Appendix

Appendix 4-AESTIMATING OPERATING LEVERAGE

W1

W2 APPENDIX 4-A ESTIMATING OPERATING LEVERAGE

W3

Appendix 4-BEARNINGS PER SHARE—ADDITIONAL ISSUES

Earnings per share is probably the most widely used indicator of corporate performance. Yetmost of those who use it do not understand how it is computed. Fewer still understand how itis affected by the issuance of convertibles, options, or other potentially dilutive securities. Inthe text we have outlined the procedures used in its calculation. In this appendix, we discusscomputational issues, disclosure requirements, and the few differences between US andIASB standards.

COMPUTATIONAL ISSUES

Weighted Average Number of Common Shares Outstanding

The denominator must reflect all stock dividends and stock splits effective during the periodand those announced after the end of the reporting period (but before the financial statementsare issued) as if they had been effective at the beginning of the reporting period. All prior pe-riods presented are restated for comparability.

Acquisitions

Shares issued in purchase method acquisitions (see Chapter 14) are included in the denomi-nator only for the period following the acquisition date. Similarly, only the postacquisitionresults of operations of the acquired firms are included in the numerator of the EPS computa-tion. Note that no restatement of prior periods is permitted for purchase method acquisitions.

The impact of the pooling method is quite different. Merged firms are considered com-bined entities for all years presented. The shares issued in the combination are assumed tohave been outstanding for all periods presented, and the results of operations for the twofirms are also combined for those periods in the EPS calculation.

Contingent Shares

Acquisitions and incentive compensation plans may require the issuance of commonshares if specific conditions, such as the passage of time, achievement of income levels, orspecified market prices of the common stock, are met. Securities whose issuance dependssolely on the passage of time are always included in the weighted average shares outstand-ing. Other contingent shares are included in the computation of basic and diluted EPS ifthe required income levels or market prices have been reached at the end of the reportingperiod.

When the issuance of contingent shares depends on the achievement of earnings targets,and when it is likely that those targets will be achieved, the computation of diluted earningsper share includes both the incremental shares and the level of income assumed to have beenachieved. These adjustments to the EPS measures are required even if the incremental sharesare to be issued at a later date.

W4 APPENDIX 4-B EARNINGS PER SHARE—ADDITIONAL SHARES

EPS Computations for Two-Class Securities

Some firms issue more than one class of common stock or have “participating” securitiesthat are entitled to share in the dividends paid on common stock. EPS computations for eachclass of nonconvertible1 two-class securities are based on an allocation of earnings accordingto dividends paid and participation rights in undistributed earnings.

Adjustments for Rights Issues

Both SFAS 128 and IAS 33 mandate the use of the ex-rights method in the computation ofbasic and diluted EPS for the bonus element (discount to market price prior to the offering)in a rights issue. Under prior US GAAP, the bonus element was ignored. The ex-rightsmethod recognizes dilution when rights are issued to buy shares below the current marketprice.

Impact of New and Proposed Accounting Standards

SFAS 144 (2001) broadened the definition of discontinued operations as discussed on pages54 and 275 of the text. This change means that, for firms disposing of unprofitable opera-tions, income from continuing operations will be higher than it would have been under prioraccounting standards. Because income from continuing operations is the “control number”used to determine whether options and convertible securities are dilutive, higher incomefrom continuing operations will result in more of these potential common shares enteringinto the computation of dilutive EPS.

In its proposed reporting for securities with characteristics of liabilities or equity or both(see Box 10-2 on page 338 of the text), the FASB intends to redefine the control number asincome from continuing operations attributable to controlling shareholders. Under currentGAAP, income allocated to minority or noncontrolling shareholders is deducted in comput-ing the income from continuing operations. Thus, companies with profitable majority-ownedsubsidiaries will report higher control numbers under this proposed standard. Again, morepotential common stock will be classified as dilutive securities.

INTERNATIONAL DIFFERENCES

As stated in the text, the FASB and IASB developed their new standards together. As a re-sult, there are few differences between the two. The most important difference is that USGAAP requires that EPS be reported for all components of net income. IASB GAAP re-quires disclosure of EPS only for net income; any other components of EPS reported, how-ever, must accord with the new standard.

Under SFAS 128, earnings from continuing operations is the “control number” used todetermine whether potential common shares are dilutive (see previous section). Thus, ac-counting changes, discontinued operations, and extraordinary items do not affect determina-tion of the dilutive effect. Under IAS 33, net income is the control number. Given the highfrequency of extraordinary items and other differences between earnings from continuingoperations and net income, it is likely that, for some firms, the dilutive effect will be differentdepending on whether they use US GAAP or IASB GAAP.

1If shares of one class are convertible into shares of another class, as is normally the case, the if-converted methodmust be used for the convertible securities if the effect is dilutive.

Appendix 6-ALIFO MEASUREMENT ISSUES

This appendix is concerned with two measurement issues that arise when the LIFO methodis used:

• Different varieties of LIFO

• Difficulties when LIFO is applied to interim earnings

Although these issues arise frequently, they are segregated within this appendix to simplifythe presentation in the chapter itself.

LIFO INVENTORY METHODS

The discussions in the chapter implicitly assume that:

• Firms account for each inventory item

• There is only one manner of applying the LIFO method of accounting

Neither assumption is correct. In practice, all but the smallest firms have far too many inven-tory items to use specific item-based costing methods efficiently. The potential for LIFO liq-uidations and the resulting loss of tax benefits are additional deterrents to the use of specificitem methods. More efficient methods of applying LIFO to inventories involve the poolingof “substantially identical” inventory units to compute unit costs and physical quantities.

Reeve and Stanga (1987) found that a majority of LIFO method companies use a singlepool, generally defined by the natural business unit, and they use the same pooling methodfor financial reporting and taxes although conformity is not required. The number of poolsused was inversely related to the magnitude of tax benefits (companies with large tax savingsfrom LIFO tended to use fewer pools).

They also reported substantial variation in the number of pools used within an industryand across all the firms in their sample. The impact on cash flows and financial statementssuggests that analysts should carefully evaluate announcements of changes in LIFO pools tounderstand the impact of the change on reported earnings.

Example: Oxford

Oxford [OXM], a clothing manufacturer, uses the LIFO method for most inventories. In fis-cal 2002, Oxford reduced the number of inventory pools used to compute LIFO from five tothree. As a result, the company avoided a LIFO liquidation that would have increased net in-come by 30% (and would have resulted in significant tax payments).1 The company statedthat one reason for the change was to “reduce the likelihood of LIFO layer liquidations.” Thechange was reported as a change in accounting principle.

Inventories may also be pooled on the basis of similarity of use, production method, orraw materials used. Liquidations are reduced because these “dollar value” LIFO methodscompute inventories using dollars, facilitating substitutions of items in the pools. Inventory

W5

1Despite the change in number of pools, Oxford reported a small LIFO liquidation for the year.

W6 APPENDIX 6-A LIFO MEASUREMENT ISSUES

layers may be priced using indices published by the Bureau of Labor Statistics or internallydeveloped indices. The differences can be substantial.

For example, during 1990 Kmart switched to internally generated indices (from the U.S.Department of Labor’s Department Store Price Index) for its U.S. merchandise inventories.The financial statement footnote stated the firm’s belief that the internal index “results in amore accurate measurement of the impact of inflation on the prices of merchandise sold in itsstores.” The change reduced its COGS by $105 million (net of tax), increasing income by$0.52 per share (32.3% of reported income for the year).

Retailers use more complex LIFO methods. Interested readers are referred to intermedi-ate and advanced accounting texts for explanations of the LIFO Retail and Dollar ValueLIFO Retail methods.

INTERIM REPORTING UNDER LIFO

As discussed in Chapter 1, interim reporting creates special problems for both financial re-porting and financial analysis. Because LIFO is a tax-based inventory method, its use createsadditional problems. The actual LIFO effect for the year cannot be known until the year iscomplete. Thus, LIFO charges for interim periods require management assumptions regard-ing both inventory quantities and prices at the end of the year.

Technological changes, fluctuations in demand, and strikes may also result in a reduc-tion in LIFO layers during the year. The application of LIFO during interim periods may re-sult in substantial distortions (income statement and balance sheet) if the factors causing theLIFO liquidations are temporary and the layers will be replenished prior to year-end.

Financial reporting for interim periods is governed by APB Opinion 28, which providesspecial inventory valuation procedures during interim periods when the firm experiences aLIFO liquidation during one or more of the first three quarters.

Permanent liquidations must be reported in the quarter of occurrence. However, whenmanagement believes that the liquidated layer(s) will be replenished before year-end, thecost of goods sold for the quarter must include the estimated cost of replacing the temporaryliquidation rather than the LIFO cost of the goods sold. The application of this method is il-lustrated using the following example:

Assumptions: All transactions occur during the second quarter

Beginning inventory (FIFO): 10 units @ $30 � $300

LIFO reserve (@ $20) � ($200)

LIFO inventory 10 units @ $10 � $100

Purchases: 20 units @ $30 � $600

Goods available for sale � $700

Sales: 21 units @ $40 � $840

Management determines that the liquidation is temporary and expects the next purchaseprice (cost to replace) to be $35. GAAP requires the use of $35 rather than the unit cost ofthe liquidated layer. COGS is reported at

Inventory is reduced by

The firm recognizes a current liability (called the LIFO base liquidation) for the differ-ence of $25, indicating that the firm has temporarily “borrowed” a unit from the base layer.2

The next purchase of inventory is used to eliminate the current liability and replenish the

$610 (20 units @ $30 and 1 unit @ $10)

$635 (20 units @ $30 plus 1 unit @ $35)

2An AICPA issues paper, “Identification and Discussion of Certain Financial Accounting and Reporting Issues Con-cerning LIFO Inventories” (AICPA, 1984), suggests that the interim liquidation may also be credited directly toinventories.

LIFO base layer. This method eliminates any distortion in reported gross profit and incomenumbers due to temporary interim period liquidations.

Year-end LIFO liquidations are permanent reductions in LIFO layers, and the reportedgross profit must include the impact of the reduction in LIFO reserves. If the foregoing sce-nario occurs during the fourth quarter, the firm would report COGS of $610 [(20 � $30) �(1 � $10)] and separately disclose the impact of the LIFO liquidation on COGS and net in-come in the footnotes.

Example: Nucor

The following example illustrates the impact of volatile prices and the procedures requiredfor interim reporting. It is based on Nucor Corp., a steel and steel products manufacturer thatuses the LIFO method of inventory accounting. Steel scrap is a major component of inventorycost, and since scrap prices can be volatile, Nucor must estimate its year-end position at theend of each interim period. That is, it must estimate both physical inventory and the price ofscrap at year-end to establish the appropriate LIFO reserve at the end of each interim period.

In 1981, scrap prices rose during the first part of the year, but declined in the secondhalf. The LIFO reserve declined for 1981 as a whole, reflecting a decline in the price of steelscrap. (At the end of 1981, the difference between the LIFO and FIFO cost of its inventorywas lower than it had been one year earlier.)

During the first two quarters, Nucor assumed that scrap prices would be higher at the endof 1981 than one year earlier and accrued additional LIFO reserves. Because of the decline insteel scrap prices late in the year, these earlier accruals were reversed in the fourth quarter.The impact of the interim changes in the LIFO reserve can be seen in the following table:

Reported Nucor Quarterly Results 1981 ($ in thousands)

Quarter I II III IV Year

Pretax income $13,087 $11,204 $4,637 $15,901 $44,829

LIFO effect 1,873 1,900 0 (5,134) (1,361)

LIFO reserve (end of period) $25,600 $27,500 $27,500 $22,366 $22,366

(12/31/80 � $23,727)

Source: Nucor, 1981 annual and interim reports.

Although the interim LIFO accruals (LIFO effect � change in reserve) were made ingood faith, in retrospect we can see that they were incorrect and distorted operating results.To correct that distortion, we can (with perfect hindsight) reallocate the decrease in the LIFOreserve for the year so that an equal amount is credited to each interim period. We can obtainthe “true” interim results by restating the LIFO impact as follows:

Adjusted Nucor Quarterly Results 1981 ($ in thousands)

Quarter I II III IV Year

Pretax income $13,087 $11,204 $4,637 $15,901 $44,829

LIFO adjustment* $12,213 $12,240 $4,340 $ (4,793) $44,820

Adjusted pretax $15,300 $13,444 $4,977 $11,108 $44,829

% Change from reported 16.9% 20.0% 7.3% (30.1)% 0

*Difference between original LIFO effect and true LIFO effect (one-fourth of annual). For example, the first quarteradjustment is $1,873 � (�$1,361/4).

The Nucor case indicates that management assumptions can play a major role in re-ported interim earnings and the application of LIFO accounting to interim periods can resultin large distortions in interim comparisons. It should also be noted that there are many waysof making interim LIFO calculations. This illustration also serves as an example of fourth-quarter adjustments that have a significant impact on reported earnings and trends reflectedduring the previous three quarters.

INTERIM REPORTING UNDER LIFO W7

Appendix 6-BTHE FIFO/LIFO CHOICE: EMPIRICAL STUDIES

As noted in the chapter, the LIFO to FIFO choice provides an ideal research topic as thechoice has

1. conflicting income and cash flow (tax effect) implications, and

2. data availability allowing for adjustment from one method to the other permitting“as-if” comparisons in research design.

Earlier research focused on market reaction to FIFO-to-LIFO switches and the motivationfor using one method as compared to the other. This line of research was consistent withthe market-based and positive accounting approaches1 to research prevalent at that time.More recently, in line with the renewed interest in security valuation issues, researchershave examined the relationship between equity valuation and alternative methods of in-ventory reporting.

Equity Valuation Issues

Jennings, Simko, and Thompson (1996) examined the contention that

1. LIFO income statements were more useful than non-LIFO statements, and

2. Non-LIFO balance sheets were more useful than LIFO balance sheets

by comparing which set of statements better explained the distribution of equity values for aset of LIFO firms. The “as if” non-LIFO statements were created by using the LIFO reservedisclosures and the methodology described in the chapter.

Their results were mixed. Consistent with their expectation, they found that LIFO-basedincome statements explained more of the variation in equity valuations than non-LIFO in-come statements. However, they found that LIFO balance sheets were more useful than theirnon-LIFO counterparts—a surprising result given that non-LIFO balance sheets are closer tocurrent (rather than outdated LIFO) costs.

Jennings et al. explained these results by noting the negative empirical relationship (re-ported earlier by Guenther and Trombley (1994))—between a firm’s value and the magni-tude of the LIFO reserve.2 They argue (and demonstrate using a theoretical model) that iffirms cannot (fully) pass on input price increases to their customers, a larger LIFO reserve in-dicates lower future profitability. In such cases, a negative relationship is expected betweenfirm value and the LIFO reserve.

Thus, the poor performance of the non-LIFO balance sheet may be explained as follows.When the LIFO reserve is added to LIFO inventory to create the non-LIFO balance sheet in-ventory, the positive relationship between value and assets may be offset by the loss of infor-

W8

1See Chapter 5 for further discussion.2This result seems anomalous because a higher LIFO reserve is indicative of higher asset values.

mation (with respect to the effects of inflation) that is provided by the LIFO inventory andLIFO reserve individually.

As the elasticity of output prices with respect to input price changes fall, the LIFO and LIFOreserve components of non-LIFO inventory have increasingly different implications for futurenet resource inflows, and loss of information through aggregation increases.3

An alternative “deferred tax” explanation for the negative relationship between firmvalue and the LIFO reserve is offered by Dhaliwal, Trezevant and Wilkins (2000). Theyargue that the LIFO reserve indicates a potential future tax liability if the inventory (or firm)is liquidated or sold.

Whichever argument is correct in explaining the negative relationship between firmvalue and the LIFO reserve, these results and those with respect to the comparison of LIFOand non-LIFO balance sheets point out the need for well-grounded economic analysis whenpreparing a research design for empirical testing.

The LIFO/FIFO Choice

As the chapter discussion indicates, there may be sound reasons for firms to stay on FIFO. Inaddition to those related to LIFO liquidations and declining prices, these reasons include bur-densome record keeping requirements, the inability to write down obsolete inventory, andthe desire to maximize taxable income when using up a tax loss carryforward.

Another reason is the desire to avoid the negative effect of LIFO on a firm’s reportedearnings. This motivation depends on whether (as discussed in Chapter 5) a market-based orfinancial contracting argument is used.

The market-based argument says that, whether or not the market is efficient and can seethrough the FIFO/LIFO choice to the real economics of the firm, managers who believe thatthe market can be fooled by lower reported earnings are reluctant to use LIFO.

Alternatively, the financial contracting approach considers the impact of the FIFO/LIFO choice on management compensation and debt covenant restrictions. The bonus planhypothesis argues that when top management compensation is based on income, the firm isless likely to use the LIFO method if the resultant lower earnings reduce their compensation.

The debt covenant hypothesis argues that the negative effect of LIFO on a firm’s re-ported income and ratios increases the probability that a firm will violate debt covenants re-garding such financial measures as working capital, net worth, income, and the dividendpayout ratio. Highly leveraged firms may be especially reluctant to use LIFO for that reason,notwithstanding the tax benefits.

Studies of the FIFO/LIFO choice generally examine the impact of the choice on firms’financial performance in terms of both market reaction and management behavior, as well asthe effect on firms’ financial statements. These studies and the hypotheses tested are affectedby both the progression in academic accounting theory and economic factors (such as higherinflation) that caused a resurgence in the adoption of LIFO in the mid-1970s.

Market-Based Research

LIFO has been permitted in the United States since before World War II, and its rate ofadoption understandably follows the rate of inflation. In the 1970s, when the rate of inflationreached double-digits, LIFO adoptions soared. Approximately 400 companies switched fromFIFO to LIFO in 1974 alone. This period coincided with heavy academic emphasis on mar-ket-based empirical research and the efficient market hypothesis, and the effect of theFIFO/LIFO switch was viewed as an ideal area for research.

Given these conditions, the functional fixation hypothesis was tested to see whether:

• The market accepts financial statements as presented and thus views the switch toLIFO unfavorably since income is depressed.

