GATX Corporation 222 West Adams Street Chicago, IL 60606-5314 2013-270 September 13, 2013 Mr. Russell Golden, Chairman Financial Accounting Standards Board 401 Merritt 7 PO Box 5116 Norwalk, CT 06856 Mr. Hans Hoogervorst, Chairman International Accounting Standards Board 30 Cannon Street London EC4M 6XH United Kingdom File Reference No. 1850-100 Submitted electronically to director@fasb.org Dear Chairman Golden and Chairman Hoogervorst: GATX Corporation provides leasing and related services to customers operating rail, marine, and other targeted assets. GATX is a leader in leasing transportation assets and controls one of the largest railcar fleets in the world. We appreciate the opportunity to comment on the FASB s Proposed Accounting Standards Update, Leases (Topic 842) ( the ED ). We do not support issuance of the ED as a final standard because we believe it adds complexity and does not represent an improvement over existing guidance. We commend and appreciate the Boards efforts to address concerns with the 2010 ED, particularly with respect to determining the lease term and lease payments. However, we believe the current ED continues to present numerous conceptual and implementation issues that would be costly for preparers to address and would not provide users with decision-useful information. We have specific concerns regarding the ED s approach to lease classification, lessee accounting, and lessor accounting, which are explained below. Finally, we offer our recommendations for improving the standard. Lease Classification We disagree with the Boards proposed approach to lease classification because we believe it will impair comparability across asset classes, across entities, and across transactions within an entity. The distinction between leases of property and leases of non-property impairs comparability across asset classes. In our view, the accounting for leases should focus on the lease contract as the unit of account, not the nature of the underlying asset. If the Boards believe leases of certain types of assets deserve special treatment, then principles should be developed to support why that treatment is justified and how to identify the types of leases eligible for that treatment. Additionally, we believe comparability across entities would be impaired due to application of the consumption approach to lease classification, as entities would likely develop different interpretations of the new terms of insignificant, major part, and substantially all. We note the Boards did not include any principles-based guidance to assist companies in determining
whether a lease is expected to result in significant consumption of the underlying asset. The only proposed implementation guidance that is provided to indicate what is meant by insignificant is the example in paragraphs 842-10-55-62 to 63 (16.67% is more than an insignificant part of the total economic life and 27.8% is more than insignificant relative to the fair value of the leased asset). Combining this example with other non-authoritative guidance in ED 1, we infer that the Boards believe the new dividing line between Type A and Type B leases is somewhere between 8% and 16.7% of the economic life of underlying asset. We recognize that the Boards may be hesitant to provide a new bright-line threshold; however, in the absence of any principled guidance or additional context in which to evaluate the relative significance of asset consumption during the lease, we don t see any way for a company to objectively and faithfully apply the guidance other than by adopting (voluntarily or involuntarily 2 ) a formal policy whereby leases exceeding the predetermined consumption ratio would be classified as Type A and leases below the consumption ratio would be classified as Type B. Over time, we believe practice would evolve such that most entities would use a commonly accepted quantitative threshold for classifying leases. While this practice would improve comparability across entities, we fail to see how this represents an improvement over existing literature. Regardless of where the new line is drawn, we also believe that comparability across transactions within an entity will also be adversely affected. Otherwise identical assets may be leased on slightly different lease terms, leading to substantially different accounting treatment for economically similar transactions. We note that lessees and lessors in the railcar leasing market commonly attempt to position the expirations of their leases to manage their exposure to market disruptions in railcar supply and/or demand. For example, a lessee wishing to lease 150 railcars for an average of 5 years may enter into three 50-car leases with terms of four, five and six years. In addition, lease terms and rates in many equipment leasing markets are primarily determined by market conditions and can be highly cyclical. Assets in high demand tend to require longer lease terms while assets in weaker markets are generally leased on shorter terms. As a result, the same asset may move between the two lease models depending on the specific market conditions at the time of lease inception, impairing comparability over time at the asset level (see further discussion below). Lessee Accounting We continue to have concerns about the proposed lessee accounting model due to the requirements to recognize expense on a front-loaded basis (for Type A leases) and to reassess and remeasure the right of use asset and lease liability upon a change