Content area
Lease contracts written in 1994 in the US have been estimated at over $140 billion. The amount of leasing activity continues to grow, particularly operating-type leases which provide a source of off-balance sheet financing. However, a recent publication by an international group of representatives from the FASB and 6 other national and international accounting standard setting bodies suggests that lease accounting should require all lease contracts to be capitalized as assets and liabilities. A study cites evidence that suggests the income statement effects may be material, and illustrates how to estimate the impact of constructive capitalization of operating leases on both operating income and net income.
Full text
SYNOPSIS: Lease contracts written in 1994 in the U.S. have been estimated at over $140 billion (London Financial Group Ltd. 1996). The amount of leasing activity continues to grow, particularly operating-type leases which provide a source of off-balance sheet financing. However, a recent publication by an international group of representatives from the FASB and six other national and international accounting standard setting bodies suggests that lease accounting should require all lease contracts to be capitalized as assets and liabilities (McGregor 1996). This suggestion has also been made by the Association for Investment Management and Research (AMIR) in a December 1993 white paper.
A previous Horizons paper by Imhoff et al. (1991) illustrated how to constructively capitalize operating leases. However, this prior study focused exclusively on the balance sheet effects for a single period, and assumed the income statement effects were negligible. The current study cites evidence that suggests the income statement effects may be material, and illustrates how to estimate the impact of constructive capitalization of operating leases on both operating income (before interest expense) and net income. Understanding both the income statement and balance sheet consequences will likely become increasingly important as the sentiment that operating leases should be capitalized continues to gain favor with users and standard setters.
Key Words: Operating leases, Off-balance sheet financing.
INTRODUCTION
An earlier paper in this journal reported on methods of estimating the off-balance sheet assets and liabilities represented by a firm's long-term operating leases (Imhoffet al. 1991) (hereafter ILW). This process, referred to as the constructive capitalization of operating leases, focused on the single-period balance sheet effects of restating operating leases as capital leases. The paper's focus was to illustrate the impact of adjusting the assets and liabilities on two key financial ratios: return on assets (ROA) and debt to equity (D/E). However, in doing so, the adjustments to both net income and operating income that would have resulted from the constructive capitalization of operating leases were assumed to be zero. The accuracy of this simplifying assumption is challenged by subsequent research demonstrating a potentially significant effect on both net income ("bottom line income") and operating income (income before interest expense and unusual items) from the constructive capitalization of operating leases. For example, ILW (1993) found a median decrease to net income of 22 percent among sample firms in the airline industry, and a median increase in operating income of 34 percent.1
This paper illustrates how to estimate the effect of constructively capitalizing operating leases on both net income and operating income, and how to identify the settings in which these income statement effects will be material. These adjustments have the potential to significantly alter a number of important measures used for predicting, comparing and evaluating the performance of the firm and its managers, particularly: Return on Assets (ROA); Return on Equity (ROE); Price to Earnings multiples (P/E); and various earnings measures. Current research and practice in which accounting data are used to value the firm may be improved by constructive capitalization of a firm's operating leases (see, for example, Frankel and Lee 1995; Imhoff and Lee 1995). As a result, understanding the income statement effects of constructive capitalization is a relevant concern for both teaching and research, related to operating leases in particular and off-balance sheet financing in general.
JUSTIFICATION FOR CONSTRUCTIVE CAPITALIZATION
Accounting policy makers have often faced the difficult policy choice of deciding between formal recognition and additional disclosure of economically relevant business transactions. A prospectus, titled "Disclosure Effectiveness," was issued by the FASB on July 31, 1995, in an effort to stimulate discussion, comment letters, and academic research on disclosure versus recognition. It seems inevitable that some transactions having an economic effect on how investors and others view the firm are not formally recognized in the financial statements but are instead depicted in the footnotes thereto. For example, the FASB recently decided to require only the footnote disclosure of the estimated value of executive stock options rather than full expense recognition in the income statement, despite the economic rationale for the latter alternative. Because accounting policy making is, at least in part, a political process (demonstrated impressively in the case of stock options), the accounting standards do not always recognize the substantive economic effects of business transactions.
