May 25, 2003

http://www.sentimentrader.com

Sunday, May 25, 2003

I will be traveling this weekend for the holiday, so this abbreviated commentary will not follow the usual extended format.

Sentiment activity didn't show a whole lot of change this week from those recently past, although there are a couple of items which should be noted. I have been saying that until a new bull market proves itself, I believe rallies should be sold (and I still believe that), but I am seeing more evidence on a longer time horizon that makes me believe that new lows are not imminent, and in fact we may be in the beginning stages of a longer-term upswing.

The weekly short-selling data out of the NYSE showed a shocking change, at least to me. During a reporting week which saw the S&P jump nearly 4%, the specialists on the NYSE actually decreased their shorting activity as a percentage of the total, driving the Specialist Short Ratio down to 29% from 38% last week. Other than after the panic of 9/11, we haven't seen this low a percentage of specialist shorting in over four years. In recent years, these spikes down in the Specialist Short Ratio have been associated almost exclusively with waterfall-type declines, which rebounded soon after. To see this type of reading now is extremely unusual, and almost seems like a reporting error. However, I don't think this development is as bullish as it may appear, and I'll outline why in just a bit.

I presented a chart a couple of weeks ago showing that the level of specialist shorting during the past year has been historically very low. At the same time, NYSE non-member (i.e. public) shorting as a percentage of the total is close to all-time highs. From 1950 through the late 1970's, this type of combination was an infallible buy signal. The six signals given during that time span coincided with long-term lows without fail.

Something changed in the 1970's, however (perhaps the emergence of the Nasdaq as a viable exchange, or increasing activity of private investment funds), and since that point the data has been less useful as an oscillator. Since that time, there have really been only three times where the moving averages of the shorting activity between the two trader categories (specialists and the public) have spiked apart in a major way, meaning the public went on sustained bouts of heavy shorting, while the specialists were buying that inventory. The first was November 1990 (ending the correction over the previous year), the second was June/July 1996 (ending the four-month consolidation the market had gone through), and the third was about a month ago. Each of these occurrences, even during the late 1960's and early 1970's, lead to multi-year bull moves. Obviously the jury is still out on the current signal, but when we see eight excellent previous signals spread out over the course of five decades, with no failures, we should take note of the possibilities.

The chart below shows these interactions between specialists and the public. It is a bit scrunched in order to fit into a reasonably-sized chart, but you should be able to make out the extremes.

I mentioned above that this week's information may not be as bullish as it first appears, and here is why. Although my primary way of looking at this data is through the percentage of shorts initiated by the specialists (i.e. the Specialist Short Ratio), the extremely unusual reading this week prompted a different look.

There are three groups of traders tracked by this data. There are the specialists, the public and other NYSE members. For the past 5 years or so, specialists and the public made up the great bulk of shorting activity, to the exclusion of other NYSE members. Usually, when we see an increase in specialist short activity, we see a decrease in public shorting. Since 1943, this relationship has a moderately strong negative correlation of -0.55 and the chances of this being due to chance is essentially zero. Therefore, we can say with a relatively high degree of confidence that when specialists short stock, the public is buying (or at least not shorting), and when the public is shorting, the specialists are buying. However, that relationship broke down during the most recent reporting period, as the ratios of both specialist shorting AND public shorting decreased. So while the Specialist Short Ratio is giving off one of the most bullish readings possible, the Public Short Ratio is giving off a reading similar to what was seen at the market peaks in 2002 and January 2003.

The difference in these ratios can be explained by other NYSE members, who shorted the greatest amount of stock (as a percentage) since 1995. These members are those who hold a seat on the NYSE, but are not specialists. This includes brokerage firms who execute orders for clients (usually institutions) and also trade for their own accounts. Going back through history, I have not found a consistent relationship between a high level of non-specialist member shorting and future market performance. That leaves us with the conflicting information between a low level of specialist shorting AND a low level of public shorting. Over time, I have found the Specialist Short Ratio to be the more consistent indicator within this data set, so I would have to tip my hat to the argument that this information is bullish, though it is tempered to a degree by the decrease in public shorting activity.

In the large S&P 500 futures contract, there has STILL not been a real change in character among the commercials or small speculators. Our Futures Balance Matrix remains extremely positive, as it has since mid-March. To reiterate yet again, this is unprecedented in the 17 years the data has been available, and is another indication that the argument that "this time is different" may actually hold some water. In the e-mini contract, commercials added a relatively large number of long contracts while being fairly quiet on the short side, thus decreasing their net short position by a fair amount. The small spec positions were basically unchanged for the third week in a row. As before, overall these positions appear bearish on the surface, but we just don't have enough of a history to reach any conclusions.

Adding a little more fuel for the bulls is the fact that the Rydex mutual fund timers are still skeptical of any upside we get. I alluded to this on Thursday, and it has been a relatively consistent pattern over the past couple of weeks. Notably, our longer-term Rydex indicators are closer to oversold than overbought. The only other time we've seen this type of phenomenon was January 2002, when the indicators reached a momentum peak in November 2001, and had worked much of that excess off by the time the market finally peaked in January. Interestingly, that was the last time assets in the money market were as low as they are now. These money market assets are about the only negative I see in the Rydex complex. As I said last week, such a low amount of assets in the money market suggest traders are relatively comfortable with their market positions (regardless of whether they are long or short). This type of comfort level suggest complacency has set in, and as we all know complacency can be a real problem for a market that has risen as steadily as this one. On a short-term basis, our RSI Spread indicator, which entered positive (for the market) territory on Tuesday, is still positive after several days of rallying. Taken in a vacuum, this suggests more upside may be in the offing.

In a rare move, the individual investors polled by AAII decreased their bullish ways in a significant manner, after the extreme in bullishness they reached the week before last. I have not been able to confirm this, but I believe AAII was collecting poll results through this past Tuesday, so Monday's large decline may have been an influence. While such a wild swing in attitudes is relatively rare, it is certainly not unprecedented for this notoriously noisy survey. I show eight other times since 1987 where the bull ratio on this survey went from 70% or greater one week and dropped by at least 25% the next, as it has this week. Interestingly, out of those 8 occurrences, the market was down the next week every time but once, with an average loss of 1.1%. The greatest return the market was able to generate was 0.9%, while it lost more than 1% on five separate occasions. From two to twelve weeks after that, the market turned in a mixed performance. I don't hang a whole lot of weight on these results, but they suggest that we may be more likely to see weakness this coming week than strength.

In the short-term (the next few days), I don't see much of an edge from a sentiment perspective and am basically neutral. Longer-term (the next few weeks), I continue to believe that this market will not be able to sustain much further upside without a more substantial correction first, and believe the risk of being short is preferable to the risk of being long. Very long-term (out several months to a year or more), I am seeing some signs that the "new bull market" people may have a leg to stand on, however until the market proves itself by completing the four steps I outlined in the April 29th daily commentary, I believe that selling rallies that are accompanied by a complacent marketplace is the higher-odds bet.

- Jason Goepfert

Disclosure: long OEX puts

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.