July 16, Financial Accounting Standards Board Technical Director 401 Merritt 7 P O Box 5116 Norwalk CT

Transcription

1 College of Business & Economics Department of Accounting PO Box Moscow, ID In State Toll Free FAX July 16, 2009 Financial Accounting Standards Board Technical Director 401 Merritt 7 P O Box 5116 Norwalk CT File Reference No Discussion Paper: Leases: Preliminary Views Dear FASB: As an accounting professor, I have been creating numerous lease-related assignments every semester for well over 20 years. During the last two years, I have been adding discussion and assignments so that students would be able to classify and account for leases under IFRS as well as US GAAP. My practical experience is therefore in the rather distant past before I entered academe. FASB Statement No. 13 was just becoming effective when I took the CPA exam in Because of the detailed rules, I have used lease questions as an opportunity to help students learn to research the authoritative literature after all, there seems to be a pretty specific answer to locate on just about any issue related to leases! This is the expertise and background I bring to my examination of the discussion paper Leases: Preliminary Views (referred to hereafter as the PV-Leases). To make sure that I thoroughly understood the tentative conclusions, I have worked through of most the examples provided including creating amortization tables and the like in Excel. I ve learned a lot about alternatives that I would prefer not to be forced to teach (and will comment on those aspects.) I d like to express my appreciation for a summer grant from the College of Business and Economics at the University of Idaho that has provided support for the many hours I ve spent commenting on this preliminary reviews document as well as the one on revenue recognition earlier this summer.

2 Comment Letter from T. Gordon July 16, 2009 Page 2 General Comment Overall, I am supportive of the FASB and IASB efforts to improve lease accounting. It is clear that an entire industry has developed to help companies use leases to obtain off-balance sheet financing. This is detrimental for those who use financial statements and may even drive unsound economic decisions. Thus, requiring the capitalization of almost all leases is the right direction to go. In fact, financial reporting would be greatly improved by retaining current standards with respect to lease terms and present value of minimum lease payments (PVMLP) and merely eliminating the possibility of having operating leases. When I began this review, I was of the tentative opinion that sticking with most of our current guidelines would be a sufficient improvement to financial reporting. In other words, the changes to accounting for contingent rentals and guarantees of residual value or the assumed extensions of leases even when the renewal option is not a bargain, seemed unnecessary. However, after many hours of playing with possible lease situations, I ve concluded that the most of the proposed changes are reasonable and possibly necessary to discourage the kind of game playing that has led to off-balance sheet financing. The proposed changes related to computing lease term and lease payments are actually rather limited and should not be too much of a burden on preparers. After all, they are already familiar with the current standards regarding capital leases because the tests must be applied to classify leases as operating or finance. However, I encourage the Boards to minimize other changes to reduce the cost of implementation (such as frequent updating of the incremental borrowing rate). As requested, I have provided detailed responses to numbered questions in the PV-Leases. Chapter 2: Scope of lease accounting standard 1. The boards tentatively decided to base the scope of the proposed new lease accounting standard on the scope of the existing lease accounting standards. Do you agree with this proposed approach? If you disagree with the proposed approach, please describe how you would define the scope of the proposed new standard. This seems to be a logical first step. To me, leases are related to physical assets but I m aware that IASB also permits the lease of intangibles (sounds like royalty or licensing arrangement to me but I haven t studied that aspect of IFRS yet). Requiring the capitalization of all assets should be an improvement to financial reporting although it is important that the changes do not adversely impact ratio analysis, particularly with respect to interest expense. (see Question 11.) 2. Should the proposed new standard exclude non-core asset leases or short-term leases? Please explain why. Please explain how you would define those leases to be excluded from the scope of the proposed new standard. All leases should be included, whether or not the right-to-use asset is related to core versus non-core physical assets. I read the arguments in paragraphs 2.16 to 2.17 and found them lacking. With respect to the arguments for treating short-term leases differently, there is perhaps more justification. However, we are well aware that the majority of leases have been considered operating (i.e., there has been a willingness to use the standards to reach a particular accounting result) and I worry whether special rules for short leases would encourage

