To: Chuzzlewit who wrote (82441 ) 11/28/1998 8:40:00 AM From: Geoff Nunn Read Replies (1) | Respond to of 176387
Hi Chuz: re returns on call options You mentioned in your post that you have realized returns in writing call options of ~2.5% per mo. in extra income. It wasn't clear to me, however, whether the 2.5% figure is gross (before transactions costs are subtracted) or net. Can you please clarify? I do assume you are only counting income derived from the sale of the options, not any appreciation in the stock. If the 2.5% is gross, then it would seem to me to be irrelevant to this discussion since the issue here is net returns on options writing. If it is net then it is an extremely impressive figure, every bit worthy of the energetic defense you continue to mount in favor of s-t options writing. <bg> You wrote,...in several of those cases I noticed that the implied volatility was greater for short-term options than for their longer-term counterparts. If the game were fair, wouldn't you expect that implied volatility be constant? Agreed. The fact that s-t options have greater IV suggests they are overpriced (that, or l-t are underpriced) and that the market isn't efficient. I personally am not a great believer in the efficient market hypothesis, and consider such mispricing to be entirely possible. It could very well be that call buyers are willing to pay a premium for s-t options because they provide greater leverage. (nevermind that such leverage presumably could be obtained at lower cost using brokerage margin). I certainly don't rule out that s-t options for whatever reason may well be more profitable to write, both net as well as gross. On the issue of implied volatility as a plug variable, I think I see it a little differently from you. As I understand it, Black /Scholes assumed that the rate of return on any stock is log-normally distributed, and the standard deviation of the distribution is the stock's volatility. Volatility was not a "plug variable" in the derivation of the model. The problem is that all the variables in the formula are observable except one - standard deviation. So, in order to apply the model a means must be developed to estimate s.d. One method to overcome the problem is to rearrange the equation and solve for s.d. This gives the implied volatility. Yes, it is a plug variable as you call it, but a useful one nonetheless. For example, it can be used the way you used it to conclude that s-t options are overpriced. Also, if it can be assumed that variability has not changed from historical norms, it can be used to decide whether options currently are overpriced or underpriced. If it can be assumed that historical option prices are correct (efficient) and that historical implied volatility equals the actual standard deviation, it can even be used to determine the present value of an option. That, at least, is my understanding based on secondary- source reading on B-S. Geoff