Jan 06, 2003

A couple of weeks ago, I presented a new way to analyze asset flows in the Rydex series of mutual funds. There was considerable interest from subscribers in seeing the information daily, so I have added it to the stable of Rydex indicators on the site as of tonight. I think the concept that the indicator captures is important, and we have another extreme reading currently, so I'm including some of the commentary from a couple of weeks ago tonight (slightly altered) as background.

I've always thought that when the majority (i.e. the public) recognized a trend, they would place their money in the highest-beta stocks or indexes in order to capitalize on what they think will be higher prices. For those of you unfamiliar, ''beta'' is a term typically used to show a stock's correlation with an index based on historical performance. For example, if a stock of a beta of 2, then if the S&P went up 1%, the stock should go up 2%. If a stock has a beta of -1, then if the S&P went up 1%, the stock should go down 1%. Most betas tend to be somewhere between 0 and 1, although many tech stocks have betas greater than 1, and "safe"' stocks usually have betas closer to 0. It is possible to have a negative beta (meaning the security will typically move counter to the index), which some of the Rydex bear funds have. As an example of what types of funds have certain types of betas, here is a sampling of a few cases from the Rydex family:

FUND	TYPE OF FUND	BETA
URSA	S&P SHORT FUND WITH NO LEVERAGE	-1.01
PRECIOUS METALS	INVOLVED IN VARIOUS ASPECTS OF PRECIOUS METALS PRODUCTION	0.26
HEALTH CARE	INCLUDES PHARMACEUTICAL COMPANIES	0.27
TITAN	S&P LONG FUND LEVERAGED 2-TO-1	2.03
ELECTRONICS	INCLUDES SEMICONDUCTOR COMPANIES	2.42

To test this theory, I created an index from the Rydex information which compares the momentum of assets flowing into the highest-beta funds to the momentum of assets going into the lowest-beta (or negative beta) funds. Those funds which had a beta close to 1.0 were ignored. Notably, the index is weighted by the beta of the funds, so that those funds with the highest beta get the highest weighting. This should serve to appropriately capture the amount of true speculation flowing into the various funds, relative to each other. Theoretically, the index should spike higher if traders are buying funds with high beta and/or pulling assets out of low beta funds. This would indicate that the majority of the traders (or the majority of assets, anyway) felt that the market was headed higher and thus chose to put their money into the highest-beta funds. It would be a measure of complacency or optimism. On the other hand, if the index was low, then that would show a greater momentum into low-beta, or "safe" funds, and would be a relative measure of fear or uncertainty.

The chart below is a current two-year representation of this index. We can see that the spikes higher (suggesting a sudden surge of speculation) do correspond quite well with short- to intermediate-term highs in the S&P. The highest spike on the chart - in March of 2002 - occurred when the general sentiment among traders was that we had suffered a retrace to the September lows, and were now ready to exceed the January high. When we approached that level, the beta chasers really took charge and shifted their assets into the highest-beta funds. When new highs didn't materialize, you can see that the momentum quickly shifted out of those funds. Likewise, in late October 2002, the talk was all about a successful retest of the July lows. When it appeared that we were going to exceed the Augusts highs (actually, we did on the Nasdaq), then the momentum once again flowed to the highest-beta funds. Not surprisingly, lower prices soon followed.

As of today, the index stands at 3.63, one of the highest readings over the past two years. The reason for the spike is that the momentum of assets flowing into the highest-beta funds (such as the leveraged index funds, Telecommunications and Electronics) is, on average, far exceeding that of the momentum into the low-beta funds (such as the leveraged short index funds, Health Care and Consumer Products). This momentum shift suggests that there is a high - even extreme - degree of optimism that the recent trend of higher prices will continue. To capture that upside, these traders are becoming aggressive in shifting their money, which is confirmed by the low level of the equity put/call ratio and relatively low level of the VIX. So in the context of a defined downtrend, we have notoriously wrong (or at least late) traders not only SAYING they're bullish (via the sentiment surveys as discussed yesterday) but also BETTING with real dollars on that opinion. Not good from an intermediate-term perspective.

