future years—say, not more than five. This evidently they will not do either, since they have already conveniently disposed of the entire sum as a 1970 special charge. The more seriously investors take the per-share earnings figures as published, the more necessary it is for them to be on their guard against accounting factors of one kind and another that may impair the true comparability of the numbers. We have mentioned three sorts of these factors: the use of special charges, which may never be reflected in the per-share earnings, the reduction in the normal income-tax deduction by reason of past losses, and the dilution fac- tor implicit in the existence of substantial amounts of convertible securities or warrants. 1 A fourth item that has had a significant effect on reported earnings in the past is the method of treating depreciation—chiefly as between the “straight-line” and the “accelerated” schedules. We refrain from details here. But as an example current as we write, let us mention the 1970 report of Trane Co. This firm showed an increase of nearly 20% in per-share earnings over 1969—$3.29 versus $2.76—but half of this came from returning to the older straight-line depreciation rates, less burden- some on earnings than the accelerated method used the year before. (The company will continue to use the accelerated rate on its income-tax return, thus deferring income-tax payments on the difference.) Still another factor, important at times, is the choice between charging off research and development costs in the year they are incurred or amortizing them over a period of years. Finally, let us mention the choice between the FIFO (first-in-first- out) and LIFO (last-in-first-out) methods of valuing inventories.* 316 The Intelligent Investor * Nowadays, investors need to be aware of several other “accounting fac- tors” that can distort reported earnings. One is “pro forma” or “as if” finan- cial statements, which report a company’s earnings as if Generally Accepted Accounting Principles (GAAP) did not apply. Another is the dilu- tive effect of issuing millions of stock options for executive compensation, then buying back millions of shares to keep those options from reducing the value of the common stock. A third is unrealistic assumptions of return on the company’s pension funds, which can artificially inflate earnings in good years and depress them in bad. Another is “Special Purpose Entities,” or affiliated firms or partnerships that buy risky assets or liabilities of the com- An obvious remark here would be that investors should not pay any attention to these accounting variables if the amounts involved are relatively small. But Wall Street being as it is, even items quite minor in themselves can be taken seriously. Two days before the ALCOA report appeared in the Wall Street Journal, the paper had quite a discussion of the corresponding statement of Dow Chemi- cal. It closed with the observation that “many analysts” had been troubled by the fact that Dow had included a 21-cent item in regu- lar profits for 1969, instead of treating it as an item of “extraordi- nary income.” Why the fuss? Because, evidently, evaluations of Dow Chemical involving many millions of dollars in the aggregate seemed to depend on exactly what was the percentage gain for 1969 over 1968—in this case either 9% or 4 1 ⁄2%. This strikes us as rather absurd; it is very unlikely that small differences involved in one year’s results could have any bearing on future average profits or growth, and on a conservative, realistic valuation of the enterprise. By contrast, consider another statement also appearing in Janu- ary 1971. This concerned Northwest Industries Inc.’s report for 1970.* The company was planning to write off, as a special charge, not less than $264 million in one fell swoop. Of this, $200 million represents the loss to be taken on the proposed sale of the railroad subsidiary to its employees and the balance a write-down of a recent stock purchase. These sums would work out to a loss of about $35 per share of common before dilution offsets, or twice its then current market price. Here we have something really signifi- Things to Consider About Per-Share Earnings 317 pany and thus “remove” those financial risks from the company’s balance sheet. Another element of distortion is the treatment of marketing or other “soft” costs as assets of the company, rather than as normal expenses of doing business. We will briefly examine such practices in the commentary that accompanies this chapter. * Northwest Industries was the holding company for, among other busi- nesses, the Chicago and Northwestern Railway Co. and Union Underwear (the maker of both BVD and Fruit of the Loom briefs). It was taken over in 1985 by overindebted financier William Farley, who ran the company into the ground. Fruit of the Loom was bought in a bankruptcy proceeding by Warren Buffett’s Berkshire Hathaway Inc. in early 2002. cant. If the transaction goes through, and if the tax laws are not changed, this loss provided for in 1970 will permit Northwest Industries to realize about $400 million of future profits (within five years) from its other diversified interests without paying income tax thereon.