THE FIFO/LIFO CHOICE: EMPIRICAL STUDIES W9

3Ross Jennings, Paul J. Simko, and Robert B. Thompson III, “Does LIFO Inventory Accounting Improve the In-come Statement at the Expense of the Balance Sheet?,” Journal of Accounting Research, (Spring 1996), p. 105.

W10 APPENDIX 6-B THE FIFO/LIFO CHOICE: EMPIRICAL STUDIES

• The market is efficient in the sense that it sees through reported data and views theswitch to LIFO positively since cash flow increases.

Proponents of the efficient market hypothesis predicted that the market would see throughthe switch and react favorably to the cash flow effects.

Surprisingly, the results were equivocal. Sunder (1973) examined a sample of firms thatchanged to LIFO in the period 1946 to 1966 and found that prior to the switch these firms ex-perienced positive abnormal returns (Figure 6B-1a). At the time of the change itself, the re-action was slightly negative or nonexistent, as investors seemed to ignore the positive cashflow effect. Moreover, the risk (beta) of firms that switched to LIFO increased in the monthssurrounding the switch.

This result was similar to that of Ball (1972), who examined the market reaction to sev-eral accounting changes, FIFO/LIFO included. The positive reaction in the year of the switchwas interpreted by some as a sign that the market anticipated the switch and had reacted priorto the actual announcement. Others felt that firms that switched had been having good yearsand could thus “afford” the negative impact of the switch, and that these studies sufferedfrom a self-selection bias.

Subsequent studies such as Eggleton et al. (1976), Abdel-khalik and McKeown (1978),Brown (1980), and Ricks (1982) extended this research by controlling for earnings-relatedvariables and focusing on the large number of firms that switched in the 1974 to 1975 period.Generally, their results confirmed a negative market reaction in the year of the switch.

Ricks, for example, used a control sample of non-LIFO adopters (matched on the basisof industry and earnings calculated “as if” the control company was also on LIFO) and com-puted the cumulative average return differences between the two groups. His results, pre-sented in Figure 6B-1b, clearly indicate better market performance for firms that did notadopt LIFO. Although these lower market returns were reversed within a year, the initial pro-longed negative reaction is difficult to understand.

One explanation for this anomalous behavior is that firms that switched to LIFO werethose most affected by inflation. Thus, the market may have reacted negatively to theadded risk (higher inflation) of these firms, explaining the lower returns and higher riskmeasures.4

The difficulty with this explanation is that the sample firms were matched by industry.Thus, we must assume that the sample firms were somehow more adversely affected by in-flation than other firms in the same industry. Biddle and Ricks (1988), discussed shortly, alsofound evidence consistent with this explanation. Implicitly, these studies help explain whyfirms stayed on FIFO; they wanted to avoid the unfavorable market reaction resulting fromthe adoption of LIFO.

Biddle and Lindhal (1982) attempted to resolve some of these issues by arguing that pre-vious studies did not consider the amount of tax savings from the LIFO adoption. Theyfound a positive association (see Figure 6B-1c) between the market reaction and the esti-mated tax savings:

The results in this study are consistent with a cash-flow hypothesis, which suggests that in-vestor reactions to LIFO adoptions depend on the present value of tax-related cash-flow sav-ings. After controlling for abnormal earnings performance, larger LIFO tax savings were foundto be (cross-sectionally) associated with larger cumulative excess returns over the year inwhich a LIFO adoption (extension) first applied.5

Biddle and Lindhal studied 311 LIFO adopters from the period 1973 to 1980. The pat-tern of abnormal returns reported is similar to Sunder’s findings (Figure 6B-1a). Neitherstudy used a control group,6 making these results not directly comparable to those of Ricks.

4This argument is consistent with the Jennings et al. (1996) explanation (discussed earlier) that the negative associa-tion between equity values and the LIFO reserve was related to the inability of firms to pass on higher input prices.5Gary C. Biddle and Frederick W. Lindahl, “Stock Price Reactions to LIFO Adoptions: The Association BetweenExcess Returns and LIFO Tax Savings,” Journal of Accounting Research, Autumn 1982, Part II, pp. 551–588.6Biddle and Lindahl instead used the size of the tax saving as a “within-group” control.

Thus, it is possible that there was some systematic but unexplained factor affecting the 1974to 1975 adoptions, and that the research results were sensitive to the research design and thetime horizon examined.

Biddle and Ricks (1988), using daily data, confirmed that there were negative excess marketreturns around the preliminary dates of firms adopting LIFO in 1974. There is little evidence of


FIGURE 6B-1 Abnormal returns: Inventory method studies. Sources:(a) 1946–1966 Adopters: Shyam Sunder, “Relationship Between Ac-counting Changes and Stock Prices: Problems of Measurement andSome Empirical Evidence,” Journal of Accounting Research, Supple-ment 1973, pp. 1–45, Fig. 2, p. 18. (b) 1974–1975 Adopters: William E.Ricks, The Market’s Response to the 1974 LIFO Adoption,” Journal of Accounting Research, Autumn 1982, pp. 367–387, Fig. 2, p. 378. (c) 1973–1982 Adopters: Gary C. Biddle and Fredrick W. Lindahl,“Stock Price Reactions to LIFO Adoptions: The Association BetweenExcess Returns and LIFO Tax Savings,” Journal of Accounting Re-search, Autumn 1982, pp. 551–588, Fig. 1, p. 569.

W12 APPENDIX 6-B THE FIFO/LIFO CHOICE: EMPIRICAL STUDIES

significant excess returns (negative or positive) near the preliminary dates of firms adoptingLIFO in other years.7

To explain the negative returns, they examined analyst forecast errors for the 1974LIFO adopters. They found that analysts significantly overestimated the earnings and didnot fully appreciate the magnitude of the impact of inflation.8 In other years, however, theerror in analyst forecasts for LIFO adopters was not significant. Further, they found thatthe negative returns were positively correlated with the forecast error, indicating that themarket (as well as analysts) was surprised by the actual reported earnings. Thus, the nega-tive returns were due to the “surprise” when the market realized that it had underestimatedthe impact of inflation. As the firms that adopted LIFO were presumably those most af-fected by inflation, the negative surprise reaction hit them hardest. In later years, however,the market learned from experience and the impact of inflation was more readily factoredinto earnings estimates.

Although these studies shed some light on the market reaction to LIFO adoption, theystill do not explain why some firms remain on FIFO. On the contrary, Biddle (1980) found

surprising the finding that many firms voluntarily paid tens of millions of dollars in additionalincome taxes by continuing to use FIFO rather than switching to LIFO.9

Contracting Theory Approach

The contracting theories of accounting choice focus on this issue. Abdel-khalik (1985) exam-ined the bonus plan hypothesis and its implicit corollary that management-controlled firms,in which ownership is widely held, are more likely to use FIFO than owner-controlled firms.The rationale for this argument was that when management is more removed from ownershipof the firm, then management compensation rather than the wealth of the firm becomes theprimary motivator for manager actions. Thus, the LIFO-induced tax savings are less impor-tant to the management-controlled firm.

Abdel-khalik found that manager-controlled FIFO firms had relatively higher income-based bonuses. On the other hand, there was no evidence that differences in compensationplans were related to the FIFO/LIFO choice. In explaining this (non)finding, Abdel-khalikhypothesized that either

1. firms switching to LIFO modify their compensation arrangements, or

2. as some executives have indicated to me, the FIFO-based income continues to beused in determining annual bonus.10

Hunt (1985) examined the bonus plan and debt convenant hypotheses. His results didnot support the bonus plan hypothesis. Contrary to expectations, he found that LIFO firmstended to be less owner-controlled. Hunt, however, did find support for the debt covenant hy-pothesis, especially with respect to the leverage and interest coverage ratios. His evidencealso indicates a threshold level of dividend payout ratios above which firms are reluctant touse LIFO.

Dopuch and Pincus (1988) examined the bonus plan, debt covenant, and taxation hy-potheses in one study and found that the taxation effect provided the best explanation for theLIFO/FIFO decision. They compared the holding gain that would have accrued to LIFOfirms had they stayed on FIFO with the holding gain for firms that remained on FIFO.

7Gary C. Biddle and William E. Ricks, “Analyst Forecast Errors and Stock Price Behavior Near the Earnings An-nouncement Dates of LIFO Adopters,” Journal of Accounting Research, Autumn 1988, pp. 169–194.8At that time, LIFO adoptions were unusual, and it took time for analysts to learn to estimate the impact. That theydid learn is evidenced by the reduced earnings forecast errors for LIFO adopters in later years.9Gary C. Biddle, “Accounting Methods and Management Decisions: The Case of Inventory Costing and InventoryPolicy,” Journal of Accounting Research, Supplement 1980, pp. 235–280.10A Rashad Abdel-khalik, “The Effect of LIFO-Switching and Firm Ownership on Executive’s Pay,” Journal of Ac-counting Research, Autumn 1985, pp. 427–447.

They found larger holding gains for LIFO firms, resulting in higher tax savings. In addi-tion, the holding gain grew as they approached the switch date. Dopuch and Pincus arguedthat this indicated

the long-term FIFO firms in our sample have not been forgoing significant tax savings, inwhich case remaining on that method is certainly consistent with FIFO being an optimal taxchoice, given other considerations. In contrast, long-term LIFO firms would have forgone sig-nificant tax savings. . . . Finally, using the long-term FIFO sample’s average holding gains as abase, our change-firms’ average holding gains became significantly larger than the FIFO aver-age as they approached the year in which they switched, and this difference continued to growsubsequently.11

Further, Dopuch and Pincus argued that financial analysts could have calculated the in-creased holding gains for the switch firms and thus anticipated the switch. Therefore, the in-conclusive findings of the market reaction studies could be a result of ignoring the “advancewarning” market agents had regarding the switch.

More recently, Jennings et al. (1992) supported this advance warning contention. Theyconstructed a model that predicted which firms in the 1974 to 1975 period were more likelyto adopt LIFO. The model accurately forecast adopting/nonadopting firms approximatelytwo-thirds of the time. Furthermore, the prior probability of adoption affected the market re-action. The less likely candidates for adoption (according to the model) had more positivemarket reactions when they adopted LIFO. Similarly, firms that were originally viewed aslikely candidates for adoption, but did not adopt, suffered negative market reaction whenthey failed to adopt LIFO.

However, in summing up the research in this area, the editor of The Accounting Reviewstated

We continue to be relatively uninformed about these issues and know little about the real rea-sons that many firms do not switch to LIFO when it appears that they would benefit by positivetax savings.12


11Nicholas Dopuch and Morton Pincus, “Evidence of the Choice of Inventory Accounting Methods: LIFO VersusFIFO,” Journal of Accounting Research, Spring 1988, pp. 28–59.12Editor’s Comments, The Accounting Review, Vol. 67, No. 2, April 1992, p. 319.

Appendix 7-ARESEARCH AND DEVELOPMENT AFFILIATES

INTRODUCTION

Because GAAP in the United States requires that all expenditures for research and develop-ment (R&D) be expensed, firms have looked for alternative methods of financing R&D thatpostpone the associated earnings charge. Alternate financing methods may also have the fol-lowing advantages:

• Targeting investors who are attracted by the risk/reward characteristics of specific projects

• Focusing management attention on specific projects by placing their development in aseparate entity.

We discuss the two most common forms of these arrangements, R&D partnerships and de-velopment companies. The drug and biotechnology industries have been the most commonusers of these techniques, perhaps because R&D is focused on the development of discretepatentable products.

Appendix Objectives

1. Examine the motivation for the establishment of R&D arrangements.

2. Show the effect of R&D arrangements on the amounts and timing of research and de-velopment expense.

3. Show the effects of R&D arrangements on reported net income, stockholders’ equity,and financial ratios.

4. Compare the effects of R&D arrangements on companies using accounting methodsthat expense all R&D with those permitting capitalization.

RESEARCH AND DEVELOPMENT PARTNERSHIPS

An R&D partnership raises funds from investors. Those funds are then used to pay the com-pany for research. Any patents or products resulting from that research belong to the partner-ship, but the company can either purchase the partnership or license the product. Thus, thecompany controls the technology without reporting the expenses resulting from researchcosts, as the “revenue” from the partnership offsets the research expense.

This arrangement has many of the attributes of an option; the firm has a call option onthe patents or products developed for the partnership, with the purchase price being the exer-cise or strike price. Shevlin (1991) treats such limited partnerships (LPs) as an option anduses option pricing theory to value the LP:

The value of the LP call option to the R&D firm may be decomposed into the present value ofthe underlying project financed by the LP (an asset) less the present value of the payments tothe limited partners if the firm exercises its option (liability).1

W14

1Terry Shevlin, “The Valuation of R&D Firms with R&D Limited Partnerships,” The Accounting Review, Jan. 1991,pp. 1–21.

SFAS 68 (1982), Research and Development Arrangements, sets criteria to distinguishtrue transfers of risk from disguised borrowings. The following are indicators that there hasnot been a true transfer of risk:

1. The company has an obligation to the partnership (or investors) regardless of the out-come of the research. Such obligation may take the form of a guarantee of partner-ship debt or granting of a “put” option to the investors.

2. Conditions make it probable that the company will repay the funds raised by the part-nership. Such conditions include the company’s need to control the technologyowned by the partnership or relationships between the company and the investors(e.g., top management invests in the partnership).

If there has not been a true transfer of risk, then the company is required to expense the ac-tual research costs and treat funds received from the partnership as borrowings.

When the requirements of SFAS 68 are met, however, the company can recognize rev-enue from the partnership to offset R&D costs. The result is, in effect, a deferral of researchcost until products are sold (and license fees paid) or the partnership is purchased. Sucharrangements are disclosed in financial statement footnotes and analysts should be alert totheir effects on reported income.

In recent years, a new vehicle has largely superseded the R&D partnership: a separatedevelopment company that sells “callable common” shares to the public. The shares are oftenpackaged with warrants of the (parent) company to make the resulting “units” more attrac-tive to investors. The new common shares are callable at prices that promise a high rate ofreturn to investors if the venture is successful. These vehicles are similar to R&D partner-ships in their effects on the firm.

Analysis of Firms with R&D Affiliates

The impact of R&D affiliates on reported financial results is favorable as research costs areoffset by “revenue” from the affiliate. Reported income would be lower if these costs werefunded by borrowing (or from the firm’s own assets). Further, obtaining those funds wouldrequire additional debt or equity capital. R&D financing arrangements permit the company toconduct research without incurring debt or equity dilution, in addition to avoiding the effectsof reporting the research costs as an expense.

There is a cost to this capital, however. When the partnership is purchased or thecallable common is called, a substantial cash payment or share issuance is required. Giventhe risk, investors in R&D affiliates require a high rate of return.

The second cost factor is the impact when the affiliate is purchased. At that time, thepurchase price must be written off as research costs.2 The resulting write-off usually exceedsthe amount of funds originally raised. But that write-off is delayed until the partnership ispurchased. In effect, these arrangements permit the deferral of research costs, but with thepenalty of a high interest factor (cost of capital).

R&D Affiliates Outside of the United States

In jurisdictions that do not require all R&D to be expensed, the incentives for alternativearrangements are weaker. Under IAS GAAP, as discussed in the chapter, research costs mustbe expensed but development costs are capitalized and amortized. Canada has similar re-quirements, as seen in the analysis of Biovail that follows. However, given the preeminenceof the United States capital market, even non-U.S. firms may use these techniques to enhancetheir earnings reported under U.S. GAAP.

RESEARCH AND DEVELOPMENT PARTNERSHIPS W15

2FASB Interpretation 4 (1975) provides that when an acquisition is accounted for under the purchase method of ac-counting, any portion of the purchase price allocated to R&D must be immediately expensed at the time of the ac-quisition. Chapter 14 contains more discussion of this issue.

W16 APPENDIX 7-A RESEARCH AND DEVELOPMENT AFFILIATES

Example: Biovail

Biovail [BVF] is a Canadian pharmaceutical company. It used several R&D affiliates to fi-nance drug development in the 1990’s. We will focus on one such arrangement, IntelligentPolymers [INP], incorporated in Bermuda.

In October 1997, there was an initial public offering of 3.7 million units at $20 per unit,resulting in net proceeds after expenses of approximately, $69.5 million.3 Each unit con-sisted of:

• One Intelligent Polymer common share

• One warrant to purchase one Biovail share at $10 per share (adjusted for subsequentstock splits) from October 1, 1999 through September 30, 2002

Biovail recorded a credit to equity of $8.244 million to reflect the value of the warrants is-sued and an equal reduction of retained earnings to record the contribution to INP. The netresult of the offering was that INP received $69.5 million of capital with no net effect onBiovail’s financial statements.

At the time of the offering, the two companies entered into a series of agreements, in-cluding the following provisions:

1. INP agreed to spend the proceeds to develop seven possible products, paying BVF toconduct the required research.

2. INP would hold the rights to products developed but Biovail would have options topurchase those rights at predetermined terms.

3. Biovail had the option to purchase all shares of INP at the following prices:

• $39.06 per share before October 1, 2000

• $48.83 per share from October 1, 2000 through September 30, 2001

• $61.04 per share from October 1, 2001 through September 30, 2002

The development agreement resulted in payments from INP to Biovail shown in the follow-ing table:

Years ended December 31 1998 1999 2000 Totals

Payments to Biovail $9.7 $33.0 $55.2 $97.9

Biovail’s related costs (6.7) (19.8) (35.2) (61.7)

Biovail gross profit $3.0 $13.2 $20.0 $36.2

Over the three-year period, INP paid Biovail approximately $98 million for research. IfBiovail had conducted the research itself, the total cost would have been nearly $62 million.The effect of forming INP was to increase Biovail’s reported pretax earnings by the amountof the payments received. The significance of these amounts can be seen from Biovail’s rev-enues (Exhibit 7A-1), which rose from $111.6 million in 1998 to $309.2 million in 2000.

