Graham is pointing out that

 * Graham is pointing out that, despite the promotional rhetoric that investors usually hear, convertible bonds do not automatically offer “the best of both worlds.” Higher yield and lower risk do not always go hand in hand. What Wall Street gives with one hand, it usually takes away with the other. An investment may offer the best of one world, or the worst of another; but the best of both worlds seldom becomes available in a single package. † According to Goldman Sachs and Ibbotson Associates, from 1998 through 2002, convertibles generated an average annual return of 4.8%. That was considerably better than the 0.6% annual loss on U.S. stocks, but substantially worse than the returns of medium-term corporate bonds (a o comparisons would demonstrate that convertible securities as a whole have relatively poor quality as senior issues and also are tied to common stocks that do worse than the general market except during a speculative upsurge. These observations do not apply to all convertible issues, of course. In the 1968 and 1969 particularly, a fair number of strong companies used convertible issues to combat the inordinately high interest rates for even first-quality bonds. But it is noteworthy that in our 20-stock sample of convertible preferreds only one showed an advance and 14 suffered bad declines.* 406 The Intelligent Investor TABLE 16-2 Price Record of Preferred Stocks, Common Stocks, and Warrants, December 1970 versus December 1968 (Based on Random Samples of 20 Issues Each) Straight Preferred Stocks Rated A or Better Rated Below A Convertible Preferred Stocks Listed Warrants Listed Common Stocks Advances 2 0 1 2 1 Declines: 0–10% 3 3 3 4 0 10–20% 14 10 2 1 0 20–40% 1 5 5 6 1 40% or more 0 0 9 7 18 Average declines 10% 17% 29% 33% 65% (Standard & Poor’s composite index of 500 common stocks declined 11.3%.) * Recent structural changes in the convertible market have negated some of these criticisms. Convertible preferred stock, which made up roughly half the total convertible market in Graham’s day, now accounts for only an eighth of the market. Maturities are shorter, making convertible bonds less volatile, and many now carry “call protection,” or assurances against early redemption. And more than half of all convertibles are now investment grade, a significant improvement in credit quality from Graham’s time. Thus, in 2002, the Merrill Lynch All U.S. Convertible Index lost 8.6%—versus the 22.1% loss of the S & P 500-stock index and the 31.3% decline in the NASDAQ Composite stock index. The conclusion to be drawn from these figures is not that convertible issues are in themselves less desirable than nonconvertible or “straight” securities. Other things being equal, the opposite is true. But we clearly see that other things are not equal in practice and that the addition of the conversion privilege often—perhaps generally—betrays an absence of genuine investment quality for the issue. It is true, of course, that a convertible preferred is safer than the common stock of the same company—that is to say, it carries smaller risk of eventual loss of principal. Consequently those who buy new convertibles instead of the corresponding common stock are logical to that extent. But in most cases the common would not have been an intelligent purchase to begin with, at the ruling price, and the substitution of the convertible preferred did not improve the picture sufficiently. Furthermore, a good deal of the buying of convertibles was done by investors who had no special interest or confidence in the common stock—that is, they would never have thought of buying the common at the time—but who were tempted by what seemed an ideal combination of a prior claim plus a conversion privilege close to the current market. In a number of instances this combination has worked out well, but the statistics seem to show that it is more likely to prove a pitfall. In connection with the ownership of convertibles there is a special problem which most investors fail to realize. Even when a profit appears it brings a dilemma with it. Should the holder sell on a small rise; should he hold for a much bigger advance; if the issue is called—as often happens when the common has gone up considerably—should he sell out then or convert into and retain the common stock?* Let us talk in concrete terms. You buy a 6% bond at 100, convertible into stock at 25—that is, at the rate of 40 shares for each $1,000 bond. The stock goes to 30, which makes the bond worth at least 120, and so it sells at 125. You either sell or hold. If you hold, hoping for a higher price, you are pretty much in the position of a comConvertible Issues and Warrants 407 * A bond is “called” when the issuing corporation forcibly pays it off ahead of the stated maturity date, or final due date for interest payments. For a brief summary of how convertible bonds work, see Note 1 in the commentary on this chapter (p. 418). mon shareholder, since if the stock goes down your bond will go down too. A conservative person is likely to say that beyond 125 his position has become too speculative, and therefore he sells and makes a gratifying 25% profit. So far, so good. But pursue the matter a bit. In many cases where the holder sells at 125 the common stock continues to advance, carrying the convertible with it, and the investor experiences that peculiar pain that comes to the man who has sold out much too soon. The next time, he decides to hold for 150 or 200. The issue goes up to 140 and he does not sell. Then the market breaks and his bond slides down to 80. Again he has done the wrong thing. Aside from the mental anguish involved in making these bad guesses—and they seem to be almost inevitable—there is a real arithmetical drawback to operations in convertible issues. It may be assumed that a stern and uniform policy of selling at 25% or 30% profit will work out best as applied to many holdings. This would then mark the upper limit of profit and would be realized only on the issues that worked out well. But, if—as appears to be true—these issues often lack adequate underlying security and tend to be floated and purchased in the latter stages of a bull market, then a goodly proportion of them will fail to rise to 125 but will not fail to collapse when the market turns downward. Thus the spectacular opportunities in convertibles prove to be illusory in practice, and the overall experience is marked by fully as many substantial losses—at least of a temporary kind—as there are gains of similar magnitude. Because of the extraordinary length of the 1950–1968 bull market, convertible issues as a whole gave a good account of themselves for some 18 years. But this meant only that the great majority of common stocks enjoyed large advances, in which most convertible issues were able to share. The soundness of investment in convertible issues can only be tested by their performance in a declining stock market—and this has always proved disappointing as a whole.* In our first edition (1949) we gave an illustration of this special 408 The Intelligent Investor * In recent years, convertibles have tended to outperform the Standard & Poor’s 500-stock index during declining stock markets, but they have typically underperformed other bonds—which weakens, but does not fully negate, the criticism Graham makes here. problem of “what to do” with a convertible when it goes up. We believe it still merits inclusion here. Like several of our references it is based on our own investment operations. We were members of a “select group,” mainly of investment funds, who participated in a private offering of convertible 41 ⁄2% debentures of Eversharp Co. at par, convertible into common stock at $40 per share. The stock advanced rapidly to 651 ⁄2, and then (after a three-for-two split) to the equivalent of 88. The latter price made the convertible debentures worth no less than 220. During this period the two issues were called at a small premium; hence they were practically all converted into common stock, which was retained by a number of the original investment-fund buyers of the debentures. The price promptly began a severe decline, and in March 1948 the stock sold as low as 73 ⁄8. This represented a value of only 27 for the debenture issues, or a loss of 75% of the original price instead of a profit of over 100%. The real point of this story is that some of the original purchasers converted their bonds into the stock and held the stock through its great decline. In so doing they ran counter to an old maxim of Wall Street, which runs: “Never convert a convertible bond.” Why this advice? Because once you convert you have lost your strategic combination of prior claimant to interest plus a chance for an attractive profit. You have probably turned from investor into speculator, and quite often at an unpropitious time (because the stock has already had a large advance). If “Never convert a convertible” is a good rule, how came it that these experienced fund managers exchanged their Eversharp bonds for stock, to their subsequent embarrassing loss? The answer, no doubt, is that they let themselves be carried away by enthusiasm for the company’s prospects as well as by the “favorable market action” of the shares. Wall Street has a few prudent principles; the trouble is that they are always forgotten when they are most needed.