3 Comment Letter from T. Gordon July 16, 2009 Page 3 companies to structure leases at a year or less even when there are strong reasons to expect renewals or the like. On the other hand, permitting the presentation of very short-term leases within current assets (rather than plant, property and equipment whether operating or investing) might alleviate certain complexities of disclosure as discussed later in the PV-Leases. In fact, an argument could be made that breaking apart the payments between interest and principal would serve little useful purpose on leases of a year or less. Instead of making specific rules, perhaps the boards can leave it to preparers to handle very short leases under the general rule that standards need not apply to immaterial items. In this case, an example might be the occasional rental of carpet cleaning equipment or the like for a few days. Elaborate accounting in such circumstances would be ridiculous! [See also my separate discussion of envisioned ways folks will try to cook the books under the proposed guidelines which appears just before the conclusion to this letter.] Chapter 3: Approach to lessee accounting 3. Do you agree with the boards analysis of the rights and obligations, and assets and liabilities arising in a simple lease contract? If you disagree, please explain why. Yes. The analysis of lease rights (asset) and lease obligations (liability) seem logical and reasonable. 4. The boards tentatively decided to adopt an approach to lessee accounting that would require the lessee to recognise: (a) an asset representing its right to use the leased item for the lease term (the right-of use asset) (b) a liability for its obligation to pay rentals. Appendix C describes some possible accounting approaches that were rejected by the boards. Do you support the proposed approach? If you support an alternative approach, please describe the approach and explain why you support it. I was not particularly impressed with any of the rejected approaches described in Appendix C of the PV-Leases document. Accordingly, I m in support of the conclusions regarding simple leases. However, the devil is in the details and I m not sure I like some of the details related to initial and subsequent re-measurement. Financial reporting would be greatly improved if we did nothing more than apply existing guidance to the determination of the initial value of the right-to-use asset (for example, include a bargain purchase option but not a fair value purchase option among the cash flows). Of course, we should get rid of the operating versus finance lease rules and the bright line aspects. I m concerned that the Boards might unnecessarily make lease accounting more complicated than it currently is, particularly if frequent complex adjustments to the lease liability and asset accounts are mandated. Most accountants are already familiar with capital or finance leases because they have to compute the present value of the minimum lease payments (PVMLP) to test one of the basic rules/principles even if the lease turns out to be operating. This is true even if the lease has a title transfer, bargain purchase option, or covers most of the assets economic life: the PVMLP must be computed to establish the lease obligation and asset amounts. In addition, some of the later material in the PV-Leases appears to be adding unnecessary complications to the determination of the initial value (particularly the tentative IASB views on weighted probability expected values). Fancy or complex computations will not necessarily add value for users since the result may be no more accurate than application of existing and well-understood standards to all leases. I complement

4 Comment Letter from T. Gordon July 16, 2009 Page 4 the Boards on already resisting the appeal of certain kinds of complications, e.g., approach (a) on page 26 and suggest that further simplifications might be warranted as will be pointed out later in this comment letter. 5. The boards tentatively decided not to adopt a components approach to lease contracts. Instead, the boards tentatively decided to adopt an approach whereby the lessee recognises: (a) a single right-ofuse asset that includes rights acquired under options (b) a single obligation to pay rentals that includes obligations arising under contingent rental arrangements and residual value guarantees. Do you support this proposed approach? If not, why? I concur with this decision. I believe there is a single asset and a single liability in a simple lease whether for equipment or something more exotic. Keeping the accounting consistent with existing standards governing debt contracts makes sense to me that is, carrying the obligation at amortized cost using the initial effective interest rate. The component approach with respect to the physical asset would imply separating the value of roofs from elevators and the like (IASB) and a leased asset is actually an intangible right-to-use rather than different asset components with possibly differing economic lives. Chapter 4: Initial measurement 6. Do you agree with the boards tentative decision to measure the lessee s obligation to pay rentals at the present value of the lease payments discounted using the lessee s incremental borrowing rate? If you disagree, please explain why and describe how you would initially measure the lessee s obligation to pay rentals. My initial reading led me to a definite NO answer because I have been very impressed by the IASB guidance which says we should use the lessor s implicit interest rate if it can be determined. Having added lease classification under IASB to my courses over the last two years, I have been fascinated to learn how much more likely one is to get a capital lease rather than an operating lease by simply using the implicit rate rather than the incremental borrowing rate. In situations where there is a title transfer, bargain purchase option, or fully guaranteed residual value, the implicit rate can be determined easily and accurately. The rate is more problematic if there is an unguaranteed residual value or in cases where the fair value of the leased asset is unknown at the inception of the lease. I explain to my students that the FASB rules came about in a time when we were pretty much stuck with using time-value of money tables the first spreadsheet program had yet to be invented and a simple four-function calculator cost me about $150 as I recall, and I don t think there were financial calculators available then, at least at a reasonable price! In other words, using the incremental borrowing rate unless the implicit rate was provided by the lessor was a practical solution in FASB Statement No. 13. The IASB had the advantage of learning from many years of FASB experience as well as a much higher level of readily available technology. An incremental borrowing rate is subjectively determined and there would be a bias to claim a high rate because that gives a lower present value. On the other hand, the lessor s implicit rate can be likened to the actual rate of return the lessor believed would be necessary from a particular lessee. It is therefore a risk-adjusted rate of return essentially equal to a lessee-specific lending rate or the rate that the lessor would have charged the lessee for borrowing money to purchase the asset.