* What will then be the real earnings of that enterprise; should they be calculated with or without provision for the nearly 50% in income taxes which it will not actually have to pay? In our opinion, the proper mode of calculation would be first to consider the indicated earning power on the basis of full income- tax liability, and to derive some broad idea of the stock’s value based on that estimate. To this should be added some bonus figure, representing the value per share of the important but temporary tax exemption the company will enjoy. (Allowance must be made, also, for a possible large-scale dilution in this case. Actually, the convertible preferred issues and warrants would more than double the outstanding common shares if the privileges are exercised.) All this may be confusing and wearisome to our readers, but it belongs in our story. Corporate accounting is often tricky; security analysis can be complicated; stock valuations are really depend- able only in exceptional cases.† For most investors it would be probably best to assure themselves that they are getting good value for the prices they pay, and let it go at that. 318 The Intelligent Investor * Graham is referring to the provision of Federal tax law that allows corpora- tions to “carry forward” their net operating losses. As the tax code now stands, these losses can be carried forward for up to 20 years, reducing the company’s tax liability for the entire period (and thus raising its earnings after tax). Therefore, investors should consider whether recent severe losses could actually improve the company’s net earnings in the future. † Investors should keep these words at hand and remind themselves of them frequently: “Stock valuations are really dependable only in exceptional cases.” While the prices of most stocks are approximately right most of the time, the price of a stock and the value of its business are almost never iden- tical. The market’s judgment on price is often unreliable. Unfortunately, the margin of the market’s pricing errors is often not wide enough to justify the expense of trading on them. The intelligent investor must carefully evaluate the costs of trading and taxes before attempting to take advantage of any price discrepancy—and should never count on being able to sell for the exact price currently quoted in the market. Use of Average Earnings In former times analysts and investors paid considerable atten- tion to the average earnings over a fairly long period in the past— usually from seven to ten years. This “mean figure”* was useful for ironing out the frequent ups and downs of the business cycle, and it was thought to give a better idea of the company’s earning power than the results of the latest year alone. One important advantage of such an averaging process is that it will solve the problem of what to do about nearly all the special charges and credits. They should be included in the average earnings. For cer- tainly most of these losses and gains represent a part of the company’s operating history. If we do this for ALCOA, the average earnings for 1961–1970 (ten years) would appear as $3.62 and for the seven years 1964–1970 as $4.62 per share. If such figures are used in conjunction with ratings for growth and stability of earn- ings during the same period, they could give a really informing picture of the company’s past performance. Calculation of the Past Growth Rate It is of prime importance that the growth factor in a company’s record be taken adequately into account. Where the growth has been large the recent earnings will be well above the seven- or ten- year average, and analysts may deem these long-term figures irrel- evant. This need not be the case. The earnings can be given in terms both of the average and the latest figure. We suggest that the growth rate itself be calculated by comparing the average of the last three years with corresponding figures ten years earlier. (Where there is a problem of “special charges or credits” it may be dealt with on some compromise basis.) Note the following calculation for the growth of ALCOA as against that of Sears Roebuck and the DJIA group as a whole. Comment: These few figures could be made the subject of a long discussion. They probably show as well as any others, derived by elaborate mathematical treatment, the actual growth of earnings Things to Consider About Per-Share Earnings 319 * “Mean figure” refers to the simple, or arithmetic, average that Graham describes in the preceding sentence. for the long period 1958–1970. But how relevant is this figure, gen- erally considered central in common-stock valuations, to the case of ALCOA? Its past growth rate was excellent, actually a bit better than that of acclaimed Sears Roebuck and much higher than that of the DJIA composite. But the market price at the beginning of 1971 seemed to pay no attention to this fine performance. ALCOA sold at only 11 1 ⁄2 times the recent three-year average, while Sears sold at 27 times and the DJIA itself at 15+ times. How did this come about? Evidently Wall Street has fairly pessimistic views about the future course of ALCOA’s earnings, in contrast with its past record. Sur- prisingly enough, the high price for ALCOA was made as far back as 1959. In that year it sold at 116, or 45 times its earnings. (This compares with a 1959 adjusted high price of 25 1 ⁄2 for Sears Roebuck, or 20 times its then earnings.) Even though ALCOA’s profits did show excellent growth thereafter, it is evident that in this case the future possibilities were greatly overestimated in the market price. It closed 1970 at exactly half of the 1959 high, while Sears tripled in price and the DJIA moved up nearly 30%. It should be pointed out that ALCOA’s earnings on capital funds* had been only average or less, and this may be the decisive factor here. High multipliers have been maintained in the stock mar- ket only if the company has maintained better than average prof- itability. 