margin vs. LEAPs: better math, more guesses:
Formulas for calculating returns:
LEAP R= (S-SP+3.5)/L
Margin R=(1.5S/B)-0.55
R= return, expressed as a multiple of the original investment. So, a 10K original investment and a R of 2.7 means you have 27K at the end, or 17K of profit, before taxes.
S= stock price when stock or LEAP sold
B= stock price when stock bought
L= price of LEAP when bought (ask call price)
SP= strike price
3.5=4-0.5= my guess to adjust for time premium (4) for deep-in-the-money LEAPS sold 6 months before expiration, and spread (0.5).
0.55= 0.50+0.05. 0.50 accounts for the borrowed money, equal to half the original capital; 0.05=interest cost on margin= 6.75% X 1.5 years X 0.5 (half my initial capital)
I am ignoring as a trivial cost the LEAP commissions, and the margin commissions and spread. The LEAP spread is not trivial. It can become a large fraction of the total cost, especially far-out-of-the-money or rolling them over or complicated things like straddles or buying at a variety of strike prices.
Assumptions:
1. I can time the bottom.
2. I can time the top.
3. from bottom to top, or at least a stock double, will be no more than 24 months.
4. I have money I can afford to lose.
5. Let's say the recent low was the bottom, and we are on the way up now. Let's say I bought tomorrow, with AMAT at 30, and held for 18 months. I'd buy Jan. 2000 LEAPS, or buy the stock and use half my money to buy more on margin, (so 1/3 of my stocks are on borrowed money). If I sold in June 1999, the LEAPS would still have 6 months till expiration.
Results:
1. If the stock doubles, to 60, then R is 2.45 for margin, 2.7 to 2.5 for LEAPs that are deep-in-the-money till 40 (10 out of the money). From there on out, LEAPs fall off sharply in return. By the time you get to a strike price of 60, R is 0.6, and that comes entirely from the time premium. 50% probability.
2. If the stock quadruples, to 120, then R is 5.45 for margin, 6.5 for 20 LEAPs (deep in the money), and increase in a linear fashion to 11.3 for 60 LEAPs (far out of the money). 25% probability.
3. If the stock goes to 45 (a 50% return over the next 18 months, which is the lowest I think it can possibly be. Again, this assumes I can time the bottom, and buy at or just after the nadir.): then margin gives the best return, at R= 1.7, and LEAPs lose badly. You have to have a strike price of 20 to equal the return from margin. 25% probability.
Conclusions:
1. Don't do LEAPs (of any kind) unless you are 90% certain the stock price will at least double in the next 24 months. With AMAT, the only way you can do this is buy at X1 to X1.5 sales, when chip prices have started to recover, but before the first hint of increasing equipment orders. I won't buy LEAPs unless AMATs price is below 25, and memory chip prices have stabilised.
2. Unless you enjoy the thrill of gambling, don't buy LEAPs more than 10 points out of the money. These strike prices work only if the underlying stock price at least triples. The odds of being able to time a triple in 18 months are not good.
3. The best risk-reward balance seems to be at a strike price 0-10 points out-of-the-money. Going further in-the-money doesn't give you better returns, unless the underlying does very badly.
4. For now, I buy nothing and sell nothing. I will probably wait until the Jan. 2001 LEAPs become available in July 1998, and buy 35 calls if the stock price is still at 25. Buying later is safer, because the pressure of .25 and .18 and 300mm and merced will eventually produce a sharp increase in orders. The semi companies who went into this downturn with a pile of cash will make those orders, and increase market share. The semi companies who went into this downturn with a pile of debt will be liquidated. The latest these orders can be put off is about the year 2000, IMnot-so-humbleO. The closer we get to 2000, with the stock still out-of-favor, the safer LEAPs will be.
Wouldn't it be great if the stock stays at 25-35 all next year, so we can buy lots of it cheaply? And then sell in 2000 at 120 or so? Things are turning out perfectly! |