* Hence that other famous dictum of the oldtimers: “Do as I say, not as I do.” Our general attitude toward new convertible issues is thus a mistrustful one. We mean here, as in other similar observations, Convertible Issues and Warrants 409 * This sentence could serve as the epitaph for the bull market of the 1990s. Among the “few prudent principles” that investors forgot were such market clichés as “Trees don’t grow to the sky” and “Bulls make money, bears make money, but pigs get slaughtered.” that the investor should look more than twice before he buys them. After such hostile scrutiny he may find some exceptional offerings that are too good to refuse. The ideal combination, of course, is a strongly secured convertible, exchangeable for a common stock which itself is attractive, and at a price only slightly higher than the current market. Every now and then a new offering appears that meets these requirements. By the nature of the securities markets, however, you are more likely to find such an opportunity in some older issue which has developed into a favorable position rather than in a new flotation. (If a new issue is a really strong one, it is not likely to have a good conversion privilege.) The fine balance between what is given and what is withheld in a standard-type convertible issue is well illustrated by the extensive use of this type of security in the financing of American Telephone & Telegraph Company. Between 1913 and 1957 the company sold at least nine separate issues of convertible bonds, most of them through subscription rights to shareholders. The convertible bonds had the important advantage to the company of bringing in a much wider class of buyers than would have been available for a stock offering, since the bonds were popular with many financial institutions which possess huge resources but some of which were not permitted to buy stocks. The interest return on the bonds has generally been less than half the corresponding dividend yield on the stock—a factor that was calculated to offset the prior claim of the bondholders. Since the company maintained its $9 dividend rate for 40 years (from 1919 to the stock split in 1959) the result was the eventual conversion of virtually all the convertible issues into common stock. Thus the buyers of these convertibles have fared well through the years—but not quite so well as if they had bought the capital stock in the first place. This example establishes the soundness of American Telephone & Telegraph, but not the intrinsic attractiveness of convertible bonds. To prove them sound in practice we should need to have a number of instances in which the convertible worked out well even though the common stock proved disappointing. Such instances are not easy to find.* 410 The Intelligent Investor * AT&T Corp. no longer is a significant issuer of convertible bonds. Among the largest issuers of convertibles today are General Motors, Merrill Lynch, Tyco International, and Roche. Effect of Convertible Issues on the Status of the Common Stock In a large number of cases convertibles have been issued in connection with mergers or new acquisitions. Perhaps the most striking example of this financial operation was the issuance by the NVF Corp. of nearly $100,000,000 of its 5% convertible bonds (plus warrants) in exchange for most of the common stock of Sharon Steel Co. This extraordinary deal is discussed below pp. 429–433. Typically the transaction results in a pro forma increase in the reported earnings per share of common stock; the shares advance in response to their larger earnings, so-called, but also because the management has given evidence of its energy, enterprise, and ability to make more money for the shareholders.* But there are two offsetting factors, one of which is practically ignored and the other entirely so in optimistic markets. The first is the actual dilution of the current and future earnings on the common stock that flows arithmetically from the new conversion rights. This dilution can be quantified by taking the recent earnings, or assuming some other figures, and calculating the adjusted earnings per share if all the convertible shares or bonds were actually converted. In the majority of companies the resulting reduction in per-share figures is not significant. But there are numerous exceptions to this statement, and there is danger that they will grow at an uncomfortable rate. The fast-expanding “conglomerates” have been the chief practitioners of convertible legerdemain. In Table 16-3 we list seven companies with large amounts of stock issuable on conversions or against warrants.† Indicated Switches from Common into Preferred Stocks For decades before, say, 1956, common stocks yielded more than the preferred stocks of the same companies; this was particularly Convertible Issues and Warrants 411 * For a further discussion of “pro forma” financial results, see the commentary on Chapter 12. † In recent years, convertible bonds have been heavily issued by companies in the financial, health-care, and technology industries. true if the preferred stock had a conversion privilege close to the market. The reverse is generally true at present. As a result there are a considerable number of convertible preferred stocks which are clearly more attractive than the related common shares. Owners of the common have nothing to lose and important advantages to gain by switching from their junior shares into the senior issue. Example: A typical example was presented by StudebakerWorthington Corp. at the close of 1970. The common sold at 57, while the $5 convertible preferred finished at 871 ⁄2. Each preferred share is exchangeable for 11 ⁄2 shares of common, then worth 851 ⁄2. This would indicate a small money difference against the buyer of the preferred. But dividends are being paid on the common at the annual rate of $1.20 (or $1.80 for the 11 ⁄2 shares), against the $5 obtainable on one share of preferred. Thus the original adverse difference in price would probably be made up in less than a year, after which the preferred would probably return an appreciably higher dividend yield than the common for some time to come. But most important, of course, would be the senior position that the common shareholder would gain from the switch. At the low prices 412 The Intelligent Investor TABLE 16-3 Companies with Large Amounts of Convertible Issues and Warrants at the End of 1969 (Shares in Thousands) Common Stock Outstanding Additional Common Stock Issuable Total Additional Common Stock Against Warrants Preferred Bonds Stock On Conversion of Avco Corp. 11,470 1,750 10.436 3,085 15,271 Gulf & Western Inc. 14,964 9,671 5,632 6,951 22,260 International Tel. & Tel. 67,393 190 48,115 48,305 Ling-Temco-Vought 4,410a 1,180 685 7,564 9,429 National General 4,910 4,530 12,170 16,700 Northwest Industriesb 7,433 11,467 1,513 12,980 Rapid American 3,591 426 1,503 8,000 9,929 a Includes “special stock.” b At end of 1970. of 1968 and again in 1970 the preferred sold 15 points higher than 11 ⁄2 shares of common. Its conversion privilege guarantees that it could never sell lower than the common package.2 Stock-Option Warrants Let us mince no words at the outset. We consider the recent development of stock-option warrants as a near fraud, an existing menace, and a potential disaster. They have created huge aggregate dollar “values” out of thin air. They have no excuse for existence except to the extent that they mislead speculators and investors. They should be prohibited by law, or at least strictly limited to a minor part of the total capitalization of a company.* For an analogy in general history and in literature we refer the reader to the section of Faust (part 2), in which Goethe describes the invention of paper money. As an ominous precedent on Wall Street history, we may mention the warrants of American & Foreign Power Co., which in 1929 had a quoted market value of over a billion dollars, although they appeared only in a footnote to the company’s balance sheet. By 1932 this billion dollars had shrunk to $8 million, and in 1952 the warrants were wiped out in the company’s recapitalization—even though it had remained solvent. Originally, stock-option warrants were attached now and then to bond issues, and were usually equivalent to a partial conversion privilege. They were unimportant in amount, and hence did no harm. Their use expanded in the late 1920s, along with many other financial abuses, but they dropped from sight for long years thereafter. They were bound to turn up again, like the bad pennies they are, and since 1967 they have become familiar “instruments of Convertible Issues and Warrants 413 * Warrants were an extremely widespread technique of corporate finance in the nineteenth century and were fairly common even in Graham’s day. They have since diminished in importance and popularity—one of the few recent developments that would give Graham unreserved pleasure. As of year-end 2002, there were only seven remaining warrant issues on the New York Stock Exchange—only the ghostly vestige of a market. Because warrants are no longer commonly used by major companies, today’s investors should read the rest of Graham’s chapter only to see how his logic works. finance.” In fact a standard procedure has developed for raising the capital for new real-estate ventures, affiliates of large banks, by selling units of an equal number of common shares and warrants to buy additional common shares at the same price. Example: In 1971 CleveTrust Realty Investors sold 2,500,000 of these combinations of common stock (or “shares of beneficial interest”) and warrants, for $20 per unit. Let us consider for a moment what is really involved in this financial setup. Ordinarily, a common-stock issue has the first right to buy additional common shares when the company’s directors find it desirable to raise capital in this manner. This so-called “preemptive right” is one of the elements of value entering into the ownership of common stock—along with the right to receive dividends, to participate in the company’s growth, and to vote for directors. When separate warrants are issued for the right to subscribe additional capital, that action takes away part of the value inherent in an ordinary common share and transfers it to a separate certificate. An analogous thing could be done by issuing separate certificates for the right to receive dividends (for a limited or unlimited period), or the right to share in the proceeds of sale or liquidation of the enterprise, or the right to vote the shares. Why then are these subscription warrants created as part of the original capital structure? Simply because people are inexpert in financial matters. They don’t realize that the common stock is worth less with warrants outstanding than otherwise. Hence the package of stock and warrants usually commands a better price in the market than would the stock alone. Note that in the usual company reports the per-share earnings are (or have been) computed without proper allowance for the effect of outstanding warrants. The result is, of course, to overstate the true relationship between the earnings and the market value of the company’s capitalization.