Lots of Odd-Lotters (Well, Sorta)
Tuesday, July 27th, 2004 9:00pm EST
Small Traders Aggressively Shorting
Bottom Line: Short sales for fewer than 100 shares at a time, a hallmark of amateur traders, have surged over the past week, recording its highest one-week average in history. That has definite bullish connotations.
A piece of data that I had followed for many years, but let slide due to continual inconsistencies, is the odd lot purchases and sales information collected by the NYSE. An “odd lot” is defined as any trade for fewer than 100 shares of stock. Since an odd lot order has a tendency to get a bad fill (not to mention the low dollar amount it takes to buy fewer than 100 shares of most stocks), those kinds of orders were universally considered to be from small, amateur traders. And as we’ve seen with many of our indicators which try to track this population, we know that when many of them are leaning the same way, the market tends to go the exact opposite direction.
While this data was quite effective in the past, changed market dynamics seemed to have forced it to go the way of the dinosaurs. Most likely, these traders migrated towards the options markets, and more recently the e-mini futures contracts. The chart below shows how odd lot volume, as a percentage of total NYSE volume, dropped off precipitously beginning in the late 1970’s.
The CBOE began trading options in 1973, with put options beginning trading in 1977, so it’s no real surprise that odd lot volume dropped off so dramatically after those markets came online. Odd lot volume regularly accounted for 5% to 10% or more of total NYSE volume prior to 1977. After 1977, it rarely reached more than 2%, and now it usually makes up only about 1%.
Even though these traders have many other alternative trading vehicles now, every once in a while we will see a spike in odd lot volume. For example, it consistently stayed about double its previous norm from June 2002 through April 2003 as the market struggled to find a bottom. It has dropped off significantly since then, but so far this year odd lot volume has been a particularly good contrary indicator.
To give a bigger-picture view of how these traders have performed, the chart below shows the difference between odd lot purchases and odd lot sales. When the line is above zero, then odd lotters have been net purchasers of stock and when it is below zero, they have been net sellers. The data as shown is a 60-day moving average.
We can see that these traders were heavy buyers in late 2000 as the market made it first real decline of the bear market. Then, in October 2002, they were net sellers to a record degree, which was nearly matched in the spring of 2003. By January of this year, the tide had turned once again and they were heavy net purchasers of stock. By May, we can see that they were selling on balance, though not to the degree seen at the major market lows in 2002 and 2003.
Now let’s zoom in and look at the data in a smaller time frame. The chart below is a 10-day average of the same data over the past year.
From the chart, it is readily apparent that these wrong-way traders were heavy sellers at the March and May lows, but heavy buyers near each of the peaks, particularly so in late June. Currently, we are seeing some of the most pessimistic behavior out of this group so far this year.
Also worth noting is that this group has been doing an enormous amount of shorting over the past few days, betting on a continuation of the downtrend. Total odd lot short sales over the past five trading days has averaged over 1.5 million shares per day, the highest in history. Why this may be notable is that anytime this group has averaged more than 1 million short sales over a 5-day period, the S&P 500 was higher after 20 days every time (19 out of 19), with an average return of 5.2%. The chart below shows these short sales, in millions, on a 5-day moving average basis.
The fact that this data is setting an all-time record is really quite remarkable. Over the past week, they’ve shorted more stock than they did in July and October 2002, February and March 2003, and March and May of this year. Assuming that this data is a reasonable proxy for small traders everywhere, it is a compelling piece of evidence that we may finally be seeing at least a short-term display of excessive pessimism.
I have not found this data to be particularly useful until only recently after a long period of being of little use. However, if there is enough interest from subscribers, I can have it added to the regular stable of indicators that are posted daily to the site. Please let me know directly if you believe you would find value in having it accessible.
Conclusion
In an intraday comment Monday morning, I noted that we were seeing signs that the market was becoming washed out, increasing the chances for a “turnaround Tuesday”. The TICK on the NYSE had hit -1000, and the last time we saw that was on June 14th (which coincided with a short-term low as we rallied for the next week). Also, as of last Friday we had 6 weeks in a row where the S&P closed lower. In its history, it has only gone 7 consecutive weeks three times, and 8 weeks once. It normally rebounded after 6 weeks, though that kind of downward momentum more often than not continued after 4 weeks or so (suggesting we most often saw a short-term rebound before the downtrend resumed to some extent). This time could obviously be different, but just looking at how the S&P has performed in the past, another down week would be a stretch.
Now that we got the rebound, many of our shortest-term measures are back to as much of an overbought condition as they were on the morning of July 21st. I certainly don’t expect the kind of dramatic selloff we got then, but it should be more difficult to make upside progress from this point. Ideally, we will see another decline over the coming days that tests the low put in on Monday. Should we see such a thing, I believe it may actually set up a good buying opportunity.
Another positive sign is that the Nasdaq 100 tracking stock, QQQ, has now traded more than 100,000,000 shares for 8 straight trading days. This is one of the longest such streaks in its history, besides mid-May of this year, and mid-July 2002. As we can see from the SPY and QQQ Liquidity Premiums, high ETF volume is associated with trader uncertainty. High volume equals high uncertainty, so this is a welcome change from the extremely low volume we saw a few weeks ago.
In the intermediate-term, it remains a difficult proposition to suggest that sentiment is overly pessimistic. The sentiment surveys are somewhat mixed, but overall still showing too many bulls; the put/call ratios are mixed; volatility remains compressed (even accounting for seasonality); and Rydex traders refuse to jump heavily onto the short side, unlike near past market lows. As I noted above, it feels as though the broader market is washed out to some extent now, and another small decline could set up a longer-lasting rebound, but from a sentiment perspective it would be unusual if we formed an intermediate-term low with the readings in place now.
Jason Goepfert
President and CEO
Sundial Capital Research, Inc.
Disclosure: no positions
This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.
© 2004 Sundial Capital Research, Inc. All Rights Reserved.