking
Mortgage banking interest rate lock commitments (IRLC) are short-term commitments to fund mortgage loan applications in process (the pipeline) for a fixed term at a fixed price. During the term of an IRLC, First Horizon Home Loans has the risk that interest rates will change from the rate quoted to the borrower. First Horizon Home Loans enters into forward sales contracts with respect to fixed rate loan commitments and futures contracts with respect to adjustable rate loan commitments as economic hedges designed to protect the value of the IRLC from changes in value due to changes in interest rates. Under SFAS No. 133 IRLCs qualify as derivative financial instruments and as such do not qualify for hedge accounting treatment. As a result, the IRLCs are recorded at fair value with changes in fair value recorded in current earnings as gain or loss on the sale of loans in mortgage banking noninterest income. See Accounting Changes section for impact of SAB No. 105 on the valuation of interest rate lock commitments. The fair value of the derivative is classified as either Other Assets or Other Liabilities, with changes in fair value included in current earnings as a component of gain or loss on the sale of loans in mortgage banking noninterest income.
First Horizon Home Loans’ mortgage loans held for sale (the warehouse) are subject to changes in fair value, which are primarily due to fluctuations in interest rates from the loan closing date through the date of sale of the loan into the secondary market. Typically, the fair value of the warehouse declines in value when interest rates increase and rises in value when interest rates decrease. To mitigate this risk, First Horizon Home Loans enters into forward sales contracts and futures contracts to provide an economic hedge against those changes in fair value on a significant portion of the warehouse. These derivatives are recorded at fair value with changes in fair value recorded in current earnings as a component of the gain or loss on the sale of loans in mortgage banking noninterest income.
To the extent that these interest rate derivatives are designated to hedge specific similar assets in the warehouse and prospective analyses indicate that high correlation is expected, the hedged loans qualify for hedge accounting under SFAS No. 133. If the regression analysis does not show consistently strong correlation, hedge accounting under SFAS No. 133 is
Page A-1
not attempted. Anticipated correlation is determined based on the historical regressions between the change in fair value of the derivatives and the change in fair value of the hedged mortgage loans. Hedges are reset daily and the statistical correlation is calculated using these daily data points which are accumulated monthly. Retrospective hedge effectiveness is measured monthly using the regression correlation at the end of the month. First Horizon Home Loans generally maintains a coverage ratio (the ratio of expected change in the fair value of derivatives to expected change in the fair value of hedged assets) of approximately 100 percent on warehouse loans accounted for under SFAS No. 133. Effective SFAS No. 133 hedging results in adjustments to the recorded value of those hedged loans which adjustments are included as a component of the gain or loss on the sale of loans in mortgage banking noninterest income.
Warehouse loans subject to SFAS No. 133 hedge accounting treatment included in loans held for sale totaled $1.2 billion and $1.5 billion on December 31, 2004 and 2003 respectively. The related derivatives valued at $(4.3) million and $(16.3) million on December 31, 2004 and 2003, respectively, are classified as other assets or other liabilities based on the net open position. For 2004, 2003, and 2002, gains of $.8 million, gains of $1.2 million and losses of $9.8 million, respectively, represent the hedge ineffectiveness for these fair value hedges that was recognized as a component of gain or loss on sale of loans in mortgage banking noninterest income.
