I guess it’s time for me to weigh in on the subject of short interest as a tool for stock picking.
The idea that derivative trading has spoiled the analytical value of short interest data has been out there ever since derivatives began to trade on a standardized basis. There is no doubt that “riskless” short selling occurs, perhaps even on a significant basis. My firm found John Succos’s statement, that the derivatives market is “ten times the size of the cash markets,” intriguing. I decided to take a closer look at the numbers to investigate this.
As of Thursday’s trading, the total market cap for stocks that have any level of equity option open interest is $21,549,556,912,940. The market cap for all equity options open interest contracts is $523,676,613 – the cash would appear to be 41,151 times the size of the derivatives in terms of market cap. Looking strictly at shares that are represented by the equity options open interest, there are 622,821,390,500 shares in stocks that have any level of equity option open interest. Thursday’s total equity option open interest represents 19,024,368,600 shares of stock – the cash would appear to be 33 times the size of the derivatives in terms of shares represented.
Let’s examine the potential that the current short interest is dominated by derivative activity. The total shares short as of the current month’s report from the NYSE and NASDAQ total 16,945,801,849 shares. Delta neutral strategies that tend that deploy short selling are mostly concerned with hedging call positions, which totaled 106,267,406 contracts and thus represent 10,626,740,600 shares. My firm's professional insiders tell us that approximately 10% of options trades are of the delta neutral variety, which brings the derivative influencing activity to a level representing 1,062,674,060 shares, or about 6% of the entire equity short interest position. So let’s not throw the baby out with the bath water just yet.
That said, we totally agree that there is little value in the observation that raw short interest statistics are at record levels.
Record levels of short selling happen quite frequently, in both bull and bear markets. The chart above shows the monthly short interest data since 1998. For all this decade the NYSE has reported record short interest 40% of the time. Does that seem extreme?
In fact it is moderate:
The above chart shows short selling data since 1931. For most of the latter half of the last century, short interest rose to a record almost every month. How can this be? Because short interest is a subset of volume:
Saying that short interest is at record levels is about as valuable as saying volume is at record levels. This is a mistake that we see time and again, especially in the media. To get meaningful data from short interest statistics, you must relate short interest to volume to determine short selling intensity. Some examples:
The above charts show the average short selling intensity of the component issues of the diversified financial and semiconductor sectors. Periods of heavy short selling do not guarantee an immediate rally in the stock price, but they do comprise setups for a rally. In fact, our model is a 2 factor model: sentiment and price action – in other words setup and trigger. Short interest is a setup for a contrary move that is confirmed by some trigger indicator that measures price action. In these charts the red arrows are periods of relatively light short interest, and thus are setups for a decline in price. Green arrows are periods of heavy short selling, and are setups for price upturns.
The above charts show similar data for 2 industry groups – retail and airlines. And of course, this data can be seen in individual stocks as well:
The point is that looking at short interest is of added value in the investment decision making process. You just have to put a little work into the analytical process.
Looking for “short squeezes” does have its seasonal aspects. In a rip roaring secular bull market, most short squeeze setups and triggers work great because a market that doubles and triples tends to take most “boats” with the advance. But a secular bear market, like the one since 2000, tends to be less fertile ground for short squeeze opportunities. That is why my firm looks for both short squeeze and long squeeze opportunities. In fact , Since March of 2002, we started what was then purely an exercise to see if we could demonstrate the value of hedging short squeezes (what we call Type 1 classification) and long squeezes (Type 4 classification.) Every week my firm determines these from a universe of the 230 largest cap stocks. Incidentally, these are issues most likely to be affected by any delta neutral options strategies by virtue of their liquidity. These are screens done each week with no intra-week fiddling. The results:
The short squeeze screens averaged a gain of 119.09% while the long squeeze screens fell -76.03%. Over this period, this screening process in real time outperformed the S&P 500 by 174.60%.