Publication: Journal of Derivatives Accounting, Vol 2. No. 2
The challenge for accounting standard setters is to develop accounting procedures that enable financial analysts to understand and assess the economic health and prospects of reporting institutions. In the U.S. the Financial Accounting Standards Board (FASB) bears this responsibility; and as a rule, accounting principles and practices that the FASB has put forth have received high marks for approaching this ideal. In the case of the mortgage industry, however, the current guidance may be falling short of the mark.
The genesis of this criticism is Financial Accounting Standard No. 133 (FAS 133), “Accounting for Derivative Instruments and Hedging Activities,” which was released in June 1998. One source of controversy surrounding this standard had to do with the fact that, with the advent of FAS 133, financial reporters had to decide if contractual arrangements that had never previously been considered to be derivatives satisfied the FASB’s new definition. Loan commitments fell into this category. Were they derivatives, or not? In March of 2002, the FASB resolved this question for one class of commitments by explicitly dictating that commitments relating to loans intended for resale were, in fact, derivatives. In making this determination, FASB left a host of related, unanswered questions. This paper endeavors to air these issues and highlight how or why inconsistent accounting treatments may have resulted. Identifying such inconsistencies should be of interest to standard-setters, accounting professionals, and analysts seeking to understand and evaluate the performance of mortgage issuing firms.
The paper is organized as follows: First, we describe the economics of the mortgage banking activity giving rise to the accounting issues examined in the paper. Second, we detail the steps in the standard-setting process that resulted in the current guidance. And finally, in the third section we lay out the possible resolutions that might reasonably come to be adopted and discuss the respective ramifications.
1. How Mortgage Originators Work
For the most part, mortgage originators have an infrastructure that enables them to issue commitments to potential borrowers, where the originator agrees to lend on the basis of terms established on the date at which the commitment is extended. Thus, the standard commitment is an instrument that conveys a right to the prospective borrower that allows him or her to secure a mortgage with pre-determined terms. Under this commitment, the borrower may elect to borrow under these terms or not, but the lender is obligated to make the loan if the borrower elects to exercise his or her right. In some cases, some linguistic gymnastics appear to be operating, where both the prospective lender and the prospective borrower may choose not to enter into the mortgage, after all. It should be appreciated, however, that the former situation is the more typical commitment design; and this former contract design is, in fact, a put option.
This paper focuses exclusively on this first type of commitment. The underlying to the option is a whole loan, which, when closed, will be an asset that the originator buys and a liability that the borrower issues or sells.1 Thus, following the extension of the commitment, the prospective borrower owns the right (but not the obligation) to sell the loan. The originator, on the other hand, has written or sold this put option.2
With the issuance of a fixed-rate commitment, often referred to as a “rate-lock,” the originator accepts a measure of interest rate risk: If mortgage rates increase and if the prospective borrower elects to exercise the right to borrow, the originator will be forced to issue the loan at a below-market rate. Clearly, this prospect would result in an economic loss to the originator. Alternatively, if mortgage rates fall, the lender stands to benefit from lending at an above market rate.
The risk associated with issuing interest rate commitments is generally addressed with a dynamic hedging process, where the prospective lender would arrange forward sale contracts (often with Fannie Mae or Freddie Mac). The size of the forward sales position would reflect the volume of loans from the pool of outstanding commitments that are expected to close; and this forward position would be adjusted upward or downward as the forecast of expected loan closures rises or falls.
Importantly, this forecast is quite sensitive to the probability of exercise, which, in turn, is sensitive to mortgage rate changes. That is, when rates rise between the issuance of the commitment and the expiration date, the option will be in-the-money and the borrower will be more likely to exercise this option, given the capacity to borrow at below-market rates; whereas when rates fall, the borrower would be more likely to ignore his or her right to exercise and seek funds from an alternative lender at a new, lower rate.3
Coincident with loans actually closing (as a consequence of exercise of the put option by the borrower), but before they are sold, originators will generally want to have a fully hedged position, where the ratio of forward sales to closed loans would be one-to-one. Thus, at any given point, when the originator has some volume of outstanding commitments and some holdings of whole loans, a portion of the overall forward sale positions would be dedicated to hedging the commitments, while the remainder would be dedicated to hedging the whole loans.
2. The Sequence of Standard-Setting Activities
To date, the evolution of accounting rules relating to mortgage commitments may be divided into four distinct phases:
(1) the passage of FAS 133, Accounting for Derivative Instruments and Hedging Activities;
(2) the issuance of Derivatives Implementation Group Issue C13, Scope Exceptions: When a Loan Commitment Is Included in the Scope of Statement 133 and FAS 149, Amendment of FASB Statement No. 133 on Derivative and Hedging Activities;
(3) the December 11, 2003, SEC speech on accounting for mortgage commitments; and
(4) the behind-the-scenes activities and subsequent issuance of SEC Staff AccountingBulletin 105 (SAB 105).
