The Historical Record
In Chapter 5 we saw the basic system earn over $100,000 in actual investments, averaging about 25% a year. Did these warrants from 1961 to 1966 offer unusual opportunities that did not exist in the past and may never exist again? We now show that if investors had used the basic system from 1946 to 1966 they could have made equally spectacular profits.
A Simplified Mechanical Strategy
We express the essence of the basic system in a simple set of rules, which we then apply to the 1946-1966 period.
Rule 1. Restrict attention to warrants listed on the American Stock Exchange which
expire in less than four years. This rule exploits the fact that short-term warrants decline
faster than long-term warrants. (See Appendix E.) We consider only listed warrants for this historical playback because it is easier to sell short listed securities than over-the-counter securities and because past price data for unlisted warrants are often unreliable or unavail- able.
selling for less than 6% of the adjusted exercise price. Also eliminate warrants when the stock is selling for more than 1.2 times the adjusted exercise price. This is a simplified inter-
pretation of Figure 6.1. If the common stock is above 1.2 times the exercise price, it is very seldom profitable to hedge. If the warrant is less than 6% of the exercise price, there is too “little” to be squeezed out by a short sale.
Rule 3. Eliminate unadjusted warrants trading for less than $1. Because $2.50 margin is required for securities selling for $2.50 or less, margin on a $1 security is actually 250%. This sharply reduces the expected percentage return from short sales of these warrants.
Rule 4. From the remaining warrants select the one with the closest expiration date and
sell short 3 warrants for every common share purchased. Use the entire purchasing power in the account under the prevailing margin rates. We assume the short sale occurred on the first
available up-tick and then the common was purchased. In actual practice an investor would not choose a mix arbitrarily. Depending upon his expectation for the common and on the position of the warrant-common point in Figure 6.1, he might sell more or fewer warrants short for every share of common purchased. But since our hypothetical investor wishes to sail away in January 1946 and not be bothered with intervening decisions, he arbitrarily chooses a mix of three to one.
Rule 5. Cover short sales and sell the common on the last day warrants trade on the
exchange. On the next trading day, start again with Rule 1. If no opportunities are available, invest in short-term treasury bills.
Table 7.1 follows the step-by-step investments resulting from this simplified strategy. We deduct commissions but omit dividends from the common stocks. For four years, between 1956
T
and 1960, no listed warrants met the criteria of the five rules. The system was actually employed then for seventeen years, during which it averaged 30% a year, before taxes.* If we assume a flat rate of 25% on profits when they are taken, the profit after taxes still aver- ages 22% per year.† Before taxes, the original investment multiplied 50 times and after 25% tax it multiplied 22 times. Our cash experiences in Chapter 5 are supported by the historical record.
We remark that a 25% tax on basic-system profits in Table 7.1 corresponds to a tax bracket higher than 25% when part of the profit was long-term capital gains. Since tax laws have changed, we illustrate with 1966 tax law. Profits on common held more than 6 months were long-term capital gains and were taxed at half of the ordinary income tax, but in no case at more than 25%. Profits from common held less than 6 months and all profits from short sales were taxed as ordinary income.
We have ignored the avalanche effect. If it were used, the basic-system profit figures cited in this chapter would be increased, often substantially.
Figure 7.1 graphs the performance of the basic system from 1946 through 1966. This is not as revealing as Figure 7.2, which contains this information on a semi-log grid. There, equal vertical distances represent equal percentage changes and a straight line represents a constant percentage increase, compounded annually. The greater the slope of the line, the greater the compound rate of increase. Since we are interested in compound rate of return, a
*This is the arithmetic average. For investors interested mainly in long-term growth, the equivalent annual compounding rate, which is 26% before taxes, is a more important figure. Elsewhere in the book we have referred to these figures of 26% and 30% by citing “more than 25% for seventeen years.”
†It is customary in stock market literature to figure rates of return before taxes, since the effect of taxes will vary with the type of investment, the investor’s situation, and with the tax laws.
semi-log grid makes it easier to compare various investments. It also allows us to see how constant the rate of return is on any investment.
In Figure 7.2 we consider an investor who purchased the same securities as called for by the basic system but did not sell short any warrants. An initial $1,000 would have increased to about
Figure 7.2. Performance of basic system vs. straight short and long positions (after commissions and before taxes).
$2,200 in the seventeen investment years. The basic system did more than 22 times as well. Selling short the warrants but not purchasing common gave a more erratic performance, earning more in some years and much
less in others. An initial $1,000 would have increased to about $22,000. The basic system did 2.8 times as well.
In Figure 7.2, which graphs the performance of all three strategies, the superiority of hedging is clear. The investor who only purchased stock had a relatively poor performance. The investor who only sold short warrants had dramatic successes mixed with the spectacu- lar losses.
This last strategy would in practice have had even greater losses than indicated. For instance, after Mack Trucks warrants were sold short at 17⁄, they rose to 35fl. This seems to indicate that the investor was wiped out! But actually, depending on how quickly his broker reacted, he would have received margin calls as the warrants rose above 22fi. We assume that he then covered part of his short position rather than ante up more money. If he received a margin call for every 1 point rise in the warrant, his loss would amount to about $20,000 because the warrant then fell from 35fl to 23/8. If his broker were not alert and he was asked
to cover only when the warrant rose more than 1 point, his loss would have been greater. In actual practice, with the warrant moving as much as 3 or 4 points in one day, he would prob- ably have suffered a more severe loss.