Yesterday, I said that as long as short-term sentiment was no longer overbought and we remained above S&P 900 and NDX 1030, I fully expected another attempt higher, perhaps taking out the recent highs. We got the rally today, and from the looks of things it may not be over quite yet. Even though we're close to our recent highs, our shortest-term sentiment measures are still neutral to mildly overbought. The STEM.MR model is sitting right on its bearish trading band, as its components are hovering right around theirs. In a very unusual development, the S&P price oscillator is near overbought while the NDX oscillator is near oversold. These two indicators almost always move in lockstep, but I think this is more a reflection of the small range in the afternoon rather than some important market dynamic between the two.

Several of our intermediate-term indicators, after becoming quite overbought at the close on Monday, began to work off that condition during the correction through yesterday. While today's action served up some readings to again push them into overbought, one day usually does not have enough power to get them into extreme territory, meaning there may be some more fuel left to carry us higher over the coming day(s). I expect more buying to come in with each successive resistance level that is broken, and that self-fulfilling journey could take us up to (or even over) the August highs in the S&P. While such a move would be a stretch sentiment-wise, it would make sense technically. However, any such move up to that level would undoubtedly bring with it extreme readings in our indicators, suggesting that what we've seen will be nothing more than another bear market rally. I continue to believe that we are seeing a topping process, and higher prices (if seen) should be viewed not as an opportunity to buy, but rather sell, hedge or sell short.

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.

I mentioned the Rydex RSI spread in today's intraday note, stating that at +85 it was the 2nd most extreme reading we've seen in its history. The only other readings above +80 were on 3/27/01, 5/22/01, 11/19/01, 3/6/02, 7/30/02 and 10/16/02. The S&P was lower 5 and 10 days after every occurrence mentioned above except for a marginal gain after the 10/16/02 reading. While I don't think it's wise to read too much into this indicator since we have a limited history (two years), and it's already one day old, I believe it still reflects poorly over the coming days.

The Investor's Intelligence survey came out with a small drop in bulls and rise in bears, but the bullish ratio (bulls / (bulls + bears) remained above 60% (63.2% to be exact). This is the 11th straight week with 6 out of 10 respondents voting that we go higher (out of those who expressed an opinion), even while the S&P only closed 1.4% higher during those 11 weeks. Interestingly, there have been 12 streaks of at least 11 straight weeks with a bullish ratio above 60% in the past 12 years. 6 of those streaks occurred before the year 2000, and 6 after. So in 9 years of one of the greatest bull market pushes in history, we saw the same number of streaks as in 3 years of one of the swiftest and most severe bear markets. Put another way, out of the 481 weeks leading up to the peak in March 2000, we saw 173 weeks with a bullish ratio above 60%. Since late March 2000, we've seen 89 weeks out of 146 with such a bullish ratio. So while the S&P 500 appreciated 356%, only 36% of the weeks showed a bullish ratio above 60%. But while the S&P 500 DECLINED 41% since then (and in one-third of the time), 61% of the weeks have shown a bullish ratio above 60%.

It is apparent from this that individual investors (or at least those sampled by this survey) have an ingrained "buy the dip" mentality that is showing no signs of letting up. For us to have essentially at least triple the number of bullish streaks as we "should" have during a severe market decline is not healthy. The common belief among bears is that it will take several months, perhaps as long as a year or two, of complete apathy on the part of investors before the bear market is over. By the looks of this survey population, we have a long, long ways to go.

On a shorter time frame, today's decline was enough to work off the overbought condition in our shortest-term indicators. The STEM.MR model, after hitting a low of 11% at the peak on Monday, has now recovered and is firmly neutral at 37%. The story is similar with each of the model components (the cumulative TICK, all-exchange put/call ratio, TRIN and VIX). The Down Pressure indicators, which dipped into sell territory on Monday, have also moved back into neutral. The price oscillators on the S&P and NDX are actually already in or near oversold territory with today's steady decline.