5 Comment Letter from T. Gordon July 16, 2009 Page 5 Note that Paragraphs 4.12 and 4.13 make valid point as to the difficulty of estimating the lessor s implicit interest rate, particularly for many leases currently classified as operating. Retaining the current IASB rule would mean that the incremental borrowing rate would be used in most cases when the leased asset must be returned to the lessor at the end of the lease term (because of the difficulty of obtaining information on residual values). I think the Boards may be trying to simplify something that is currently no problem at all for preparers (see Paragraph 4.17) who already have a (limited) choice for picking between two rates to use. The implicit rate, when available, is more valid than a subjectively determined incremental borrowing rate. I think financial statements users would be harmed rather than helped by this particular desire to simplify current standards. Two issues have modified my original opinion. First, if we end up re-measuring the liability with a new discount rate, there would be no way to get or estimate a new implicit interest rate. Of course, I m strongly recommending against changing the discount rate for subsequent measurements: if we decide to use the incremental borrowing rate, it should NOT be reassessed in subsequent reporting periods. (In other words, I prefer the FASB approach in Paragraph 5.24 to the IASB approach in Paragraph 5.25.) However, a more interesting problem arose as I got into exploring what lessor accounting will look like in the future. In the case of what are now operating leases, the value of the asset may not decline much between inception and termination of the lease. For real estate, the residual value might even be higher. Neither of these situations would prevent computation of an implicit interest rate. (See last example in my Appendix E.) However, that rate could be rather high and, in the cases where the asset is projected to increase in value, the interest expense for the lessee would exceed the rent payment leading to the nonsensical situation of negative reductions to principal in the lease obligation amortization table. Thus, the implicit rate could be higher than the incremental borrowing rate and that information (if disclosed to the lessee) might become a popular way to cook the books. Accordingly, a modification of the current IASB guidance would work better: Use the implicit interest rate if determinable unless it is higher than the incremental borrowing rate. 7. Do you agree with the boards tentative decision to initially measure the lessee s right-of-use asset at cost? If you disagree, please explain why and describe how you would initially measure the lessee s right-of-use asset. Yes. I agree but would prefer a different interest rate rule (as discussed in my answer to Question 6). A lease is the right to use one or more tangible assets (at least under US GAAP). The accounting on the asset side should parallel the accounting for a purchased asset of the same type. In other words, the accounting for the liability side would be consistent with accounting for other loans which is generally done using the historic interest rate. However, a lease versus purchase situation has one very important difference. For a lease, we can only estimate the value using present value techniques applied to anticipated cash outflows. For an outright purchase, the initial value can be known with certainty. In many cases in the PV- Leases document, the Boards have talked about the right-to-use asset becoming more or less valuable as options are exercised or declined (for example, see Paragraph 6.51). I think this is the wrong language to be using. Instead, we should be talking about subsequent measurement in the light of changes in accounting estimates. When the estimated lease payments change, we discover the initial measurement was incorrect, not that the actual value has changed. At the

6 Comment Letter from T. Gordon July 16, 2009 Page 6 beginning of the lease, all the possible options were already known. Therefore, it can t be worth more or less later merely because available options were later accepted or declined. However, with changes in assumptions, we do learn that our initial estimate of the initial value was incorrect. This recommended change in language is important and should also guide how we account for leases. For example, the analogy to purchased assets suggests that total depreciation expense should change when the estimates involved change. This does not happen in any of the three methods discussed (prospective, catch-up or retrospective) in the PV-Leases as described in my Appendix A, Table 1. None of the methods ends up with total depreciation expense equal to the revised value of the right-to-use asset. I have proposed a slightly different catch-up method that corrects this problem (see my Appendix B). Chapter 5: Subsequent measurement 8. The boards tentatively decided to adopt an amortised cost-based approach to subsequent measurement of both the obligation to pay rentals and the right-of-use asset. Do you agree with this proposed approach? If you disagree with the boards proposed approach, please describe the approach to subsequent measurement you would favour and why. This is a reasonable approach for both the liability and asset components of a lease and is similar to what we currently do for capital leases. However, it will become burdensome if frequent re-measurements are required, such as revision of the incremental borrowing rate (as envisioned by IASB) or for minor changes in estimated contingent rentals or residual value guarantees. In my hours of fiddling with the latter two issues, anything other than the proposed FASB method of taking the change in liability to P&L becomes unnecessarily complicated and time consuming because of the necessary changes in estimated depreciation expense as the right-of-use asset value changes. If we also had to change the incremental borrowing rate (which would surely be different at the end of every reporting period), the process would become exceedingly complex. Accordingly, I m in strong support of the basic decision for an amortized cost approach -- assuming frequent adjustments are not mandated. My examples 9 and 9-RV Excel files explored both the IASB and the FASB approach at length (amortization tables, journal entries and t-accounts) but more on this later. 9. Should a new lease accounting standard permit a lessee to elect to measure its obligation to pay rentals at fair value? Please explain your reasons. Maybe. If a fair value option is available for other financial assets and liabilities, it wouldn t make much sense to prohibit equivalent treatment for leases. The problem is that neither the asset nor liability side is not easily re-measured with level 1 or 2 market inputs. In addition, the fair value option was intended to facilitate an easier way of hedging and, in this case, the asset drops in value (through depreciation) as the obligation drops in value (through principal payments). Therefore, the hedge is basically automatic. However, if the boards see fit to simplify the re-measurement decisions, then it might make sense for companies that want to do a lot of fancy estimates of fair value be given that option. I m having a hard time seeing that financial statement users would benefit given the difficulties in arriving at fair values for leased assets.