In 1999, Biovail paid INP $25 million for the rights to one developed drug. On Septem-ber 29, 2000, Biovail exercised its option to purchase all INP shares, for a total price (includ-ing bank debt) of $204.9 million. The purchase resulted in a write-off of in-process researchand development (IPRD) of $208.4 million. The write-off resulted in an operating loss forthe year of $78 million. The IPRD was far above the actual research expenditures. Howeverthe creation of INP had the effect of delaying the recognition of these costs in Biovail’s fi-nancial statement. It also reduced Biovail’s risk; if the INP research had not been successful,Biovail would not have exercised its option.4

We can see the cost of capital implicit in the creation of INP by examining the invest-ment from the investor point of view. Ignoring (for the moment) the Biovail warrants in-

3All dollar amounts in this section are United States dollars even though Biovail is Canadian.4It is also possible that Biovail would have exercised its option even in the event of failure in order to maintain fullcontrol of its proprietary technology.

cluded in the offering, investors bought INP shares for $20 each. The call prices shownabove provide rates of return of 25% per annum. As Biovail shares rose substantially, tradingabove $45 per share in November 2000, the actual return (including the gain in the Biovailwarrants) was even higher. Of course, investors took the risk that the INP research would nothave produced marketable drugs.5

In economic terms, the Intelligent Polymers capital came at a high price to Biovail.However, the risk reduction may have made the cost of capital acceptable relative to othersources of capital available at that time. The INP arrangement also resulted in postponedrecognition of the research costs associated with the development of these drugs. As IPRDwrite-offs are often seen as “non-recurring” costs, it is uncertain how the financial marketsvalue firms with such charges.

Comparison of Biovail Financial Statements: U.S. vs. Canadian GAAP

Under Canadian GAAP, IPRD and the acquisition cost of drug rights are capitalized andamortized over the useful life of the products. Both the $25 million paid to INP in 1999 andthe cost of acquiring INP in 2000 resulted in asset recognition (rather than being expensedunder U.S. GAAP). Biovail had written off more than $105 million of IPRD in 1999 fromanother R&D arrangement. The difference between the treatment of these transactions be-tween U.S. and Canadian GAAP can be seen in Exhibit 7A-1.

RESEARCH AND DEVELOPMENT PARTNERSHIPS W17

EXHIBIT 7A-1. BIOVAILFinancial Data under United States and Canadian GAAPAll data in $US thousands, except per share

Years Ended December 31 1998 1999 2000

United States GAAP

Revenue $111,657 $172,464 $ 309,170Operating income (loss)* 45,303 (40,160) (78,032)Net income (loss) 41,577 (109,978) (147,796)Earnings per share (diluted) 0.38 (1.07) (1.16)

Total assets $198,616 $467,179 $1,107,267Long-term obligations 126,835 137,504 438,744Convertible securities 299,985Common equity 49,888 267,336 237,458

Common shares outstanding 99,444 124,392 131,461

*Includes IPRD charges $105,700 $ 208,400

Canadian GAAP

Revenue $ 98,836 $165,092 $ 311,457Operating income 35,145 64,117 116,223Net income 31,419 52,080 81,163Earnings per share (diluted) 0.29 0.47 0.57

Total assets $199,919 $635,137 $1,460,967Long-term obligations 126,835 137,594 438,744Shareholders’ equity 19,091 391,794 839,110

Common shares outstanding 99,444 124,392 131,461

Note: While the treatment of IPRD and the cost of acquired drugs are the principal differences between U.S. andCanadian GAAP, the data also reflect other differences.Source: Biovail 10-K, December 31, 2000

5See footnote 4; for this reason, the risk may not have been excessive.


The principal differences are:

1. Under Canadian GAAP, there is a progressive improvement in both operating andnet income, as well as earnings per share. Under U.S. GAAP, the IPRD write-offs re-sult in operating and net losses for both 1999 and 2000.

2. Canadian GAAP assets exceed those under U.S. GAAP, reflecting the capitalizationof drug acquisition costs.

3. Canadian GAAP equity exceeds that under U.S. GAAP, mainly due to the differencein net income.

The ratio effects of these differences are the subject of Problem 7A-1.

Conclusion

The Biovail—Intelligent Polymers example illustrates the effects of research and develop-ment arrangements on the financial statements of the sponsoring company. Such arrange-ment can have major impacts on the amount and timing of reported net income, as well asthe balance sheet and cash flow statements. While ultimately, company valuation depends onresearch (and subsequent marketing) outcomes, the analyst should carefully consider the ef-fect of such arrangements on the financial statements of affected companies.

PROBLEMS

7A-1. [Ratio effects of differences in accounting for R&D arrangements].

A. Using the data in Exhibit 7A-1, calculate the following ratios for Biovail for 1998through 2000 under both United States and Canadian GAAP:

(i) Return on sales (net income margin)

(ii) Return on equity

(iii) Asset turnover

(iv) Equity per common shareNote: use year-end amounts for balance sheet data.

B. Discuss the differences between both the level and trend of the ratios computed inpart A.

C. The price of Biovail shares rose from less than $9 per share at the end of 1997 tonearly $39 per share at the end of December 2000. Discuss which set of ratios ap-pears to be reflected in the market performance of Biovail shares. Discuss anyother factors that may have affected the price of Biovail shares during this timeperiod.

D. State which of the two methods of accounting for IPRD (immediate write-off ver-sus capitalization and amortization) comes closest to recognition of the economicimpact of the acquisition of drug rights. Justify your choice.

E. Discuss the limitations of the method chosen in part D.

F. Discuss whether the accounting for internal drug research expenditures shoulddiffer from that for acquired drug rights.

7A-2. [Analysis of R&D Arrangements] In September 1997 ALZA (acquired by Johnsonand Johnson in June, 2001) contributed $300 million to Crescendo Pharmaceuticals, anewly created company. ALZA formed Crescendo to help fund the development ofnew pharmaceutical products. Crescendo and ALZA entered into the following agree-ments:

1. Crescendo was required to spend virtually all of its available funds to fund the de-velopment (by ALZA) of seven possible new products.

2. ALZA granted Crescendo a worldwide license to use ALZA technology in con-nection with product development activities. Crescendo paid ALZA a specified li-cense fee.

3. Crescendo granted ALZA options to license products developed, exercisable on acountry-by-country basis after clearance from the Federal Drug Administration(FDA) or appropriate foreign regulatory body. ALZA also had the right to pur-chase Crescendo’s right to receive license fees. Both the license fee and the pur-chase price were based on predetermined formulas.

4. ALZA had the right to purchase all Crescendo shares until January 31, 2002 at aprice equal to the greater of:

(i) $100 million

(ii) The market value of 1 million ALZA shares

(iii) $325 million less all amounts paid to ALZA by Crescendo under the agree-ment, and

(iv) A formula based on license fees paid to Crescendo by ALZA over the previ-ous four calendar quarters.

ALZA could purchase Crescendo shares for cash, ALZA shares, or a combination ofthe two. The option deadline would be extended if Crescendo had not yet expendedall of its funds.

On September 29, 1997, ALZA contributed $300 million to Crescendo, of which$247 was recorded as a “non-recurring” expense. Crescendo shares were distributedto ALZA’s shareholders and debenture holders as a dividend.

Over the next three years, Crescendo made the following payments to ALZA:


Payments for research $95.0 $90.5 $68.3

Technology fees 10.7 6.7 2.7

Administrative service fees 0.2 0.2 0.2

ALZA paid the following to Crescendo for three drugs that had been successfullydeveloped:

Drug license fees $2.4 $4.5

On November 13, 2000 ALZA paid $100 million to acquire all outstandingshares of Crescendo. $45.7 million of the purchase price was allocated to developedproducts as deferred product acquisition costs and $9.4 million was expensed asIPRD.

Exhibit 7AP-1 contains financial data on ALZA for the three years ended Decem-ber 31, 2000.

Use the data provided to answer the following questions.

A. Prepare income statements for ALZA for the years 1998 through 2000 assumingthat the Crescendo transactions had not taken place.

B. Calculate the percentage change in each of the following from 1998 to 1999 andfrom 1999 to 2000, using reported data:

(i) Revenues

(ii) Expenses

(iii) Operating income

(iv) Pretax income

C. Calculate the percentage change in each of the following from 1998 to 1999 andfrom 1999 to 2000, using adjusted data from part A:

(i) Revenues

(ii) Expenses


(iv) Pretax income

PROBLEMS W19

D. Calculate the effect of the adjustments in part A on each of the following for thethree years ended December 31, 2000:

(i) Revenues

(ii) Expenses


(iv) Operating margin

(v) Pretax income

(vi) Pretax margin

(vii) Times interest earned

E. Describe the effect of the Crescendo transactions on each of the following, usingthe results of parts A through D:

(i) ALZA’s reported growth rate for 1999 and 2000

(ii) ALZA’s reported profitability for 1998–2000

(iii) The volatility of ALZA’s profitability for 1998–2000

(iv) ALZA’s reported return on equity for 1998–2000.

Hint: Consider the effect of the Crescendo transactions on ALZA’s equity.

F. Discuss the benefits and drawbacks to ALZA of the Crescendo transactions, usingthe results of parts A through E.

G. Considering the Crescendo transactions as a whole, justify the analytical adjust-ments in this problem.


EXHIBIT 7AP-1. ALZA CORP.Financial DataAll data in $millions, except per share


Net sales $ 289.4 $ 448.0 $607.2Royalties, fees, and other 233.1 227.1 281.2Research and development 124.4 120.8 100.1

Total revenues $ 646.9 $ 795.9 $ 988.5

Costs of products shipped (125.7) (158.4) (180.2)Research and development (182.8) (183.6) (190.8)Selling and administrative (141.9) (259.0) (349.4)Merger-related charges — (45.7) —In-process R & D 00000— 00000— 00.(12.4)

Total expenses $ (450.4) $ (646.7) $ (732.8)

Operating income 196.5 149.2 255.7Interest and other income 26.4 41.6 59.0Interest expense 00.(56.7) 00.(58.1) 00.(58.0)

Pretax income $ 166.2 $ 132.7 $ 256.7Income tax expense 00.(57.9) 00.(41.7) 00.(26.0)Net income* $ 108.3 $ 91.0 $ 230.7

*before cumulative effect of accounting change

Appendix 7-BANALYSIS OF OIL AND GAS DISCLOSURES

INTRODUCTION

The two acceptable accounting methods used for oil and gas exploration: the successful ef-forts method (SE) and full cost method (FC)1 are illustrated in Exhibit 7-1. The choice be-tween these methods has significant effects on reported financial statements. Thesedifferences can be summarized as follows:

• SE firms, by expensing dry hole costs, have lower carrying costs of oil and gas re-serves than FC firms.

• SE firms have lower earnings than FC firms when exploration efforts are rising.

• SE firms have lower cash from operations than FC firms (unless explicitly adjustedfor, as in the case of Texaco).

APPENDIX OBJECTIVES

1. Examine the motivation for use of the successful efforts and full cost methods.

2. Describe the motivation and effects of changes between the two accounting methods.

3. Analyze the supplementary disclosures regarding oil and gas reserves, showing howthey can be used to gain insight into the:

• changes in reserve quantities over time.

• cost of finding new reserves.

• level and trend of present value of reserves, a proxy for the fair value of oil and gasreserves.

4. Show how to adjust present values for subsequent price changes.

5. Adjust stockholders’ equity and the debt-to-equity ratio for the difference betweenthe carrying cost and present value of oil and gas reserves.

Motivations for Accounting Choice

The differential effects on financial statements demonstrated in Exhibit 7-1 as well as the il-lustration in Box 7-1 help explain why some firms prefer the SE method and others the FCmethod. Empirical evidence as to these preferences is provided in Box 7B-1.

Small firms generally prefer the FC method; large firms tend to be indifferent. For largerfirms, with relatively stable exploration budgets and relatively constant success ratios (pro-ductive to total expenditures) across a “portfolio” of exploration projects, the year-to-yearvariability of dry hole expense is low. Amortization of past expenditures is large, reflecting alarge reserve base. As a result, the difference between the two methods is small.

Additionally, larger oil companies are often diversified into the refining and distributionsegments of the oil business. Income from these sources dampens the variability of exploration

W21

1Both methods are described on pages 244–246.

W22 APPENDIX 7-B ANALYSIS OF OIL AND GAS DISCLOSURES

income. Large oil companies tend to use the SE method as well because it is perceived to bemore conservative.2

For smaller companies, however, the differential impact of these two accounting meth-ods can be considerable. Year-to-year variations in spending and success ratios mean that dryhole expense can vary greatly. Under SE accounting, this variability is transmitted directly tothe income statement. Further, smaller companies (especially if growing rapidly) have smallreserve bases and low amortization of past capitalized costs. Dry hole costs from currentdrilling activities often exceed the amortization of the capitalized costs of past drilling.

BOX 7B-1SE Versus FC Choice of Methods: Empirical Evidence

A number of research studies* have examined characteristics offirms using SE versus FC accounting. Malmquist (1990) tested therelationship between the following characteristics and firm choice.

1. Size

The larger the firm, the less likely it will choose FC for severalreasons. First, large firms prefer income-reducing alternativessuch as SE to avoid earning “windfall profits,” especially whenprices are rising, given the political sensitivity of energy prices.

Second, large firms have more drilling activities occurring si-multaneously, creating a portfolio effect and thereby decreasingincome variability. Third, in addition to the risks associated withexploration, oil companies are subject to the risks associatedwith marketing and refining. The larger the proportion of thefirm’s activities in marketing and refining, the lower the impactof SE because its effect is limited to the income associated withexploration. As large firms tend to be more diversified, they haveless incentive to opt for FC.

Using sales as a proxy for size (political costs) and the ratioof exploration costs to market value as well as the ratio of pro-duction costs to market value to measure the various aspects re-lated to size, Malmquist found them all to be significant inexplaining the accounting choice. Higher sales and a larger pro-portion of production costs made the firm more likely to chooseSE. Conversely, the larger the exploration cost proportion, themore likely the firm was to choose FC.

2. Difficulty of Raising Capital in the Equity and Debt Markets

SE companies report lower assets than FC companies. There-fore, securities underwriters may be hesitant (or find it diffi-cult) to sell the securities of firms having low or negative netbook value (equity) levels. Borrowing may also be more diffi-cult for firms with high and variable debt/equity ratios. More-over, for debt already in existence, there is a higher probability

of technical violation of debt/equity-related debt covenants.Malmquist’s study confirmed that firms with higher debt/eq-uity ratios are less likely to choose SE.

3. Management Compensation Contracts

Earnings-based management compensation contracts are af-fected by the choice of accounting method. Opportunistic man-agers may choose full costing to increase the level of theircompensation and decrease its variability. Malmquist notes“there are strong disincentives and limits placed on such behav-ior by the managerial labor market.” No apparent relationshipbetween the choice of accounting method and the presence of anearnings-based compensation contract was observed.

These results are consistent with some (but not all) ofDeakin’s (1989) findings. Analyzing firms that lobbied for FCand the reasons given by those firms for lobbying, Deakin foundthat, on average, they had characteristics consistent with thestated reasons. The reasons given by the firms were:†

1. The expected impact on cost of capital and access to capitalmarkets

2. The potential of the proposed elimination of the FC method toaffect accounting income-based management incentive con-tracts

3. The perceived effect on future drilling activity

4. The effect of rate regulation‡

To some extent, generalizing from Deakin’s sample of com-panies, which lobbied for a particular accounting method, to thegeneral population of firms, is fraught with danger as the samplemay be biased. Taking the time and effort to lobby can be an in-dication that these firms are the ones most likely to be affectedby the choice. Thus, Deakin’s finding that the presence of man-agement incentive contracts was associated with firms that lob-bied for FC in contrast to Malmquist, who did not find such arelationship, may reflect their different samples.

*See, for example, Steven Lilien and Victor Pastena, “Determinants of Intra-Method Choice in the Oil and Gas Industry,” Journal of Accounting andEconomics, 1982, pp. 145–170 and Edward B. Deakin III, “An Analysis of Differences Between Non-Major Oil Firms Using Successful Efforts andFull Cost Methods,” The Accounting Review, Oct. 1979, pp. 722–734.†Edward B. Deakin III, “Rational Economic Behavior and Lobbying on Accounting Issues: Evidence from the Oil and Gas Industry,” The AccountingReview, Jan. 1989, pp. 137–151.‡The last reason applied primarily to regulated companies that were required by rate-making authorities to use FC accounting procedures.

2A more detailed analysis of the financial reporting effects of SE versus FC on firms under different environments isprovided by Sunder (1976).

Smaller companies are also less diversified as they concentrate on exploration. Widely fluc-tuating patterns of earnings growth are considered a drawback for firms attempting to obtainexternal (equity or debt) financing. This problem is further exacerbated because, under suc-cessful efforts, the balance sheet shows lower assets and equity, thus hurting reported sol-vency ratios. As a result, smaller companies tend to use the FC method of accounting.

Changing Accounting Methods

The FC method has one drawback, however. When the price of oil causes the value of the re-serves to fall below book value, the SEC requires that companies using the FC method writedown properties whose carrying cost exceeds the present value of future cash flows of theproved reserves attributable to that property.3 Companies using the SE method are requiredto use the less stringent measure of undiscounted future cash flows.4 In the 1980s, when the price of oil fell drastically, some companies that had previously chosen FC accounting(presumably to report higher income) were forced to take large write-offs, reducing reportedincome.

One method of avoiding such large write-offs was to change reporting methods from FCto SE, reducing the carrying amount of reserves. The change to or from the FC method is oneof those cases where retroactive adjustment for accounting changes is mandatory; all prioryears presented must be restated and the cumulative effect reported as an adjustment to thebeginning retained earnings. Note that this does not change the value of the reserves; it onlychanges the carrying amount on the balance sheet. Sonat changed its accounting method sev-eral times, reflecting changing industry conditions (see Problem 7B-1).

Adoption of the SE method of accounting requires the expensing of capitalized dry holecosts, lowering reported income. On the other hand, the amortization of previously capital-ized costs is also reduced, increasing reported earnings. The balance between increased ex-pensing of current year expenditures and reduced amortization of past expendituresdetermines the net effect on earnings for any given year.