Perhaps one of the longest-running controversies in which accounting standards are widely acknowledged to fall short of capturing the economic substance of a relatively common transaction is in the area of operating leases. Since long before the origin of either the FASB or the APB (Accounting Principles Board), finance texts have been suggesting ways to estimate the amount of debt represented by uncapitalized long-term lease commitments (Cottle et al. 1988). The resources and related commitments associated with operating leases are most often instrumental in generating revenues and profits. Hence, while they are not recognized as assets or liabilities in GAAP-based financial statements, their economic significance is often estimated and reflected in the way outsiders view the firm. Evidence from the accounting literature clearly suggests that the economic implications of off-balance sheet assets and/or liabilities such as operating leases, are not ignored by investors and creditors (e.g. ILW 1993; Bowman 1980; Dhaliwal 1986). Therefore, markets suggest that adjusting GAAP-based financial statements for such unrecorded assets and/or liabilities better approximates the economic position and performance of the business entity. Fortunately, disclosures in the notes to the financial statements frequently provide information that facilitates making the economic adjustments for those users who feel they are relevant and appropriate.
Note that the remainder of the paper is based on the assumption that the user is interested in estimating the complete financial statement effects of constructively capitalizing operating leases. This approach is consistent with the December 1993 position paper issued by the Association for Investment Management and Research (AIMR), and more recently supported in a Special Report, Accounting for Leases: A New Approach, (July 1996) by an international group comprised of senior members of standard setting bodies of Australia, Canada, New Zealand, the United Kingdom, the United States and the International Accounting Standards Committee (McGregor 1996). Partial adjustments to only the balance sheet (as in ILW 1991) would violate the clean surplus relation assumed by most contemporary valuation and financial reporting models (e.g., Ohlson 1996). Similarly, standard profitability and performance evaluation ratios such as return on assets or return on equity would be internally inconsistent through the use of different accounting measurement devices in the numerator and denominator. While it is commonly assumed in finance and accounting research that long-term operating lease commitments are effectively equivalent to recorded assets and liabilities (Sharpe and Nguyen 1995; Bowman 1980; Ely 1991; Abdelkhalik 1981; Imhoff and Thomas 1988), some would disagree. To the extent that one believes only a portion of the operating leases represent unrecorded assets and liabilities, the procedures illustrated here are relevant to obtain consistent balance sheet and income statement treatment of the those leases the analyst does wish to constructively capitalize. However, since we know of no method for determining how to distinguish operating lease cash flows that should or should not be constructively capitalized, we proceed under the assumption that they all represent off-balance sheet debt and assets.2
ESTIMATING BALANCE SHEET EFFECTS
In the earlier Horizons paper (ILW 1991), the balance sheet effects of constructively capitalizing operating leases were illustrated in some detail. BecaUse the income statement effects described in this paper are necessarily linked to the balance sheet adjustments, the latter are briefly revisited here, with additional focus on the key assumptions supporting these estimates. However, our focus in this paper is on illustrating the income statement consequences of constructive capitalization, and it may be necessary for the reader to refer to the earlier paper to obtain the more complete justification of some of the balance sheet adjustments made here when illustrating the income statement consequences of interest.
To illustrate the general concepts underlying constructive capitalization, we use a numerical analysis of Southwest Airlines throughout most of the paper. Table 1 provides a brief profile of the reported financial statements of Southwest Airlines for the period 1990 to 1994. Table 2 provides a summary of Southwest's footnote disclosures pertaining to operating leases for the years 1994 (panel A) and 1993 (panel B). SFAS No. 13 requires firms to disclose their minimum future cash flow commitments. The estimation of the offbalance sheet liability represented by operating leases calls for computation of the present value of these future cash flows, referred to as PVOL (present value of operating lease commitments). To compute PVOL, two fundamental assumptions are required:
1) The appropriate interest rate for discounting the minimum cash flows.
2) The assumed pattern and duration of cash flows occurring more than five years into the future (since they are reported as a lump sum).
Interest Rate
The original Horizons paper used a ten percent interest rate to discount future cash flows in computing PVOL. This rate has also been used by Standard & Poors (S&P) in estimating PVOL for standardized bond rating review purposes.3 Subsequent research (ILW 1993) used firm-specific interest rates but did not elaborate on how those rates were obtained. The following discussion provides two methods for estimating the interest rate to be used in calculating PVOL.
The appropriate interest rate from a conceptual point of view is the weighted average, implicit interest rate for each firm's portfolio of operating leases which is the weighted average of the marginal interest rates in effect when the leases were originally signed.4 If we rely on GAAP for guidance here, the rate suggested for capital leases is the lower of the incremental borrowing rate or the lessor's implicit rate of return on the lease, if known. The estimation of the interest rate using the conceptual approach also calls for some consideration of the average age of the lessee's portfolio of operating leases and the historical interest rates relevant for each lessee. Note that the higher the interest rate, the lower will be PVOL.