7 Comment Letter from T. Gordon July 16, 2009 Page Should the lessee be required to revise its obligation to pay rentals to reflect changes in its incremental borrowing rate? Please explain your reasons. If the boards decide to require the obligation to pay rentals to be revised for changes in the incremental borrowing rate, should revision be made at each reporting date or only when there is a change in the estimated cash flows? Please explain your reasons. NO. As stated in response to Question 6, I prefer the use of the lessor s implicit rate (when feasible) and amortization based on the historic rather than current rates (see discussion under Question 7). I d like to see the re-measurement occur (if we really need one) only when projected cash flows change substantially due to different conclusions regarding acceptance of renewal options or the like. 1 Even then, current practice (for example, creditor accounting for troubled receivables) retains the historical interest rate. Changes in incremental borrowing rates happen with regularity as economic conditions drive interest rates up or down. This is true for all debt but, unless the fair value option has been adopted, other payables and receivables continue to be accounted for using the original effective interest rate. Therefore, I m strongly opposed to re-measurement of lease obligations when the incremental borrowing rate changes. It adds unnecessary complexity to the accounting for leases. The historic discount rate should be retained and used as illustrated with the catch-up method in the examples provided in Chapter 5. See also example in my Appendix B. I would also rather avoid the complications of current standards for changes in estimated asset retirement obligations (AROs). In FASB Statement No. 143, downward changes are at the historic rate but increases in future costs are measured at the credit-adjusted risk free rate in existence at the re-measurement date. I think a lease situation is quite different because all of the variations in cash flows were possibilities from the beginning. For an ARO, increases in projected costs may arise from new laws or discoveries. In addition, we might end up with the complication of using a weighted-average incremental borrowing rate even for downward revisions! So, to reiterate, I m strongly in favor of the use of the original historic discount rate to account for lease obligations including remeasurements trigger by revised estimates of lease term or cash flows. 11. In developing their preliminary views the boards decided to specify the required accounting for the obligation to pay rentals. An alternative approach would have been for the boards to require lessees to account for the obligation to pay rentals in accordance with existing guidance for financial liabilities. Do you agree with the proposed approach taken by the boards? If you disagree, please explain why. My initial reaction to reading Chapter 5 was to leave the accounting method up to the preparer as would be the case for changes in other accounting estimates. After many hours of exploring variations of Example 3, I became convinced that specifying the method is a good idea because of the extreme variations that could result from optional choices between the prospective, retrospective and catch-up methods. In some cases, application of the prospective and retrospective methods results in negative total interest expense being recorded over the life of the lease. Both of these methods would tend to distort any ratio analysis that uses interest 1 Unfortunately, leases with contingent rentals will probably need re-measurement frequently every time actual contingent rent differs from expected. So let s keep that as simple as possible as discussed under Question 21.