* 414 The Intelligent Investor * Today, the last remnant of activity in warrants is in the cesspool of the NASDAQ “bulletin board,” or over-the-counter market for tiny companies, where common stock is often bundled with warrants into a “unit” (the contemporary equivalent of what Graham calls a “package”). If a stockbroker ever offers to sell you “units” in any company, you can be 95% certain that warrants are involved, and at least 90% certain that the broker is either a thief or an idiot. Legitimate brokers and firms have no business in this area. The simplest and probably the best method of allowing for the existence of warrants is to add the equivalent of their market value to the common-share capitalization, thus increasing the “true” market price per share. Where large amounts of warrants have been issued in connection with the sale of senior securities, it is customary to make the adjustment by assuming that the proceeds of the stock payment are used to retire the related bonds or preferred shares. This method does not allow adequately for the usual “premium value” of a warrant above exercisable value. In Table 16-4 we compare the effect of the two methods of calculation in the case of National General Corp. for the year 1970. Does the company itself derive an advantage from the creation of these warrants, in the sense that they assure it in some way of receiving additional capital when it needs some? Not at all. Ordinarily there is no way in which the company can require the warrant-holders to exercise their rights, and thus provide new capital to the company, prior to the expiration date of the warrants. In the meantime, if the company wants to raise additional common-stock funds it must offer the shares to its shareholders in the usual way— which means somewhat under the ruling market price. The warrants are no help in such an operation; they merely complicate the situation by frequently requiring a downward revision in their own subscription price. Once more we assert that large issues of stock-option warrants serve no purpose, except to fabricate imaginary market values. The paper money that Goethe was familiar with, when he wrote his Faust, were the notorious French assignats that had been greeted as a marvelous invention, and were destined ultimately to lose all of their value—as did the billion dollars worth of American & Foreign Power warrants.* Some of the poet’s remarks apply Convertible Issues and Warrants 415 * The “notorious French assignats” were issued during the Revolution of 1789. They were originally debts of the Revolutionary government, purportedly secured by the value of the real estate that the radicals had seized from the Catholic church and the nobility. But the Revolutionaries were bad financial managers. In 1790, the interest rate on assignats was cut; soon they stopped paying interest entirely and were reclassified as paper money. But the government refused to redeem them for gold or silver and issued massive amounts of new assignats. They were officially declared worthless in 1797. equally well to one invention or another—such as the following (in Bayard Taylor’s translation): Faust: Imagination in its highest flight Exerts itself but cannot grasp it quite. Mephistopheles (the inventor): If one needs coin the brokers ready stand. The Fool (finally): The magic paper ...! 416 The Intelligent Investor TABLE 16-4 Calculation of “True Market Price” and Adjusted Price/Earnings Ratio of a Common Stock with Large Amounts of Warrants Outstanding (Example: National General Corp. in June 1971) 1. Calculation of “True Market Price.” Market value of 3 issues of warrants, June 30, 1971 $94,000,000 Value of warrants per share of common stock $18.80 Price of common stock alone 24.50 Corrected price of common, adjusted for warrants 43.30 2. Calculation of P/E Ratio to Allow for Warrant Dilution Before After Warrant Dilution (1970 earnings) Warrant Company’s Our A. Before Special Items. Dilution Calculation Calculation Earned per share $ 2.33 $ 1.60 $ 2.33 Price of common 24.50 24.50 43.30 (adj.) P/E ratio 10.5
15.3
18.5
B. After Special Items. Earned per share $ .90 $ 1.33 $ .90 Price of common 24.50 24.50 43.30 (adj.) P/E ratio 27.2
18.4
48.1
Note that, after special charges, the effect of the company’s calculation is to increase the earnings per share and reduce the P/E ratio. This is manifestly absurd. By our suggested method the effect of the dilution is to increase the P/E ratio substantially, as it should be. Practical Postscript The crime of the warrants is in “having been born.”* Once born they function as other security forms, and offer chances of profit as well as of loss. Nearly all the newer warrants run for a limited time—generally between five and ten years. The older warrants were often perpetual, and they were likely to have fascinating price histories over the years. Example: The record books will show that Tri-Continental Corp. warrants, which date from 1929, sold at a negligible 1/32 of a dollar each in the depth of the depression. From that lowly estate their price rose to a magnificent 753 ⁄4 in 1969, an astronomical advance of some 242,000%. (The warrants then sold considerably higher than the shares themselves; this is the kind of thing that occurs on Wall Street through technical developments, such as stock splits.) A recent example is supplied by Ling-Temco-Vought warrants, which in the first half of 1971 advanced from 21 ⁄2 to 121 ⁄2—and then fell back to 4. No doubt shrewd operations can be carried on in warrants from time to time, but this is too technical a matter for discussion here. We might say that warrants tend to sell relatively higher than the corresponding market components related to the conversion privilege of bonds or preferred stocks. To that extent there is a valid argument for selling bonds with warrants attached rather than creating an equivalent dilution factor by a convertible issue. If the warrant total is relatively small there is no point in taking its theoretical aspect too seriously; if the warrant issue is large relative to the outstanding stock, that would probably indicate that the company has a top-heavy senior capitalization. It should be selling additional common stock instead. Thus the main objective of our attack on warrants as a financial mechanism is not to condemn their use in connection with moderate-size bond issues, but to argue against the wanton creation of huge “paper-money” monstrosities of this genre. Convertible Issues and Warrants 417 * Graham, an enthusiastic reader of Spanish literature, is paraphrasing a line from the play Life Is a Dream by Pedro Calderon de la Barca (1600–1681): “The greatest crime of man is having been born.” COMMENTARY ON CHAPTER 16 That which thou sowest is not quickened, except it die. —I. Corinthians, XV:36. THE ZEAL OF THE CONVERT Although convertible bonds are called “bonds,” they behave like stocks, work like options, and are cloaked in obscurity. If you own a convertible, you also hold an option: You can either keep the bond and continue to earn interest on it, or you can exchange it for common stock of the issuing company at a predetermined ratio. (An option gives its owner the right to buy or sell another security at a given price within a specific period of time.) Because they are exchangeable into stock, convertibles pay lower rates of interest than most comparable bonds. On the other hand, if a company’s stock price soars, a convertible bond exchangeable into that stock will perform much better than a conventional bond. (Conversely, the typical convertible—with its lower interest rate—will fare worse in a falling bond market.)1 418 1 As a brief example of how convertible bonds work in practice, consider the 4.75% convertible subordinated notes issued by DoubleClick Inc. in 1999. They pay $47.50 in interest per year and are each convertible into 24.24 shares of the company’s common stock, a “conversion ratio” of 24.24. As of year-end 2002, DoubleClick’s stock was priced at $5.66 a share, giving each bond a “conversion value” of $137.20 ($5.66
24.24). Yet the bonds traded roughly six times higher, at $881.30—creating a “conversion premium,” or excess over their conversion value, of 542%. If you bought at that price, your “break-even time,” or “payback period,” was very long. (You paid roughly $750 more than the conversion value of the bond, so it will take nearly 16 years of $47.50 interest payments for you to “earn back” that con- From 1957 through 2002, according to Ibbotson Associates, convertible bonds earned an annual average return of 8.3%—only two percentage points below the total return on stocks, but with steadier prices and shallower losses.2 More income, less risk than stocks: No wonder Wall Street’s salespeople often describe convertibles as a “best of both worlds” investment. But the intelligent investor will quickly realize that convertibles offer less income and more risk than most other bonds. So they could, by the same logic and with equal justice, be called a “worst of both worlds” investment. Which side you come down on depends on how you use them. In truth, convertibles act more like stocks than bonds. The return on convertibles is about 83% correlated to the Standard & Poor’s 500- stock index—but only about 30% correlated to the performance of Treasury bonds. Thus, “converts” zig when most bonds zag. For conservative investors with most or all of their assets in bonds, adding a diversified bundle of converts is a sensible way to seek stock-like returns without having to take the scary step of investing in stocks directly. You could call convertible bonds “stocks for chickens.” As convertibles expert F. Barry Nelson of Advent Capital Management points out, this roughly $200 billion market has blossomed since Graham’s day. Most converts are now medium-term, in the seven-to10-year range; roughly half are investment-grade; and many issues now carry some call protection (an assurance against early redemption). All these factors make them less risky than they used to be.3 Commentary on Chapter 16 419 version premium.) Since each DoubleClick bond is convertible to just over 24 common shares, the stock will have to rise from $5.66 to more than $36 if conversion is to become a practical option before the bonds mature in 2006. Such a stock return is not impossible, but it borders on the miraculous. The cash yield on this particular bond scarcely seems adequate, given the low probability of conversion. 