I believe the powerful thrusts up over the past week should be respected in the short-term. I try to avoid discussing technical analysis in this forum, but I believe that as long as we remain above a 50% retracement of the recent move (equating to 900 on the S&P and 1030 on the NDX), the short-term trend should be considered up. Considering that short-term sentiment is no longer overbought, I fully expect to see some sort of rally attempt as long as we are above those levels, perhaps even taking out the recent highs. However, from an intermediate-term perspective, I believe that if such a rally occurs, it would once again serve to set up a short opportunity lasting at least several weeks.

We have just a few changes in our situation tonight, so tonight's comments will be brief.

The STEM model, after hitting a one-year low yesterday, set a new three-year low tonight with a reading of 19%. This has not only set an absolute low, but also a relative low, as the model is 6% beneath its lower trading band. I've pointed out several times now that these spikes of optimism tend to be short-lived, so this week's model extremes should serve notice that we are likely in a topping process.

The Rydex mutual fund timers have begun jumping on board the recent move, as our RSI spread hit +58 yesterday. I consider anything over +60 to be extreme, so we're close but not quite there. Since December 31st, these traders have moved $230 million out of the S&P and NDX bearish funds (leveraged and non-leveraged) and put $226 million into the bullish funds. This is a 24% increase in bullish assets and a 16% drop in bearish assets. If we simply combine these into a single number, we could say that there was a net 40% bullish swing (24% rise in bullish funds + 16% drop in bearish funds = 40% total swing). This is one of the largest swings in the past two years. The chart below only shows the past year since the swings are so noisy. I've circled each of the past instances where the swing reached 30% or greater, and it's obvious the market didn't too terribly well after any of the occurrences.

I showed the 10-day equity put/call ratio yesterday, and the trend lower continued today, with the ratio losing another couple of percent to reach .57. This matches the level we hit on 11/27 and 11/29 before the most recent decline. On a relative and absolute level, we have reached a point that has coincided with every intermediate-term top during the past two years.

On the bright side, today's consolidation session served to relieve much of the pressure off of some of our shortest-term indicators such as the cumulative TICK and the price oscillators. While this is a positive for the very near-term (basically meaning tomorrow), the fact that we reached the levels we did must urge caution over the next 3-10 days.

I can see some wiggle room here for yet another attempt higher, and most of our indicators that are reaching extremes have not started to roll over in earnest (suggesting that they have reached a peak or trough), but at this point I am viewing any rally at all as an opportunity to initiate short positions at a better price. With an outlook of 3-10 days, my view is strongly in favor of lower prices.

This weekend I mentioned that the CFTC delayed the release of the Commitments of Traders data until today on account of the holidays. The most recent reporting period, which ended on 12/31/02, showed another rather positive change in commercial and small speculator net positions in the large S&P 500 contract. During the down week, the commercials covered a net 6,800 contracts from their net short position, while the small specs decreased their net longs by a net 16,800 contracts. The position changes over the past two weeks have swung the three-month stochastics for both groups into more favorable territory. The three-month Futures Balance Matrix (which measures the relative level of long-side commitments between the commercials and small specs) has moved from extremely low levels in early December to 48% now. The higher the number, the more long-side momentum the commercials have compared to the small specs - a bullish sign. While 48% is nowhere near bullish, the fact that we are no longer below 10% is something of a positive in the intermediate-term.

Also somewhat positive is the most recent lowrisk.com sentiment survey, released today and covering the past week (through Sunday). The bullish percentage dropped 10% to 32%, and the 4-week average dropped 4% to 40%. While the drop in bullishness is welcome, the moving average is still only neutral. This survey (on a moving average basis) has a close correlation to the more widely-following Investor's Intelligence survey, so we should also see a drop in bullishness there. Of course, with the market performance we've seen over the past few weeks, we certainly should see some kind of drop in bullishness.

