May 11, 2003

http://www.sentimentrader.com

Sunday, May 11, 2003

In the last weekly comment two weeks ago, I reiterated that a move over 905 on the S&P would likely lend enough support to technical-type traders to push us into the 930/940 level, but any such move would just set up a better opportunity to hedge longs and/or initiate longer-term short positions. I've been favoring longer-term shorts since about the 900 level, so obviously the stair-step 4% rise since then hasn't helped. While many of our intermediate-term indicators are now entering extreme territory for the first time for this move, others are staying stubbornly bullish (for the market). However, I believe the weight of the evidence continues to support the short side for those with a longer time horizon.

____________________________________

Indicator: COMMITMENTS OF TRADERS

Status: BULLISH (int-term), BEARISH (short-term)

Comment: The data from the full S&P 500 futures contract has been right-on during this move, as our Futures Balance Matrix indicator turned maximum bullish on March 18th and hasn't looked back since. For the first time since then, it has moved off its maximum positive reading this week, but only slightly, as the small speculators increased their net long position for the first time in nearly two months. As I've mentioned several times, the recent activity here has been unprecedented in its bullishness. Perhaps the only comparable action in terms of the magnitude and time of the position changes is what happened in late 1990 and Spring 1994, both of which of course coincided with multi-year moves higher. We're limited with the amount of weight we can put on such instances, however, since there are only two data points, and we were in a confirmed bull market at the time.

We also didn't have the e-mini contract, which I believe now has to be considered with total outstanding positions nearing 1.25 million contracts. This is still dwarfed by the large contract when total nominal dollar value is taken into account, but the growth of the e-mini has been exponential and I think it's value as reflection of trader psychology will only improve with time. That said, the commercial and small spec positions in this contract had been hitting record extremes for seven weeks in a row - the commercials had been becoming progressively more short while the small specs have been adding net longs. So far, the small specs have been uncharacteristically correct. During the latest reporting period, which ended this past Tuesday, commercial traders were buying into the steady market rise while small specs were selling it. This is not typical behavior of these two groups of traders, as normally it would be reversed - commercials tend to sell into rallies and buy declines, while small specs do the opposite. To be perfectly honest, I'm not quite sure what to make of this activity, and am backing off giving this data much emphasis until its track record becomes more defined.

Bottom Line: The action in the full S&P contract is still not giving any solid reason to doubt the continuation of the rally off the March lows. The positioning in the e-mini contract remains a concern, but due to its lack of record, I find it hard to give it too much weight.

____________________________________

Indicator: PUT/CALL RATIOS

Status: BEARISH

Comment: Once again on Thursday, a large trader waded into QQQ options, skewing the equity put/call ratio. For some reason, the CBOE continues to lump these options in with equity options, even though they are index options for all intents and purposes. It seemed to me as though this type of activity is happening at an increasing pace, so I went back as far as the data allows (2001) to determine if, indeed, QQQ options are beginning to have an inordinate affect on the overall equity volume figures.

The chart below shows the total number of days for each quarter that option volume on the QQQ's exceeded 15% of total equity option volume:

We can see that although there were some days with QQQ volume spikes prior to 2002, since that time it has steadily picked up steam. Since the first quarter of 2002, there has been a progressive climb in QQQ option volume spikes relative to total equity option volume. It has gotten to the point now where we see one of these "Q skews" about once every three days. Due to the increasing number of these spikes, I believe it is more useful to now look at the equity put/call ratio sans QQQ options.

The chart below shows the current 10-day average of the equity p/c ratio minus QQQ options:

Since this data has been made available about two years ago, we can see that the 0.80 level has appeared to mark pessimistic, high-put-volume extremes and it's no coincidence that a market low often soon followed. Conversely, when put volume shrank to about half that of call volume, pushing the ratio to around 0.55, it has tended to coincide with market weakness. Currently, this 10-day average is on the lower end of its range, after having already pierced that 0.55 level a few days ago. As I convert the charts on the site and build up the chart archive, I will be adding this "Q-less" equity put/call ratio to the regular site updates.

While the equity ratio has been declining recently, the OEX put/call ratio has been steadily rising. As you may know, this ratio has been more successfully interpreted in a non-contrarian manner than a contrarian one, meaning high OEX put volume tends to be bearish for the market instead of bullish. Currently, the 10-day average of the OEX p/c ratio is showing one of the highest readings in the past year. All evidence is pointing to OEX traders building up long put positions, either hedging existing long positions or speculating that we are about to see some weakness in the market. We can tell this because put open interest on the OEX is growing exponentially in relation to calls, and OEX options are typically not sold to open, since they can be exercised at any time. This means that if there is high put volume, and open interest is increasing, then we can be reasonably sure that these traders are building up positions to bet on market weakness. Also, call open interest is anemic, as currently there are the fewest call contracts open the week before an expiration since January, and March 2002 before that. These traders have a decent record at calling market turns, and I would rather trade with them than against them.

Bottom Line: Wrong-way equity option traders have been more aggressive in their call activity than put activity, brining the "Q-less" equity put/call ratio down to a point that has coincided with market weakness in the past couple of years. Simultaneously, OEX traders, who tend to be right much more often than wrong, are building up positions to bet on market weakness.