2 Like many of the track records commonly cited on Wall Street, this one is hypothetical. It indicates the return you would have earned in an imaginary index fund that owned all major convertibles. It does not include any management fees or trading costs (which are substantial for convertible securities). In the real world, your returns would have been roughly two percentage points lower. 3 However, most convertible bonds remain junior to other long-term debt and bank loans—so, in a bankruptcy, convertible holders do not have prior It’s expensive to trade small lots of convertible bonds, and diversification is impractical unless you have well over $100,000 to invest in this sector alone. Fortunately, today’s intelligent investor has the convenient recourse of buying a low-cost convertible bond fund. Fidelity and Vanguard offer mutual funds with annual expenses comfortably under 1%, while several closed-end funds are also available at a reasonable cost (and, occasionally, at discounts to net asset value).4 On Wall Street, cuteness and complexity go hand-in-hand—and convertibles are no exception. Among the newer varieties are a jumble of securities with acronymic nicknames like LYONS, ELKS, EYES, PERCS, MIPS, CHIPS, and YEELDS. These intricate securities put a “floor” under your potential losses, but also cap your potential profits and often compel you to convert into common stock on a fixed date. Like most investments that purport to ensure against loss (see sidebar on p. 421), these things are generally more trouble than they are worth. You can best shield yourself against loss not by buying one of these quirky contraptions, but by intelligently diversifying your entire portfolio across cash, bonds, and U.S. and foreign stocks. 420 Commentary on Chapter 16 claim to the company’s assets. And, while they are not nearly as dicey as high-yield “junk” bonds, many converts are still issued by companies with less than sterling credit ratings. Finally, a large portion of the convertible market is held by hedge funds, whose rapid-fire trading can increase the volatility of prices. 4 For more detail, see www.fidelity.com, www.vanguard.com, and www. morningstar.com. The intelligent investor will never buy a convertible bond fund with annual operating expenses exceeding 1.0%. Commentary on Chapter 16 421 UNCOVERING COVERED CALLS As the bear market clawed its way through 2003, it dug up an old fad: writing covered call options. (A recent Google search on “covered call writing” turned up more than 2,600 hits.) What are covered calls, and how do they work? Imagine that you buy 100 shares of Ixnay Corp. at $95 apiece. You then sell (or “write”) a call option on your shares. In exchange, you get a cash payment known as a “call premium.” (Let’s say it’s $10 per share.) The buyer of the option, meanwhile, has the contractual right to buy your Ixnay shares at a mutually agreed-upon price— say, $100. You get to keep the stock so long as it stays below $100, and you earn a fat $1,000 in premium income, which will cushion the fall if Ixnay’s stock crashes. Less risk, more income. What’s not to like? Well, now imagine that Ixnay’s stock price jumps overnight to $110. Then your option buyer will exercise his rights, yanking your shares away for $100 apiece. You’ve still got your $1,000 in income, but he’s got your Ixnay—and the more it goes up, the harder you will kick yourself.1 Since the potential gain on a stock is unlimited, while no loss can exceed 100%, the only person you will enrich with this strategy is your broker. You’ve put a floor under your losses, but you’ve also slapped a ceiling over your gains. For individual investors, covering your downside is never worth surrendering most of your upside. 1 Alternatively, you could buy back the call option, but you would have to take a loss on it—and options can have even higher trading costs than stocks. CHAPTER 17 Four Extremely Instructive Case Histories The word “extremely” in the title is a kind of pun, because the histories represent extremes of various sorts that were manifest on Wall Street in recent years. They hold instruction, and grave warnings, for everyone who has a serious connection with the world of stocks and bonds—not only for ordinary investors and speculators but for professionals, security analysts, fund managers, trustaccount administrators, and even for bankers who lend money to corporations. The four companies to be reviewed, and the different extremes that they illustrate are: Penn Central (Railroad) Co. An extreme example of the neglect of the most elementary warning signals of financial weakness, by all those who had bonds or shares of this system under their supervision. A crazily high market price for the stock of a tottering giant. Ling-Temco-Vought Inc. An extreme example of quick and unsound “empire building,” with ultimate collapse practically guaranteed; but helped by indiscriminate bank lending. NVF Corp. An extreme example of one corporate acquisition, in which a small company absorbed another seven times its size, incurring a huge debt and employing some startling accounting devices. AAA Enterprises. An extreme example of public stock-financing of a small company; its value based on the magic word “franchising,” and little else, sponsored by important stock-exchange houses. Bankruptcy followed within two years of the stock sale and the doubling of the initial inflated price in the heedless stock market. 422 The Penn Central Case This is the country’s largest railroad in assets and gross revenues. Its bankruptcy in 1970 shocked the financial world. It has defaulted on most of its bond issues, and has been in danger of abandoning its operations entirely. Its security issues fell drastically in price, the common stock collapsing from a high level of 861 ⁄2 as recently as 1968 to a low of 51 ⁄2 in 1970. (There seems little doubt that these shares will be wiped out in reorganization.)* Our basic point is that the application of the simplest rules of security analysis and the simplest standards of sound investment would have revealed the fundamental weakness of the Penn Central system long before its bankruptcy—certainly in 1968, when the shares were selling at their post-1929 record, and when most of its bond issues could have been exchanged at even prices for wellsecured public-utility obligations with the same coupon rates. The following comments are in order: 1. In the S & P Bond Guide the interest charges of the system are shown to have been earned 1.91 times in 1967 and 1.98 times in 1968. The minimum coverage prescribed for railroad bonds in our textbook Security Analysis is 5 times before income taxes and 2.9 times after income taxes at regular rates. As far as we know the validity of these standards has never been questioned by any investment authority. On the basis of our requirements for earnings after taxes, the Penn Central fell short of the requirements for safety. But our after-tax requirement is based on a before-tax ratio of five times, with regular income tax deducted after the bond interest. In the case of Penn Central, it had been paying no income taxes to speak of for the past 11 years! Hence the coverage of its interest charges before taxes was less than two times—a totally inadequate figure against our conservative requirement of 5 times. Four Extremely Instructive Case Histories 423 * How “shocked” was the financial world by the Penn Central’s bankruptcy, which was filed over the weekend of June 20–21, 1970? The closing trade in Penn Central’s stock on Friday, June 19, was $11.25 per share—hardly a going-out-of-business price. In more recent times, stocks like Enron and WorldCom have also sold at relatively high prices shortly before filing for bankruptcy protection. 2. The fact that the company paid no income taxes over so long a period should have raised serious questions about the validity of its reported earnings. 3. The bonds of the Penn Central system could have been exchanged in 1968 and 1969, at no sacrifice of price or income, for far better secured issues. For example, in 1969, Pennsylvania RR 41 ⁄2s, due 1994 (part of Penn Central) had a range of 61 to 741 ⁄2, while Pennsylvania Electric Co. 43 ⁄8s, due 1994, had a range of 641 ⁄4 to 721 ⁄4. The public utility had earned its interest 4.20 times before taxes in 1968 against only 1.98 times for the Penn Central system; during 1969 the latter’s comparative showing grew steadily worse. An exchange of this sort was clearly called for, and it would have been a lifesaver for a Penn Central bondholder. (At the end of 1970 the railroad 41 ⁄4s were in default, and selling at only 181 ⁄2, while the utility’s 43 ⁄8s closed at 661 ⁄2.) 4. Penn Central reported earnings of $3.80 per share in 1968; its high price of 861 ⁄2 in that year was 24 times such earnings. But any analyst worth his salt would have wondered how “real” were earnings of this sort reported without the necessity of paying any income taxes thereon. 5. For 1966 the newly merged company* had reported “earnings” of $6.80 a share—in reflection of which the common stock later rose to its peak of 861 ⁄2. This was a valuation of over $2 billion for the equity. How many of these buyers knew at the time that the so lovely earnings were before a special charge of $275 million or $12 per share to be taken in 1971 for “costs and losses” incurred on the merger. O wondrous fairyland of Wall Street where a company can announce “profits” of $6.80 per share in one place and special “costs and losses” of $12 in another, and shareholders and speculators rub their hands with glee!† 424 The Intelligent Investor * Penn Central was the product of the merger, announced in 1966, of the Pennsylvania Railroad and the New York Central Railroad. † This kind of accounting legerdemain, in which profits are reported as if “unusual” or “extraordinary” or “nonrecurring” charges do not matter, anticipates the reliance on “pro forma” financial statements that became popular in the late 1990s (see the commentary on Chapter 12). 6. A railroad analyst would have long since known that the operating picture of the Penn Central was very bad in comparison with the more profitable roads. For example, its transportation ratio was 47.5% in 1968 against 35.2% for its neighbor, Norfolk & Western.