320 The Intelligent Investor TABLE 12-1 ALCOA Sears Roebuck DJIA Average earnings 1968–1970 $4.95 a $2.87 $55.40 Average earnings 1958–1960 2.08 1.23 31.49 Growth 141.0% 134.0% 75.0% Annual rate (compounded) 9.0% 8.7% 5.7% a Three-fifths of special charges of 82 cents in 1970 deducted here. * Graham appears to be using “earnings on capital funds” in the traditional sense of return on book value—essentially, net income divided by the company’s tangible net assets. Let us apply at this point to ALCOA the suggestion we made in the previous chapter for a “two-part appraisal process.”* Such an approach might have produced a “past-performance value” for ALCOA of 10% of the DJIA, or $84 per share relative to the closing price of 840 for the DJIA in 1970. On this basis the shares would have appeared quite attractive at their price of 57 1 ⁄4. To what extent should the senior analyst have marked down the “past-performance value” to allow for adverse developments that he saw in the future? Frankly, we have no idea. Assume he had rea- son to believe that the 1971 earnings would be as low as $2.50 per share—a large drop from the 1970 figure, as against an advance expected for the DJIA. Very likely the stock market would take this poor performance quite seriously, but would it really establish the once mighty Aluminum Company of America as a relatively unprofitable enterprise, to be valued at less than its tangible assets behind the shares?† (In 1971 the price declined from a high of 70 in May to a low of 36 in December, against a book value of 55.) ALCOA is surely a representative industrial company of huge size, but we think that its price-and-earnings history is more unusual, even contradictory, than that of most other large enter- prises. Yet this instance supports to some degree, the doubts we expressed in the last chapter as to the dependability of the appraisal procedure when applied to the typical industrial company. Things to Consider About Per-Share Earnings 321 * See pp. 299–301. † Recent history—and a mountain of financial research—have shown that the market is unkindest to rapidly growing companies that suddenly report a fall in earnings. More moderate and stable growers, as ALCOA was in Graham’s day or Anheuser-Busch and Colgate-Palmolive are in our time, tend to suffer somewhat milder stock declines if they report disappointing earnings. Great expectations lead to great disappointment if they are not met; a failure to meet moderate expectations leads to a much milder reac- tion. Thus, one of the biggest risks in owning growth stocks is not that their growth will stop, but merely that it will slow down. And in the long run, that is not merely a risk, but a virtual certainty. COMMENTARY ON CHAPTER 12 You can get ripped off easier by a dude with a pen than you can by a dude with a gun. —Bo Diddley THE NUMBERS GAME Even Graham would have been startled by the extent to which compa- nies and their accountants pushed the limits of propriety in the past few years. Compensated heavily through stock options, top execu- tives realized that they could become fabulously rich merely by increasing their company’s earnings for just a few years running. 1 Hun- dreds of companies violated the spirit, if not the letter, of accounting principles—turning their financial reports into gibberish, tarting up ugly results with cosmetic fixes, cloaking expenses, or manufacturing earn- ings out of thin air. Let’s look at some of these unsavory practices. AS IF! Perhaps the most widespread bit of accounting hocus-pocus was the “pro forma” earnings fad. There’s an old saying on Wall Street that every bad idea starts out as a good idea, and pro forma earnings pre- sentation is no different. The original point was to provide a truer pic- ture of the long-term growth of earnings by adjusting for short-term deviations from the trend or for supposedly “nonrecurring” events. A pro forma press release might, for instance, show what a company would have earned over the past year if another firm it just acquired had been part of the family for the entire 12 months. 322 1 For more on how stock options can enrich corporate managers—but not necessarily outside shareholders—see the commentary on Chapter 19. But, as the Naughty 1990s advanced, companies just couldn’t leave well enough alone. Just look at these examples of pro forma flim- flam: • For the quarter ended September 30, 1999, InfoSpace, Inc. pre- sented its pro forma earnings as if it had not paid $159.9 million in preferred-stock dividends. • For the quarter ended October 31, 2001, BEA Systems, Inc. pre- sented its pro forma earnings as if it had not paid $193 million in payroll taxes on stock options exercised by its employees. • For the quarter ended March 31, 2001, JDS Uniphase Corp. pre- sented its pro forma earnings as if it had not paid $4 million in payroll taxes, had not lost $7 million investing in lousy stocks, and had not incurred $2.5 billion in charges related to mergers and goodwill. In short, pro forma earnings enable companies to show how well they might have done if they hadn’t done as badly as they did. 2 As an intelligent investor, the only thing you should do with pro forma earn- ings is ignore them. HUNGRY FOR RECOGNITION In 2000, Qwest Communications International Inc., the telecommuni- cations giant, looked strong. Its shares dropped less than 5% even as the stock market lost more than 9% that year. But Qwest’s financial reports held an odd little revelation. In late 1999, Qwest decided to recognize the revenues from its telephone directories as soon as the phone books were published—even though, as anyone who has ever taken out a Yellow Pages advertisement knows, many businesses pay for those ads in monthly installments. Commentary on Chapter 12 323 2 All the above examples are taken directly from press releases issued by the companies themselves. For a brilliant satire on what daily life would be like if we all got to justify our behavior the