8 Comment Letter from T. Gordon July 16, 2009 Page 8 expense to analyze liquidity and debt coverage. As shown in my Appendix A, I ve carefully weighed the pros and cons of the three methods in settings where the lease term increases (like Example 3) and where the lease term decreases. It is exceedingly difficult to handle the prospective and retrospective methods when estimated lease term decreases since technology like Excel and calculators were not designed to compute negative interest something that makes no sense in any real world setting! As an instructor, I would also evaluate these methods as very hard to teach. My laboriously reached conclusion: catch-up methods using the original interest rate should be the only permitted method. As mentioned earlier and assuming that I ve properly implemented the catch-up method as intended by the Boards, adoption of the alternative catch-up method illustrated in my Appendix B could ameliorate the identified problems with little increase in complexity unless revisions become very frequent. 12. Some board members think that for some leases the decrease in value of the right-of-use asset should be described as rental expense rather than amortisation or depreciation in the income statement. Would you support this approach? If so, for which leases? Please explain your reasons. I had a hard time figuring out exactly what the question means but I assume it goes along with Example 2. The approach in this example would book the rent payment as rent expense while amortizing both the obligation and the right-to-use asset by the same amount and recording no interest expense. While I can see that this a best of both worlds approach, it would only be reasonable for what would now be called an operating lease. However, retaining a distinction between operating and finance leases would add unnecessary complexity to the lessee accounting standard (although I m becoming convinced that it is probably important to retain the distinction on the lessor side). My main objection to the Example 2 approach: a lease is often, if not always, a financing decision. The company could purchase the asset instead of leasing it. Therefore, it makes sense to divide the lease payments into principal and interest components so that the accounting is consistent between the two financing mechanisms. On the other hand, I am not necessarily opposed to the use of a mortgage method of depreciation which produces a declining amount of depreciation expense. A lessee can arrive at the exact same constant effect on the income statement by adopting this depreciation method. However, the expense components should be labeled correctly! If companies are serious about wanting this even impact on the bottom line, they could adopt this currently obscure mortgage-method. 2 I see no particular problem with permitting this method along with all the others like straight-line or double-declining balance that companies already have available. The complexity of the method might naturally limit its use given that it would probably not be acceptable for tax returns. However, when I prepared the third example in Appendix E for a lessor, I found it to be as easy as pie! Adopting mortgage-method depreciation while reporting interest expense, the Example 2 journal entry at the end of the first year would be: 2 As I understand it, this depreciation method was often used by regulated utilities. The material is no longer covered in intermediate accounting texts even though it is an excellent management accounting tool since it facilitates the evaluation of subunits on a return on investment basis consistent with capital budgeting decisions.

9 Comment Letter from T. Gordon July 16, 2009 Page 9 End of Year 1, Example 2 (page 25 in PV Leases): debit credit Interest expense 11,180 Obligation for lease payments 23,820 Cash 35,000 Depreciation expense 23,820 Accumulated depreciation 23,820 Note that the impact on the profit and loss statement is equal to the required rental payment on an annual basis (and therefore effectively equivalent to having an operating lease on the profit and loss statement with a capital lease on the balance sheet. 3 But let s call the expenses by the right name! Chapter 6: Leases with options 13. The boards tentatively decided that the lessee should recognise an obligation to pay rentals for a specified lease term, ie in a 10-year lease with an option to extend for five years, the lessee must decide whether its liability is an obligation to pay 10 or 15 years of rentals. The boards tentatively decided that the lease term should be the most likely lease term. Do you support the proposed approach? If you disagree with the proposed approach, please describe what alternative approach you would support and why. The most-likely lease term should work and the weighted probability method of computing the most-likely lease term is probably an acceptable alternative for situations where informed probabilities are feasible. However, I believe that the lease term used should be one that is actually possible. For example, the expected lease term for Example 5 turns out to be years so it should be rounded up to 15 since the lease must be 5, 10, 15, 20 or 25 years. Most of the time, management will make its best guess, without any explicit probabilities available. If a company has sufficient information to apply probabilities to the various possible lease terms, then it should be permitted to use that analysis to help it selected its most likely length for the lease. There doesn t seem to be a specific question related to Paragraph 6.41 so I m going to make comments here. According to this paragraph, the Boards have tentatively decided that the lessee s intentions and past practices would not be considered when determining the lease term. I m having a hard time seeing how this statement is consistent with the examples included in the PV-Leases document. As a case in point, where would the probabilities in Example 5 come from if not derived from lessee specific experience? The entity is described as a mature business with experience in expanding to new locations. If the Boards decide to exclude lesseespecific expectations regarding renewals, I m guessing that most leases will be capitalized at a value based on the minimum lease. In other words, we will end up with a PVMLP just like have 3 Of course, over the lease term, any approach will have equality between total depreciation expense plus total interest expense and total cash flows for rental payments. So, one might consider this mortgage-method depreciation simply a variation of other accounting methods that could be used to recognize lease obligations and right-to-use assets.

