In last Sunday’s New York Times, there was an article regarding a paper done at MIT comparing a hypothetical portfolio of stocks - buying the 20% of stocks with the lowest put-call ratios while shorting the top 20% of those with the highest put call ratio - the positions were adjusted weekly based on the changes in data. They reported a whopping 62% average annual return over the 12 years from 1990 thru 2001. Perhaps Vic, you can send them a copy of each of your books (to the N.Y.T. and to the authors) to help explain why this article is at best ballyhoo and at worst manipulation of those who know no better- the most glaring key omission - they excluded transaction costs (which I imagine would be high with a weekly rebalancing based on P/C data). In addition, their data set seems somewhat small - listed options have been traded for much longer periods so why did the authors choose this period only? Why did they exclude transaction costs? And why would an obscure academic paper be mentioned in the N.Y.T.? Ah yes, the Chicago Board of Exchange just happens to be starting a subscription service at the rate of $600/month for deep datasets. I have not read the original paper — I plan to find and go through it — however I have read multiple sources commenting on it, and it apparently is gaining traction. I do believe a comprehensive dataset from the options exchange would be a valuable research tool, however this seems to fall in the “too good to be able to be realized” section of market lore.

To read the article …


Resources & Links