One relatively easy way to estimate the appropriate interest rate is to consider the implicit rate in the firms capital leases.5 While some firms voluntarily disclose the weighted average interest rate for their capital leases, it is possible to estimate this rate even when it is not disclosed. For example, in panel A of table 2, Southwest Airlines reports a capital lease payment of $26,282 scheduled for 1995. The footnote also reveals that the current portion of the $195,756 capital lease obligation is $9,542. This $9,542 is equivalent to the principal portion of the $26,282 payment for 1995, implying 1995 interest expense of $16,740 ($26,282 - $9,542). Interest of $16,740 on the entire present value (principal) of $195,756 implies an interest rate of about 8.5% ($16,740 / $195,756 = .0855). Because somewhat more ownership risk remains with the lessor in the case of operating leases, we might expect the interest rate for operating leases to be slightly higher.
Another way to estimate the appropriate interest rate is to consider the recognized debt, the recognized interest expense, and the footnote disclosures for long-term debt. The implied interest rate found by dividing interest expense by the book value of interest-bearing debt provides an estimate of the weighted average interest rate for all types of interestbearing debt. While activities such as the sale of trade receivables or conversion features of debt may confound this estimation, as long as interest expense is not net of interest income or capitalized interest, it provides a reasonable estimate.
Duration of Cash Flows
The duration of the future cash flows is somewhat more ambiguous than the interest rate assumption. Knowing the economic lives of the assets being leased helps in estimating the total number of future periods, but the operating leases often include leases for land plus structures (common examples include restaurants, hotels, supermarkets, retail clothing stores and aircraft hangars) in combination with leases for machinery and equipment. These two categories of assets have significantly different economic lives, making a portfolio of operating leases that includes both difficult to evaluate.
ILW (1991) suggest a procedure that takes the fifth future year's minimum cash payment and divides it into the "beyond five years" out total to approximate how many years the payments would continue at the level of the fifth year's payment.fi This estimate is then rounded up to at least the next whole year. In those cases where the total number of future periods is greater than 15, one or two more years might be added. While the actual pattern of annual cash flows beyond year 5 can not be observed, we expect the normal case would show some "decay" in the cash payments beyond year 5 just as they do for the first five years. As a result, we round up and suggest that a year or two be added to anticipate the effects of the decay. Our (unreported) efforts to predict this decay and compute the present value of the "thereafter" lump sum based on such a prediction failed to materially effect the estimation of PVOL. This is easily understood by considering the nearly trivial present value of cash flows that are 15 to 30 years into the future. Slight errors in the estimate of the number of future years will normally have an unimportant effect on the estimate of PVOL.7
Results for Southwest Airlines
The estimated present value of the minimum future operating lease payments, PVOL, provides the foundation for computing all of the other balance sheet adjustments. As explained in detail in an earlier paper (ILW 1991), the subsequent estimates of the unrecorded asset, deferred taxes and shareholders' equity are all related to the estimated unrecorded liability. Measurement of the unrecorded asset calls for an estimation of the weighted average age of the assets under operating lease, the remaining life of the operating leases (also required for measuring PVOL), and the method of depreciation. Since the unrecorded asset value at the end of the operating lease must be zero, salvage value is not an issue for the lessee, and should always be assumed to equal zero. For the Southwest Airlines illustration provided here, we use a nine percent interest rate and we estimate the unrecorded asset to be 75 percent of PVOL.8 While these "generic" interest rate and asset balance assumptions have important implications for the income statement effects illustrated in this paper, they will not always be the most appropriate assumptions to employ. The earlier paper (ILW 1991), which focused on these balance sheet measurement issues, provides more detailed guidance on the balance sheet consequences of operating leases. For illustrative purposes, we use these "generic" estimates for the unrecorded assets and use a 40 percent tax rate to depict the unrecorded balance sheet effects of operating leases for any given year as in figure 1.
Note this analysis assumes a 40 percent tax rate for the excess of the depreciation expense plus the interest expense associated with capital leases versus the rent expense recorded for operating leases. Table 3 provides a summary of these estimates for Southwest Airlines for each year in the period 1990-1994. The average remaining lease period estimated for each year ranges from 16 to 18 years.