Each of these respective phases is discussed, in turn.
Issuance of FAS 133 : In June of 1998, the FASB issued Statement of Financial Accounting Standards No. 133, Accounting for Derivative Instruments and Hedging Activities. The critical passage relating to mortgage commitments is paragraph 291, which states that a loan commitment would not meet the definition of a derivative “if it (a) requires the holder to deliver a promissory note that would not be readily convertible to cash and (b) cannot readily be settled net.” This paragraph simply restates the obvious — that in order for an instrument to be deemed a derivative under FAS 133, the definition of a derivative (in paragraphs 6–9) must be satisfied. The crucial issue is whether the net settlement condition of the definition (paragraph 6c) is satisfied; and, as of the release date for FAS 133, this determination required a judgment call, where reasonable people might disagree.
An additional problem arose relating to fees and costs that are often charged when commitments are originated. Without a doubt, these charges reflect at least some portion of the value of the mortgage commitment being issued — or perhaps even the entire value — but practice tends to ignore this reality. In fact, despite collecting these fees, commitment issuers often claim that they issue commitments “for free,” reflecting the unsupportable perspective that, somehow, these receipts on the part of the issuers are entirely independent of the value of the options that they are writing.
Issuance of DIG Issue C13 and FAS 149: At least some measure of ambiguity relating to commitments was eliminated with the issuance of DIG Issue C13, Scope Exceptions: When a Loan Commitment Is Included in the Scope of Statement 133, in March 2002. This content was later incorporated into FAS 149, Amendment of FASB Statement No. 133 on Derivative and Hedging Activities. According to these pronouncements, loan commitments for mortgage loans to be held for resale were to be accounted for as derivatives — but only by the mortgage originator. That is, this treatment was not extended to the potential borrower. In addition, FAS 149 amended FAS 133 by adding paragraph 10(i), which scoped out all other types of loan commitments, for both issuer and holder.
The determination that loan commitments for mortgages to be held for resale had to be treated as derivatives brought to the fore the question of how to measure the fair value of these contracts. Unfortunately, practice did not have a consensus view. Instead, three divergent views were expressed. One group argued that the fair value of these commitments is zero at the issue date, because these commitments are issued at no charge.4 A second group recognized that these commitments are option contracts, believing them to be liabilities for the issuer at inception. And finally, the third view focused on the eventual sale and servicing cash flows associated with the eventual loan, and concluded that the commitment should be recorded as an asset by the issuer, with an offsetting credit to revenue.
December 11, 2003, SEC Speech: A resolution, of sorts, was provided by a speech delivered on December 11, 2003 at the Thirty-First AICPA National Conference on Current SEC Developments, by Eric Schuppenhauer, a Professional Accounting Fellow from the SEC’s Office of the Chief Accountant. Although the speech contained the usual disclaimer about not being an official statement of SEC position,5 speeches from this conference are generally broadly disseminated by the major accounting firms within days of being presented, and it was posted to the SEC website. As a consequence, the speech effectively became de facto guidance for SEC registrants.
This speech contained two conclusions of particular interest related to accounting for loan commitments for mortgages to be held for resale: (1) Upon the issuance of a commitment, a liability should be recognized (a credit) and the offsetting debit should be recorded as an expense in the income statement; and (2) the derivative should always be reported as a liability until expiration or termination of the commitment. In other words, the liability recognized at inception should never be permitted to morph into an asset. Although many disagreed with one or both of these requirements, the speech did provide the kind of direction that had been entirely lacking up to that point. Still, this guidance was hardly definitive. The speech included the indication that the SEC expected FASB to continue to deliberate the issue, and the SEC would “look forward to the Board’s guidance.”
The mortgage banking community took issue with the conclusions of the Schuppenhauer speech. The primary trade association, the Mortgage Bankers Association (MBA), along with representatives from the large accounting firms, met with the SEC to present their alternative views on the economics of and appropriate accounting for these loan commitments.
The MBA referenced Emerging Issues Task Force (EITF) issue 02-3, Issues Involved in Accounting for Derivative Contracts Held for Trading Purposes and Contracts Involved in Energy Trading and Risk Management Activities, which indicates that the exchange price for such a contract is normally its initial fair value. EITF 02-3 further precludes recognition of a dealer profit or loss absent “quoted market prices or other current market transactions.” (paragraph 4) Based on this guidance, the MBA argued against Schuppenhauer’s position of recording a liability and loss at inception of the commitment. Although the MBA recognized that EITF 02-3 refers specifically to energy derivatives, they argued that the underlying principle was valid and it should be applied in this case, as well.