* 7. Along the way there were some strange transactions with peculiar accounting results.1 Details are too complicated to go into here. Conclusion: Whether better management could have saved the Penn Central bankruptcy may be arguable. But there is no doubt whatever that no bonds and no shares of the Penn Central system should have remained after 1968 at the latest in any securities account watched over by competent security analysts, fund managers, trust officers, or investment counsel. Moral: Security analysts should do their elementary jobs before they study stock-market movements, gaze into crystal balls, make elaborate mathematical calculations, or go on all-expense-paid field trips.† Ling-Temco-Vought Inc. This is a story of head-over-heels expansion and head-overheels debt, ending up in terrific losses and a host of financial problems. As usually happens in such cases, a fair-haired boy, or “young genius,” was chiefly responsible for both the creation of the great empire and its ignominious downfall; but there is plenty of blame to be accorded others as well.‡ Four Extremely Instructive Case Histories 425 * A railroad’s “transportation ratio” (now more commonly called its operating ratio) measures the expenses of running its trains divided by the railroad’s total revenues. The higher the ratio, the less efficient the railroad. Today even a ratio of 70% would be considered excellent. † Today, Penn Central is a faded memory. In 1976, it was absorbed into Consolidated Rail Corp. (Conrail), a federally-funded holding company that bailed out several failed railroads. Conrail sold shares to the public in 1987 and, in 1997, was taken over jointly by CSX Corp. and Norfolk Southern Corp. ‡ Ling-Temco-Vought Inc. was founded in 1955 by James Joseph Ling, an electrical contractor who sold his first $1 million worth of shares to the pub- The rise and fall of Ling-Temco-Vought can be summarized by setting forth condensed income accounts and balance-sheet items for five years between 1958 and 1970. This is done in Table 17-1. The first column shows the company’s modest beginnings in 1958, when its sales were only $7 million. The next gives figures for 1960; the enterprise had grown twentyfold in only two years, but it was still comparatively small. Then came the heyday years to 1967 and 1968, in which sales again grew twentyfold to $2.8 billion with the debt figure expanding from $44 million to an awesome $1,653 million. In 1969 came new acquisitions, a further huge increase in debt (to a total of $1,865 million!), and the beginning of serious trouble. A large loss, after extraordinary items, was reported for the year; the stock price declined from its 1967 high of 1691 ⁄2 to a low of 24; the young genius was superseded as the head of the company. The 1970 results were even more dreadful. The enterprise reported a final net loss of close to $70 million; the stock fell away to a low price of 71 ⁄8, and its largest bond issue was quoted at one time at a pitiable 15 cents on the dollar. The company’s expansion policy was sharply reversed, various of its important interests were placed on the market, and some headway was made in reducing its mountainous obligations. The figures in our table speak so eloquently that few comments are called for. But here are some: 426 The Intelligent Investor lic by becoming his own investment banker, hawking prospectuses from a booth set up at the Texas State Fair. His success at that led him to acquire dozens of different companies, almost always using LTV’s stock to pay for them. The more companies LTV acquired, the higher its stock went; the higher its stock went, the more companies it could afford to acquire. By 1969, LTV was the 14th biggest firm on the Fortune 500 list of major U.S. corporations. And then, as Graham shows, the whole house of cards came crashing down. (LTV Corp., now exclusively a steelmaker, ended up seeking bankruptcy protection in late 2000.) Companies that grow primarily through acquisitions are called “serial acquirers”—and the similarity to the term “serial killers” is no accident. As the case of LTV demonstrates, serial acquirers nearly always leave financial death and destruction in their wake. Investors who understood this lesson of Graham’s would have avoided such darlings of the 1990s as Conseco, Tyco, and WorldCom. TABLE 17-1 Ling-Temco-Vought Inc., 1958–1970 (In Millions of Dollars Except Earned Per Share) A. Operating Results 1958 1960 1967 1969 1970 Sales $ 6.9 $143.0 $1,833.0 $3,750.0 $374.0 Net before taxes and interest 0.552 7.287 95.6 124.4 88.0 Interest charges .1 (est.) 1.5 (est.) 17.7 122.6 128.3 (Times earned) (5.5
) (4.8
) (54
) (1.02
) (0.68
) Income taxes 0.225 2.686 35.6 cr. 15.2 4.9 Special items dr. 40.6 dr. 18.8 Net after special items 0.227 3.051 34.0 dr. 38.3 dr. 69.6 Balance for common stock 0.202 3.051 30.7 dr. 40.8 dr. 71.3 Earned per share of common 0.17 0.83 5.56 def. 10.59 def. 17.18 B. Financial Position Total assets 6.4 94.5 845.0 2,944.0 2,582.0 Debt payable within 1 year 1.5 29.3 165.0 389.3 301.3 Long-term debt .5 14.6 202.6 1,500.8 1,394.6 Shareholders’ equity 2.7 28.5 245.0† def. 12.0* def. 69.0* Ratios Current assets/current liabilities 1.27
1.45
1.80
1.52
1.45
Equity/long-term debt 5.4
2.0
1.2
0.17
0.13
Market-price range 28–20 1691⁄2–109 973⁄4–241⁄8 291⁄2–71⁄8 * Excluding debt-discount as an asset and deducting preferred stock at redemption value. † As published. cr.: credit. dr.: debit. def.: deficit. 1. The company’s expansion period was not without an interruption. In 1961 it showed a small operating deficit, but—adopting a practice that was to be seen later in so many reports for 1970— evidently decided to throw all possible charges and reserves into the one bad year.* These amounted to a round $13 million, which was more than the combined net profits of the preceding three years. It was now ready to show “record earnings” in 1962, etc. 2. At the end of 1966 the net tangible assets are given as $7.66 per share of common (adjusted for a 3-for-2 split). Thus the market price in 1967 reached 22 times (!) its reported asset value at the time. At the end of 1968 the balance sheet showed $286 million available for 3,800,000 shares of common and Class AA stock, or about $77 per share. But if we deduct the preferred stock at full value and exclude the good-will items and the huge bond-discount “asset,”† there would remain $13 million for the common—a mere $3 per share. This tangible equity was wiped out by the losses of the following years. 3. Toward the end of 1967 two of our best-regarded banking firms offered 600,000 shares of Ling-Temco-Vought stock at $111 per share. It had been as high as 1691 ⁄2. In less than three years the price fell to 71 ⁄8.‡ 428 The Intelligent Investor * The sordid tradition of hiding a company’s true earnings picture under the cloak of restructuring charges is still with us. Piling up every possible charge in one year is sometimes called “big bath” or “kitchen sink” accounting. This bookkeeping gimmick enables companies to make an easy show of apparent growth in the following year—but investors should not mistake that for real business health. † The “bond-discount asset” appears to mean that LTV had purchased some bonds below their par value and was treating that discount as an asset, on the grounds that the bonds could eventually be sold at par. Graham scoffs at this, since there is rarely any way to know what a bond’s market price will be on a given date in the future. If the bonds could be sold only at values below par, this “asset” would in fact be a liability. ‡ We can only imagine what Graham would have thought of the investment banking firms that brought InfoSpace, Inc. public in December 1998. The stock (adjusted for later splits) opened for trading at $31.25, peaked at 4. At the end of 1967 the bank loans had reached $161 million, and a year later they stood at $414 million—which should have been a frightening figure. In addition, the long-term debt amounted to $1,237 million. By 1969 combined debt reached a total of $1,869 million. This may have been the largest combined debt figure of any industrial company anywhere and at any time, with the single exception of the impregnable Standard Oil of N.J. 5. The losses in 1969 and 1970 far exceeded the total profits since the formation of the company. Moral: The primary question raised in our mind by the LingTemco-Vought story is how the commercial bankers could have been persuaded to lend the company such huge amounts of money during its expansion period. In 1966 and earlier the company’s coverage of interest charges did not meet conservative standards, and the same was true of the ratio of current assets to current liabilities and of stock equity to total debt. But in the next two years the banks advanced the enterprise nearly $400 million additional for further “diversification.” This was not good business for them, and it was worse in its implications for the company’s shareholders. If the Ling-Temco-Vought case will serve to keep commercial banks from aiding and abetting unsound expansions of this type in the future, some good may come of it at last.* The NVF Takeover of Sharon Steel (A Collector’s Item) At the end of 1968 NVF Company was a company with $4.6 million of long-term debt, $17.4 million of stock capital, $31 million of sales, and $502,000 of net income (before a special credit of $374,000). Its business was described as “vulcanized fiber and plastics.” The management decided to take over the Sharon Steel Corp., Four Extremely Instructive Case Histories 429 $1305.32 per share in March 2000, and finished 2002 at a princely $8.45 per share. * Graham would have been disappointed, though surely not surprised, to see that commercial banks have chronically kept supporting “unsound expansions.” Enron and WorldCom, two of the biggest collapses in corporate history, were aided and abetted by billions of dollars in bank loans. which had $43 million of long-term debt, $101 million of stock capital, $219 million of sales, and $2,929,000 of net earnings. The company it wished to acquire was thus seven times the size of NVF. In early 1969 it made an offer for all the shares of Sharon. The terms per share were $70 face amount of NVF junior 5% bonds, due 1994, plus warrants to buy 11 ⁄2 shares of NVF stock at $22 per share of NVF. The management of Sharon strenuously resisted this takeover attempt, but in vain. NVF acquired 88% of the Sharon stock under the offer, issuing therefore $102 million of its 5% bonds and warrants for 2,197,000 of its shares. Had the offer been 100% operative the consolidated enterprise would, for the year 1968, have had $163 million in debt, only $2.2 million in tangible stock capital, $250 million of sales. The net-earnings question would have been a bit complicated, but the company subsequently stated them as a net loss of 50 cents per share of NVF stocks, before an extraordinary credit, and net earnings of 3 cents per share after such credit.