in expected lease term. We continue to disagree that most leases should reflect the front-loaded expense pattern of Type A leases for lessees. In our view, many (if not most) operating leases today are viewed as shortterm rental contracts, rather than financed purchases. As a result, we believe a straight-line expense recognition pattern should be the presumptive approach for leases of all assets, not just leases of property. Consistent Mr. Linsmeier s comments in paragraph BC358, we believe the 1 We note the ED also provides non-authoritative guidance of the Boards intended meaning of insignificant in paragraphs BC46 (42.9% consumption on a car lease is more than insignificant); BC47 (4% consumption on an office leases is insignificant); BC62 (Type B accounting approach is not appropriate for 20 year lease of an aircraft or vessel with an unspecified economic life); and BC125 (40% consumption on a truck lease is more than insignificant, 8% consumption on a railcar lease is insignificant, 16% consumption on a railcar may be more than insignificant). 2 We believe that the absence of authoritative guidance outlining what constitutes insignificant (either quantitatively or qualitatively) also exposes companies to auditor judgments (or potentially changing judgments) with no basis on which to support the company s own conclusions. 2
right-of-use asset represents a new kind of asset in accounting. As a result, we do not believe the accounting for the right-of-use asset should necessarily follow existing accounting requirements for physical or intangible assets. We continue to disagree with the requirement to reassess and remeasure the lease asset and lease liability upon changes in the lease term or lease payments. We believe the reassessment requirements would be costly for preparers to implement and maintain and would provide information of marginal value. Further, we have concerns with the proposed implementation guidance with respect to the evaluation of whether a significant economic incentive to exercise a renewal option exists. For example, paragraph 842-10-55-4(a) states a lessee should consider the contractual terms and conditions for the optional periods compared with current market rates. It is not clear to us whether this implies that a significant economic incentive would be assumed to exist in a case where the lease payments on the optional periods were below current market rates, but equal to projected market rates for the optional periods. Additionally, we are unclear how to apply the guidance in paragraph 842-10-55-5 that a change in market-based factors (such as market rates to lease a comparable asset) should not, in isolation, trigger reassessment. In practice, we believe it would be difficult to separate the market-based factors from the other factors noted in paragraph 842-10-55-4. We are also concerned that the inherently subjective nature of the estimates required would further impair the usefulness and comparability of the information provided. As leases represent contracts that are signed at a single point in time, we believe remeasurement should occur only upon a modification or revision of the agreement (including a decision to exercise an existing renewal option). We also believe the separate income statement and cash flow statement presentation requirements increase complexity (cost) without providing decision-useful information (benefits). Lessor Accounting In our view, the ED proposes a more fundamental and significant change for lessor accounting than for lessee accounting, despite the fact that few considered the lessor model to be deficient or broken. In the ED, lessee lease classification determines whether the expense would be recognized on a front-loaded or straight-line basis; but regardless of the expense recognition pattern, the right of use asset and the lease liability would be recognized on the balance sheet. On the other hand, lessor classification determines whether the lessor continues to recognize the asset or applies the partial sale accounting of the Type A lease approach and derecognizes the leased asset. This knife-edged approach would result in economically similar transactions being accounted for under drastically different methods and lessors with a mix of Type A and Type B leases would be required to report their financial results based on this split model. As noted above, lease terms and rates in the railcar leasing market are highly cyclical. Under the proposed approach, we believe assets will frequently move between Type A and Type B treatment depending market conditions at the time of lease inception. We evaluated our renewal activity (consisting of approximately 20,000 railcars annually) over the last three years and found that the majority of renewals in 2010 had a term representing less than 10% of the asset s economic life. In 2011, the majority of renewal terms represented between 10% and 15% of the economic life. In 2012, the majority of renewal terms represented more than 15% of the economic life. The fluctuation between Type A and Type B models would create significant swings in net income and the up-front profit recognized at lease commencement as assets coming off Type B leases (and typically depreciated on a straight line basis over the economic life) are likely to have lower net book values than assets coming off Type A leases (which would have a carrying value of the estimated residual value adjusted for any deferred profit). Not only would tracking and monitoring this switching between lease types be an extraordinary burden for lessors (as new information systems and processes would need to be developed and 3