It happens that contrary guidance has been given in connection with loss recognition, as well. Consider paragraph 11 of Financial Interpretation (FIN) 45, Guarantor’s Accounting and Disclosure Requirements for Guarantees, Including Guarantees of Indebtedness of Others. This guidance identifies a situation where a liability and a corresponding loss should be recorded at the inception of issuing a guarantee. In this case, as in the energy derivative and mortgage origination cases, the discussion pertains to circumstances where no initial or other consideration is exchanged. As FIN 45 scopes out derivatives accounted for at fair value under FAS 133, neither the guidance referenced by the MBA nor contrary guidance from FIN 45 specifically applies to the issue at hand, leaving room for divergent interpretations.
The MBA also reacted to the proposition that commitments could never be assets. In support of their position, they cited the case where interest rates fell after the commitment was issued. In this case, the lender stood to issue loans at above market rates, thereby allowing for a windfall gain. Failing to record this commitment as an asset under these circumstances would seem to be dismissive of an economic reality.
Release of SAB 105 : On March 9, 2004 the SEC issued Staff Accounting Bulletin 105, Application of Accounting Principles to Loan Commitments. The first interpretive response in the SAB is that cash flows from mortgage servicing rights should not be included in the calculation of the amount to record at inception of the loan commitment. Moreover, no other internally-developed intangible assets should be recorded as part of the loan commitment derivative.6 The import of this bulletin was that it struck at the heart of the issue for those reporting entities that perceived their commitments to be assets.
3. Implications for Future Standard Setting
The current state of affairs is one where practitioners have no explicit guidance that details the correct approach for accounting for commitments that qualify as derivatives; but the history to date suggests that one of four distinct approaches might reasonably be followed. We identify these approaches as the Schuppenhauer approach, the modified Schuppenhauer approach, the Kawaller-Teets approach, and the Mortgage Bankers approach. Each is discussed, in turn:
1. The Schuppenhauer Approach — The mortgage commitment is recognized by the prospective lender to be a written put option, where the underlying instrument is the mortgage loan. At inception, the commitment is recorded as a liability, at some non-zero value. The corresponding debit entry as of the date that the commitment is issued is treated as an expense item.
Over time, the commitment is marked-tomarket with changes flowing through earnings. The commitment’s value, however, is constrained to be at least zero, such that the liability will never morph into an asset. Put another way, this approach essentially disregards the potential for the issuer to realize a windfall gain if the option is exercised when it is out-of the- money.
Ultimately, the commitment is either exercised or not. If the commitment is notexercised, it expires worthless and its value change, again, would flow through earnings. The net gain from writing off the liability would exactly offset the originally booked expense (i.e., the original debit), albeit perhaps not in the same accounting period(s). If the commitment is exercised, the lender (a) marks the commitment to market (through earnings), and (b) records the loan as an asset and closes the commitment liability. If market interest rates had risen since the issuance of the commitment, the commitment would be in- the-money when it is closed; and the loan would be issued — and initially carried at a corresponding discount. On the other hand, if interest rates had fallen, such that the loan is issued with an above-market rate, the commitment would be written off at a zero value, generating an income effect exactly equal and opposite in total to the original debit expense. Because of the lower limit of zero relating to the value of the commitment liability, the loan would necessarily have to be recorded at par. In the former case, then, the impact of the rate change on the commitment is immediately recognized coincidentally with the issuance of the loan, while in the latter case, the impact is realized over the entire life of the loan. This approach appears inconsistent with the guidance of EITF 02-3, but consistent with the guidance in FIN 45. To the extent that commitments are issued to generate servicing assets and/or other customer relationship assets, this approach is also consistent with current guidance that prohibits recognizing servicing rights or internally-developed intangible assets unless evidenced by sale to an external party.
2. The Modified Schuppenhauer Approach —This approach is identical to the first, with the single exception that the value of the commitment is not constrained; or in other words, when interest rates fall (sufficiently) this commitment (which started as a liability) may actually become an asset. Thus, unlike the original Schuppenhauer approach, this modified approach does reflect the possibility that the issuer will reap a windfall gain when out-of-the-money commitments are exercised. The benefit or the modified Schuppenhauer approach is that it results in the consistent outcome with respect to the recognition of an income effect related to the commitment coincidentally with the extension of an off-par — irrespective of whether the loan is at a premium to par or a discount.
3. The Kawaller-Teets Approach — This approach is identical to the modified Schuppenhauer approach except that in this case the corresponding debit item on the issue date is treated as an asset. That is, the prospective lender does expect to be paid for issuing the commitment — in the form of a component of future interest payments for loans that actually close. In essence, then, this asset would be considered to be akin to a receivable.7 Ultimately, this asset would be closed out (with a zero value) either when the commitment is exercised or when the commitment expires. Unfortunately, this approach may be criticized for being inconsistent with current guidance on asset recognition.
