Short-term Decline Should be Just That

Bottom Line: Watching whether traders are buying into new long positions or covering existing shorts can give us some information on whether market moves are more likely to be sustained or fizzle out.

Virtually every day, we all hear financial commentators attribute market action to how people are trading. For example, on yesterday’s relatively large rally, most commentators that I read stated confidently that “it was nothing but short covering”. There are two problems with that: 1) Most of them don’t know that, and 2) they don’t have any solid idea of what it means going forward. Fortunately, there may be a way to help us find out some of the truth behind the theories.

One of the most exciting developments in sentiment analysis in recent years is the release of the daily asset levels in the Rydex funds. I discuss them frequently, and they have been good guides since their inception. Watching the flow of real money into and out of speculative funds gives us a glimpse into traders’ mindsets that was not possible before. By watching where traders transfer their money when the market has an up day, we can get an idea of whether traders in general were buying the market or covering existing short positions.

The table below outlines how these traders have reacted to days when the S&P 500 closes higher, and when the market is in an uptrend or a downtrend (defined as an upward or downward sloping 200-day moving average, respectively). When the S&P closes higher than the day before, and the bullish Rydex funds gain more in assets than the bearish funds lose, then we can consider the traders to be buying into the rally. On the other hand, when the S&P closes higher but the bearish funds lose more in assets than the bullish ones gain, then we can classify these traders as covering more shorts than buying new positions, on balance. For each of the four possibilities, the table also gives the performance in the S&P 500 thirty, sixty and ninety days later.

Market is in…	Traders are…	30 Days Later		60 Days Later		90 Days Later
Market is in…	Traders are…	Avg Reg	% Pos	Avg Reg	% Pos	Avg Reg	% Pos
Downtrend	Covering	4.8%	83%	7.5%	80%	10%	74%
Uptrend	Covering	-0.5%	47%	1.1%	48%	3.3%	60%
Downtrend	Buying	-3.5%	33%	-4.9%	30%	-7.0%	24%
Uptrend	Buying	1.2%	61%	-1.2%	49%	-1.7%	41%

From the table, we can see two sets of results that are quite noticeable. First, when these traders were COVERING their shorts into a DOWNTREND, then the S&P enjoyed very positive performance going forward. On the other hand, when they were BUYING new long positions into a DOWNTREND, then the S&P suffered greatly.

When the market was in an uptrend, something interesting also occurred. When traders were covering their shorts, the market had tepid performance out to 60 days or so, then picked up later on. However, when they were doing actual buying into new long positions, then the market was more positive in the short-term and actually negative in the longer-term.

I think these are important distinctions. Bouts of short-covering are not necessarily bad – we can see from the evidence that when Rydex traders were aggressively covering short positions when the market was in a downtrend, it often signaled that a low of some sort was relatively close at hand. Even when they were covering shorts into an uptrend, it did not bode particularly ill for the market. Some of the worst performance occurred when traders were aggressively buying into new long positions. This is the opposite of what we normally read.

I know this can be a bit confusing, as there are four variables we’re trying to look at, so let’s look at a chart from 2002 – 2003. The highlights below show those times the traders were covering their shorts while the market was in a downtrend (green highlights) and those times they were buying into a downtrend (red highlights).

This is a good example of what the table above showed us - when traders were buying into a downtrend, we often saw further declines going forward, but when they were scrambling to cover their shorts, then the S&P was close to some sort of a low more often than not.

On the chart above, “100%” on the blue line means that over the prior 10 days, these traders had been buying into new longs more than they were covering their shorts on every one of the up days in the market. Conversely, “-100%” means that on every one of the up days, traders were covering their shorts more than they were buying into new longs.

Currently, Rydex traders are covering a hefty amount of shorts into an uptrend, which as we can see from the table above means that we have historically seen a bit of weak performance in the shorter-term but better performance longer-term. This is one of the weakest edges out of all of them, so personally I wouldn’t read too much into it. However, I do think watching information like this gives us insight into what traders are actually doing, instead of relying on poor anecdotal evidence from commentators that haven’t a clue about what they’re saying.

Bottom Line: Looking at the rate of change in the bull ratio of a popular sentiment survey finds that its population has reversed course to such a large degree that it has corresponded well with other major lows over the past 7 years.

On August 17^th, I talked about using the rate of change in OEX option open interest to get a better read on what it was telling us. Using absolute or relative levels in our measures can be useful, but so can watching how quickly or how slowly those indicators change over time. One other application of this way of looking at our data is highlighting an interesting signal in the Investor’s Intelligence sentiment survey.

The latest results showed a stiff decline in the number of bullish respondents, to the point where we can now consider them historically low. The number of bears also rose, and the combination of the two pushed the bull ratio (bulls / (bulls + bears)) to its lowest point since April 2003. More significantly than that, however, may be the swift descent in bullish opinion.

Over the past 8 weeks, the bull ratio has declined by more than 25%. The chart below shows other instances of such swift two-month declines in bullishness (please note that the scale in the bottom half of the chart is inverted).

Six months after such extremes, the S&P was higher every time, with an impressive average six-month gain of 15% (which includes three out of five unique instances occurring during the bear market). Looking at the entire history of the survey back to 1969, this signal was not perfect. There was a failed signal in early 1973, a very bad one in May 1974, and another “early” one in August 1981. But overall, especially recently, such a quick about-face from this population has resulted in positive markets a good deal of the time.

I noted a divergence between the VXN volatility measure and the NDX in the last comment, and on top of that as of yesterday the VXN had declined for 9 consecutive days. In its history, going back to 2001, that had only been matched one other time, which was the end of May 2004. There have been a handful of streaks of longer than 5 consecutive days of VXN declines, but they did not result in any consistent performance going forward. Two of them (May 2001 and March 2002) lead to severe weakness going forward, while two others (September 2001 and April 2003) kicked off stupendous rallies. I suppose one could suggest that the current situation looks like a mini-March 2002, but I wouldn’t read too much into that. In any event, the divergence I noted last time is still in effect and is still what I would consider a short-term negative for the market.

The broader market is doing a fine job of confirming the likelihood of a longer-term low, but it is doing a lousy job of allowing low-risk points of entry. I continue to feel that any short-term decline which puts us into an oversold condition should serve as a good spot to establish or add to longer-term long positions, and I also continue to feel that it is not especially high-odds to be aggressively long for short-term traders. The negatives I mentioned last time are still with us, and while it seems as though “everyone” is expecting prices to take a breather, I don’t like relying on anecdotal evidence to form trading decisions. The data we look at suggests lower prices are more likely than sustainably higher ones in the short-term, so that’s how we will approach the market in the coming days.

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.