The rationale is simple. As outlined in a paper by Pan and Poteshman, insiders with advance information of important news will try to capitalize on that info short term. They could buy the underlying stock, but they can leverage their information more efficiently by buying puts or calls. Theoretically, if there were some way of calculating which stock options had the highest ratio of put to call volume, one could track the insider information leading the market.
Pan and Poteshman discovered that indeed, there is profitable information present in the put/call ratio: for the week following the changes in put/call volume ratio, stocks behaved in a non-random fashion.
This wouldn't matter if the data on puts and calls weren't publicly available, but it is. The CBOE started marketing a database product that reports the new open volume of puts and calls for each stock option traded by CBOE. Prior to this, the "put/call ratio" was calculated using summarized data. Now, any investor can, for $600 per month, download data on every stock option every day.
An enterprising investor could write a computer program to analyze all that data and determine which stocks had the highest and lowest put/call ratios. A hypothetical portfolio constructed from these ratios outperformed the S&P 500 for twenty years. This portfolio bought or held long stocks with low put/call ratios, and sold stocks with high put/call ratios. This involves quite a bit of trading per week. The return of 62% has been called, "seductive."
At that rate of return, if you had invested $10,000 in the strategy twenty years ago, you'd have something like $150 million today!
Sound too good to be true? I thought so too. I've spent the last two months thinking about the strategy and, more importantly, writing computer programs. I emailed a few times with the creator of the strategy, Dr. Poteshman at the University of Illinois.
The caveats:
1. The strategy uses a database that costs $600 per month. A high price if you're just casually checking things out, but not that much if you plan on making 62% per year.
2. The strategy costs a lot in transaction costs. I figure between 175 and 350 trades PER WEEK will be needed. If you get an account on Interactive Brokers, at $1 per trade, that's still $16,000 per year just in commissions.
The upside:
Volatility is less than with a mutual fund, since half the portfolio is held as short stocks. If the market tanks, the short side goes up. Hence the low likelihood of a down year.
I have written thousands of lines of code in the Perl language to parse the put-call options databases. Last night was my breakthrough:
| Date | Portfolio | VFINX | Diff |
| 9/11/2006 | 0.98192 | 1.7 | -42% |
| 9/18/2006 | 0.6227 | -0.0164 | 3897% |
| 9/25/2006 | 0.53689 | 0.393 | 37% |
| 10/2/2006 | 2.1552 | 1.51 | 43% |
| 10/9/2006 | 1.9833 | 1.37 | 45% |
| 10/16/2006 | 1.142 | 0.594 | 92% |
| 10/23/2006 | 1.1298 | 0.0867 | 1203% |
| 10/30/2006 | 1.0337 | 0.181 | 471% |
| 11/6/2006 | 0.4643 | 0.401 | 16% |
| 11/13/2006 | 1.7942 | 1.21 | 48% |
| Total | 11.84 | 7.43 | 59% |
| Annualized | 61.59 | 38.63 |
The portfolio beat the S&P 500 by 59%, even though the S&P was on its way to having a huge year during the same period. Not too many funds can claim that they did that.
I figure you need about $100,000 or more to make it worth your while to do this strategy, otherwise your gains get eaten up in expenses. I'm good for part of that, but if anyone is interested in becoming partners, I'm open to suggestions.