What is the effect of the accounting change on cash flow? There is no effect on actualcash flow as the change to the successful efforts method merely reallocates cash flows for fi-nancial reporting purposes. (For income tax purposes, oil and gas companies expense themaximum allowable; the accounting change has no impact on tax return income.)

However, components of reported cash flows may be affected by the accounting change.Lower reported capital expenditures are offset over time by lower reported operating cashflows. Once again, we see how the classification of cash flow components is affected by ac-counting choice.

SFAS 69: DISCLOSURES REGARDING OIL AND GAS RESERVES

A major drawback of both accounting methods is the lack of correspondence between the re-ported cost of a producing oil or gas field and its economic value. Although this is true of vir-tually all fixed assets, it is especially true of oil- and gas-producing assets because, even atthe time of drilling, there may be little relationship between the expenditures and results. Anexpenditure of millions of dollars can result in a dry hole. Alternatively, a small expenditurecan result in a discovery of oil or gas worth many times its cost.

Neither method provides truly relevant data as to the value of reserves. This shortcom-ing is addressed by the disclosure requirements of SFAS 69 (1982), which requires extensiveinformation about the results of operations for oil and gas activities and disclosure of a stan-dardized measure of proved oil and gas reserves. Additional summary disclosures of theseactivities by equity method investees and minority interests are also required.

SFAS 69: DISCLOSURES REGARDING OIL AND GAS RESERVES W23

3The comparison of the carrying value of reserves with their present value is sometimes referred to as the ceilingtest.4See David B. Pariser and Pierre L. Titard, “Impairment of Oil and Gas Properties,” Journal of Accountancy, Dec.1991, pp. 52–62.


Disclosure of Physical Reserve Quantities

Texaco’s 1999 financial statements (included in the website and CD that accompany the text)contain a section entitled, “Supplemental Oil and Gas Information.” Table I provides data onthe physical quantities of Texaco’s proved oil and gas reserves, including:

1. Separate disclosure of oil and gas reserves

2. Separate disclosure by geographic area

3. Separate disclosure of the reserves of equity affiliates5

4. Reconciliation of the year-to-year change in proved reserves

5. Disclosure of proved developed reserves

These data describe the company’s physical reserves at each balance sheet date. The first twofeatures listed help the user understand the nature of the reserves. For example, oil reservesin the United States have different economic characteristics than gas reserves in Africa. Sep-arate disclosure of the reserves of equity method affiliates aids the evaluation of the invest-ment in such companies.

The reconciliation is one of the most significant features as it enables us to understandhow estimated reserves change from year to year as a result of:

1. Production, which reduces reserves

2. Discoveries, which increase reserves

3. Purchases and sales of reserves

4. Revisions of estimates

5. Price changes, which can make reserves economically feasible to produce, or not6

Each of these disclosures provides useful data because physical quantities can be related tocash flows. For example, the cost of finding reserves can be derived by comparing explo-ration expenditures with reserves discovered. This is considered an important measure ofmanagement ability.

Revisions, as noted by Clinch and Magliolo (1992),7 are important indicators of the“quality” of management estimates. Companies reporting predominantly downward revi-sions are viewed with some skepticism, reflecting the apparent overoptimism of past esti-mates. Investors prefer positive surprises, that is, upward revisions of estimated reserves.

Texaco’s disclosures show that worldwide oil reserves increased over the three-year pe-riod, from 2,704 million barrels at December 31, 1996 to 3,480 million barrels at December31, 1999. Most of the increase was in the United States and “Other East” geographic areas.Gas reserves also rose, with the United States and “Other East” (the largest percentage in-crease) again accounting for the gain.

5See Chapter 13 for a discussion of the equity method.6For example, in 1985, Atlantic Richfield removed 8.3 trillion cubic feet (trillion � billion MCF) of natural gas re-serves located in northern Alaska from its estimate of proved reserves, reducing its domestic gas reserves by morethan 50%. The company explained that this change was prompted by a review of economic factors, especially thesignificant drop in oil and gas prices in that year. In its 1999 10-K, the company stated that:

ARCO is actively evaluating various technical options for commercializing North Slope gas. . . . Signifi-cant technical uncertainties and existing market conditions still preclude gas from such potential projectsbeing included in ARCO’s reserves.

7Clinch and Magliolo argue that the value-relevance (informativeness) of the SFAS 69 data depends on the reliabil-ity investors attach to it. As data are subject to constant revision, reliability suffers. They found that although themarket did not find reserve data to be value-relevant, production data were found to be informative. Production data,they argue, are more objective as they reflect actual actions taken by management rather than just estimates. Further,they found, for the subset of firms whose quantity estimates appeared more reliable (less revision of estimates), thatproved reserve data were also value-relevant. (Greg Clinch and Joseph Magliolo, “Market Perceptions of ReserveDisclosures Under SFAS No. 69,” The Accounting Review, Oct. 1992, pp. 843–861.)

The reconciliations give us additional insights regarding the reserve increases:

• Revisions have generally been positive.8

• Improved recovery estimates also consistently made positive contributions to esti-mated reserve quantities. These gains may reflect newer technologies that permithigher recovery from existing oil and gas wells.

• Texaco purchased oil reserves in the U.S. in 1997 (Monterey Resources) and gas re-serves in the “Other East” in 1998 and 1999.

• Discoveries and extensions, however, were below production levels in all three yearsfor oil and all but 1997 for gas

The data can also be used to measure the reserve life (end-of-year reserves divided by pro-duction) of Texaco’s reserves, by type and geographic segment. The computations below in-dicate that Texaco’s reserve lives increased over the period as U.S. oil production andworldwide gas production failed to increase with reserves. Reserve lives in the United Statesare higher that in other areas for oil, but lower for gas.

Reserve Lives in Years

United States Worldwide

1997 1998 1999 1997 1998 1999

Oil reserves 1,767 1,824 1,782 3,267 3,573 3,480

Production 157 144 144 317 351 336

Ratio 11.25 12.67 12.38 10.31 10.18 10.36

Gas reserves 4,022 4,105 4,205 6,242 6,517 8.108

Production 643 633 550 839 879 786

Ratio 6.26 6.48 7.65 7.44 7.41 10.32

Data from Table I; oil in millions of barrels, gas in billions of cubic feet

Disclosure of Capitalized Costs

Table IV reports the balance sheet carrying cost of the disclosed reserves and Table V thecurrent year costs incurred.

When reviewing Table IV (Capitalized Costs), note that:

• Capitalized costs depend on the accounting method followed: Companies using theFC method will capitalize more exploration cost than companies employing the SEmethod. Notice that the capitalized costs of equity affiliates are disclosed separately,just as their reserve quantities are disclosed separately.

• Costs are net of accumulated depreciation, amortization, and valuation allowances;different accounting choices in these areas will affect the net carrying cost.

• Costs of unproved properties and support facilities are separately disclosed.

• Capitalized costs are aggregated for oil and gas, unlike reserve quantities.

These data give analysts a balance sheet cost to match against the physical reserves with alloil and gas reserves combined into one measure, usually termed barrel of oil equivalent(BOE). Quantities (of oil and gas reserves disclosed in Table I) can be combined into units of


8There is a typographic error in the 1999 gas reserve change data. The worldwide revisions should be 915 and thetotal changes 1,591; the negative signs are in error.


BOE based on either energy equivalence (1 barrel of oil � 6 MCF of gas)9 or the basis of rel-ative price.10

Once this has been done, the balance sheet cost per BOE can be computed. At December31, 1999, the calculation for Texaco’s reserves is (in millions of barrels):

The capitalized cost per BOE is

Note that part of the capitalized cost represents outflows for unproved properties (for whichno reserves have yet been estimated) and for support facilities. This calculation, therefore,overstates the capitalized cost per BOE.

With two years of data, we can look at the trend of capitalized cost per BOE as well asvariations by geographic area:

Capitalized Cost per BOE Equivalent

December 31 United States Europe Other East Equity Worldwide

1998

Oil reserves 1,824 419 598 684 3,573

Gas reserves 4,105 964 477 151 6,517

BOE 2,508 580 678 709 4,659

Capitalized costs $8,086 $1,436 $1,278 $1,072 $12,190

Costs per BOE 3.22 2.48 1.89 1.51 2.62

1999

Oil reserves 1,782 427 670 546 3,480

Gas reserves 4,205 962 1,866 134 8,108

BOE 2,483 587 981 568 4,831

Capitalized costs $7,933 $1,459 $2,056 $1,178 $13,038

Costs per BOE 3.20 2.48 2.10 2.07 2.70

Data from Tables I and IV. Oil reserves and BOE in millions of barrels, gas reserves in billion cubic feet, capitalizedcosts in $millions

This table indicates that Texaco’s unit carrying costs are below even the cyclical lowpoints of recent oil prices (approximately $10 per barrel). Low capitalized costs are ex-pected, given the use of successful efforts accounting. These amounts represent the costs thatTexaco must amortize as oil and gas reserves are produced; low capitalized costs equate tolow amortization and higher operating earnings. Low capitalized costs also indicate that therisk of impairment write-downs is minimal.

$BOE

� $13,0384,831

� $2.70

� 4,831 � 3,480 � 1,351

� 3,480 � 8,108

6 BCF (billion cubic feet)

No. of BOE � No. of Barrels of Oil � BOE Equivalent of Gas Reserves

9Natural gas is measured in MCF (thousand cubic feet).10In recent years, in the United States, gas has usually sold at a lower relative price than its energy equivalent wouldsuggest. The relationship changes over time. In 2000, natural gas prices rose more rapidly than oil prices. Whilesome analysts combine oil and gas reserves based on relative price, such calculations may require frequent revision.

The geographic differences are revealing. Capitalized costs per BOE are significantlylower in the Other East segment and for Texaco’s equity affiliate (also “Other East”). Higherfinding costs in the United States and Europe have driven exploration efforts for companiessuch as Texaco increasingly to areas with lower costs.

These data also reflect historical costs, well below the cost of finding new reserves. Thecapitalized cost per BOE, moreover, is only a crude means of comparing the cost of reservesfor different companies. It reflects both the accounting method used and the “efficiency” infinding oil (the finding cost per BOE). Companies that use the SE method and have low find-ing costs have a low capitalized cost per BOE. Companies using the FC method or recordinghigher finding costs have higher capitalized cost per BOE.

The capitalized cost per BOE can also be compared with the market value of oil and gasreserves, as revealed by market transactions. If the capitalized cost is higher than transactionprices, this indicates that the balance sheet amount is overstated; if transaction prices arehigher, the reverse is true.

However, using the capitalized cost per BOE is, at best, only an approximation of thevalue of reserves. It is deficient because it fails to recognize the following factors:

1. Reserves in different geographic markets vary in value.

2. Oil reserves have different values from natural gas reserves of equivalent energycontent.

3. The cost of producing reserves (bringing them to the surface) may vary with location.

4. A barrel of oil produced today is more valuable (assuming constant pricing) than oneproduced in five years because of the time value of money.

5. Tax rates vary by jurisdiction and, within jurisdictions, may vary by location andtype of resource.

For these reasons, the aggregation of all reserves by physical quantities does not capturethe market value of reserves. Fortunately, better data are available.

Analysis of Finding Costs

Table V, “Costs Incurred,” reports Texaco’s exploration costs. This table includes all expen-ditures, regardless of whether they are capitalized or expensed, making the data comparableamong companies with different accounting methods.

These expenditures can be compared with reserves found to compute the actual per unitfinding cost. Although annual finding costs are volatile, over longer time periods they mea-sure management’s proficiency in discovering reserves. Texaco’s 1999 finding cost was$4.37 per BOE,11 well above both the carrying cost of reserves and the finding costs over thefive-year period ending in 1999. Finding costs can be compared by geographic area and overtime, although we have not done so here.

Disclosure of Present Value Data

Table II, “Standardized Measure,” reports the estimated future cash flows of the specific re-serves owned by the firm. The following elements are presented:

1. Future cash inflows. Based on a year-by-year schedule of planned unit production,multiplied by current price levels, that is, future gross revenues based on currentprices. Companies are not permitted to assume price changes, unless provided for bya firm contract, which may then be incorporated in the computation. These calcula-tions use proved developed reserves only.

2. Future production costs. Also based on current prices. Production costs include allexpenditures required to bring the oil or gas to market.


11Reported in Table V of Texaco’s 1999 Supplemental Oil and Gas Information.


3. Future development costs. Include the cost at current price levels of additional wellsand other production facilities that may be required to produce the reserves.

4. Future income tax expense. The estimated tax liabilities assuming that the forecastcash flows actually take place.

The net of these amounts, net future cash flows before discount, is a forecast of net cashflows from existing oil and gas reserves. These data must also be adjusted to reflect the timevalue of money by discounting to present value. SFAS 69 requires that all firms use a dis-count rate of 10%. The objective is comparability; the “correct” discount rate will vary overtime and, perhaps, from firm to firm.

The result is a net present value of the after-tax12 cash flows expected from the firm’s re-serves. Note that these data are provided separately for reserves in different geographicareas, but with oil and gas combined.

Companies providing these data routinely state that the standardized measure is not marketvalue and suggest that the data have limited usefulness. Nonetheless, the data are widely usedin the analysis of companies with oil and gas reserves and, in practice, are a useful approxima-tion of market value. Despite some limitations, the data are far more representative of marketvalues than the cost shown on the balance sheet, regardless of the accounting method used.13

Using Present Value Disclosures

How can the data be used? One simple adjustment is to replace the capitalized cost of re-serves with the net present value (standardized measure). This is one step in preparing a cur-rent value balance sheet (see Chapter 17) or computing adjusted net worth. Before makingthis adjustment, the following issues should be considered:

1. Have prices changed since the balance sheet date? If so, the present value data mustbe adjusted to current prices, for example, a 10% increase in oil prices increases fu-ture cash flows by 10%. (Because oil and gas prices do not always move together,use a weighted-average based on the composition of reserves.)

2. Costs may also be adjusted. Although hard data are difficult to come by, industrysources can provide a rough guide as to changes in production and development costs.

3. Do economic or other factors suggest a need for assumptions of future pricechanges? Some analysts construct their own price scenarios and make their owncomputations of future cash flows.

4. Is 10% the right discount rate? The discount rate is a function of the general level ofinterest rates and the relative riskiness of the firm’s reserves. Adjustments may be re-quired. A higher discount rate, of course, reduces the net present value calculation; alower rate increases the present value.

5. Should pretax or after-tax net present values be used? The answer depends on the taxstatus of the firm and purpose of the analysis.14 In a liquidation analysis, for example,

12Texaco deducts tax payments from net cash flows (both undiscounted) and then discounts the after-tax cash flows.We can estimate the discounted income taxes by using the ratio of the discounted pretax cash flows to the undis-counted cash flows. (This assumes a constant tax rate.)

Some firms deduct the present value of tax payments from the net present value of pretax cash flows. The re-sult is the same, but this latter case permits more accurate calculation of the pretax net present value.13Surprisingly, early empirical studies did not seem to bear this out. Harris and Ohlson (1987) and Shaw and Wier(1993), for example, found that SFAS 69 disclosures had weak explanatory power for stock prices and that book valuemeasures outperformed the standardized present value measure. More recently, however, Boone (2002) demonstratedthat the valuation models used in the previous studies were misspecified and, for the valuation model used in hisstudy, the present value measure exhibited significantly more explanatory power than the historical cost measure.14Disclosures for firms with significant reserves outside of North America and Europe frequently show very high incometax rates for these reserves. These high rates reflect the fact that royalties in many countries are a percentage of the grossvalue of the oil or gas produced. Accounting for these royalties as income taxes obtained better income tax treatment inthe United States. This suggests that net present value data for such reserves should always be used on an after-tax basis.

Texaco’s “Other East” clearly fits the category just described, with an estimated tax rate of 64% in 1999[$7,665/($7,665 � $4,323)].

when all cash flows are evaluated on a pretax basis, pretax present values would beused for consistency.

Example: Texaco

To illustrate, we use the data provided by Texaco and the following assumptions:

1. No change in prices or costs

2. A 10% discount rate

3. Pretax net present values for U.S. reserves but after-tax present values for foreign re-serves.15 The data provided can be used to adjust Texaco’s equity at December 31,1998 and 1999, for the difference between the present value of its oil and gas re-serves and the carrying amount:

Years Ended December 31

Standardized Measure 1998 1999

United States* $ 4,879 $15,604

Europe 1,382 4,990

Other areas** $ (1,116 $23,909

Total $ 9,375 $26,502

Carrying amount $12,190 $13,038

Excess $ (2,815) $13,464

Reported equity $11,833 $12,042

Adjusted equity $ 9,018 $25,506

% change �24% 112%

Total debt $ 7,291 $ 7,647

Debt-to-equity ratio

Reported 0.62 0.64

Adjusted 0.81 0.30

*Using United States 1999 as an example, $15,604 was calculated asthe net present value ($11,352) plus the estimated present value of in-come tax payments ($4,252). The later is estimated by applying theratio, ($11,352/$22,168) � ($8,304) and assuming a constant rate.**Sum of Other West, Other East, and Affiliate (after-tax) present values.

This adjustment more than doubles Texaco’s equity at December 31, 1999; for 1998the adjustment reduces equity by 24% because of low oil and gas prices on that date. Theadjustment sharply reduces Texaco’s debt-to-equity ratio in 1999. Varying the discountrate or making assumptions about changes in prices or costs would also lead to differentadjustments.

The adjustment of net worth is not an end in itself, but one step in the analysis of a firm.Although equity after adjustment is not a precise measure of the market value of Texaco’snet assets, it is a better measure than the historical cost of those assets. Chapter 17 discussesthe usefulness of equity adjustments in greater detail.

Adjustments for Subsequent Price Changes

In 2000, natural gas prices rose sharply from the year-end 1999 levels. As a result the De-cember 31, 1999 present value data no longer reflected the economic value of Texaco’s re-serves. Exhibit 7B-1 shows the assumptions and calculations required to adjust the 1999standardized value of U.S. reserves for subsequent price changes.