implemented), but the bifurcated accounting and reporting model would adversely impact the usefulness of lessor financial statements. We disagree with the Type A approach for most equipment leases because we do not believe the business model of many equipment lessors (and most railcar lessors) is primarily about providing finance to lessees as stated in paragraph BC73(b)(1) The BC provides that from an economic perspective, and subject to market constraints, a lessor would generally price a lease to ensure that it obtains a desired return on its total investment in the underlying asset and also to recover an amount representing the portion of the underlying asset that the lessee is expected to consume during the lease term. This may be true in certain industries where specific lease terms are typically negotiated between a lessee and a lessor. However, in many industries, such as railcar leasing, both lease rates and lease terms are primarily driven by those market constraints. As a result, we rarely price a particular railcar lease in order to obtain a desired return or recover an amount of the asset the lessee is expected to consume. We believe our business model more closely resembles the model described in paragraph BC73(b)(2). We manage our assets throughout their entire economic lives in order to generate cash flows from the underlying asset on an ongoing basis, which typically involves periods longer than any one lease term. Lease terms typically range from three to ten years and the assets typically have economic lives of at least 30 years. As a result, railcars are generally leased out numerous times to one or more lessees and the residual value will generally far outweigh the lease receivable on a given lease. In addition, we generally perform the servicing on the assets while they are on lease not only for the benefit of the lessee, but also in order to protect our investment in the leased asset over its life, which generally extends beyond the lease term. This ongoing involvement in the servicing of the asset is generally included in the overall rental rate, as opposed to be being a separately priced service. We also believe lessor accounting results for Type A leases would be highly sensitive to estimates of the asset s residual value at the end of the lease term, which is generally an unobservable input subject to significant judgment. For a given fair value, low estimates of residual value would generally result in recognition of more profit upon lease commencement (as well as upon commencement of any subsequent leases) and recognition of lower interest income during the term of the lease. Higher estimates of residual value would result in recognition of less profit on lease commencement and greater interest income during the term of the lease. We believe that basing lessor accounting on subjective and unobservable estimates would be excessively costly for lessors to implement and would not result in decision-useful information. Conclusions In light of these concerns, we do not support issuance of the ED as a final standard. We have previously expressed support for the right-of-use approach for lessee accounting. However, the difficulty encountered by the Boards in developing an acceptable standard has caused us to question whether the right-of-use approach is operational or beneficial. In paragraph BC37, the Boards note the lack of a single, widely accepted view of the economics of a leasing transaction. This lack of consensus was also evident in the comments made recently by the FASB s Investor Advisory Council ( IAC ). In light of that feedback as well as recent research findings that users consider note disclosures similarly to recognized amounts, we believe the Boards should consider whether they could more effectively provide users with decision-useful information on leases through a disclosure-only project. However, if the Boards decide to continue the current project path, we recommend development of a model that is less complex and would provide more useful information for financial statement users. For example, we believe an approach consistent with the alternative views expressed by Mr. Linsmeier would be simpler to apply and would result in more decision-useful information. Alternatively, we would support an approach to classifying all leases based the 4
proposed lease classification guidance for leases of property (paragraph 842-10-25-7), which we understand is intended to be similar to the existing classification criteria in IAS 17, Leases. In addition, we would support inclusion of guidance identifying the business model of the lessor or the lessee s business purpose for entering into the lease as factors to consider when determining lease classification. We believe this approach would represent an improvement to the ED as it would eliminate complexity from a dual classification model and would be simpler to apply, given the similarity to the existing criteria, which is well-understood in practice. We appreciate the opportunity to share our views and are available to assist the Boards and provide any additional information that may be needed. Sincerely, /s/ William M. Muckian Senior Vice President, Controller and Chief Accounting Officer GATX Corporation /s/ Michael J. Maffei Director of Accounting Policy, Planning and Research GATX Corporation 5