4. The Mortgage Bankers Approach —The value of the commitment is assumed to be zero when the commitment is issued. This value will change over time, however, but only if mortgage rates change from the rate stipulated by the commitment. If rates rise, the commitment will become a liability to the issuer, while if rates fall, the commitment will become an asset. In effect, this approach measures the intrinsic value of the option contract without the traditional lower bound of zero, along with the associated probability of exercise. As with the first method, if the commitment is exercised, a final mark to market is made (through earnings), and the loan is recorded as an asset to the lender at its then prevailing market value. If the commitment is not exercised, the commitment expires worthless and this final value change, again, flows through earnings.
In effect, this approach gives license to the idea that the time value of the option (i.e., the commitment) may be ignored, such that in the general case (i.e., when the commitment rate differs from the market rate), the value of the commitment would be measured as the intrinsic value of the option, multiplied by the expected probability of exercise. The approach is consistent with the EITF 02-3 prohibition on loss recognition at inception of the derivative, which, of course, makes it inconsistent with the FIN 45 requirement to recognize a loss at the inception of some guarantees issued for no cash or other consideration.
As indicated, each of these above alternatives is flawed in that each contradicts either the spirit or the letter of existing GAAP. Unfortunately, that’s the consequence of a mixed attribute accounting model and a rules-based accounting system. Regardless, it is precisely because the “correct” approach is not obvious that further guidance is so critical. The FASB’s deliberation may come down to simply choosing the least offensive or the easiest to implement alternative; and, in the judgment of the current authors, the fourth alternative may very well satisfy this criterion.
Of the four choices provided, the last choice offers three critical advantages over the other three. First, it reduces the modeling uncertainty to the determination of the assumed fraction of commitments that will be exercised. While some degree of uncertainty exists in connection with this concern, by and large, mortgage issuers seem to be able to make judgments on the basis of past history with a reasonable degree of confidence.
Secondly, in the case where interest rates are stable over the life of the commitment, prior to the expiry (or exercise) of the commitment, no income volatility would be recorded under the fourth approach. In the other three methods, on the other hand, the gains from the commitment will be equal and opposite to the losses associated with the original debit item, but the two income effects will not necessarily be realized in the same accounting period. A presentation of income that shows this volatility (i.e., alternative 1–3) may very well distract analysts, given that these effects are (a) purely transitory and (b) very short-lived.
And finally, this fourth alternative also offers the advantage that it allows FASB to finesse the issue of any fees or origination charges that may be received by the prospective lender. With this choice, whether these charges reflect part or all of the fair value of the put option is irrelevant, as the put option is never recognized on the balance sheet, per se. The current treatment of these receipts, which views them as being separate and distinct from the commitments’ fair values could be continued without modification. It would seem, however, that if the any of the other three alternatives were chosen, a more exacting resolution of the relationship between these fees and the fair value of the commitment would be desirable — if not necessary.
1 The identification of the loan as the underlying instrument of the option reflects an economic (as opposed to accounting) orientation to option nomenclature. For an economist, the underlying is the good that might be purchased or sold when the option is exercised. In paragraph 57a of FAS 133, however, FASB defines the underlying of a derivative to be the price of that good, rather than the good, itself. In the case of mortgage loan commitments, then, in accounting terms, the underlying would be the specified interest rate on the prospective mortgage.
2 For readers who have difficulty accepting the idea that by extending the commitment, the prospective lender is selling a put, consider the transaction from the perspective of the prospective borrower. This entity has (and thus owns) a right that has value; and this right clearly meets the definition of an asset. Therefore, the counterparty to this prospective borrower (i.e., the prospective lender) must coincidentally record a liability. (Although the put option is an asset from the perspective of the potential borrower, paragraph 10(i) excludes this option from the scope of FAS 133/149 for the potential borrower.)
3 For traditional options, exercise when the option is out-of the-money generally does not occur. With home purchases, however, failure to exercise an out-of-the-money option might delay the purchase of the home, or otherwise put the purchase in jeopardy. Thus, in these cases, seemingly uneconomic decisions occur with some frequency.
4 This claim seems to have been made without regard to whether origination fees and charges were collected.
5 The following disclaimer was taken from the speech discussed here: “As a matter of policy, the Securities and Exchange Commission disclaims responsibility for any private publication or statement of any SEC employee or Commissioner. This speech expresses the author’s views and does not necessarily reflect those of the Commission, the Commissioners, or other members of the staff.”
6 In addition to the question of servicing cash flows, SAB 105 also discussed the disclosures that should be provided relating to loan commitments and whether this guidance should be applied retroactively.
7 The hesitancy to call this asset a “receivable” stems from a definitional concern. That is, to qualify as a traditional receivable, the holder would have to expect to receive subsequent payment. In this case, however, receipts only arise in connection with commitments that are exercised, and the prospective lender is not able to determine, in advance, which commitments will be exercised and which will not be.
Heidi Lindahl, Marketing Manager