* First Comment: Among all the takeovers effected in the year 1969 this was no doubt the most extreme in its financial disproportions. The acquiring company had assumed responsibility for a new and top-heavy debt obligation, and it had changed its calculated 1968 earnings from a profit to a loss into the bargain. A measure of the impairment of the company’s financial position by this 430 The Intelligent Investor * In June 1972 (just after Graham finished this chapter), a Federal judge found that NVF’s chairman, Victor Posner, had improperly diverted the pension assets of Sharon Steel “to assist affiliated companies in their takeovers of other corporations.” In 1977, the U.S. Securities and Exchange Commission secured a permanent injunction against Posner, NVF, and Sharon Steel to prevent them from future violations of Federal laws against securities fraud. The Commission alleged that Posner and his family had improperly obtained $1.7 million in personal perks from NVF and Sharon, overstated Sharon’s pretax earnings by $13.9 million, misrecorded inventory, and “shifted income and expenses from one year to another.” Sharon Steel, which Graham had singled out with his cold and skeptical eye, became known among Wall Street wags as “Share and Steal.” Posner was later a central force in the wave of leveraged buyouts and hostile takeovers that swept the United States in the 1980s, as he became a major customer for the junk bonds underwritten by Drexel Burnham Lambert. step is found in the fact that the new 5% bonds did not sell higher than 42 cents on the dollar during the year of issuance. This would have indicated grave doubt of the safety of the bonds and of the company’s future; however, the management actually exploited the bond price in a way to save the company annual income taxes of about $1,000,000 as will be shown. The 1968 report, published after the Sharon takeover, contained a condensed picture of its results, carried back to the year-end. This contained two most unusual items: 1. There is listed as an asset $58,600,000 of “deferred debt expense.” This sum is greater than the entire “stockholders’ equity,” placed at $40,200,000. 2. However, not included in the shareholders’ equity is an item of $20,700,000 designated as “excess of equity over cost of investment in Sharon.” Second Comment: If we eliminate the debt expense as an asset, which it hardly seems to be, and include the other item in the shareholders’ equity (where it would normally belong), then we have a more realistic statement of tangible equity for NVF stock, viz., $2,200,000. Thus the first effect of the deal was to reduce NVF’s “real equity” from $17,400,000 to $2,200,000 or from $23.71 per share to about $3 per share, on 731,000 shares. In addition the NVF shareholders had given to others the right to buy 31 ⁄2 times as many additional shares at six points below the market price at the close of 1968. The initial market value of the warrants was then about $12 each, or a total of some $30 million for those involved in the purchase offer. Actually, the market value of the warrants well exceeded the total market value of the outstanding NVF stock— another evidence of the tail-wagging-dog nature of the transaction. The Accounting Gimmicks When we pass from this pro forma balance sheet to the next year’s report we find several strange-appearing entries. In addition to the basic interest expense (a hefty $7,500,000), there is deducted $1,795,000 for “amortization of deferred debt expense.” But this last is nearly offset on the next line by a very unusual income item Four Extremely Instructive Case Histories 431 indeed: “amortization of equity over cost of investment in subsidiary: Cr. $1,650,000.” In one of the footnotes we find an entry, not appearing in any other report that we know of: Part of the stock capital is there designated as “fair market value of warrants issued in connection with acquisition, etc., $22,129,000.” What on earth do all these entries mean? None of them is even referred to in the descriptive text of the 1969 report. The trained security analyst has to figure out these mysteries by himself, almost in detective fashion. He finds that the underlying idea is to derive a tax advantage from the low initial price of the 5% debentures. For readers who may be interested in this ingenious arrangement we set forth our solution in Appendix 6. Other Unusual Items 1. Right after the close of 1969 the company bought in no less than 650,000 warrants at a price of $9.38 each. This was extraordinary when we consider that (a) NVF itself had only $700,000 in cash at the year-end, and had $4,400,000 of debt due in 1970 (evidently the $6 million paid for the warrants had to be borrowed); (b) it was buying in this warrant “paper money” at a time when its 5% bonds were selling at less than 40 cents on the dollar—ordinarily a warning that financial difficulties lay ahead. 2. As a partial offset to this, the company had retired $5,100,000 of its bonds along with 253,000 warrants in exchange for a like amount of common stock. This was possible because, by the vagaries of the securities markets, people were selling the 5% bonds at less than 40 while the common sold at an average price of 131 ⁄2, paying no dividend. 3. The company had plans in operation not only for selling stock to its employees, but also for selling them a larger number of warrants to buy the stock. Like the stock purchases the warrants were to be paid for 5% down and the rest over many years in the future. This is the only such employee-purchase plan for warrants that we know of. Will someone soon invent and sell on installments a right to buy a right to buy a share, and so on? 4. In the year 1969 the newly controlled Sharon Steel Co. changed its method of arriving at its pension costs, and also 432 The Intelligent Investor adopted lower depreciation rates. These accounting changes added about $1 per share to the reported earnings of NVF before dilution. 5. At the end of 1970 Standard & Poor’s Stock Guide reported that NVF shares were selling at a price/earning ratio of only 2, the lowest figure for all the 4,500-odd issues in the booklet. As the old Wall Street saying went, this was “important if true.” The ratio was based on the year’s closing price of 83 ⁄4 and the computed “earnings” of $5.38 per share for the 12 months ended September 1970. (Using these figures the shares were selling at only 1.6 times earnings.) But this ratio did not allow for the large dilution factor,* nor for the adverse results actually realized in the last quarter of 1970. When the full year’s figures finally appeared, they showed only $2.03 per share earned for the stock, before allowing for dilution, and $1.80 per share on a diluted basis. Note also that the aggregate market price of the stock and warrants on that date was about $14 million against a bonded debt of $135 million—a skimpy equity position indeed. AAA Enterprises History About 15 years ago a college student named Williams began selling mobile homes (then called “trailers”).† In 1965 he incorpoFour Extremely Instructive Case Histories 433 * The “large dilution factor” would be triggered when NVF employees exercised their warrants to buy common stock. The company would then have to issue more shares, and its net earnings would be divided across a much greater number of shares outstanding. † Jackie G. Williams founded AAA Enterprises in 1958. On its first day of trading, the stock soared 56% to close at $20.25. Williams later announced that AAA would come up with a new franchising concept every month (if people would step into a mobile home to get their income taxes done by “Mr. Tax of America,” just imagine what else they might do inside a trailer!). But AAA ran out of time and money before Williams ran out of ideas. The history of AAA Enterprises is reminiscent of the saga of a later company with charismatic management and scanty assets: ZZZZ Best achieved a stock-market value of roughly $200 million in the late 1980s, even though its purported industrial vacuum-cleaning business was little more than a telephone and a rented office run by a teenager named Barry Minkow. ZZZZ Best went bust and Minkow rated his business. In that year he sold $5,800,000 of mobile homes and earned $61,000 before corporate tax. By 1968 he had joined the “franchising” movement and was selling others the right to sell mobile homes under his business name. He also conceived the bright idea of going into the business of preparing income-tax returns, using his mobile homes as offices. He formed a subsidiary company called Mr. Tax of America, and of course started to sell franchises to others to use the idea and the name. He multiplied the number of corporate shares to 2,710,000 and was ready for a stock offering. He found that one of our largest stock-exchange houses, along with others, was willing to handle the deal. In March 1969 they offered the public 500,000 shares of AAA Enterprises at $13 per share. Of these, 300,000 were sold for Mr. Williams’s personal account and 200,000 were sold for the company account, adding $2,400,000 to its resources. The price of the stock promptly doubled to 28, or a value of $84 million for the equity, against a book value of, say, $4,200,000 and maximum reported earnings of $690,000. The stock was thus selling at a tidy 115 times its current (and largest) earnings per share. No doubt Mr. Williams had selected the name AAA Enterprise so that it might be among the first in the phone books and the yellow pages. A collateral result was that his company was destined to appear as the first name in Standard & Poor’s Stock Guide. Like Abu-Ben-Adhem’s, it led all the rest.