10 Comment Letter from T. Gordon July 16, 2009 Page 10 under current standards, with minor differences mostly limited to contingent rentals and guarantees of residual value. Is that what the Boards really want? Of course, it is probably what is going to happen anyway since the lessee s bias will be to minimize the lease obligation whenever possible. 14. The boards tentatively decided to require reassessment of the lease term at each reporting date on the basis of any new facts or circumstances. Changes in the obligation to pay rentals arising from a reassessment of the lease term should be recognised as an adjustment to the carrying amount of the right-of-use asset. Do you support the proposed approach? If you disagree with the proposed approach, please describe what alternative approach you would support and why. Would requiring reassessment of the lease term provide users of financial statements with more relevant information? Please explain why. Periodic reassessment on the basis of new facts or circumstances is a reasonable approach for lease accounting (particularly if the retrospective and prospective methods of implementing the change are prohibited). As stated elsewhere, I believe that the reassessment should always be at the original discount rate. For some situations like minor changes in expected cash flows for a guarantee of residual value, the reassessment would not be particularly relevant for financial statement users, particularly in the early years of the lease when substantial uncertainty exists about future values of used assets. Accordingly, revisions should be avoided unless have a substantial affect. In fact, I discovered that leases with contingent rentals are going to be troublesome there is no way that I can see that one could avoid adjustments to at least the lease obligation for even minor differences between expected and actual contingent rentals. In fact, changing projections for all future contingent rentals to match the actual amount incurred for the most recent period would not be any more difficult than leaving future rentals at the originally expected amount. However, the Boards could keep the process a bit simpler if the right-to-use asset is not revised each time a contingent rent payment differs from the expected amount. (Please see my Appendix C for a complete example with additional comments). If the right-of-use asset is not adjusted, the total cash flows over the lease term would be equal to the total depreciation expense plus total interest expense plus total gains or losses recognized over the lease term. Given the frequent ups and downs that might be expected for contingent rentals based on sales, it is likely that the gains and losses will cancel out over the lease term. I d like to suggest the Boards consider permitting the use of my retroactive catch-up method for dealing with situations where the expected lease term or payments under the lease term change from the original estimate. In arriving at my new method, I first spent a good bit of time on Examples 6 through 8 trying to understand the intended accounting impact. In addition, I played around with variations of Example 3 that had both increased and decreased lease terms. My recommended method is a slight variation of the catch-up method but has certain advantages. With my catch-up method, the original present value of expected lease payments (PVELP) liability amortization and right-to-use asset depreciation schedules are converted to the revised PVELP liability amortization and right-to-use depreciation schedules at the date when it becomes apparent that the lease term or total cash flows will be substantially different than the original estimate (see my Appendix B for an example). The main difference in my approach is that interest expense is adjusted so that, by the end of the actual lease term, the total interest expense and total depreciation expense recognized will be equal to what would have

11 Comment Letter from T. Gordon July 16, 2009 Page 11 been recognized had the actual lease term and actual payments been known at inception. Thus my method is retroactive in that is the if only we had known approach to a change in accounting estimate. This alternative approach is consistent with accounting standards regarding a change in accounting estimate. The initial value of the right-to-use asset is an estimate based on the expectations regarding payments to be made under the terms of the lease. When the estimate changes, it does not mean that the right-to-use asset has increased or decreased in value. Rather, it means that our original estimate of the value was incorrect. Therefore, we make changes to the current and future periods to correct the erroneous original estimate. As discussed in my response to Question 7, hange in accounting estimate language should be used in the final standard. Much of the language in the PV-Leases document seems to imply that the right-to-use asset suddenly became more or less valuable as projected payments change. This language is flawed in my opinion since the option to renew or purchase the asset existed when the lease was signed. The original measurement did not find truth and neither does the re-measurement. We are merely revising an estimate; therefore, the accounting should be consistent with existing guidance on changes in accounting estimates. 15. The boards tentatively concluded that purchase options should be accounted for in the same way as options to extend or terminate the lease. Do you agree with the proposed approach? If you disagree with the proposed approach, please describe what alternative approach you would support and why. After doing lots of examples, I am forced to agree. A purchase option, like a renewal option, is either exercised or not. Therefore it should be included in the minimum lease payments when it is more likely than not to occur. However, there may still be room for guidance regarding the inclusion of a bargain purchase option. However, a bargain purchase option should no longer trump a bargain renewal option (as it does under current US GAAP). For example, the renewal option might be $5 per month for a year and the purchase option might be $100. If the purchase option were a bargain, current rules would have the lease term always end at the date the bargain purchase option becomes available even though it would be cheaper to renew for a year and more logical is only a year of useful life remained. Chapter 7: Contingent rentals and residual value guarantees Contingent rentals 16. The boards propose that the lessee s obligation to pay rentals should include amounts payable under contingent rental arrangements. Do you support the proposed approach? If you disagree with the proposed approach, what alternative approach would you recommend and why? I agree. In reaching this conclusion, I carefully considered variations in Example 9 including both the proposed FASB and IASB approaches to initial and subsequent measurement. My initial reaction was a preference for current GAAP which basically ignores most contingent rentals as of the inception of the lease. However, I can see that abuses would arise if leases were structured to have a higher percentage of contingent rentals as compared to fixed rental payments. So I believe we do need to include an estimate of contingent rental payments in the computation of the initial value for the obligation to make lease payments and the right-of-use asset. The initial estimate could be based on the current level of sales or usage but that