Combining the table 1 balance sheet data with the table 3 PVOL estimates, we can see the adjustments for operating leases are material for Southwest Airlines, as they are for most major air carriers. Unrecorded assets represented by operating leases grew from about 25 percent of reported assets in 1990 to about 33 percent in 1994. Unrecorded liabilities have an even greater impact, with PVOL increasing from about 55 percent of 1990's total recognized liabilities to about 78 percent in 1994. The balance sheet effects of constructive capitalization (as reported in the lower part of table 3) are relatively easy to generalize. They systematically increase the amount of assets utilized to generate profits (as represented by the increased height of the adjusted bar each year), and they systematically increase the proportion of debt capital (vs. equity capital) used to finance the firm (as represented by the relative proportions of the upper and lower sections of the bars). The adjusted balance sheet measures depict both a larger and a substantially more levered, riskier entity.
ESTIMATING THE INCOME STATEMENT EFFECTS
While the income statement effects of constructively capitalizing operating leases are not always as significant as the balance sheet effects, they do have the potential to materially change some measures commonly used to predict, compare and evaluate a firm's economic performance. Two common performance measures are ROA and ROE. Also, PB and P/E are popular comparative market multiples that are usually materially affected by adjustments to net assets and earnings, respectively. Various earnings measures are the focus of most economic forecasts of firm performance as well as elements of the previous noted ratios.
To understand the impact of operating leases on the income statement, consider the comparison between operating and capital leases illustrated in table 4, panel A. This hypothetical example is based on a lease with the following characteristics:
Life of lease = 20 years
Implied interest = 10%
Annual Payment = $10,000/year each year-end
Down Payment = $0
Tax Rate = 40%
Assuming the common practice of "no initial lease payment" allows the analyst to consider the unrecorded asset and unrecorded liability to be equal at the inception of the lease. While the analysis is also applicable for operating leases where there is an initial payment upon signing the lease, the additional complexity and reduction in the generalizability of the examples makes them less attractive for illustration purposes. The present value of this lease at its inception is $85,136. This represents the present value of the asset's property rights under the control of the lessee. If the lessee treats this as a capital lease, it records an asset and a liability of $85,136. Straight-line depreciation expense of $4,257 ($85,136/20 years) is recorded each year, if capitalized. The first year's interest expense for the capital lease is $8,514 ($85,136 x ten percent). If the lessee treats the lease as an operating lease, rent expense simply equals the annual cash payment of $10,000.
Impact on Net Income
Panel B of table 4 shows how to convert net income from an operating lease situation to net income that would have been reported had the lease been capitalized. In this example, the net income under the capital lease treatment is $1,663 less than under the operating lease treatment. This $1,663 difference is the net income adjustment needed to constructively capitalize the operating lease. The higher reported net income combined with lower total assets and lower total liabilities recorded under operating lease treatment results in a more favorable depiction of both the firm's performance and its riskiness.
These net income adjustments may be more easily and directly computed by simply taking the change in the balance sheet adjustments to shareholders' equity from one year to the next. Table 5 illustrates this point by carrying the above example to a second year. Panel A of table 5 illustrates the lessee's Year 2 income statements under both operating and capital lease treatments. In panel B of table 5, the adjustments that would be made to the lessee's balance sheets for both years are provided. Note that the change in the retained earnings adjustment in panel B ($3,236 $1,663 = $1,573) is equal to the Year 2 difference in net income attributable to capital versus operating lease treatment illustrated in panel A ($30,000 - $28,427 = $1,573). As a result, we see that a series of balance sheet adjustments alone can provide the impact of constructive capitalization on net income. In particular, income statement details are not required to compute the impact of constructive capitalization on net income as long as comparative balance sheet adjustments are made. The change in retained earnings, along with other balance sheet changes, will also be affected by the trend in operating leasing. As the amount of nominal dollars committed to operating leases increases (decreases) the amount of additional unrecorded assets will increase (decrease) while the amount of the decrease in both unrecorded deferred taxes and retained earnings will increase (decrease). Because of the effects of general and specific inflation, even when the number of units of items (e.g., square feet of floor space or number of B747 aircraft) under operating leases remain stable, the nominal dollar commitments for operating leases tends to increase in most cases. (The lease payments for a replacement B747 are much higher than an old B747 whose lease is expiring.) For companies with real growth in operating leases, the increases in assets and decreases in deferred taxes and retained earnings are greater still. For those companies experiencing real decreases in their "assets" financed through operating leases, there will be a decline in assets, and the change in both deferred taxes and retained earnings (therefore, net income) may be positive. Empirically speaking, such companies are the exceptions, since the popularity of operating leases continues to grow.