First Horizon Home Loans also enters into interest rate contracts (including swaps, swaptions, and mortgage forward sales contracts) to hedge against the effects of changes in fair value of its MSR due solely to changes in the benchmark interest rate (10-year LIBOR swap rate). All capitalized MSR are hedged for economic purposes with the vast majority of MSR routinely qualifying for hedge accounting. For purposes of measuring effectiveness of the hedge, time decay and recognized net interest income, including changes in value attributable to changes in spot and forward prices, if applicable, are excluded from the change in value of the related derivatives. Interest rate derivative contracts used to hedge against interest rate risk in the servicing portfolio are designated to specific risk tranches of servicing. First Horizon Home Loans enters into hedges of the MSR to minimize the effects of loss in value of MSR associated with increased prepayment activity that generally results from declining interest rates. In a rising interest rate environment, the value of the MSR generally will increase while the value of the hedge instruments will decline. Hedges are reset at least monthly and more frequently, as needed, to respond to changes in interest rates or hedge composition. Generally, a coverage ratio approximating 100 percent is maintained on hedged MSR. Prior to acquiring a new hedge instrument, First Horizon Home Loans performs a prospective evaluation of anticipated hedge effectiveness by reviewing the historical regression between the underlying index of the proposed hedge instrument and the mortgage rate. At the end of each hedge period, the change in the fair value of the hedged MSR asset due to the change in benchmark interest rate is calculated and becomes an historical data point. Retrospective hedge effectiveness is determined by performing a regression analysis of all collected data points over a rolling 12 month period. Effective hedging under SFAS No. 133 results in adjustments to the recorded value of the MSR. These basis adjustments, as
Page A-2
well as the change in fair value of derivatives attributable to effective hedging, are included as a component of servicing income in mortgage banking noninterest income.
MSR subject to SFAS No. 133 hedges totaled $970.9 million and $732.7 million on December 31, 2004 and 2003, respectively. The related derivatives, valued at $79.0 million and $204.2 million on December 31, 2004 and 2003, respectively, were classified as other assets or other liabilities based on a net open position. For 2004, 2003, and 2002, the increase in value of the hedge instruments exceeded the decrease in value of the hedged MSRs by $47.9 million, $115.1 million and $100.7 million, respectively; these amounts were recognized in mortgage banking noninterest income. Included in these amounts were $46.5 million, $95.4 million and $87.5 million in 2004, 2003 and 2002, respectively, representing recognized net interest income and time decay on various hedges, which were excluded for purposes of measuring hedge effectiveness. The remaining ineffectiveness was $1.4 million, $19.7 million and $13.2 million in 2004, 2003, and 2002, respectively. Pursuant to SFAS No. 133, the basis in MSR that qualify for hedge accounting are adjusted for the impact of hedge performance.
First Horizon Home Loans uses different MSR stratification for purposes of determining hedge effectiveness pursuant to SFAS No. 133 and impairment testing pursuant to SFAS No. 140. The hedge results under SFAS No. 133 are allocated at a loan level and the loans are then aggregated into the SFAS No. 140 strata. This adjusted MSR basis is subsequently compared to the full fair value of the MSR to test for asset impairment. MSR basis is reduced to the extent that adjusted basis exceeds fair value. This reduction in basis as a result of impairment is a component of servicing income in mortgage banking noninterest income.
First Horizon Home Loans utilizes derivatives (including swaps, swaptions, and mortgage forward sales contracts) that change in value inversely to the movement of interest rates to protect the value of its interest-only securities as an economic hedge. Changes in the fair value of these derivatives are recognized currently in earnings in mortgage banking noninterest income as a component of servicing income. Interest-only securities are included in trading securities with changes in fair value recognized currently in earnings in mortgage banking noninterest income as a component of servicing income.
Interest Rate Risk Management
First Horizon National Corporation has entered into pay floating, receive fixed interest rate swaps to hedge the interest rate risk of certain large institutional certificates of deposit, totaling $149.7 million on December 31, 2004, and certain long-term debt obligations, totaling $1.2 billion and $502.7 million on December 31, 2004 and 2003, respectively. These swaps qualify as fair value hedges and meet all conditions necessary to assume effectiveness has been established and no ineffectiveness exists. The fair market value of these swaps is recorded on the balance sheet in other assets and other liabilities. The balance sheet impact of these swaps was $13.9 million included in other assets and $11.2 million included in other liabilities
Page A-3
on December 31, 2004 and $11.1 million included in other assets on December 31, 2003. Interest paid or received for these swaps is recognized as an adjustment of the interest expense of the liabilities whose risk is being managed.
In addition, First Horizon National Corporation has entered into certain interest rate swaps and caps as a part of our relationship with loan customers. These derivatives are not designated as hedging instruments and are entered into as part of a product offering to our commercial customers with customer derivatives paired with offsetting market instruments that, when completed, are designed to eliminate market risk. These contracts do not qualify for hedge accounting and are measured at fair value with gains or losses included in current earnings in noninterest income.
Page A-4