12 Comment Letter from T. Gordon July 16, 2009 Page 12 wouldn t always be available. So an estimated most likely amount should be included in the computation of PVELP. I think that either a best estimate or a weighted probability expected amount would be acceptable. I don t think the weighted probability approach should be required but it also should not be precluded. 17. The IASB tentatively decided that the measurement of the lessee s obligation to pay rentals should include a probability-weighted estimate of contingent rentals payable. The FASB tentatively decided that a lessee should measure contingent rentals on the basis of the most likely rental payment. A lessee would determine the most likely amount by considering the range of possible outcomes. However, this measure would not necessarily equal the probability-weighted sum of the possible outcomes. Which of these approaches to measuring the lessee s obligation to pay rentals do you support? Please explain your reasons. I think either approach should be permitted. If reasonable probabilities are available for a range of outcomes (more than mere guesses to implement an accounting requirement!) then the lessee should be able to use the expected value technique to estimate the most likely contingent rentals. However, the method should not be required. Note that a symmetrical probability distribution will produce the middle number anyway something I accidentally discovered in making up an example for an asset retirement obligation! 18. The FASB tentatively decided that if lease rentals are contingent on changes in an index or rate, such as the consumer price index or the prime interest rate, the lessee should measure the obligation to pay rentals using the index or rate existing at the inception of the lease. Do you support the proposed approach? Please explain your reasons. This seems consistent with current GAAP and therefore it would be okay with me. However, one could estimate regular increases in price indices and that could be used to forecast an increasing series of payments. Conceivably, that could give rise to a more accurate initial estimate of the value of the right-to-use asset. The same cannot be said for contingent rentals based on interest rates like LIBOR. These change frequently with economic conditions and move both up and down. To keep it simple, therefore, I think it might be wise to always use the current index to determine the initial value of the PVELP (right-to-use asset and obligation). Please note that an index-based estimate of future contingent rents will be just as subject to error as any other type of estimated contingent rents such as those based on sales or usage. Accordingly, one can expect every period adjustments to the lease obligation, even when the projected future amounts have not changed. This is another argument for the FASB rather than the IASB approach discussed as described in Paragraphs 7.31 and See example in my Appendix C which more fully demonstrates the complications and also suggests an even easier alternative. 19. The boards tentatively decided to require remeasurement of the lessee s obligation to pay rentals for changes in estimated contingent rental payments. Do you support the proposed approach? If not, please explain why. Contingent rentals (in many cases like the example in my Appendix C) will almost never be exactly the amount predicted at inception. This raises a number of severe complications for the subsequent accounting for leases with contingent rentals. I suppose that one could simply have

13 Comment Letter from T. Gordon July 16, 2009 Page 13 a gain/loss account for contingent rentals that differ from the expected amount (with no adjustment to either the liability or the asset). However, I believe the Boards are leaning towards revision in at least the lease obligation. As far as I can tell, revisions in the lease obligation will be required with every payment that differs from the originally expected amount (that means adjustments with essentially every payment!) In addition, differences in the past may lead to different expectations about future contingent rentals and therefore cause additional revisions in the estimated liability because of changes in future expected payments. I would like to see it made plain that changes to the right-to-use asset should only be triggered by major differences in projected contingent rentals. Otherwise, the accounting will be too complex even though the added complexity will not make the financial statements much more meaningful to users. Instead of changes to just the liability account, the Boards could consider simply accounting for insignificant differences between expected and actual contingent rentals via a gain or loss account on the income statement. Please see the end of my Appendix C example for an illustration. 20. The boards discussed two possible approaches to recognising all changes in the lessee s obligation to pay rentals arising from changes in estimated contingent rental payments: (a) recognise any change in the liability in profit or loss (b) recognise any change in the liability as an adjustment to the carrying amount of the right-of-use asset. Which of these two approaches do you support? Please explain your reasons. If you support neither approach, please describe any alternative approach you would prefer and why. After testing out both approaches on variations of Example 9, I have a strong preference for the FASB approach (a). This approach was relatively easy to implement with a simple modification to the lease amortization schedule. In contrast, it took me quite awhile to get something working to illustrate the IASB approach (b). In my example, I have differences between actual and expected (without changing future amounts) and also a change in future amounts triggered by a second year of differences between expected and actual contingent payments. For this analysis, I used the catch-up method, as I understand it. Depreciation was simple straight-line, no salvage value. However, the IASB approach (b) has to adjust depreciation expense each time the right to-use asset is adjusted. Therefore, this second approach is more complicated and time consuming. (It was even harder when I tried to apply my alternate catch-up method.) The conclusion I reached based on my examples: many adjustments will be relatively immaterial and of little interest or benefit to financial statement users. From my multiple hours of fiddling with the two approaches, my preference for (a) is exceedingly strong! In fact, I d like to see something even simpler that could be used for immaterial differences between estimate and actual. Certainly, major changes in estimates need to be recognized, probably in both the obligation and the right-to-use asset accounts. Please see my Appendix C example and discussion. Residual value guarantees 21. The boards tentatively decided that the recognition and measurement requirements for contingent rentals and residual value guarantees should be the same. In particular, the boards tentatively decided not to require residual value guarantees to be separated from the lease contract and accounted for as derivatives. Do you agree with the proposed approach? If not, what alternative approach would you recommend and why?