Impact on Operating Income
Measuring the impact of constructive capitalization on operating income is more complex than measuring the bottom-line effects and draws on the details provided in tables 4 and 5 for our hypothetical example. As noted in the memo to panel A of table 4, operating income before interest expense but after taxes may be computed by adding back the aftertax cost of interest expense to net income. This measure of operating profit is, theoretically, the most desirable measure for computing return on assets (ROA) as well as operating profitability of the entity. It eliminates the effects of how the firm is financed from both operating income and return on assets (investment) and is an appropriate measure of performance used frequently for managerial compensation purposes and for comparisons across business entities or their various divisions.
The constructive capitalization adjustment to operating income requires the elimination of rent expense and the insertion of depreciation (but not interest) expense into the first subtotal illustrated in panel A of tables 4 and 5 (followed by application of the appropriate tax rate to this revised subtotal). However, as a practical matter, the depreciation expense that would have been reported had the leases been capitalized is rarely disclosed. Furthermore, depreciation expense would be difficult to estimate without knowing the future cash flows and estimated lease lives at the inception of the leases.
Fortunately, the equivalent adjustments may be obtained by the relatively simple computation shown in figure 2. Although the hypothetical example illustrated in figure 2 provides details of the income statements for both operating and capital lease treatments, these details are rarely provided in financial statement disclosures.9 Yet the computations illustrated here indicate how both net income and operating income may be adjusted for the constructive capitalization of operating leases using only the comparative balance sheet adjustments as described in ILW 1991.
Impact of Adjustments on ROA and ROE Performance Measures
Both the income statement and the balance sheet adjustments will influence the two most common relative performance measures: ROA and ROE. ROA permits the analyst to abstract away from how the firm (or business unit) is financed and to focus on the level of profit generated by a given set of productive resources. Moreover, ROA is often broken into its two components to emphasize that it is based on both profitability and efficiency, as shown in figure 3.
In figure 3, the first component measures profitability by the percentage of each dollar of sales (or revenues) that flows through to operating income after taxes but before interest expense and other non-operating charges and their related tax consequences. The second component measures efficiency by what is commonly referred to as the total asset turnover ratio. Performance goals established in terms of ROA place emphasis on both the profitability of operations and the amount of assets needed to achieve that profitability. ROA may be increased by requiring fewer assets to achieve the same level of profitability or by increasing profitability from a given quantity of assets. Because the "claimants" side of the balance sheet reflecting the proportions of debt and equity capital are ignored, comparisons of ROA measures may be made within and between firms without having to adjust for leverage. In other words, this measure of income is addressed to all claimants as opposed to just the residual claimants, who are the focus of measures of return on equity (ROE) and bottom line income.
The effects of constructively capitalizing operating leases on the two components of ROA are systematic: profitability always increases and efficiency always decreases. Profitability is affected by replacing rent expense with depreciation expense on the unrecorded asset. Since rent expense must include the equivalent of interest plus principal on the unrecorded operating lease asset and liability, it will always exceed depreciation expense.10 Hence, using operating leases to avoid recognizing capital lease assets and liabilities will always result in lower operating profits. However, the efficiency, or asset turnover, ratio will always be higher when operating leases are used instead of capital leases. The impact of constructive capitalization on ROA in total is ambiguous, since among other things it depends on the level of ROA before considering the operating versus capital lease choice. However, in firms where the nominal amount of cash flow commitments for operating leases are growing significantly, the reduction in ROA from constructively increasing assets will dominate the increase in operating income, resulting in a lower ROA after constructively capitalizing operating leases.ll In general, the greater the nominal dollar growth in operating lease commitments, the higher will be the impact on ROA from off-balance sheet lease financing activities.
The impact of constructive capitalization on ROE, like ROA, is somewhat ambiguous. While the capitalization of operating leases will systematically result in a smaller denominator (average shareholders' equity), the impact on net income in the numerator is unclear. If the operating leases are in approximately the first 65 percent of their weighted average total life (i.e., less than about two-thirds expired), capitalizing them will decrease net income.12 Since this is normally the case for most portfolios of operating leases, capitalization is usually expected to decrease net income. Hence, as with ROA, the impact of constructively capitalizing operating leases on ROE depends to a large extent on the magnitude of the unadjusted ROE ratio, and the size of the adjustment to net income relative to the size of the adjustment to average shareholders' equity.
In the previously considered hypothetical example illustrated in tables 4 and 5, the adjustments to ROA and ROE may be summarized as shown in figure 4.