15See footnote 14.


EXHIBIT 7B-1. TEXACO—UNITED STATESAdjustments to Present Values for Subsequent Price ChangesAmounts in $ millions except for reserve quantities (oil in millions of barrels, gas in billions ofcubic feet)

A. Future Cash Inflows

Quantity Unit Price Cash Flows

December 31, 1999Crude oil and natural gas liquids 1,361 $25.60 $34,842Natural gas 3,388 2.33 $47,894

$42,736December 31, 2000 Estimated

Crude oil and natural gas liquids 1,361 $26.80 $36,475Natural gas 3,388 9.77 $33,101

$69,576B. Standardized Measure

Reported Adjusted Explanation

Future cash inflows $ 45,281 $ 69,576 Part AFuture production costs (10,956) (12,052) 20% higherFuture development costs $1(3,853) $1(4,238) 20% higherPretax net cash flow $ 30,472 $ 53,286Future income tax expense $1(8,304) $(14,521) Same rateNet future cash flows $ 22,168 $ 38,765Discount (10% rate) $(10,816) $(18,914) Same rateStandardized measure $ 11,352 $ 19,851

C. Discussion

The objective is to recompute the standardized measure using price changes at a later period. In part A,we estimate the future cash flows associated with Texaco’s U.S. reserves, using reserve quantities fromTable I of the 1999 supplementary data and prices obtained from the futures market at December 31,1999. Our computed future cash flows of $42.7 billion is nearly 6% below the $45.3 billion shown inTable II. The difference must be due to different prices as the standardized measure must use provedreserves.

We estimate future cash flows at December 31, 2000 using the same reserve quantities but withprices at December 31, 2000. These calculations produce future cash flows of $69.6 billion, 63% higherthan the December 31, 1999 level.

In part B, we adjust each component of the standardized measure to estimated levels at December31, 2000. The future cash inflows come from part A. We assume 20% increases in both future produc-tion costs and future development costs, on the assumption that the cost of drilling equipment and ser-vices rises with higher oil and gas prices. We assume the same tax rate (27.25%). We also assume thesame production time pattern so that the % discount is unchanged. These calculations produce a 75%increase in the standardized measure for U.S. oil and gas reserves, to $19.8 billion. The actual standard-ized value (see Exhibit 7BP-1) at December 31, 2000 was just under $18 billion. The major reason forthe difference was that reserves declined during 2000, reducing future cash flows to the followingamounts:

December 31, 2000 Actual Quantity Unit Price Cash Flows

Crude oil and natural gas liquids 1,202 $ 26.80 $ 32,214Natural gas 3,299 9.77 $332,231

$ 64,445

Lower reserves reduces future cash flows and, therefore, lowers the standardized value.

PROBLEMS W31

Changes in Present Values

Table III is a reconciliation of changes in the standardized measure, akin to the reconcili-ation of reserve quantities. But these data are richer as they include the impact of suchfactors as:

• Changes in prices and costs

• Accretion of discount (the passage of time reduces the discount period)

• Expenditures that reduce future required cash flows

• Changes in estimates

• Purchases and sales of reserves

• Effect of production

The standardized measure of Texaco’s oil and gas reserves declined by nearly one-third in1997 and more than half in 1998, but soared to a higher level at December 31, 1999. The rec-onciliation provides the following insights:

1. Changing prices and costs were the major factor accounting for the sharp decline inthe standardized measure in 1997 and 1998 and its recovery in 1999.

Over the three-year period, the price effect was slightly negative.

2. Texaco’s quantity revisions were positive each year, suggesting that the company’sestimates have been conservative.

3. Timing effects were negative each year, suggesting that Texaco’s production ratewas below previous forecasts.16

Summary and Conclusion While the supplemental oil and gas data mandated bySFAS 69 must be used with care, they provide considerable useful information regardingthe firm’s exploratory activities and the value of its reserves. These data are far morecomparable among firms than reported financial data as most are unaffected by account-ing methods.

PROBLEMS

7B-1. [Changes between full cost and successful efforts methods] Sonat [SNT], a diversifiedenergy company, announced the following accounting change when it reported its re-sults for the third quarter of 1998:

Sonat Exploration Company [Sonat subsidiary] changed from successful efforts to full costaccounting because its future capital spending will be focused significantly more on explo-ration activity than in the past. Full cost accounting, which amortizes rather than expensesdry-hole exploration and other related costs, provides a more appropriate method of match-ing revenues and expenses. Exploration activity has increased from 6 percent of 1995 capi-tal spending, or $27 million, to an estimated 33 percent of 1998 capital spending, orapproximately $175 million. . . .

The adoption of the full cost method is expected to increase 1998 and 1999 normalizedearnings from levels that would have been reported under successful efforts accounting and,more important, will reduce earnings volatility from quarter-to-quarter and year-to-yeargoing forward. . . . The change to full cost accounting will not materially affect the com-pany’s cash flow from operations.

Sonat has restated all prior period statements . . . all previous charges related to theimpairment of Sonat Exploration’s assets . . . were reversed, which significantly raisedthe book value of those properties as well as Sonat’s stockholders’ equity. The full costmethod, however, requires quarterly ceiling tests17 to insure that the carrying value of as-sets on the balance sheet is not overstated. . . . The end result of the full cost conversion

16Postponing production reduces the net present value by increasing the discount factor.17Authors’ note: see footnote 3 to this appendix and the related text.


is that both the book value of Sonat Exploration’s properties and Sonat’s stockholders’equity are at higher levels than if it had continued with the successful efforts method ofaccounting.18

Note 2 to Sonat’s annual report for the year ended December 31, 1998 reports the fol-lowing effects of the accounting change and restatement of prior periods:

Effect on 1996 1997 1998

Net income ($thousands) 18,006 130,584 (258,351)

Earnings per share, fully diluted .16 1.17 (2.35)

The 1998 income statement reports ceiling test charges of $1,035,178 thousand. Re-tained earnings at January 1, 1996 were increased by $199,196 thousand for the ac-counting change.

A. Explain each of the following benefits from the accounting change stated in theSonat press release:

(i) Increased normalized earnings

(ii) Reduced earnings volatility

(iii) Higher book value of exploration properties

(iv) Higher stockholders’ equity

B. Compute the effect of the accounting change on Sonat’s stockholders’ equity atDecember 31, 1998.

C. Describe the effect of the accounting change on each of the following Sonat ratiosfor 1998:

(i) Debt-to-equity ratio

(ii) Asset turnover

(iii) Book value per share

D. Explain why the accounting change was not expected to materially affect Sonat’scash from operations.

E. Given your answers to parts A through D, evaluate Sonat’s decision to change ac-counting method.

F. The accounting change took place during a period of declining energy prices. De-scribe the risk of making the accounting change and illustrate that risk using thedata provided.

G. Sonat had changed from the full cost method to successful efforts in 1991, a pre-vious period of energy price declines. Describe the effect of that fact on your viewof the 1998 accounting change.

7B-2. [Analysis of Supplementary Oil and Gas Data] Exhibit 7BP-1 contains the supple-mental oil and gas data from Texaco’s 2000 annual report. Use this exhibit, and thedata for 1999 and prior years from Texaco’s 1999 annual report, to answer the follow-ing questions.

A. Compute Texaco’s reserve lives in years for 2000, for both oil and gas:

(i) In the United States

(ii) Worldwide

B. Discuss whether production trends mirror the reserve trends over the four yearsended December 31, 2000.

C. Compute Texaco’s capitalized cost per BOE for 2000:

(i) In the United States

(ii) Worldwide

18Sonat press release, October 22, 1998.

PROBLEMS W33

EXHIBIT 7BP-1. TEXACO 2000Supplemental Oil And Gas Information

Note: These disclosures omit text and tables that duplicate the 1999 disclosures.

Table I—Net Proved ReservesNet Proved Reserves of Crude Oil and Natural Gas Liquids (millions of barrels)

Consolidated Subsidiaries Equity

Affiliate AffiliateUnited Other Other —Other —Other World-States West Europe East Total West East Total wide

As of December 31, 1999* 1,782 55 427 670 2,934 — 546 546 3,480Discoveries & extensions 39 — 21 9 69 374 — 374 443Improved recovery 25 — — 39 64 — 14 14 78Revisions (21) — 9 30 18 — 37 37 55Net purchases (sales) (135) (52) (44) — (231) — — — (231)Production (130) (3) (44) (78) (255) — (52) (52) (307)

Total changes (222) (55) (58) — (335) 374 (1) 373 38Developed reserves 1,202 — 219 559 1,980 — 282 282 2,262Undeveloped reserves 358 — 150 111 619 374 263 637 1,256

As of December 31, 2000* 1,560 — 369 670 2,599 374 545 919 3,518

*Includes net proved NGL reservesAs of December 31, 1998 250 — 68 22 340 — 6 6 346As of December 31, 1999 250 — 74 134 458 — 1 1 459As of December 31, 2000 219 — 67 162 448 — 1 1 449

Net Proved Reserves of Natural Gas (billions of cubic feet)


Affiliate AffiliateUnited Other Other —Other —Other World-States West Europe East Total West East Total wide

As of December 31, 1999 4,205 941 962 1,866 7,974 — 134 134 8,108Discoveries & extensions 585 — — — 585 33 4 37 622Improved recovery 5 — — — 5 — — — 5Revisions 121 12 43 164 340 — 8 8 348Net purchases (sales) 8 (58) (11) — (61) — — — (61)Production (494) (95) (81) (36) (706) — (24) (24) (730)

Total changes 225 (141) (49) 128 163 33 (12) 21 184Developed reserves 3,299 738 573 977 5,587 — 121 121 5,708Undeveloped reserves 1,131 62 340 1,017 2,550 33 1 34 2,584

As of December 31, 2000 4,430 800)* 913 1,994 8,137)* 33 122 155 8,292)*

*Additionally, there are approximately 302 BCF of natural gas in Other West which will be available from production during the period 2005–2016 undera long-term purchase associated with a service agreement.


EXHIBIT 7BP-1 (continued)

The following chart summarizes our experience in finding new quantities of oil and gas to replace our production. Our reserve replace-ment performance is calculated by dividing our reserve additions by our production. Our additions relate to new discoveries, existing re-serve extensions, improved recoveries, and revisions to previous reserve estimates. The chart excludes oil and gas quantities frompurchases and sales.

Worldwide United States International

Year 2000 172% 76% 267%Year 1999 111% 99% 124%Year 1998 166% 144% 191%3-year average 150% 109% 192%5-year average 146% 108% 189%

Table II—Standardized MeasureConsolidated Subsidiaries

United Other Other(Millions of Dollars) States West Europe East Total

As of December 31, 2000Future cash inflows from sale of oil & gas,

and service fee revenue $ 67,115 $ 1,559 $ 10,549 $ 15,512 $ 94,735Future production costs (13,107) (252) (2,074) (2,768) (18,201)Future development costs (3,588) (30) (1,244) (1,280) (6,142)Future income tax expense (17,024) (612) (2,238) (6,681) (26,555)

Net future cash flows before discount 33,396 665 4,993 4,783 43,83710% discount for timing of future cash flows (15,407) (259) (1,778) (2,239) (19,683)

Standardized measure ofdiscounted future net cash flows $ 17,989 $ 406 $ 3,215 $ 2,544 $ 24,154

Equity

Affiliate Affiliate—Other —Other World-

(Millions of Dollars) West East Total wide

As of December 31, 2000Future cash inflows from sale of oil & gas,

and service fee revenue $ 3,917 $ 7,873 $ 11,790 $ 106,525Future production costs (273) (2,853) (3,126) (21,327)Future development costs (406) (694) (1,100) (7,242)Future income tax expense (1,101) (2,189) (3,290) (29,845)

Net future cash flows before discount 2,137 2,137 4,274 48,11110% discount for timing of future cash flows (1,431) (809) (2,240) (21,923)

Standardized measure ofdiscounted future net cash flows $ 706 $ 1,328 $ 2,034 $ 26,188


Table III—Changes in the Standardized MeasureWorldwide Including Equity in Affiliates

(Millions of Dollars) 2000 1999 1998

Standardized measure � beginning of year $ 18,710 $ 5,487 $ 12,057Sales of minerals-in-place (3,990) (352) (160)

14,720 5,135 11,897

Changes in ongoing oil and gas operations:Sales and transfers of produced oil and gas,

net of production costs during the period (7,345) (4,276) (3,129)Net changes in prices, production, and development costs 11,389 22,036 (11,205)Discoveries and extensions and improved recovery, less related costs 4,543 1,821 728Development costs incurred during the period 2,043 1,598 1,770Timing of production and other changes 670 (517) (1,170)Revisions of previous quantity estimates 668 301 852Purchases of minerals-in-place 901 895 48Accretion of discount 3,120 881 1,916Net change in discounted future income taxes (4,521) (9,164) 3,780

Standardized measure—end of year $ 26,188 $ 18,710 $ 5,487

Table IV—Capitalized CostsConsolidated Subsidiaries Equity

Affiliate AffiliateUnited Other Other —Other —Other World-

(Millions of Dollars) States West Europe East Total West* East Total wide

As of December 31, 2000Proved properties $18,213 $137 $3,295 $3,699 $25,344 $ 66 $1,370 $1,436 $26,780Unproved properties 1,026 98 58 655 1,837 68 265 333 2,170Support equipment and facilities 257 81 28 135 501 42 906 948 1,449

Gross capitalized costs 19,496 316 3,381 4,489 27,682 176 2,541 2,717 30,399Accumulated depreciation,

depletion, and amortization (12,084) (92) (1,821) (1,508) (15,505) (1) (1,349) (1,350) (16,855)

Net capitalized costs $ 7,412 $224 $1,560 $2,981 $12,177 $175 $1,192 $1,367 $13,544

*Existing costs were transferred from a consolidated subsidiary to an affiliate at year-end 2000.

Table V—Costs Incurred

On a worldwide basis, in 2000 we spent $3.62 for each BOE we added. Finding and development costs averaged $3.74 for the three-year period 1998–2000 and $3.92 per BOE for the five-year period 1996–2000.


Affiliate AffiliateUnited Other Other —Other —Other World-

(Millions of Dollars) States West Europe East Total West East Total wide

For the year ended December 31, 2000Proved property acquisition $ 138 $ — $ — $ 276 $ 414 $ — $ — $ — $ 414Unproved property acquisition 5 12 — — 17 — — — 17Exploration 227 62 18 287 594 — 19 19 613Development 716 121 334 677 1,848 — 169 169 2,017

Total $1,086 $195 $352 $1,240 $2,873 $ — $188 $188 $3,061

PROBLEMS W35



Table VI—Unit Prices

Average sales prices are calculated using the gross revenues in Table VII. Average lifting costs equal production costs and thedepreciation, depletion, and amortization of support equipment and facilities, adjusted for inventory changes.

Average Sales Prices

Affiliate AffiliateUnited Other Other —Other —OtherStates West Europe East West East

Crude oil (per barrel)2000 $26.20 $22.74 $26.86 $22.81 $ — $21.521999 14.97 14.12 17.15 15.33 — 13.241998 10.40 9.65 11.73 9.61 — 9.81

Natural gas liquids (per barrel)2000 18.73 — 17.93 — — —1999 10.86 — 12.53 — — —1998 8.99 — 11.89 — — —

Natural gas (per thousand cubic feet)2000 3.67 1.13 2.49 1.23 — —1999 2.07 .77 1.99 .18 — —1998 1.93 .92 2.42 .38 — —

Average lifting costs (per barrel of oil equivalent)

Affiliate AffiliateUnited Other Other —Other —OtherStates West Europe East West East

2000 $5.05 $2.94 $5.08 $3.03 $ — $5.061999 4.01 2.87 6.15 3.45 — 3.951998 4.07 1.86 5.24 3.65 — 2.68

Table VII—Results of OperationsConsolidated Subsidiaries

United Other Other(Millions of Dollars) States West Europe East Total

For the year ended December 31, 2000Gross revenues from:

Sales and transfers, including affiliate sales $ 4,460 $ — $ 869 $ 1,440 $ 6,769Sales to unaffiliated entities 545 190 591 315 1,641

Production costs (1,070) (46) (375) (232) (1,723)Exploration costs (130) (62) (18) (152) (362)Depreciation, depletion, and amortization (723) (18) (221) (147) (1,109)Other expenses (190) (27) (2) (88) (307)

Results before estimated income taxes 2,892 37 844 1,136 4,909Estimated income taxes (972) (48) (269) (945) (2,234)

Net results $ 1,920 $ (11) $ 575 $ 191 $ 2,675


Equity

Affiliate Affiliate—Other —Other World-

(Millions of Dollars) West East Total wide

For the year ended December 31, 2000Gross revenues from:

Sales and transfers, includingaffiliate sales $ — $831 $831 $7,600Sales to unaffiliated entities — 50 50 1,691

Production costs — (223) (223) (1,946)Exploration costs — (14) (14) (376)Depreciation, depletion,

and amortization — (129) (129) (1,238)Other expenses — (2) (2) (309)

Results before estimated income taxes — 513 513 5,422Estimated income taxes — (258) (258) (2,492)

Net results $ — $255 $255 $2,930

Source: Texaco 2000 Annual Report

PROBLEMS W37

D. Discuss the trend, over 1998–2000, in Texaco’s capitalized cost per BOE, and ex-plain how changes in reserve quantities and capitalized costs may have affectedthat trend.

E. Review the data in Tables II and III and discuss the effect of each of the followingfactors on the change in the standardized value over the four years ended Decem-ber 31, 2000:

(i) Price changes

(ii) Revision of estimated reserve quantities

(iii) Income taxes

F. Discuss, based on your answers to part E, the extent to which Texaco replaced theeconomic value of its reserves over the four years ended December 31, 2000.

G. Texaco’s reported debt at December 31, 2000 was $7,191 million with reportedequity of $13,444.

(i) Compute Texaco’s equity adjusted to replace the carrying cost of reserveswith the standardized value.

