Convertibles and warrants

Why it matters

Not the best of both worlds. Convertibles are sold as giving the safety of a bond and the upside of a stock. Graham points out that a device cannot make a bargain better for both sides: the buyer usually gives up quality or yield, and the company gives up part of its shareholders' future gains.

Timing. Graham notes that most convertibles are issued late in bull markets; Zweig adds that even weak companies' shares then look good enough to make conversion attractive. When the market then falls, the conversion right loses its appeal, and sometimes the bond's safety is in doubt too.

• In Graham's 1946 sample, new convertible and participating preferred shares fell about 30% on average to their lows, against about 9% for straight (non-convertible) ones.
• From 1968 to 1970, his sample of 20 convertible preferreds fell an average of 29%, while the S&P 500 fell 11.3%.

The holder's dilemma. Even a winning convertible poses a question: sell on a small rise, or hold for more? Graham's own example is Eversharp's convertible bonds, bought at par. The stock rose until the bonds were worth over 220. The bonds were called and converted, and some of the original buyers kept the stock, which then collapsed. The bonds' value fell to about 27, a loss of three-quarters of the original price. Hence the old Wall Street rule Graham quotes: "Never convert a convertible bond." Once you convert, you lose the bond's prior claim on interest and become a shareholder, usually after a big rise.

Graham's verdict. Look "more than twice" at any new convertible. The ideal is a strongly secured convertible, into a stock that is itself attractive, at a price only a little above the market. Such cases are more often found among older issues than new ones.

Warrants. Graham calls them close to fraud. A warrant takes part of the value of the existing shares and puts it on a separate certificate, so the common stock is worth less while warrants are outstanding. Reported earnings per share often ignore this. His simplest correction: add the warrants' market value to the shares' market value when judging the price.

Zweig's commentary (2003). Convertibles "behave like stocks" more than bonds: he reports that their returns were about 83% correlated with the S&P 500 and only about 30% with Treasury bonds. Citing a convertibles specialist, he notes that many are now medium-term, about half are investment-grade, and many carry call protection, which makes them less risky than in Graham's day. Small lots are costly to trade, so a low-cost convertible fund is the practical route; he warns against any convertible fund with yearly expenses over 1%. In a footnote he adds that warrants have become rare among major companies.