The impact of selecting operating leases over capital leases in periods of rapid growth are best depicted by Year 1 capitalization adjustments for the hypothetical example. Notice the relatively high ratios of adjustments to both ROA and ROE. For ROA in Year 1, the ratio of the OI adjustment to the average asset adjustment was 8.52%. If ROA before considering these adjustments was less than 8.52%, the adjustments would improve ROA. Put another way, if ROA before considering leases was less than 8.52%, structuring incremental leases as capital leases would improve ROA. However, as we see in Year 2, the OI and average asset adjustments generate a much lower ratio of 4.38%. This lower rate is more representative of adjustments for actual firms and suggests that if unadjusted ROAs exceed 4.38%, the lessee will report higher ROAs if incremental leases are structured as operating leases.
The ROE effects for Year 1 are similarly very large; a negative 200 percent for our hypothetical example. While the Year 2 adjustments to ROE are more representative of what might be observed, they still result in a -64 percent ratio. An unadjusted ROE ratio in excess of 64 percent for Year two will improve with the adjustments, while unadjusted ROEs below 64 percent will deteriorate when operating leases are constructively capitalized. Since ROE is about ten percent to 20 percent for most firms, it is very likely that operating leases would be selected over capital leases (or other purchase arrangements) as a way to increase a firm's ROE. The effects of constructively capitalizing operating leases will usually result in a reduction in a firm's ROE, ceteris paribus, where operating lease commitments are significant and are growing on a nominal basis.
Application to Southwest Airlines
Using the approach described above, we illustrate the adjustments to net income and operating income for Southwest Airlines in panels A and C of table 6. These adjustments can be computed independently based on data for Southwest in tables 1 and 3.13
One important observation from comparing the adjustments to net income and operating income in panels A and C of table 6 is the potentially more significant and systematic downward effect on net income. This result is generalizable in that the constructive capitalization of operating leases will normally reduce bottom line earnings, or earnings to common shareholders-the focus of most analysts' forecasts. As a result, we might expect sophisticated analysts to employ economic adjustments to earnings in evaluating and forecasting a firm's earnings performance. Future research might examine the relevance of such economic adjustments to earnings in evaluating market reactions to earnings releases.l4 Such adjustments could also influence research on the persistence of earnings.
Panels B and D of table 6 provide comparisons of the unadjusted and adjusted ROE and ROA measures, respectively. Note that both ratios in all years are lower after constructively capitalizing operating leases. If Southwest's performance for either executive compensation or external evaluation purposes were to be based on ROE and/or ROA, decision makers might prefer to base those evaluations on the adjusted ratios for Southwest Airlines. The unadjusted ratios appear to systematically overstate the performance of Southwest Airlines each year from 1990-1994. For ROE, the percentage overstatement from using unadjusted ratios ranges from a low of four percent in 1993 ((16.17-15.48)/16.17) to a high of 86 percent in 1991 ((4.36-.59)/4.36). For ROA, the apparent overstatement from using unadjusted ratios ranges from a low of 27 percent (1993) to a high of 51 percent (1991). These differences are both material and systematic in the case of Southwest Airlines. However, we know that similar adjustments made to a larger sample of firms will yield both increases and decreases in the unadjusted ROE and ROA ratios (see ILW 1993, table 1).
Distortions from Ignoring Income Effects
In panels B and D of table 6 we also illustrate the partially-adjusted ROE and ROA ratios that would have resulted for Southwest Airlines if income statement effects had been ignored as in the prior paper (ILW 1991). Note that the effects of ignoring bottom line net income differences results in all of the partially-adjusted ROE ratios (panel B) being higher than the unadjusted ROE ratios, while the fully-adjusted ROE ratios (panel B) were all lower than the unadjusted ratios. In partial adjustments to ROE (where the income statement effects are ignored), the impact on shareholders' equity is always negative, reducing the denominator. As a result, ROE will always improve when only the balance sheet is adjusted. When net income (the numerator) is also adjusted, the impact on ROE is more ambiguous, and will depend on both the unadjusted ROE value (high or low) and the ratio of the adjustment to net income to the adjustment to average shareholders' equity. In the Southwest Airlines case, the reductions in net income are large enough to outweigh the reductions in shareholders' equity, providing an interesting example of how misleading adjustments might be when income statement effects are ignored.
For companies with significant growth in operating leases, such as Southwest Airlines, the income statement effects of operating versus capital lease treatments will be material, and ROE will tend to decrease with constructive capitalization. Put another way, reported ROE will tend to be higher if growth in operating activity is financed by operating leases instead of capital leases or other purchase arrangements.
The impact of adjustments to operating income for measuring ROA were not as severely affected as were the ROE measures. Panel D of table 6 reveals that the partially-adjusted ROA ratios were all lower than both the unadjusted and fully-adjusted ROA ratios. While the impact on operating income is, of course, positive in each year (see panel C of table 6), it is relatively small compared to the large increase in average total assets resulting from constructive capitalization. The adjusted ROA ratios are dominated by the increasing effect of the adjustments on average total assets and are uniformly lower after adjustments based largely on the off-balance sheet assets added to the denominator.