(ii) Compute Texaco’s debt-to-equity ratio using both reported and adjusted eq-uity.

(iii) Discuss the effect of the adjustment on the trend of Texaco’s debt-to-equityratio over the period 1998 to 2000.

(iv) Describe the effect of the adjustment on Texaco’s asset turnover ratio.

H. The equity adjustment would appear to reduce Texaco’s return on equity.

(i) Discuss how you could adjust income, using the standardized measure, tocompute a current cost return on equity.

(ii) Explain how current cost ROE would be superior to reported ROE as a per-formance measure.

(iii) Describe one drawback to using current cost ROE as a performance measure.

Appendix 8-AANALYSIS OF CHANGING PRICESINFORMATION

APPENDIX OBJECTIVES

1. Distinguish between general inflation and specific price changes.2. Describe and illustrate the constant dollar and current cost methods of adjustment for

changing prices.3. Show how corporate disclosures can be used to adjust financial statements for the ef-

fects of price changes.4. Show how corporate disclosures regarding capital expenditures can provoke ques-

tions intended to provide insights into corporate strategy.

INTRODUCTION

Price changes have pervasive effects on financial statements, and good analysis must recog-nize those effects and incorporate them into valuation decisions. Before discussing these is-sues, it is important to distinguish between two types of price change: general inflation andspecific price change.

General inflation refers to price changes for an economy as a whole. Indices such as theconsumer price index in the United States attempt to measure the impact of price changes onthe broad population. Specific price changes refer to the prices of specific goods and servicesthat are the inputs and outputs of firms in a given industry.

ANALYSIS OF GENERAL INFLATION

From the financial analysis point of view, the impact of general inflation is that the purchas-ing power of capital is continuously eroded. Analytically, there is a well-developed methodof dealing with this phenomenon, constant dollar accounting, also called general price levelaccounting or purchasing power accounting.1 Its goal is to measure the impact of changes inpurchasing power (general inflation) on the financial capital of the firm.

In the simple model depicted in Exhibit 8A-1, the firm invests its capital in inventory atthe start of the first year and sells that inventory at the end of the year. At the beginning ofthe next year, it again invests its capital (obtained from the sale of inventory one day earlier)in inventory. For simplicity, we assume that there are no markups and no expenses other thancost of goods sold.

The historical cost (or nominal dollar) model recognizes as income the difference betweenthe proceeds of sale and the cost of inventory for each year. The total income over the three-yearperiod is $331, the difference between beginning capital ($1,000) and ending capital ($1,331).

W38

1Accounting Principles Board (APB) Statement 4 (1969), Financial Statements Restated for General Price-LevelChanges.

This model, however, does not recognize the decline in the real value of money or finan-cial capital due to inflation. In this case, the purchasing power of $1,000 declines (at the rateof 25%) to $800 ($1,000/1.25) in one year.

Constant Dollar Method

The constant dollar method recognizes this effect by restating all monetary amounts intounits of constant purchasing power at a designated base period, which can be any period oftime (all of 2001) or point in time (January 1, 2001). The base fixes the yardstick used tomeasure purchasing power.

In our example, January 1, 2001 is the base so that all cash flows will be restated intounits of January 1, 2001 purchasing power.

The nominal dollar cash flow of $1,100 was received at December 31, 2001. Inflation re-duces the purchasing power of those dollars to only 80% (1/1.25) of the purchasing power atthe base date of January 1, 2001. Thus, we must divide the cash flow by the relevant index(1.25) to obtain revenues in January 1, 2001 dollars.

Cost of goods sold (COGS) resulted from a cash outflow at January 1, 2001, and, there-fore, requires no restatement. In constant dollar terms, therefore, net income for 2001 equals

2001 Income � $880 � $1,000 � $(120)

2001 Sales ($1/1/01)� $1,1001.25

� $880

ANALYSIS OF GENERAL INFLATION W39

EXHIBIT 8A-1. ACCOUNTING FOR CHANGING PRICES

Assumptions: Capital at January 1, 2001, is $1000.Each January 1, the firm will invest entire capital in inventory.Each December 31, the firm will sell entire inventory.Price of inventory is $100 per unit at January 1, 2001, and rises at 10% per annum.The general price level (CPI-U) rises at 25% per annum. Base period is January 1, 2001 � 100.

Historical Cost Model

Cost ofYear Sales Goods Sold Income

2001 $1100 $1000 $ 1002002 1210 1100 $(1102003 1331 1210 $(121

Total $ 331

Constant Dollar Model

(January 1, 2001 dollars)2001 880 1000 $(120)2002 774 880 $(106)2003 681 774 $1(93)

Total $(319)

Current Cost Model

2001 1100 1100 02002 1210 1210 $(0002003 1331 1331 $(000

Total $ 0

CPI-U: January 1, 2001 100 December 31, 2001 125January 1, 2002 125 December 31, 2002 156.25January 1, 2003 156.25 December 31, 2003 195.31

W40 APPENDIX 8-A ANALYSIS OF CHANGING PRICES INFORMATION

In purchasing power terms, the firm’s capital has declined. This results from the fact thatits inventory rose in price by less than the rate of inflation.

For 2002, we compute income in the same manner. The December 31, 2002 cash inflowhas lost purchasing power over a two-year period and the January 1, 2002 cash outflow mustbe adjusted for one year’s inflation:

Thus

The calculations for 2003 are similar, resulting in

Over the three-year period, the constant dollar method reports a loss of $319 in purchas-ing power of the firm’s capital. At the end of 2003, the firm has $1,331, the proceeds of in-ventory sold at December 31, 2003. But in units of 1/1/01 purchasing power, the firm’scapital is only $681 ($1,331/1.9531), whereas its original capital was $1,000.

Note that these computations use the company’s actual cash flows but the price index isfor the economy as a whole. The calculations do not take into account the specific pricechanges faced by the firm. This feature of the constant dollar method is both its strength andits weakness.

Advantages and Disadvantages of Constant Dollar Method

The constant dollar method involves very simple calculations and the erosion of purchasingpower is a simple economic concept. The method facilitates audits because it is objective asthe only choice involved is that of the inflation index, and given the same data, the resultswill always be the same, contributing to ease of verifiability. For these reasons, corporate fi-nancial statement preparers and auditors have generally supported use of the constant dollarmethod to disclose the impact of inflation.

From the standpoint of financial analysis, however, the constant dollar method has a sig-nificant drawback: Constant dollar data do not have any apparent usefulness. Although lossof purchasing power is a useful economic concept, it has limited application in the financialworld. Stock prices, interest rates, and other financial data are stated in nominal currencyunits, not real (purchasing power) units.

ANALYSIS OF FIRM-SPECIFIC INFLATION

Contributing to the lack of utility of constant dollar data is their lack of specificity; they treatall companies identically regardless of the composition of their assets and liabilities. For datathat relate to specific companies, analysts prefer the current cost method.

Current Cost Method

The current cost2 method ignores general inflation in favor of the specific price and costchanges faced by the individual firm. It starts with the idea that income, when properly mea-sured, must include a provision for the replacement of capacity used during the period.3 Oth-erwise, income is overstated as it includes the consumption of capacity.4

2003 Income ($1/1/01) � $681 � $774 � $(93)

2002 Income ($1/1/01) � $774 � $880 � $(106)

2002 Sales ($1/1/01) � $1,2101.5625

� $774 � COGS � $1,1001.25

� $880

2Current cost is the term used in SFAS 33 and other FASB standards. Previous accounting literature used such termsas replacement cost, current value, and fair value. The distinction among these terms is often more theoretic thanreal and varies with the user. For simplicity, we ignore these distinctions throughout the appendix.3J. R. Hicks, Value and Capital, 2nd ed. (Oxford: Chaundon Press, 1946), p. 176.4This concept was more fully developed in Chapter 2.

It follows that the provision for the cost of replacing capacity must be made at currentprices. Although application of this principle is difficult in practice, it is essential in theory.If a firm has used up a machine and must replace it to remain in business, it is the cost ofbuying the new machine that is relevant, not the original cost of the worn-out one.

The current cost method, therefore, measures income by matching revenues with operat-ing costs, including the cost of replacing inventory sold and fixed assets used up during theperiod.

Exhibit 8A-1 applies this principle to our model company. At the end of 2001, the firmhas $1,100 as proceeds of sales. To remain in business, the firm must purchase new inven-tory on January 1, 2002. The cost of that new inventory will be 1,100 (10 @ $110 per unit),as prices have risen by 10% since January 1, 2001. Under the current cost method, therefore,there was no income earned in 2001:

The firm can purchase 10 units of inventory, the same as its “capacity” one year earlier.The firm has neither a profit nor a loss for 2001 but has simply maintained its physical capi-tal (capacity to do business). This contrasts with the constant dollar method, which is con-cerned with maintaining financial capital.

2002 and 2003 results are the same. There is no income in current cost terms because thefirm has simply maintained its physical capital.

Disadvantages of Current Cost

As compared with the constant dollar method, the current cost method is more complex: thefirm must estimate the cost to replace each type of inventory and each category of fixed as-sets. We discuss the difficulty of estimating current costs shortly. These estimates requirejudgements about how the firm will replace used up capacity, adding subjectivity and a lackof reliability to the results. Because of these factors, current cost data are more expensive andtime-consuming to prepare and audit than constant dollar data. For all these reasons, finan-cial statement preparers and auditors have mostly opposed the presentation of current costdata in financial statements. In some cases, however, corporations have stated that they findsuch data useful when managing their business.

For financial analysis, however, current cost data are greatly preferred to constant dollardata. The main reason is the relevance of such data to the operations of specific firms.

Accounting Series Release 190

The high rate of inflation in the 1970s and large specific price changes in some industries ledthe Securities and Exchange Commission to issue Accounting Series Release (ASR) 190(1976) requiring large firms to disclose the replacement cost of inventory and fixed assets aswell as cost of goods sold and depreciation expense computed on a replacement cost basis.Disclosures were first required in 1976.

At about the same time, the FASB placed inflation accounting on its agenda and issuedSFAS 33 in 1979, at which time the SEC withdrew ASR 190.

SFAS 33 Requirements

SFAS 33, Financial Reporting and Changing Prices, the first U.S. accounting standard to re-quire disclosure of the impact of changing prices, was a hybrid; it attempted to combine boththe current cost and constant dollar methods into one standard. In theory, the two approachescan be combined. Data adjusted for specific price changes can then be further adjusted forchanges in purchasing power. The resulting complexity, however, made use of this data diffi-cult for financial analysts.

SFAS 33 provided for review after five years. SFAS 89 (1989) made the SFAS 33 dis-closure requirements voluntary. This action resulted from three factors. First, the rate of in-flation subsided greatly in the 1980s, making the issue of general inflation effects lessimportant. Second, preparers and auditors complained that the costs of compliance with

2001 Income � $1,100 � $1,100 � 0

ANALYSIS OF FIRM-SPECIFIC INFLATION W41


SFAS 33 were too high. Finally, little or no benefit could be traced to the disclosures. Be-cause of the voluntary nature of SFAS 89, the disclosures are rarely provided.

Problems with SFAS 33 Disclosures

The data disclosed under the provisions of SFAS 33 received little use, we believe, for thefollowing reasons:

1. It was unclear whether companies should attempt to measure the market value, thereproduction cost, or the replacement cost of existing capacity. Each of these choicesresults in a different measure of cost and a different set of problems.

2. Market value is often difficult to estimate because many productive assets are cus-tomized or unique. Although market values can be estimated for office buildings, forexample, there is no active market for steel mills. Curiously, the FASB did not per-mit the disclosure of market values in lieu of current cost for such assets as oil andgas properties, timberland, and real estate, for which active markets do exist.5

3. Reproduction cost is an estimate of the cost to build existing facilities at currentprices. However, it is hard to price machines that are no longer being manufactured(having been replaced by newer models or machines using different productionprocesses). Use of reproduction cost also assumes that the firm would replace its ex-isting capacity with exactly the same mix of factory sizes and locations.

Replacement cost is, in theory, the cost of replacing existing productive capacity. Suchan estimate must, first, define whether capacity should be measured in physical units (tons ofsteel or pairs of shoes) or financial units (dollars of revenue). Second, the firm must decidewhat mix of geographic locations and plant capacities it would construct if it were to replaceits facilities today. Finally, the firm must estimate what production processes, raw and inter-mediate materials, and markets it would pursue if it could “start from scratch.”

The computations become increasingly speculative as one moves from the market valueof assets to reproduction cost to replacement cost. In many cases, companies complied withSFAS 33 by simply applying construction and machinery cost indices to the historical cost offixed assets.

Problems with Current Cost Depreciation

SFAS 33 also required that companies providing current cost data disclose depreciation ex-pense on a current cost basis. At first glance, this is a simple exercise; companies simplyapply their existing depreciation methods and lives to their estimated current cost of fixedassets.

The difficulties in defining current cost carry over to the definition of current cost depre-ciation expense. In addition, the interpretation of current cost depreciation expense is subjectto another problem. Replacement of historical cost depreciation with current cost deprecia-tion assumes that the operating costs of the firm are unaffected by the “replacement” process.It assumes that more expensive new machines and processes are no more cost efficient thanthe original machines and processes.

That assumption is, of course, absurd in most cases. In theory, therefore, the operatingcosts of the firm should be adjusted to reflect the greater efficiency of the new equipment.Such adjustments are subjective when made by the firm; a financial analyst outside the firmcannot begin to make them.

Because of the subjectivity of the data, lack of comparability of disclosures by compet-ing firms, difficulty of interpreting the data, and lack of a well-defined way of incorporatingthe data into investment decision models, use of the current cost data provided by SFAS 33was limited. Perhaps for that reason there is little evidence that current cost data impacted fi-nancial markets.

5SFAS 39, Mining and Oil and Gas, SFAS 40, Timberlands, and SFAS 41, Income Producing Real Estate, were allissued in 1980 as supplements to SFAS 33.

Adjusting Financial Statements for Changing Prices

Given the voluntary nature of changing prices disclosures under SFAS 89, the analysis of theimpact of changing prices must be done by each analyst. As we believe that constant dollarcalculations are of use only under limited circumstances (see the following section), we de-vote our attention to adjustments for specific price changes. As the effects of changing priceson inventories are dealt with in Chapter 6, we concern ourselves here only with the effects onfixed assets.

Changing prices for fixed assets have two primary effects on financial statements:

1. Since fixed assets are carried at cost (net of accumulated depreciation), their carryingamount does not reflect the current cost. Thus, the assets and the net worth are under-stated if prices have risen (the normal case).

2. Depreciation expense is also understated because it is based on the understated carry-ing amount of the fixed assets. Depreciation expense, which should be a measure ofthe capacity used up during the period, is instead just an arbitrary allocation of pastcash flows. Understatement of depreciation expense results in the overstatement ofreported earnings.

Adjustments to Fixed Assets

Some non-U.S. companies disclose asset values used for insurance or tax assessment purposes.

Example: Holmen

Footnote 10 of Holmen’s financial statements shows the assessed tax values of properties inSweden in 1998 and 1999. Exhibit 8A-2 shows how these data can be used to adjust tangibleassets and shareholders’ equity.

For each fixed asset category, we have computed the excess of tax values over carryingvalues for the four years ended in 2000. This procedure underestimates the difference as thetax values exclude properties outside of Sweden. Most of the excess relates to Holmen’s for-est properties. If these properties had not been revalued in prior years, the cost (acquisitionvalue) of these properties would be a very misleading indicator of their worth.

The total excess value was SKr 6.3 billion at the end of 1997, but declined sharply in1998. No explanation is provided, but we note that Modo Paper was spun off as a separatecompany in 1998, removing its fixed assets from the analysis. In 1999, the excess value re-lated to forest properties increased but the excess related to buildings declined. In 2000 therewas a small increase in the excess values.

Exhibit 8A-2 shows that adjustment for the excess of assessed tax values over carryingvalue increases tangible assets by as much as 32.5% (1997) and stockholders’ equity by asmuch as 38.8% (1997).

In the absence of company-provided data, the analyst must use other sources of informa-tion to make adjustments. In some cases, data on the cost of capacity are available from in-dustry sources; this is more likely to be true for relatively homogeneous industries such aspaper, oil refining, and chemicals. Cost per ton of capacity data for such industries is fre-quently cited in trade publications or can be gleaned from company contacts.

Another possible source of data is actual construction. Companies frequently report thecost and capacity of new plants. Such data from the company or its competitors can be usedto estimate the current cost of existing facilities.

Yet another approach is the use of construction cost statistics. If the year of constructionof a plant is available, the historical cost can be indexed to estimate the current constructioncost of the same facility.

For real estate assets, current land and construction cost data are frequently included inindustry publications. The analyst can use this data to estimate the current cost of construc-tion for factories, warehouses, and so forth. For some categories of real estate, especially in-come-producing properties (office buildings, shopping centers, hotels), publicly availablemarket value estimates should be used as the measure of current costs as market value is



more relevant than reproduction cost. Acquisitions accounted for under the purchase methodresult in the restatement of acquired fixed assets to their fair value or current cost.

In some cases, industry-specific disclosures are available. For example, see the discussionof the disclosures of the net present value of oil and gas reserves discussed in Appendix 7-B.

All these approaches require estimates. The lack of precision does not mean that the ex-ercise is not worthwhile. Remember that estimates are present in the reported financial state-ments as well.

Using Current Cost Asset Values

The main use for current cost asset data is to prepare a current cost balance sheet. The histor-ical cost of all assets and liabilities should be replaced with the current cost (market value) ofthose assets. As compared with the historical cost balance sheet, a current cost balance sheet

EXHIBIT 8A-2. HOLMENTangible Fixed AssetsAmounts in SKr millions

Years Ended December 31

1997 1998 1999 2000

Forest and Agricultural Property

Acquisition values 285 310 309 312Accumulated depreciation — — — —Accumulated revaluations 14,275 14,275 14,268 14,268Net carrying value 4,560 4,585 4,577 4,580

Assessed tax values* 8,474 6,050 6,699 7,026

*Sweden only

Excess tax values 3,914 1,465 2,122 2,446

Buildings, Other Land, Etc.