same way companies adjust their reported earnings, see “My Pro Forma Life,” by Rob Walker, at http://slate. msn.com/?id=2063953. (“. . . a recent post-workout lunch of a 22-ounce, bone-in rib steak at Smith & Wollensky and three shots of bourbon is treated here as a nonrecurring expense. I’ll never do that again!”) Abracadabra! That piddly-sounding “change in accounting principle” pumped up 1999 net income by $240 million after taxes—a fifth of all the money Qwest earned that year. Like a little chunk of ice crowning a submerged iceberg, aggressive revenue recognition is often a sign of dangers that run deep and loom large—and so it was at Qwest. By early 2003, after reviewing its previ- ous financial statements, the company announced that it had prema- turely recognized profits on equipment sales, improperly recorded the costs of services provided by outsiders, inappropriately booked costs as if they were capital assets rather than expenses, and unjustifiably treated the exchange of assets as if they were outright sales. All told, Qwest’s revenues for 2000 and 2001 had been overstated by $2.2 billion—including $80 million from the earlier “change in accounting principle,” which was now reversed. 3 CAPITAL OFFENSES In the late 1990s, Global Crossing Ltd. had unlimited ambitions. The Bermuda-based company was building what it called the “first inte- grated global fiber optic network” over more than 100,000 miles of 324 Commentary on Chapter 12 3 In 2002, Qwest was one of 330 publicly-traded companies to restate past financial statements, an all-time record, according to Huron Consulting Group. All information on Qwest is taken from its financial filings with the U.S. Securities and Exchange Commission (annual report, Form 8K, and Form 10-K) found in the EDGAR database at www.sec.gov. No hindsight was required to detect the “change in accounting principle,” which Qwest fully disclosed at the time. How did Qwest’s shares do over this period? At year-end 2000, the stock had been at $41 per share, a total market value of $67.9 billion. By early 2003, Qwest was around $4, valuing the entire com- pany at less than $7 billion—a 90% loss. The drop in share price is not the only cost associated with bogus earnings; a recent study found that a sam- ple of 27 firms accused of accounting fraud by the SEC had overpaid $320 million in Federal income tax. Although much of that money will eventually be refunded by the IRS, most shareholders are unlikely to stick around to bene- fit from the refunds. (See Merle Erickson, Michelle Hanlon, and Edward May- dew, “How Much Will Firms Pay for Earnings that Do Not Exist?” at http:// papers.ssrn.com.) cables, largely laid across the floor of the world’s oceans. After wiring the world, Global Crossing would sell other communications compa- nies the right to carry their traffic over its network of cables. In 1998 alone, Global Crossing spent more than $600 million to construct its optical web. That year, nearly a third of the construction budget was charged against revenues as an expense called “cost of capacity sold.” If not for that $178 million expense, Global Crossing—which reported a net loss of $96 million—could have reported a net profit of roughly $82 million. The next year, says a bland footnote in the 1999 annual report, Global Crossing “initiated service contract accounting.” The company would no longer charge most construction costs as expenses against the immediate revenues it received from selling capacity on its net- work. Instead, a major chunk of those construction costs would now be treated not as an operating expense but as a capital expenditure— thereby increasing the company’s total assets, instead of decreasing its net income. 4 Poof! In one wave of the wand, Global Crossing’s “property and equipment” assets rose by $575 million, while its cost of sales increased by a mere $350 million—even though the company was spending money like a drunken sailor. Capital expenditures are an essential tool for managers to make a good business grow bigger and better. But malleable accounting rules permit managers to inflate reported profits by transforming nor- Commentary on Chapter 12 325 4 Global Crossing formerly treated much of its construction costs as an expense to be charged against the revenue generated from the sale or lease of usage rights on its network. Customers generally paid for their rights up front, although some could pay in installments over periods of up to four years. But Global Crossing did not book most of the revenues up front, instead deferring them over the lifetime of the lease. Now, however, because the networks had an estimated usable life of up to 25 years, Global Cross- ing began treating them as depreciable, long-lived capital assets. While this treatment conforms with Generally Accepted Accounting Principles, it is unclear why Global Crossing did not use it before October 1, 1999, or what exactly prompted the change. As of March 2001, Global Crossing had a total stock valuation of $12.6 billion; the company filed for bankruptcy on January 28, 2002, rendering its common stock essentially worthless. . Another is the dilu- tive effect of issuing millions of stock options for executive compensation, then buying back millions of shares to keep those options from reducing the value of the common. of these factors: the use of special charges, which may never be reflected in the per-share earnings, the reduction in the normal income-tax deduction by reason of past losses, and the dilution. from returning to the older straight-line depreciation rates, less burden- some on earnings than the accelerated method used the year before. (The company will continue to use the accelerated rate on its