In the December 29th weekly commentary, I said that a rally of 5%-8% would likely set up a good intermediate-term shorting opportunity. The S&P and NDX have both rallied approximately 6% from the level we were at then, so we're right in the ballpark here. And from the looks of some of our indicators, that opportunity is about to arrive (or already has).

In this past Sunday's commentary, I said that another little bump up of a couple percent would move some of the indicators into the bearish camp from neutral. That has now happened, as you can include the breadth indicators, put/call ratios and STEM model with the other bearish indications.

In our breadth indicators, the 10-day a/d line that I discussed this weekend has now moved to .58 from .56. Anything above .55 can be considered overbought, and .60 can be considered "maximum" overbought, so we're very close here. We'll be dropping a very large positive reading from the average tomorrow, so it's unlikely we'll become more overbought at this point. If we look at a 5-day average, then our current level is the 4th most overbought reading we've seen in 40 years. That's pretty remarkable considering we're in a bear market and not exploding off of a panic low. In addition, our 10-day up volume indicator is now overbought at .56 (the same .55 and .60 levels can be used for this indicator). This weekend I mentioned the Down Pressure indicators on the S&P and NDX, and as of today both are firmly overbought, with the S&P at 21% and the NDX at 15%.

In the put/call ratios, the 10-day OEX ratio is approaching its bearish trading band, while the 10-day equity-only ratio is also extremely close to its bearish band. If we forget the standard deviation bands and just draw a common-sense trendline across the troughs of the equity put/call ratio over the past two years, we can plainly see that every time these poorly-capitalized traders became enthusiastic enough to drive the 10-day average down to the line, the OEX (S&P 100) had terrible trouble making any headway. The fact that we are once again down to this trendline has to be considered troublesome. Obviously, the ratio could become even more extreme, or this could be the one time when it doesn't work at all, but trading is about assigning probabilities to various outcomes. If you place your bets in alignment with the highest probability for success and place stop loss orders for the times that it doesn't go your way, you will outperform 99% of your competition over time. Today, those probabilities are against becoming aggressive on the long side.

This weekend, I said that the STEM model was at its lowest absolute level since early and late November, late August and early March (each resulting in short-term market highs). We've now bested all of those prior instances with a current reading of 23%, pushing the model into strong negative territory along with our shortest-term STEM.MR model (which closed at a very low 11%). This is yet another sign that we are about to go through a topping process.

Adding to the records are the price oscillators on the S&P and NDX, which both hit 80% (a two-year high) late this afternoon. To review, these oscillators take an average "score" of the past 13 half-hourly bars. Bars which open near the low and close near the high get high scores, and vice-versa. The fact that both indicators reached 80% show that there was an extremely steady uptrend with no rest, a condition which precedes at least a small breather with very high probability.

In the intraday note, I said that the 39-period cumulative TICK reached a new three-year high this afternoon. For those of who you didn't read the note, the 39-period TICK sums up the past 39 half-hourly TICK readings, which equates to three trading days. The prior high was +18,373 on 1/7/02 (coinciding with last year's opening flurry and subsequent selloff), and our new high as of the close today is +21,165. We've had triple the number of TICK readings greater than 2 standard deviations from average than we statistically should, which obviously is also highly unusual.

We're at a juncture now where the major indexes (and by default the majority of their largest underlying components) have not scored any type of new high. They have not broken out of congestion or overcame any major trendlines. Yet the underlying breadth of the market is overbought, extremely so in some respects, and trader sentiment regarding that overbought condition is complacent as far as we know with the information available to us. While I respect the fact that the indexes have put in a good showing over the past few days, the concerns listed above have put me on guard for an imminent reversal. I guess it would make technical sense if we went up and tagged the December 2nd highs before turning down with force, but my studies show that even that magnitude of a move from here would be highly unusual. Therefore, I am now concentrating on short setups with a time frame of at least several days, and recommend that those who trade from the long side (by preference or charter mandate) hedge themselves aggressively against downside risk.