* This gives a special reason to select it as a harrowing example of 1969 new financing and “hot issues.” Comment: This was not a bad deal for Mr. Williams. The 300,000 shares he sold had a book value in December of 1968 of $180,000 and he netted therefor 20 times as much, or a cool $3,600,000. The underwriters and distributors split $500,000 between them, less expenses. 434 The Intelligent Investor went to jail. Even as you read this, another similar company is being formed, and a new generation of “investors” will be taken for a ride. No one who has read Graham, however, should climb on board. * In “Abou Ben Adhem,” by the British Romantic poet Leigh Hunt (1784–1859), a righteous Muslim sees an angel writing in a golden book “the names of those who love the Lord.” When the angel tells Abou that his name is not among them, Abou says, “I pray thee, then, write me as one that loves his fellow men.” The angel returns the next night to show Abou the book, in which now “Ben Adhem’s name led all the rest.” 1. This did not seem so brilliant a deal for the clients of the selling houses. They were asked to pay about ten times the book value of the stock, after the bootstrap operation of increasing their equity per share from 59 cents to $1.35 with their own money.* Before the best year 1968, the company’s maximum earnings had been a ridiculous 7 cents per share. There were ambitious plans for the future, of course—but the public was being asked to pay heavily in advance for the hoped-for realization of these plans. 2. Nonetheless, the price of the stock doubled soon after original issuance, and any one of the brokerage-house clients could have gotten out at a handsome profit. Did this fact alter the flotation, or did the advance possibility that it might happen exonerate the original distributors of the issue from responsibility for this public offering and its later sequel? Not an easy question to answer, but it deserves careful consideration by Wall Street and the government regulatory agencies.† Subsequent History With its enlarged capital AAA Enterprises went into two additional businesses. In 1969 it opened a chain of retail carpet stores, and it acquired a plant that manufactured mobile homes. The results reported for the first nine months were not exactly brilliant, but they were a little better than the year before—22 cents a share against 14 Four Extremely Instructive Case Histories 435 * By purchasing more common stock at a premium to its book value, the investing public increased the value of AAA’s equity per share. But investors were only pulling themselves up by their own bootstraps, since most of the rise in shareholders’ equity came from the public’s own willingness to overpay for the stock. † Graham’s point is that investment banks are not entitled to take credit for the gains a hot stock may produce right after its initial public offering unless they are also willing to take the blame for the stock’s performance in the longer term. Many Internet IPOs rose 1,000% or more in 1999 and early 2000; most of them lost more than 95% in the subsequent three years. How could these early gains earned by a few investors justify the massive destruction of wealth suffered by the millions who came later? Many IPOs were, in fact, deliberately underpriced to “manufacture” immediate gains that would attract more attention for the next offering. cents. What happened in the next months was literally incredible. The company lost $4,365,000, or $1.49 per share. This consumed all its capital before the financing, plus the entire $2,400,000 received on the sale of stock plus two-thirds of the amount reported as earned in the first nine months of 1969. There was left a pathetic $242,000, or 8 cents per share, of capital for the public shareholders who had paid $13 for the new offering only seven months before. Nonetheless the shares closed the year 1969 at 81 ⁄8 bid, or a “valuation” of more than $25 million for the company. Further Comment: 1. It is too much to believe that the company had actually earned $686,000 from January to September 1969 and then lost $4,365,000 in the next three months. There was something sadly, badly, and accusingly wrong about the September 30 report. 2. The year’s closing price of 81 ⁄8 bid was even more of a demonstration of the complete heedlessness of stock-market prices than were the original offering price of 13 or the subsequent “hot-issue” advance to a high bid of 28. These latter quotations at least were based on enthusiasm and hope—out of all proportion to reality and common sense, but at least comprehensible. The year-end valuation of $25 million was given to a company that had lost all but a minuscule remnant of its capital, for which a completely insolvent condition was imminent, and for which the words “enthusiasm” or “hope” would be only bitter sarcasms. (It is true the year-end figures had not been published by December 31, but it is the business of Wall Street houses associated with a company to have monthly operating statements and a fairly exact idea of how things are going.) Final Chapter For the first half of 1970 the company reported a further loss of $1 million. It now had a good-sized capital deficit. It was kept out of bankruptcy by loans made by Mr. Williams, up to a total of $2,500,000. No further statements seem to have been issued, until in January 1971 AAA Enterprises finally filed a petition in bankruptcy. The quotation for the stock at month-end was still 50 cents a share bid, or $1,500,000 for the entire issue, which evidently had no more than wallpaper value. End of our story. Moral and Questions: The speculative public is incorrigible. In 436 The Intelligent Investor financial terms it cannot count beyond 3. It will buy anything, at any price, if there seems to be some “action” in progress. It will fall for any company identified with “franchising,” computers, electronics, science, technology, or what have you, when the particular fashion is raging. Our readers, sensible investors all, are of course above such foolishness. But questions remain: Should not responsible investment houses be honor-bound to refrain from identifying themselves with such enterprises, nine out of ten of which may be foredoomed to ultimate failure? (This was actually the situation when the author entered Wall Street in 1914. By comparison it would seem that the ethical standards of the “Street” have fallen rather than advanced in the ensuing 57 years, despite all the reforms and all the controls.) Could and should the SEC be given other powers to protect the public, beyond the present ones which are limited to requiring the printing of all important relevant facts in the offering prospectus? Should some kind of box score for public offerings of various types be compiled and published in conspicuous fashion? Should every prospectus, and perhaps every confirmation of sale under an original offering, carry some kind of formal warranty that the offering price for the issue is not substantially out of line with the ruling prices for issues of the same general type already established in the market? As we write this edition a movement toward reform of Wall Street abuses is under way. It will be difficult to impose worthwhile changes in the field of new offerings, because the abuses are so largely the result of the public’s own heedlessness and greed. But the matter deserves long and careful consideration.* Four Extremely Instructive Case Histories 437 * The first four sentences of Graham’s paragraph could read as the official epitaph of the Internet and telecommunications bubble that burst in early 2000. Just as the Surgeon General’s warning on the side of a cigarette pack does not stop everyone from lighting up, no regulatory reform will ever prevent investors from overdosing on their own greed. (Not even Communism can outlaw market bubbles; the Chinese stock market shot up 101.7% in the first half of 1999, then crashed.) Nor can investment banks ever be entirely cleansed of their own compulsion to sell any stock at any price the market will bear. The circle can only be broken one investor, and one financial adviser, at a time. Mastering Graham’s principles (see especially Chapters 1, 8, and 20) is the best way to start. COMMENTARY ON CHAPTER 17 The wisdom god, Woden, went out to the king of the trolls, got him in an armlock, and demanded to know of him how order might triumph over chaos. “Give me your left eye,” said the troll, “and I’ll tell you.” Without hesitation, Woden gave up his left eye. “Now tell me.” The troll said, “The secret is, ‘Watch with both eyes!’” —John Gardner THE MORE THINGS CHANG E... Graham highlights four extremes: • an overpriced “tottering giant” • an empire-building conglomerate • a merger in which a tiny firm took over a big one • an initial public offering of shares in a basically worthless company The past few years have provided enough new cases of Graham’s extremes to fill an encyclopedia. Here is a sampler: LUCENT, NOT TRANSPARENT In mid-2000, Lucent Technologies Inc. was owned by more investors than any other U.S. stock. With a market capitalization of $192.9 billion, it was the 12th-most-valuable company in America. Was that giant valuation justified? Let’s look at some basics from Lucent’s financial report for the fiscal quarter ended June 30, 2000:1 438 1 This document, like all the financial reports cited in this chapter, is readily available to the public through the EDGAR Database at www.sec.gov. A closer reading of Lucent’s report sets alarm bells jangling like an unanswered telephone switchboard: • Lucent had just bought an optical equipment supplier, Chromatis Networks, for $4.8 billion—of which $4.2 billion was “goodwill” (or cost above book value). Chromatis had 150 employees, no customers, and zero revenues, so the term “goodwill” seems inadequate; perhaps “hope chest” is more accurate. If Chromatis’s embryonic products did not work out, Lucent would have to reverse the goodwill and charge it off against future earnings. • A footnote discloses that Lucent had lent $1.5 billion to purchasers of its products. Lucent was also