14 Comment Letter from T. Gordon July 16, 2009 Page 14 I think these two items could easily be approached in the same way: for initial measurement, use the most likely amount to be paid (with the weighed probability approach neither required nor precluded) and for subsequent measurement, using approach (a) in Question 20 to adjust the lease obligation but not the right-to-use asset. I see no reason for the residual value guarantee to be separated from the rest of the lease contract, particularly if separation triggers accounting for the guarantee as a derivative (shudder!). I created an Example 9-RV (similar to Example 9 in the PV-Leases document but with a residual value guarantee and no contingent rentals) to test out the two approaches for handling changes in estimated residual value guarantees. The process was much easier than the one for contingent rentals because it is less likely to require frequent revisions. Determining the estimated difference between the residual value of the leased asset and the guaranteed amount can be calculated using the weighted probability approach, if reasonable probabilities are available. It isn t that complicated! The important complications would result from frequent adjustments to the right-to-use asset and depreciation expense if the IASB approach (b) in Question 20 becomes required. With respect to having a different treatment for changes in lease term versus changes in estimates of originally uncertain dollar amounts (contingent rents and residual value guarantees), it is true that both changes logically affect the original estimate of the initial value of the right-to-use asset and therefore the lease obligation. The justification for the different approach is that the cash flow changes are determined by uncertainty of dollar amounts for contingent rentals and residual value guarantees but the cash flow changes for changes in the lease term are related to future decisions as to whether options will be exercised. In addition, changes in expected lease term should be relatively infrequent, especially as compared to the differences between actual and expected contingent rentals that may occur with every single payment! I would not be opposed to handling changes in projected guarantees of residual value in the same way we would use for changes in lease term it would be perfectly logical and the present value of cash flows to be paid in the distant future would rarely be material enough to trigger adjustments. However, the frequency of changes that are inherent in accounting for contingent rentals argues for an exception to the adjust every period rule. Therefore I strongly support differential treatment for at least contingent rentals. Chapter 8: Presentation 22. Should the lessee s obligation to pay rentals be presented separately in the statement of financial position? Please explain your reasons. What additional information would separate presentation provide? I believe that lease obligations (as accounted for per the PV-Leases document) are substantially different from most financial liabilities and therefore need to be separately presented in the financial statements. First, financial liabilities generally specify fixed payments on known dates. In the case of leases, some portion of the payments may be contingent on sales, usage or various indices. In addition, we will be using most likely lease terms rather than contractually known payment dates. If financial liabilities have guarantees equivalent to residual value guarantees or something equivalent to purchase options, those elements are generally separated out as derivatives (which I wouldn t want for leases!). If the lease

15 Comment Letter from T. Gordon July 16, 2009 Page 15 obligations are not separately displayed on the face of the financial statements, the amount should be disclosed in the notes to the financial statements along with the related carrying amount of right-to-use assets. As mentioned earlier (see Question 9), for purposes of disclosure, fair values could be readily computed at with a Level 3 approach if that information were deemed essential to financial statement users. However, I m personally of the opinion that the value would be pretty soft and therefore not very meaningful. In addition, I m opposed to commingling interest on leases with other interest expense particularly if the prospective or retrospective methods are permitted for implementing remeasurement of lease obligations and right-to-use assets when expectations about cash flows change. As shown in my Appendix A, these methods would severely distort the annual relationship between interest and financial liabilities since the rates may be excessively high or low (even negative). 4 Since many ratios use interest expense in computing debt coverage, the financial statement user should have the choice to include or exclude leases from the analysis. Currently, we have schedules listing the rentals due over the next five years (at least in US GAAP). I would like to see this schedule maintained and expanded to clarify which payments are fixed and which are estimates of contingent rentals. It might look something like this: Projected payments under leases Minimum fixed rentals under existing contracts Planned payments under unexercised renewal and purchase options Estimated contingent rentals and residual value guarantees , , Thereafter 1,510 1, ,520 3,050 4, Less interest 3,290 Lease obligation 4,230 4 The catch-up method described in the PV-Leases document would also result in greatly increased or decreased amounts for interest expense in the periods following an adjustment. Even my proposed retroactive catch-up method would potentially distort interest expense although only in the period when an adjustment is made. See the example in my Appendix B.