EVIDENCE ON THE INCOME EFFECTS OF CONSTRUCTIVE CAPITALIZATION
How sizable are the income effects of constructive capitalization for actual U.S. corporations? In this section we report some recent evidence to demonstrate that the income effects of off-balance sheet lease financing can materially alter impressions about the relative financial performance of firms.
Panel A of table 7 compares the performance of a single company, K-Mart Corporation, across two consecutive fiscal years, 1995 and 1994. Based on the recognized results of operations, it appears that K-Mart's profitability was substantially worse in 1995 than in 1994. Recognized profit margins based on after-tax operating income and after-tax operating return on assets (ROA) fell by 38 percent and 35 percent, respectively. However, after adjusting for the effects of operating leases on both the reported income statement amounts and total assets, the decline in these ratios is much less severe-12 percent and six percent, respectively.
The effects of operating lease adjustments on bottom line profitability measures are equally dramatic. ROE ratios based on reported net income show a 39 percent decrease between 1994 (7.06% ROE) and 1995 (4.29%). However, after adjusting for the effects of operating leases, ROE is actually nine percent higher in 1995 (5.99%) than in 1994 (5.51%). Moreover, the figure 5 summary of potential ROE comparisons demonstrates how ignoring the income effects of operating leases while adjusting only the balance sheet amounts (as in ILW 1991) would not have revealed the actual improvement in ROE performance.
Panel B of table 7 demonstrates that income statement effects of operating leases can also significantly affect cross-sectional financial statement analysis as well as the intertemporal comparisons illustrated in panel A. In panel B we compare results of operations for two retail grocery store chains, Weis Markets, Inc. (fiscal year ending December 25, 1993) and National Convenience Stores (fiscal year ending June 30, 1994). Relative to either total revenue or average total assets, Weis Markets' reported after-tax operating income appears to be substantially higher than National Convenience Stores. The unadjusted operating margin and ROA ratios are higher for Weis Markets by 179 percent and 69 percent, respectively. However, after adjusting for operating leases these differences are cut in half to only 84 percent and 29 percent, respectively.
Adjusting only the balance sheet and ignoring the income statement effects (as in ILW 1991) would, in this example, further distort the comparative analysis. For example, adjusting only reported assets in the ROA ratio would produce the comparisons shown in figure 6.
Adjusting reported ROA for only the offbalance sheet assets (as suggested in ILW 1991) substantially increases the difference in the companies' performance measures. However, this partial adjustment technique ignores the substantial income statement consequences of constructive capitalization (see footnote 1). The intercompany difference in ROA after fully adjusting for operating leases (29 percent) is only half as large as the unadjusted financial statement results would suggest (69 percent). Thus, in this example, partially adjusting ROA as in ILW (1991) is actually worse than making no adjustment at all!
The comparative ROE measures in panel B are also significantly affected by the operating lease adjustments. On an unadjusted basis, Weis Markets appears to have a nine percent higher ROE (10.29% vs. 9.43%). However, after adjusting for operating leases ROE is actually 35 percent lower for Weis Markets than for National Convenience Stores (10.12% vs. 15.50%). These results indicate that the effects of off-balance sheet operating leases on standard profitability measures can be both substantial in amount and unpredictable in direction. Ignoring the existence of operating leases or only partially adjusting financial performance measures as in ILW (1991) can provide materially misleading results. Although the examples in tables 7 may not be representative of all operating lessees, they do point out the potential dangers of ignoring the income statement effects of constructive capitalization.
IMPLICATIONS
Converting GAAP-based financial statements to more economically relevant information for purposes of predicting, comparing and evaluating firm performance often calls for adjusting recognized amounts for off-balance sheet assets and liabilities and their related consequences on income measurements. Examples include LIFO reserves and off-balance sheet post-retirement obligations. Here we illustrate how previously documented adjustments for the off-balance sheet financing implicit in a firm's operating leases can also be used to adjust common measures of corporate performance. Using only a series of balance sheet adjustments for operating leases, we provide a method of adjusting both operating income and net income to estimate their amounts had the firm's operating leases been treated as capital leases. In certain industries (including airlines, retail groceries, retail clothing chains, fast-food restaurant chains, hotel and motel chains, railroad and trucking), we find operating leases to be a material source of both productive assets and their related financing. In cases where the use of operating leases is increasing or decreasing rapidly, it is important to estimate both their balance sheet and income statement effects. The effects on financial statement amounts and ratios based thereon may be readily estimated using balance sheet adjustments for two or more consecutive years.