Acquisition values 4,373 5,065 3,341 3,689Accumulated depreciation (2,432) (2,608) (1,640) (1,756)Accumulated revaluations 11,108 11,108 11,104 11,104Net carrying value 2,049 2,565 1,805 2,037

Assessed tax values* 4,496 4,012 2,701 3,149

*Sweden only

Excess tax values 2,447 1,447 896 1,112

Total excess values 6,361 2,912 3,018 3,558

Total tangible assets 19,551 20,707 14,825 16,129Adjusted tangible assets 25,912 23,619 17,843 19,687

% increase 32.5% 14.1% 20.4% 22.1%

Stockholders’ equity 16,375 18,377 15,883 17,014Adjusted equity 22,736 21,289 18,901 20,572

% increase 38.8% 15.8% 19.0% 20.9%

Source: Holmen Annual Reports, 1998–2000

provides a better measurement of the net assets available to management. These data can beused to make a better evaluation of management’s use of available resources, the borrowingcapacity of the firm, security for creditors, and the liquidation value of the company. Theseissues will be discussed more fully in Chapter 17.

Estimating Current Cost Depreciation

Once the current cost of fixed assets is estimated, the next step is to estimate depreciation ona current cost basis. The current cost of fixed assets should be amortized over the estimatedeconomic life of the assets, allowing for salvage values. The arbitrarily chosen depreciationmethod and lives used for financial reporting purposes may not be adequate for this purpose.For analysis purposes, the choice of depreciation method, lives, and salvage values should becarefully considered.

It is important to look at overall corporate trends. If real output is static, then one canargue that all capital expenditures have been made to replace used up capacity. As SFAS 14requires (see Chapter 13) the disclosure of capital expenditures and depreciation expensefor each reportable segment, this analysis can be done for each segment of a multiindustrycompany.

Some companies disclose the cost of major capital projects, allowing the analyst to“back into” an estimate of “maintenance” expenditures. Other firms provide approximatedata regarding the purpose of current capital expenditures. The portion allocated to the“maintenance of existing capacity” may be a good proxy for current cost depreciation.6 Re-member that the goal is to estimate the cost to replace capacity used up during the accountingperiod.

Example: Mead.

Exhibit 8A-3 contains financial statement data for 1996–2000 for Mead [MEA], a majorpaper producer. This capital expenditures analysis breaks out the components of capitalspending: growth, maintenance, cost-effectiveness, and environmental. Over the five-yearperiod, capital expenditures declined from 215% of depreciation (1996) to 74% (2000).


EXHIBIT 8A-3. MEADCapital Expenditures Analysis

Amounts in $millions Years Ended December 31

1996 1997 1998 1999 2000 Totals

Growth* $225.4 $139.7 $158.6 $ 40.2 $ 41.2 $ 605.1Maintenance 73.9 150.8 79.1 42.8 56.4 403.0Cost-effective 96.0 122.6 117.3 102.1 83.9 521.9Environmental 0033.4 0024.2 0029.0 0027.8 0024.4 0.0138.8

Total $428.7 $437.3 $384.0 $212.9 $205.9 $1,668.8

*Including related environmental

Depreciation Expense $199.2 $238.4 $260.3 $263.2 $276.4 $1,237.5

Ratio to depreciation expense:Total capital expenditure 215% 183% 148% 81% 74% 135%Non-growth capital expenditure 102% 125% 87% 66% 60% 86%

Source: Data from Mead Annual Reports and Fact Books.

6In IAS 7, Cash Flow Statements, the IASB recommends that companies disclose the portion of capital expendituresrequired to maintain capacity.


When growth is excluded, the ratio of capital spending to depreciation expense declines from102% in 1996 to 60% in 2000.

These data raise a number of interesting questions for an analyst to pursue:

1. Mead’s “maintenance” expenditures declined sharply from 1997 to 2000 to levels farbelow depreciation expense. While these data suggest that Mead is not truly main-taining its operating capacity, we believe that “cost-effectiveness” and “environmen-tal” expenditures should be included.

2. Even so, over the five-year period, non-growth expenditures were only 86% of de-preciation expense.7 This suggests that, even by our expanded definition, operatingcapacity is being reduced. There may be lines of business that have insufficient prof-itability or growth potential to warrant new investment.

3. Growth expenditures also declined sharply over this five-year period (they had risenrapidly from pre-1996 levels). These changes may reflect industry conditions, capitalconstraints, or strategic decisions by management.

4. Mead has made significant “cost-effectiveness” investments during this period, costreductions are presumably being realized currently, whereas depreciation is under-stated by the use of historical cost. Such expenditures should increase reported in-come as a result.

These are examples of how financial analysis can suggest lines of inquiry about fundamentalbusiness issues. Analysis of segment data and discussion with management should providesome answers to these questions.

Use of Current Cost Depreciation

Estimates of current cost depreciation should be used to adjust reported income to currentcost. Along with the adjustment to last-in, first-out (LIFO) when applicable (see Chapter 6),the replacement of historic cost depreciation by current cost will produce a better measure ofsustainable income.8

Current cost data should also be used to adjust ratios so that they are better measuresof management performance. When prices are increasing, the use of current cost data re-duces the computed return on equity (ROE) as income is reduced (higher depreciation) andequity is increased (higher asset values). If the current cost ROE is very low, for example,it tells us that the company might be better off selling its assets and either reinvesting theproceeds in other assets, providing higher returns, or distributing them to stockholders forreinvestment.

Using Constant Dollar Data

Although we have stated that constant dollar data are generally not useful for financial analy-sis, there are some applications. Constant dollar data can be used to look at investment re-turns from the investor point of view.

An investor should measure the performance of an investment relative to inflation, notin absolute terms. Investors defer current consumption to obtain higher future consump-tion. In highly inflationary societies, the instinct to save is stifled if nominal rates of returnare below the inflation rate. Under these conditions, consumption deferred is consumptionreduced.

To measure the impact of changing prices on the investor, deflate returns by a measureof purchasing power such as the consumer price index. The index for the investor, not the in-vestment, should be used. For example, an investment in General Motors’ shares by a Cana-dian investor must be evaluated by deflating the returns (translated into Canadian dollars) by

7In 1994, Mead lengthened its depreciation lives, reducing depreciation expense. This change increases the ratio ofcapital expenditures to depreciation expense.8Sustainable income is defined and discussed in Chapter 2.

the Canadian consumer price index. This can be done using the constant dollar method illus-trated in Exhibit 8A-1.

The constant dollar method is also widely used in highly inflationary economies, espe-cially when their financial systems are indexed to inflation. In many cases, the constant dol-lar method (sometimes in modified form) is used to produce the primary financial statementsfor financial and/or tax reporting.

Although the analysis of such statements is beyond the scope of this text, we will pro-vide one caveat. Unless the input and output prices of the firm subjected to analysis are fullyindexed, the constant dollar method will not provide a satisfactory basis for analysis. Soundinvestment decisions require an understanding of the effects of the specific price changesfaced by the firm.

International Accounting Standards

IAS 15 (1989) provided for voluntary disclosures similar to those of ASR 190 (previouslydiscussed). IAS 29 (1989), Financial Reporting in Hyperinflationary Economies, requires ad-justment of financial statements of companies operating in hyperinflationary economiesusing the constant dollar method. The principal provisions of IAS 29 are:

1. The currency unit at the balance sheet date must be used as the unit of measure. Allnonmonetary assets and liabilities must be restated to that unit, using the methodol-ogy illustrated in Exhibit 8A-1.

2. Balance sheet items carried at current cost are not restated.

3. Losses (gains) on net monetary assets (liabilities) are included in net income for theperiod.

4. The portion of borrowing cost that represents the premium for inflation must be ex-pensed when debt is indexed for inflation.

5. Income and cash flow statement items must be restated to the same unit of measureused for the balance sheet.

While IAS 29 does not state when an economy is considered to be hyperinflationary, it sug-gests cumulative three-year inflation of 100% as a guide.9

Although the accounting for hyperinflationary economies under IAS 29 is quite differentfrom the treatment under U.S. GAAP (described in Chapter 15),10 the IAS treatment is ac-ceptable under SEC rules for foreign companies filing in the United States; reconciliation toU.S. GAAP is not required.

Concluding Remarks

With the adoption of SFAS 89, changing prices disappeared as an accounting issue. Yetprices continue to change. While general inflation has remained at low levels in virtually allindustrialized countries, the prices of specific commodities continue to fluctuate.

Thus, financial analysis requires identification of the effects of significant price changes.Some of these effects can be dealt with summarily. For example, it is relatively easy to usean index of retail prices to compute the effect of inflation on department store sales. It ismore complex (and more difficult) to discern the effect of a change in oil prices on an oil re-finer’s profit margins, turnover ratios, and return on equity. The objective of this appendix,and the material on the effect of price changes in Chapters 6 through 8, was to provide toolsto permit such analysis.


9Cumulative three-year inflation of 100% is the criterion for hyperinflationary treatment under SFAS 52, as de-scribed in Chapter 15.10Under IAS GAAP, hyperinflation is dealt with by inflation-adjusting the subsidiary financial statements; U.S.GAAP adjusts via the choice of currency used to translate the subsidiary financial statements into the reportingcurrency.


PROBLEMS

8A-1. [Income, cash flow, and ratio effects of current cost adjustments] Use the data in Ex-hibit 8A-2 and the Holmen financial statements to answer the following questions.

A. Estimate current cost depreciation for 1999.

B. Compute Holmen’s net income for 1999 after adjustment for current cost depre-cation.

C. Describe the effect of the adjustment in part A on Holmen’s cash from operations.

D. Compute each of the following ratios for 1999 using both reported and currentcost data. Discuss your results

(i) Fixed asset turnover

(ii) Total asset turnover

(iii) Return on average equity

Appendix 11-ASECURITIZATION: SFAS 140 REPORTINGAND DISCLOSURE REQUIREMENTS—SEARS

INTRODUCTION

SFAS 140 (2000) amended SFAS 125 (1996) by changing the conditions under which secu-ritizations could be treated as sales of receivables. The principal modifications concerned (a)the criteria used to designate qualifying special purpose entities (transferees purchasing secu-ritized assets) and (b) conditions under which the transferor retains effective control over thetransferred assets. SFAS 140 requires significant new disclosures regarding securitized as-sets. SFAS 140 applied to transfers of financial assets occurring after March 31, 2001. Earlyadoption was prohibited. Sears adopted SFAS 140 on April 1, 2001.

These changes and the new disclosure provisions are illustrated using Sears’ disclosuresfrom its 2000 and 2001 annual reports. Some of these data were reported in 1999, as part ofthe Management Discussion and Analysis.

Part A: Disclosures—Sears 2000 Annual Report

NOTE 3—CREDIT CARD SECURITIZATIONS

The Company utilizes credit card securitizations as a part of its overall funding strategy.Under generally accepted accounting principles, if the structure of the securitization meetscertain requirements, these transactions are accounted for as sales of receivables.

Summary of Securitization Process

As part of its domestic credit card securitizations, the Company transfers credit card receiv-able balances to a Master Trust1 (“Trust”) in exchange for certificates representing undividedinterests in such receivables. Balances transferred from the Company’s credit card portfolioto the Trust become securities upon transfer. These securities are accounted for as available-for-sale securities. The allowance for uncollectible accounts that related to the transferred re-ceivables is amortized over the collection period to recognize income on the transferredbalances on an effective yield basis. This resulted in additional revenues of $60 and $75 mil-lion in 2000 and 1999, respectively, and did not affect 1998 revenues. The Trust securitizesbalances by issuing certificates representing undivided interests in the Trust’s receivables tooutside investors. In each securitization transaction the Company retains certain subordi-nated interests that serve as a credit enhancement to outside investors and expose the Com-pany’s Trust assets to possible credit losses on receivables sold to outside investors. Theinvestors and the Trust have no recourse against the Company beyond Trust assets.

W49

1The Master Trust is the qualifying special purpose entity referred to in the appendix introduction.

W50 APPENDIX 11-A SECURITIZATION: SFAS 140 REPORTING AND DISCLOSURE REQUIREMENTS—SEARS

In order to maintain the committed level of securitized assets, the Company reinvestscash collections on securitized accounts in additional balances. These additional investmentsresult in increases to the interest-only strip and credit revenues. As of December 30, 2000,the Company’s securitization transactions mature as follows:

Millions

2001 $1,046

2002 1,403

2003 2,020

2004 1,519

2005 and thereafter 1,846

Retained Interest in Transferred Credit Card Receivables

The Company’s retained interest in transferred credit card receivables consists of investorcertificates (undivided interests in or claims on cash flows of the Trust’s receivables) held bythe Company, interest-only strips (the company’s rights to residual, future cash flows afterthe outside investors have received the contractual return), contractually required seller’s in-terest (credit enhancement or support provided by Sears), and excess seller’s interest (receiv-ables available for future securitizations) in the credit card receivables in the Trust. Retainedinterests at year-end are as follows:

Millions 2000 1999

Subordinated interests:

Investor certificates held by the Company $1,161 $ 960

Unsubordinated interests:

Contractually required seller’s interest 898 760

Excess seller’s interest 992 1,455

Interest-only strip 136 67

Less: Unamortized transferred allowance for uncollectible accounts $3,182 $3,231

Retained interest in transferred credit card receivables $3,105 $3,211

The Company intends to hold the investor certificates and contractually required seller’s in-terest to maturity. The excess seller’s interest is considered available-for-sale. Due to the re-volving nature of the underlying credit card receivables, the carrying value of the Company’sretained interest in transferred credit card receivables approximates fair value and is classi-fied as a current asset.

Securitization Gains

Due to the qualified status of the Trust, the issuance of certificates to outside investors is con-sidered a sale for which the Company recognizes a gain and an asset for the interest-onlystrip. The interest-only strip represents the Company’s rights to future cash flows arisingafter the investors in the Trust have received the return for which they contracted. The Com-pany also retains servicing responsibilities for which it receives annual servicing fees ap-proximating 2% of the outstanding balance. The Company recognized incremental operatingincome from net securitization gains of $68, $11, and $58 million in 2000, 1999, and 1998,respectively.

The Company measures its interest-only strip and the related securitization gains usingthe present value of estimated future cash flows. This valuation technique requires the useof key economic assumptions about yield, payment rates, charge-off rates, and returns totransferees. Approximately 22% of the Company’s outstanding securitizations offer vari-

able returns to investors with contractual spreads over LIBOR ranging from 16 to 53 basispoints.

As of December 30, 2000, the interest-only strip was recorded at its fair value of $136million. The following table shows the key economic assumptions used in measuring theinterest-only strip and securitization gains. The table also displays the sensitivity of the cur-rent fair value of residual cash flows to immediate 100 and 200 basis point adverse changesin yield, payment rate, charge-off, and discount rate assumptions:

Effects of AdverseChanges

Millions Assumptions 100 bp 200 bp

Yield (annual rate) 19.85% $36 $71

Principal payment rate (monthly rate) 5.26% $20 $35

Gross charge-off rate (annual rate) 7.4% $36 $71

Residual cash flows discount rate (annual rate) 12.0% $ 1 $ 2

These sensitivities are hypothetical and should be used with caution. As the figures indicate,changes in fair value assumptions generally cannot be extrapolated because the relationshipof the change in assumption to the change in fair value may not be linear. Also, in this table,the effect of a variation in a particular assumption on the fair value of the retained interest iscalculated without changing any other assumption; in reality, changes in one factor may re-sult in changes in another, which might magnify or counteract the sensitivities.

Managed Portfolio Data

A summary of the domestic year-end securitized receivables and other domestic credit cardreceivables managed together with them follows:

Millions 2000 1999

Securitized balances $ 7,834 $ 6,579

Retained interest in transferred credit card receivables (1) 3,051 3,175

Owned credit card receivables 16,175 17,068

Other customer receivables $ 16,(59) $ 26,(37)

Managed credit card receivables $27,001 $26,785

Net charge-offs of managed credit card receivables $ 1,323 $ 1,713

Delinquency rates at year-end 7.56% 7.58%

(1) The 2000 and 1999 retained interest amounts exclude reserves of $82 and $31 million, respectively, and interest-only strip balances of $136 and $67 million, respectively, related to the transfer of credit card receivables into theTrust.

Securitization Cash Flow Data

The table below summarizes certain cash flows that the Company received from and paid tothe securitization trust during 2000:

Millions

Proceeds from new securitizations $2,620

Proceeds from collections reinvested in previous securitizations 3,547

Servicing fees received 200

Purchase of charged-off balances, net of recoveries (522)

Source: Sears 2000 Annual Report

NOTE 3—CREDIT CARD SECURITIZATIONS W51

W52 APPENDIX 11-A SECURITIZATION: SFAS 140 REPORTING AND DISCLOSURE REQUIREMENTS—SEARS

Part B: Discussion

The first part of Note 3 discussed Sears’ policies regarding the securitization of credit cardreceivables. While not clearly stated, Sears apparently retains all of the effective credit riskof these receivables.

Sears discloses the maturity of the securitizations, which extend out for more than fiveyears. Nonetheless, Sears reports all of its interest in these receivables (and all receivablesowned) as current assets. Thus, the current ratio of 1.82 overstates the liquidity of Sears’ bal-ance sheet.

Next, Sears reports its interest in the securitized receivables in several categories:

• Subordinated interests retained by Sears

• Contractual interest

• Excess interest

• Interest-only strip

• Unamortized allowance of uncollectible accounts

As the securitizations meet the requirements of SFAS 125 for sale recognition, Sears recog-nizes gains when the sales take place. In 2000, such gains were $68 million. The amount ofthe gain, and valuation of the interest-only strip depends on the following assumptions:

1. Yield on the sold receivables

2. Monthly customer payment rate

3. Annual charge-off (bad debt) rate

4. Rate used to discount residual cash flows

The table discloses Sears’ assumptions, which can be compared with those of other compa-nies, and the effect of adverse deviations on the valuation of residual cash flows.