on the hook for $350 million in guarantees for money its customers had borrowed elsewhere. The total of these “customer financings” had doubled in a year—suggesting that purchasers were running out of cash to buy Lucent’s products. What if they ran out of cash to pay their debts? • Finally, Lucent treated the cost of developing new software as a “capital asset.” Rather than an asset, wasn’t that a routine business expense that should come out of earnings? Commentary on Chapter 17 439 FIGURE 17-1 Lucent Technologies Inc. For the quarter ended . . . June 30, 2000 June 30, 1999 Income Revenues 8,713 7,403 Income (loss) from continuing operations (14) 622 Income (loss) from discontinued operations (287) 141 Net income (301) 763 Assets Cash 710 1,495 Receivables 10,101 9,486 Goodwill 8,736 3,340* Capitalized software development costs 576 412 Total assets 46,340 37,156 All numbers in millions of dollars. * Other assets, which includes goodwill. Source: Lucent quarterly financial reports (Form 10-Q). CONCLUSION: In August 2001, Lucent shut down the Chromatis division after its products reportedly attracted only two customers.2 In fiscal year 2001, Lucent lost $16.2 billion; in fiscal year 2002, it lost another $11.9 billion. Included in those losses were $3.5 billion in “provisions for bad debts and customer financings,” $4.1 billion in “impairment charges related to goodwill,” and $362 million in charges “related to capitalized software.” Lucent’s stock, at $51.062 on June 30, 2000, finished 2002 at $1.26—a loss of nearly $190 billion in market value in two-and-a-half years. THE ACQUISITION MAGICIAN To describe Tyco International Ltd., we can only paraphrase Winston Churchill and say that never has so much been sold by so many to so few. From 1997 through 2001, this Bermuda-based conglomerate spent a total of more than $37 billion—most of it in shares of Tyco stock—buying companies the way Imelda Marcos bought shoes. In fiscal year 2000 alone, according to its annual report, Tyco acquired “approximately 200 companies”—an average of more than one every other day. The result? Tyco grew phenomenally fast; in five years, revenues went from $7.6 billion to $34 billion, and operating income shot from a $476 million loss to a $6.2 billion gain. No wonder the company had a total stock-market value of $114 billion at the end of 2001. But Tyco’s financial statements were at least as mind-boggling as its growth. Nearly every year, they featured hundreds of millions of dollars in acquisition-related charges. These expenses fell into three main categories: 1) “merger” or “restructuring” or “other nonrecurring” costs, 2) “charges for the impairment of long-lived assets,” and 3) “write-offs of purchased in-process research and development.” For the sake of brevity, let’s refer to the first kind of charge as MORON, the second as CHILLA, and the third as WOOPIPRAD. How did they show up over time? 440 Commentary on Chapter 17 2 The demise of the Chromatis acquisition is discussed in The Financial Times, August 29, 2001, p. 1, and September 1/September 2, 2001, p. XXIII. As you can see, the MORON charges—which are supposed to be nonrecurring—showed up in four out of five years and totaled a whopping $2.5 billion. CHILLA cropped up just as chronically and amounted to more than $700 million. WOOPIPRAD came to another half-billion dollars.3 The intelligent investor would ask: • If Tyco’s strategy of growth-through-acquisition was such a neat idea, how come it had to spend an average of $750 million a year cleaning up after itself? • If, as seems clear, Tyco was not in the business of making things— but rather in the business of buying other companies that make things—then why were its MORON charges “nonrecurring”? Weren’t they just part of Tyco’s normal costs of doing business? • And with accounting charges for past acquisitions junking up every year’s earnings, who could tell what next year’s would be? Commentary on Chapter 17 441 FIGURE 17-2 Tyco International Ltd. Fiscal year MORON CHILLA WOOPIPRAD 1997 918 148 361 1998 0 0 0 1999 1,183 335 0 2000 4175 99 0 2001 234 120 184 Totals 2,510 702 545 All figures are as originally reported, stated in hundreds of millions of dollars. “Mergers & acquisitions” totals do not include pooling-of-interests deals. Source: Tyco International annual reports (Form 10-K). 3 When accounting for acquisitions, loading up on WOOPIPRAD enabled Tyco to reduce the portion of the purchase price that it allocated to goodwill. Since WOOPIPRAD can be expensed up front, while goodwill (under the accounting rules then in force) had to be written off over multi-year periods, this maneuver enabled Tyco to minimize the impact of goodwill charges on its future earnings. In fact, an investor couldn’t even tell what Tyco’s past earnings were. In 1999, after an accounting review by the U.S. Securities and Exchange Commission, Tyco retroactively added $257 million in MORON charges to its 1998 expenses—meaning that those “nonrecurring” costs had actually recurred in that year, too. At the same time, the company rejiggered its originally reported 1999 charges: MORON dropped to $929 million while CHILLA rose to $507 million. Tyco was clearly growing in size, but was it growing more profitable? No outsider could safely tell. CONCLUSION: In fiscal year 2002, Tyco lost $9.4 billion. The stock, which had closed at $58.90 at year-end 2001, finished 2002 at $17.08—a loss of 71% in twelve months.4 A MINNOW SWALLOWS A WHALE On January 10, 2000, America Online, Inc. and Time Warner Inc. announced that they would merge in a deal initially valued at $156 billion. As of December 31, 1999, AOL had $10.3 billion in assets, and its revenues over the previous 12 months had amounted to $5.7 billion. Time Warner, on the other hand, had $51.2 billion in assets and revenues of $27.3 billion. Time Warner was a vastly bigger company by any measure except one: the valuation of its stock. Because America Online bedazzled investors simply by being in the Internet industry, its stock sold for a stupendous 164 times its earnings. Stock in Time Warner, a grab bag of cable television, movies, music, and magazines, sold for around 50 times earnings. In announcing the deal, the two companies called it a “strategic merger of equals.” Time Warner’s chairman, Gerald M. Levin, declared that “the opportunities are limitless for everyone connected to AOL Time Warner”—above all, he added, for its shareholders. Ecstatic that their stock might finally get the cachet of an Internet 442 Commentary on Chapter 17 4 In 2002, Tyco’s former chief executive, L. Dennis Kozlowski, was charged by state and Federal legal authorities with income tax fraud and improperly diverting Tyco’s corporate assets for his own use, including the appropriation of $15,000 for an umbrella stand and $6,000 for a shower curtain. Kozlowski denied all charges. darling, Time Warner shareholders overwhelmingly approved the deal. But they overlooked a few things: • This “merger of equals” was designed to give America Online’s shareholders 55% of the combined company—even though Time Warner was five times bigger. • For the second time in three years, the U.S. Securities and Exchange Commission was investigating whether America Online had improperly accounted for marketing costs. • Nearly half of America Online’s total assets—$4.9 billion worth— was made up of “available-for-sale equity securities.” If the prices of publicly-traded technology stocks fell, that could wipe out much of the company’s asset base. CONCLUSION: On January 11, 2001, the two firms finalized their merger. AOL Time Warner Inc. lost $4.9 billion in 2001 and—in the most gargantuan loss ever recorded by a corporation—another $98.7 billion in 2002. Most of the losses came from writing down the value of America Online. By year-end 2002, the shareholders for whom Levin predicted “unlimited” opportunities had nothing to show but a roughly 80% loss in the value of their shares since the deal was first announced.5 CAN YOU FLUNK INVESTING KINDERGARTEN? On May 20, 1999, eToys Inc. sold 8% of its stock to the public. Four of Wall Street’s most prestigious investment banks—Goldman, Sachs & Co.; BancBoston Robertson Stephens; Donaldson, Lufkin & Jenrette; and Merrill Lynch & Co.—underwrote 8,320,000 shares at $20 apiece, raising $166.4 million. The stock roared up, closing at $76.5625, a 282.8% gain in its first day of trading. At that price, eToys (with its 102 million shares) had a market value of $7.8 billion.6 Commentary on Chapter 17 443 5 Disclosure: Jason Zweig is an employee of Time Inc., formerly a division of Time Warner and now a unit of AOL Time Warner Inc. 6 eToys’ prospectus had a gatefold cover featuring an original cartoon of Arthur the aardvark, showing in comic style how much easier it would be to What kind of business did buyers get for that price? eToys’ sales had risen 4,261% in the previous year, and it had added 75,000 customers in the last quarter alone. But, in its 20 months in business, eToys had produced total sales of $30.6 million, on which it had run a net loss of $30.8 million—meaning that eToys was spending $2 to sell every dollar’s worth of toys. The IPO prospectus also disclosed that eToys would use some proceeds of the offering to acquire another online operation, BabyCenter, Inc., which had lost $4.5 million on $4.8 million in sales over the previous year. (To land this prize, eToys would pay a mere $205 million.) And eToys would “reserve” 40.6 million shares of common stock for future issuance to its management. So, if eToys ever made money, its net income would have to be divided not among 102 million shares, but among 143 million—diluting any future earnings per share by nearly one-third. A comparison of eToys with Toys “R” Us, Inc.—its biggest rival—is shocking. In the preceding three months, Toys “R” Us had earned $27 million in net income and had sold over 70 times more goods than eToys had sold in an entire year. And yet as Figure 17-3 shows, the stock market valued eToys at nearly $2 billion more than Toys “R” Us. CONCLUSION: On March 7, 2001, eToys filed for bankruptcy protection after racking up net losses of more than $398 million in its brief life as a public company. The stock, which peaked at $86 per share in October 1999, last traded for a penny.