If the different tax treatment accorded to long-term and short-term capital gains were taken into account, the after-tax profits from the basic system, and from a long position only, would compare still more favorably with the profits from a short position only. Though a long position benefits the most by this, the gains do not change its position as inferior rela- tive to the others.
The Potential Future for the Basic System
Corporations enjoy tax advantages by issuing warrants. As more managers become aware of these advantages, we expect warrants
to be issued more often, ensuring many opportunities for employment of the basic system. The tax advantages are detailed in the Lybrand, Ross Bros. & Montgomery newsletters of June 1965 and September 1966. As an example, assume a corporation wishes to issue a bond with a “sweetener” (see the section on Convertible Bonds, Chapter 10). It may attach warrants to the bond or it may give the bondholder the right to exchange his bond for a fixed number of common shares. Such a bond is called convertible. In either case, the corporation is selling a straight bond plus an option on its common. But when it attaches warrants to the bond, it enjoys special tax considerations. Suppose that the face value of the bond is $1,000, redeemable in twenty years, that there are warrants attached, and that the warrants have a value of about $300. For tax purposes, the corporation has issued a “package” containing a bond and warrants. It has received $700 for the bond and $300 for the warrants. But when it redeems the bond in twenty years, it must pay the holder $1,000, for a loss of $300 on the bond. This loss may be amortized over twenty years, allowing substantial savings to the cor- poration. This amortization is not allowed if the corporation instead issued a convertible bond.
Successful use of the basic system requires more than a large crop of warrants—it also requires that the premium paid for warrants remain near the levels attained in the period 1946-1966. If, for instance, warrants become very “cheap,” the expected return from selling them short might decrease substantially. In this event a variation of the basic system (reverse hedging) explained in Chapter 8, might consistently yield better than average returns.
Performance Through the 1929 Crash
Though few economists believe we will again experience a disaster like the 1929 crash, some readers may wonder how the basic system would have performed then. Let’s glance at the early warrant market.
In 1911, American Power & Light issued notes with warrants attached. This was prob- ably the first American warrant ([8] p. 656). The price history of these, and of most over-the- counter warrants, is almost impossible to reconstruct.
The first listed warrants were probably those of Phillips Petroleum Company and White Oil Company. Both traded in 1923 on the New York Curb Exchange (now the American Stock Exchange). By June 28, 1929, at least 22 warrants were listed on either the New York Stock Exchange or the Curb Exchange. (Warrants do not meet the present listing require- ments of the New York Stock Exchange and none have been traded there since World War II.) Table 7.2 lists the warrants which traded on that date as reported in the Commercial and
Financial Chronicle. Dozens of common and preferred stocks also traded with warrants
attached. Figure 7.3 shows that these early warrant premiums compare with premiums after 1945.
Suppose an investor had discovered the basic system on June 28, 1929. How would he have survived the worst collapse of all time? Applying the simple rules set out at the start of this chapter, he would have purchased American Commonwealth Power common stock at 23fi and sold short 3 times as many warrants at 7fi. Margin regulations did not exist at that time and it would have been possible for him to enter this transaction on 10% margin.
But assume he was very conservative and used 50% margin. These warrants traded for the last time on June 27, 1930, and sold at 7/32 while the common sold at 24⁄. In one year, when the Standard &
Figure 7.3. Warrant-stock relationships June 28, 1929. (See Table 7.2 for sources and notes.)
Poor’s index of industrial stocks fell 35%, the basic system returned almost 100%. This investor not only survived the worst stock market crash in history—he doubled his money.
The reader may reasonable object that this incredible performance was due at least in part to the fact that, despite a 35% decline in Standard & Poor’s index, the price of American Com-
Table 7.2. Listed Warrants on the New York Stock Exchange and Curb Exchange on June 28, 1929.
monwealth Power actually rose slightly. But we note that we would still have had a profit after costs even if the common had fallen from the initial price of 23fi to 2!
There are three other warrants in Table 7.2 which meet the criteria of Rules 1–3. They are Aeronautical Industries, Curtiss-Wright, and General Cable. Even if General Cable fell to zero, it would have produced a profit of about 90 – 42 = 48 on an investment of (90 + 42) x 50%, or 66, a profit of about 73%. If Curtiss-Wright fell to zero, the profit would have been 60 – 30 on an investment of (60 + 30) x 50%, or 30/45, which is 67%. Aeronautical Industries could have produced a loss if the common fell from 18 to less than 6. However, Aeronautical Industries had a smaller premium and a later expiration date than the other three highly suc- cessful candidates and would easily have been rejected in favor of them.
It is pointless to continue in this land of might-have-been. There can be no doubt now that this simple-minded application of the basic system would have yielded extraordinary profits during bad times and good. Note that these profits could have been made using the rigid rules 1 through 5. No consideration was given to refinements of these rules or to the pyramiding of investments.
Chapter 8