Next, Sears reports its total domestic (U.S.) credit card receivables, showing theamounts securitized and the retained interest in those securitized receivables separately. Italso reports the charge-off rate for the year (note the decline in 2000) and the year-end delin-quency rate. These data can be compared with those of similar companies.

Finally, Sears reports the cash flows associated with its securitization activities. Theseamounts are not reported in the company’s statement of cash flows. Sears received more than$2.6 billion from securitizations in 2000 and reinvested more than $3.5 billion of collectionsin previous securitizations. The company received $200 million of servicing fees, consistentwith the 2% fee reported earlier in Note 3. Sears spent $522 million to repurchase charged-off balances, net of recovery of earlier repurchased receivables.

These disclosures provide the analyst with a reasonable understanding of the importanceof securitization as a source of financing for Sears. Comparisons can be made with otherfirms, especially once more years of data are accumulated.

Part C: Disclosures—Sears 2001 Annual Report

Note 3—Credit Card Receivables

The addition of previously uncommitted assets to the securitization trust in April 2001 re-quired the Company to consolidate the securitization structure for financial reporting pur-poses on a prospective basis. Accordingly, the company recognized approximately $8.1billion of previously unconsolidated securitized credit card receivables and related securiti-zation borrowings in the second quarter of 2001. In addition, approximately $3.9 billion ofassets were reclassified to credit card receivables from retained interests in transferred creditcard receivables. The Company now accounts for securitizations as secured borrowings.

In connection with the consolidation of the securitization structure, the Company recog-nized a non-cash, pretax charge of $522 million to establish an allowance for uncollectibleaccounts related to the receivables, which were previously considered sold or accounted foras retained interests in transferred credit card receivables.

Accounting for Securitizations—SFAS 125

Prior to April 2001, the issuance of certificates to outside investors was considered a sale ofreceivables for which the Company recognized a gain on the sale. The Company recognizedincremental operating income of $40, $128, and $86 million in 2001, 2000, and 1999, re-spectively, from net securitization activity.

The Company’s retained interests were recorded by the Company at the date of the saleto the trusts by allocating the original carrying amounts of the credit card receivables held bythe Company between the sold interests and the retained interests based on their relative fairvalues. Management used certain assumptions in determining the fair value of its retained in-terests. Key assumptions used in the first quarter of 2001 and in fiscal 2000 were a yield of19.85%, a monthly principal payment rate of 5.26%, a discount rate of 12.0%, and an annualcharge-off rate of 7.40%.

Securitization Cash Flow Data

The table below summarizes certain cash flows that the Company received from and paid tothe securitization trust in 2000. Cash flow data has not been provided for 2001 as the securi-tization trust was consolidated beginning in the second quarter.

Millions 2000

Proceeds from new securitizations $2,620

Proceeds from collections reinvested in previous securitizations 3,547

Servicing fee received 200

Purchase of charged-off balances, net of recoveries (522)

ANALYSIS OF CONSOLIDATED FINANCIAL CONDITION

The Company has significant financial capacity and flexibility due to the quality and liquidityof its assets, principally its credit card receivables. As such, the Company has the ability toaccess multiple sources of capital.

A summary of the Company’s credit card receivables at year-end is as follows:

Millions 2001 2000 1999

Domestic:

Managed credit card receivables $27,599 $27,001 $26,785

Securitized balances sold — (7,834) (6,579)

Retained interest in transferred credit card — (3,051) (3,175)

receivables(1)

Other customer receivables $27,640 $16,159 $17,037

Domestic owned credit card receivables $27,639 $16,175 $17,068

Sears Canada credit card receivables $21,682 $11,828 $11,725

Consolidated owned credit card receivables $29,321 $18,003 $18,793

(1)The 2000 and 1999 retained interest amounts exclude reserves of $82 and $31 million, respectively, and interest-only strip balances of $136 and $67 million, respectively, related to the transfer of credit card receivables into theTrust.

As of year-end 2000 and 1999, the credit card receivables balance of $18.0 billion and$18.8 billion, respectively, excluded credit card receivables transferred to a securitizationMaster Trust (“Trust”). Through its subsidiary, SRFG, Inc., the Company obtains fundingby selling securities backed by a portion of the receivables in the Trust. In addition to the re-ceivables in the Trust, which support securities sold to third parties, the Company transfers

ANALYSIS OF CONSOLIDATED FINANCIAL CONDITION W53

W54 APPENDIX 11-A SECURITIZATION: SFAS 140 REPORTING AND DISCLOSURE REQUIREMENTS— SEARS

additional receivables to the Trust to have receivables readily available for future securitiza-tions. As discussed in Note 3 of the Company’s Consolidated Financial Statements, theCompany consolidated its Master Trust beginning in the second quarter of 2001, subsequentto the adoption of SFAS No. 140. The Company continues to utilize securitizations as a keyfunding source.

CAPITAL RESOURCES

Total borrowings outstanding at the end of 2001 and 2000 were $25.6 billion and $25.7 bil-lion, respectively. Total borrowings, including debt reflected on the balance sheet and in-vestor certificates related to credit card receivables issued through securitizations, were asfollows:

% of % of % ofMillions 2001 Total 2000 Total 1999 Total

Short-term borrowings $ 3,557 13.9% $ 4,280 16.7% $ 2,989 12.2%

Long-term debt(1) $22,078 186.1% $13,580 152.8% $15,049 161.1%

Securitized balances sold(2) $22,0— — 7,834 30.5% 6,579 26.7%

Total borrowings $25,635 100.0% $25,694 100.0% $24,617 100.0%

(1)Includes capitalized lease obligations.(2)Included in long-term debt in 2001 due to the change in securitization accounting; the securitization trust was notconsolidated in 2000 and 1999 (see Note 3 of the Notes to the Consolidated Financial Statements).Source: Sears 2001 Annual Report

Part D: Discussion

Adoption of SFAS 140 requires consolidation of receivables previously considered sold.Sears notes (1) the impact of non-recognition of any gain on sale on operating income and(2) the impact on reported leverage. Although the change in operating income is not signifi-cant, reported income better reflects the earnings process and the impact of charge-offs. Re-ported leverage shows a significant increase. Exhibit 11A-1 shows an increase in reportedleverage to approximately 419% from 286% that would have been reported had Sears contin-ued to report the securitizations as sales. The inclusion of receivables and related borrowingsalso better reflects the liquidity and the interest coverage.

EXHIBIT 11A-1Sears: Impact of SFAS 140

Capitalization at 12/31/01

Amounts in $millions Pro Forma* As Reported

Short-term debt $ 6,714 $ 6,714Long-term debt 10,778 18,921Total debt 17,492 25,635Stockholders’ equity 6,119 6,119Debt-equity ratio 286% 419%

*Pro Forma assumes that SFAS 140 was not adopted on April 1

Appendix 18-ARATIOS USED IN CREDIT AND EQUITY RISKPREDICTION MODELS

Chapter 18 discusses research that examined the utility of accounting (and other financial)measures in risk evaluation and prediction. The exhibits provided in this appendix list the ex-planatory independent variables (financial risk measures) used in the key research studies inthis area. The topics covered by the exhibits are:

Exhibits 18A-1(a) and (b) Bankruptcy Prediction Models

Exhibit 18A-2 Bond Ratings Prediction Models

Exhibit 18A-3 Beta Prediction Models

The exhibits, except for Exhibit 18A-1(b), are all similar in layout detailing the specific vari-ables used in each of the studies. Exhibit 18A-1(b) [adapted from Reilly (1991) and work byGentry, Newbold, and Whitford (1994)], on the other hand, summarizes the findings of four-teen studies that focused on bankruptcy prediction.

As noted in the chapter, for the most part, the ratios found to be useful in the researchcorrespond to the categories (activity, liquidity, solvency, and profitability) that we haveused throughout the book. Additional new indicators are primarily measures of earnings vari-ability and size.

W55

W56 APPENDIX 18-A RATIOS USED IN CREDIT AND EQUITY RISK PREDICTION MODELS

EXHIBIT 18A-1(a)Independent Variables Used in Bankruptcy Prediction Models

Ohlson (1980) Altman et al. (1977) Deakin (1972) Altman (1968)

Activity Four asset categories Sales to total assetsdivided by sales:

(1) Current assets(2) Quick assets(3) Working capital(4) Cash

Liquidity Current ratio Current ratio Current ratioQuick ratioCash ratioFour asset categoriesdivided by total assets:(1) Current assets(2) Quick assets(3) Working capital

Working capital to (4) Cash Working capital tototal assets total assets

Leverage and Solvency Liabilities to assets Equity (market) to Capital Debt to assets Equity (market) toTimes Interest earned debt (book)

Funds from operations Funds from operationsto total liabilities to debt

Dummy variableindicating if networth is negative

Profitability Return on assets Return on assets Return on assets Return on assetsDummy variable indicating Retained earnings to Retained earnings

if net income was total assets to total assetsnegative in last two years

Earnings variability Percentage change in net Standard error ofincome return on assets

Size Total Assets Total Assets

RATIOS USED IN CREDIT AND EQUITY RISK PREDICTION MODELS W57

EXHIBIT 18A-1(b)Summary of Most Useful Ratios for Predicting Failure

Category/Ratios Number of Studies in Which the Ratio Was Significant

Financial LeverageCash Flow/Total Debt 7Total Debt/Total Assets 6Retained Earnings/Total Assets 5

Short-term LiquidityNet Working Capital/Total Assets 6Current Assets/Current Liabilities 6Cash/Sales 2Cash/Current Liabilities 4

ProfitabilityNet Income/Total Assets 5EBIT/Total Assets 4

ActivityQuick Assets/Sales 2

Adapted from Frank K. Reilly, “Using Cash Flows and Financial Ratios to Predict Bankruptcies,” Analyzing Invest-ment Opportunities in Distressed and Bankrupt Companies, Charlottesville, VA: The Institute of Chartered Finan-cial Analysts, 1991, Table 1, P.25

W58 APPENDIX 18-A RATIOS USED IN CREDIT AND EQUITY RISK PREDICTION MODELS

EXHIBIT 18A-2Independent Variables Used in Bond Ratings Prediction Models

Kaplan and Pinches and Pogue andBelkaoui Belkaoui Urwitz Mingo Soldovsky Horrigan West(1983) (1980) (1979) (1973) (1969) (1966) (1966)

Activity and Current ratio Current ratio Working capitalliquidity to sales

Sales to equity

Leverage and Long-term debt Long-term debt Long-term debt Total debt to Debt to capital Equity to debt Debt to equitysolvency to capital to capital to assets assets (market

values)Short-term debt Short-term Long-term debt

to capital debt to to equitycapital

Fixed charge Fixed charge Times interest Times interest Times interestcoverage coverage earned earned earned

Cash flow to Cash flow to debtinvestment infixed assets andinventory plus dividends

Profitability Return on assets Return on Return on Operating profitassets assets

Earnings Accounting beta Years of Coefficient ofvariability consecutive variation—

dividends ROACoefficient of Coefficient of

variation—net variation—netincome income

Size Total assets Total assets Total assets Issue size Total assets Total assets Bonds outstanding

Subordination 0-1 dummy 0-1 dummy 0-1 dummy 0-1 dummy 0-1 dummy

Market-based Price to net Price to net Market betabook value book value

Other Coefficient of Industry Period of variation—total dummy solvencyassets variable

RATIOS USED IN CREDIT AND EQUITY RISK PREDICTION MODELS W59

EXHIBIT 18A-3Independent Variables Used in Beta Prediction Models

Predictive and Explanatory Explanatory

Rosenberg Mandelker Balland and and

Hochman McKibben Beaver et al. Rhee Bildersee Lev Brown(1983) (1973) (1970) (1984) (1975) (1974) (1968)

Earnings Operating Accounting beta * OLE Variable AccountingVariability Risk (operating cost % beta

income) (v)

Financial Debt to capital FLE Debt to equityRisk Preferred equity

to common equity

Total Standard Standard Risk† deviation deviation

earnings/ earnings/priceprice

Growth Dividend yield Asset growth

Dividends Dividend payout

Liquidity Current ratio

*See Exhibit 18-8 in text.†Earnings variability can be measured as the sum of operating risk and financial risk.

Appendix 19-AMULTISTAGE GROWTH MODELS

The original formulation of the discounted models discussed in the chapter is presented below:

Theoretically, by predicting each year individually, any assumed growth rate of dividends orearnings payout (even zero dividends) can be accommodated. From a practical point of view,of course, one would not attempt to forecast individual periods over a very long horizon.

One palatable approach is to forecast the near future individually and then impose an as-sumption as to the appropriate valuation after that period. Recall that the preceding expres-sion is equivalent to

This is the present value of the dividends over the first n years plus the discounted value atthe end of year n.

For example, assume that you forecast a firm’s net income over the next three years asyear 1 � 100, year 2 � 120, and year 3 � 150. The firm’s k � 20% and its r � 10%. To usethe preceding equation, one must derive a terminal value for the firm at the end of year 3.You may at this point decide to make some general assumptions. One assumption might bethat from the third year on the firm will experience growth of 8%. The implicit forecast foryear 4’s earnings is (1.08 � $150) � $162, and the terminal value at the end of year 3 (if weuse the constant growth model presented earlier) is equal to

The value now will be equal to

VALUING A NONDIVIDEND-PAYING FIRM

A firm paying zero dividends can also be modeled along these lines. A firm that pays zerodividends reinvests everything in the firm. Its growth rate is equal to [1 � k]r* � r* since k � 0. Assume that a firm having an r* of 25% for the next five years does not plan to paydividends for those five years. If its present earning level is $10, its earnings in year 5 willequal $10(1.25)5 � $30.5. From year 6 and on, assume that its r* will be 20% and the firm willpay dividends at a rate k � 60%. Its growth rate will therefore equal (1 � 60%) � 20% � 8%.Earnings in year 6 will equal $30.5(1.08) � $32.9. The firm’s value at the beginning of year6 will be equal to

0.6 � $32.90.1 � 0.08

� $987

� $91 � $99 � $133 � $1,217 � $1,520

P0 � $100(1.1)

� $120(1.1)2

� $150(1.1)3

� $1,620

(1.1)3

P3 � 0.2 � $1620.10 � 0.08

� $1,620

P0 � kE1

(1 � r) �

kE2

(1 � r)2 � � � � �

kEn

(1 � r)n � Pn

(1 � r)n

P0 � ��

i�1

kEi

(1 � r)i

W60

The value today will be equal to the $987 discounted (back five years) to the beginningof year 1 or $987/(1.10)5 � $613.

SHIFTING GROWTH RATE PATTERNS

Variations of this approach assume a certain level of growth over some initial phase and dif-ferent growth rates after the initial phase (Figure 19A-1).

The finite growth model (Figure 19A-1a) assumes that the firm will experience growthof g � (1 � k)r* for n years. After that point, the abnormal investment opportunities of r* �r will not exist. The value of equity for such a firm will equal

Other models commonly referred to as three-phase models assume (Figure 19A-1b) aninitial (phase 1) high abnormal growth rate ga for a number of years that tapers off (in phase2) to a long-term (phase 3) normal growth pattern of gn. The calculations for these models

P0 � E1

r � E1

r �g � r (1 � k)r � g �1 � �1 � g

1 � r�n��

SHIFTING GROWTH RATE PATTERNS W61

FIGURE 19A-1a–c. Simplified three-phase model(Fuller and Hsia, 1984). Source: Russel J. Fullerand Chi-Cheng Hsia, “A Simplified CommonStock Valuation Model,” Financial Analysts Jour-nal, September–October 1984, pp. 49–56 (FigureB, p. 50, and Figure E, p. 53).

W62 APPENDIX 19-A MULTISTAGE GROWTH MODELS

are somewhat complex. Fuller and Hsia (1984) simplified these models by assuming agrowth pattern as depicted in Figure 19A-1c. They start with initial above-normal growth,but assume that it converges gradually to a stable long-term growth pattern. If we stay withthe definitions of ga as the initial growth pattern and gn as the long-term growth pattern to bereached within n years, the value of the equity is equal to

P0 � kE0

r � gn �(1 � gn) � n

2 (ga � gn)�

Appendix 19-BTHE EBO AND TERMINAL VALUEASSUMPTIONS

The terminal value calculations in the chapter assume that ROE remains constant after pe-riod T, at r or some other level. Figure 4-3, however, indicates that it is more likely for ROEto converge asymptotically to a steady-state level. This appendix presents valuation formulaethat can be used when the rate of convergence can be modeled as an autoregressive process,that is,

or

where ROEt converges to the steady-state level The equation indicates that, in each pe-riod, the gap between the actual ROE and the steady-state level narrows as a function of theautoregressive parameter c.Under these assumptions, the valuation formula becomes

where g represents the assumed growth rate in book value. Explicit forecasts of earnings(ROE and book value) are made for T periods, followed by the terminal value calculation inthe braces.

In the chapter, we note that, even if abnormal earnings were to continue indefinitely( � r) because of a special situation such as patent protection, it is unlikely that similarhigher returns could be earned on new projects. Thus, the abnormal earnings would not growas book value increases. Setting g � 0 yields

If competitive pressures force abnormal profits to zero, then at steady state, � r and thevaluation formula becomes

P0 � B0 � �T

j�1 (ROEj � r)Bj

(1 � r) j � � BT�1

(1 � r)T �(ROET �r)c

1 � r � c ��

ROE

P0 � B0 � �T

j�1 (ROEj � r)Bj

(1 � r) j � � BT�1

(1 � r)T �(ROET � ROE)c

1 � r � c �

(ROE � r)r ��

ROE

P0 � B0 � �T

j�1 (ROEj � r)Bj�1

(1 � r) j � � BT�1

(1 � r)T �(ROET � ROE)c

1 � r � c(1 � g) �

(ROE � r)r � g ��

ROE.

ROEt � ROE � c(ROEt�1 � ROE) � � where 0 � c � 1

(ROEt � ROE) � c(ROEt�1 � ROE)

W63

Appendix

Documents

Transcript of Appendix