Prior research suggests that some decision makers appear to ignore the impact of off-balance sheet financing on commonly used measures of firm performance. For example, the management compensation literature has reported that measures of income (OI and NI) or income deflated by investment (ROE and ROA) are the most commonly used accounting data in determining management bonuses (Antle and Smith 1986; Holthausen et al. 1995; Healy 1985; Healy et al. 1987; Bloedorn and Chingos 1991). Yet it appears as though the effects of off-balance sheet financing on recognized accounting data, while material in their magnitude, are being ignored in setting management bonuses (ILW 1993). As illustrated here, commonly reported disclosures permit adjustments to accounting performance measures which could make them more economically relevant for decision making, including decisions regarding management compensation, as well as those concerning fundamental analysis of firm value. Moreover, the adjustment process illustrated here has potentially important implications for a wide body of research dealing with earnings prediction and measurement. Future empirical research may provide additional insight into the processes by which heterogeneous user groups rely on reported or adjusted balance sheet and income statement data in making economic decisions-including using the techniques described here to further investigate the role of off-balance sheet financing through operating leases in affecting evaluations of firm profitability and financial position.
Submitted June 1996 Accepted February 1997
Corresponding author: Eugene Imhoff Email: [email protected]
REFERENCES
Abdel-khalik, A.R. 1981. The Economic Effects on Lessees of FASB Statement No. 13: Accounting for Leases. Stamford, CT: FASB.
Antle, R., and A. Smith. 1986. An empirical investigation into relative performance evaluation of corporate executives. Journal of Accounting Research (Spring): 1-39. Association for Investment Management and Research (AIMR). 1993. Financial Reporting in the 1990s and Beyond. Charlottesville, VA: AIMR.
Bloedorn, J. D. and P. T. Chingos. 1991. Executive Pay and Company Performance. Scottsdale, AZ: American Compensation Association.
Bowman, R. 1980. The debt equivalence of leases: An empirical investigation. The Accounting Review (April): 237-252.
Cottle, S., R. F. Murray, and F. E. Block. 1988. Graham and Dodd's Security Analysis. New York, NY:
McGraw-Hill.
Dhaliwal, D. 1986. Measurement of financial leverage in the presence of unfunded pension obligations. The Accounting Review (October): 651-661.
Ely, K. M. 1991. Interindustry differences in the relation between compensation and firm performance variables. Journal of Accounting Research (Spring): 37-58.
Frankel, R., and C. M. C. Lee. 1995. Accounting valuation, market expectation, and the book-to-market effect. Working paper, University of Michigan.
Healy, P. M. 1985. The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics (April): 85-107.
, S. H. Kang, and K. G. Palepu. 1987. The effect of accounting procedure changes on CEO's cash salary and bonus compensation. Journal of Accounting and Economics (April): 7-34. Holthausen, R. P., D. F. Larcker, and R. G. Sloan. 1995. Annual bonus schemes and the manipulation of
earnings. Journal of Accounting and Economics (February): 29-74. Imhoff, E. A., Jr., and J. K. Thomas. 1988. Economic consequences of accounting standards: The lease disclosure rule change. Journal of Accounting and Economics (December): 277-310. , and C. M. C. Lee. 1995. The biotech industry. Case study, University of Michigan. and .1991. Operating leases: Impact of constructive capitalization. Accounting Horizons (March): 51-63.
, R. C. Lipe, and D. W. Wright. 1993. The effects of recognition versus disclosure on shareholder risk and executive compensation. Journal of Accounting, Auditing and Finance (Fall): 335-368. London Financial Group Leasing Report. 1996. World Leasing Yearbook. London: Euromoney Publications. McGregor, W. 1996. Accounting for Leases: A New Approach. Norwalk, CN: Financial Accounting Foundation.
Ohlson, J. A. 1990. A synthesis of security valuation theory and the role of dividends, cash flows, and earnings. Contemporary Accounting Research (Spring): 648-676.
Sharpe, S. A., and H. H. Nguyen. 1994. Capital market imperfections and the incentive to lease. Federal Reserve Board Working Paper.
Eugene A. Imhoff, Jr. is Professor at University of Michigan, Robert C. Lipe is Associate Professor at University of Colorado at Boulder and David W. Wright is Associate Professor at University of Michigan.
Copyright American Accounting Association Jun 1997