____________________________________

Indicator: RYDEX RATIOS

Status: NEUTRAL to BEARISH

Comment: For the first time in a long time, the activity here has been - for the most part - useless. Our shortest-term RSI Spread indicator has been effective when it has reached extremes, but other than that the rest have not been a good guide. They turned bearish (for the market) in early or mid-April, and have pretty much stayed there since then.

One indicator that only recently has dipped into bearish (for the market) territory is the percentage of assets in the money market. At 32%, this level of money market assets is the lowest since December 2001. This indicator can be effective because it is a direct reflection of how confident traders are feeling. When they are uncertain about market direction, they move their money to the sidelines for protection. This is why we often see very high asset levels in the money market near intermediate-term market lows. For example, in July 2002, the percentage of assets in the money market rose to over 50%. Now that traders are feeling relatively comfortable with where the market is going, or at least that it won't make some sudden adverse move against them, they have put their money to work in the various funds that Rydex offers. The only time this type of complacency did not result in lower equities prices since the Dynamic funds were created was in the immediate aftermath of 9/11. The average money market yield has not decreased markedly during this time, which would be the most logical alternative explanation for such an exodus from the money market that Rydex offers.

Bottom Line: The only notable development in this complex is the low level of assets in the money market. This happens when traders are confident of market direction, so they take money out of the money market and place it in a fund that depicts their market outlook. When this confidence reaches such a high level that the percentage of assets in the money market goes as low as it is now, it has typically resulted in market weakness not long after.

____________________________________

Indicator: SENTIMENT SURVEYS

Status: BEARISH

Comment: Three out of the four major sentiment surveys are now exhibiting what can be considered extreme bullishness (negative for the market). This is the fifth distinct time since 2000 that we have had at least three surveys in negative territory at the same time (the others were August 2000, May 2001, January 2002 and November 2002). Each past occurrence signified the end of the rally, or close to it.

I also keep a de-trended ratio for the Market Vane survey, which is simply the current reading divided by the 52-week average. This is to correct for trends in the data depending on the market environment. Currently, that indicator is at 1.57, meaning the current reading is 57% greater than the 52-week average. The indicator has breached the 1.50 level only four times in 20 years - August 1984, January 1985, September 2000 and November 2002. Surprisingly, even during those initial stages of the bull market in the mid-1980's, these high Market Vane readings lead to a flat to down market for several months after they were given. Only when the readings came back down to neutral territory did the market resume its advance.

Bottom Line: All of the sentiment surveys I track are now showing a high amount of optimism. Three of them can now be considered extreme. When such occurrences have happened in the past, it has been difficult for the market to sustain its upward momentum, which has obviously been the case during the bear market as well.

____________________________________

Indicator: AIM MODEL

Status: BEARISH

Comment: After seeing the Investor's Intelligence readings this week, I thought that there was a decent chance this model would drop below the 40% level. We didn't quite make it there, as the current reading is 42.8% - the lowest since January 2002 and the fourth-lowest since 2000. Such readings have been a near-perfect intermediate-term sell signal for the past three years, so how the market reacts in the next few weeks will let us know whether we really are seeing a different market environment than we have since 2000.

Bottom Line: This model is unequivocally bearish, as it has reached one of the lowest readings in three years. In a market that is in anything other than a defined uptrend (unlike the downtrend or trading range we're currently in), such overheated optimism cannot be considered positive.

____________________________________

Indicator: VOLATILITY MEASUREMENTS

Status: BEARISH

Comment: Not much to say here, as things really haven't changed much over the past two weeks. The common volatility measurements (VIX and VXN) continue to lose weight, and their relative position remains a constant warning sign that there is a distinct lack of uncertainty being priced into the market. I have gone over various VIX measurements ad nauseam for the past couple of weeks, and a quick check of the charts on the site will tell you not much has changed.

Bottom Line: It is undeniable that, especially recently (meaning the past five years or so), high volatility has been a precursor to a tradable low while low volatility has lead to lower prices with high consistency. Currently, we have a very low-volatility situation according to several different measurements, so the conclusion is that for the intermediate-term, shorts are favored.

____________________________________

Indicator: BREADTH RATIOS

Status: BEARISH

Comment: All of our breadth measurements are once again overbought, some grossly so. For the first time in nearly a year, the new high/new low ratio has joined that list. If we create a "super-breadth" indicator, which combines a 21-day average of the advance/decline line, up volume, and new highs/new lows, our current overbought reading has only been surpassed twice in the past five years - January 2001 and May 2001. Even going back to 1965, our current reading is nearly 1.5 standard deviations from the norm. In those nearly 40 years, there were a couple of instances where the indicator stayed pinned in overbought territory for a long time (1982 and 1995), but typically it formed a peak and came right back down again, coinciding with the market taking something of a breather. For the market to continue pushing higher from this point requires truly extraordinary circumstances.

Our shorter-term breadth indicators are also mostly overbought, with the exception of the Down Pressure readings in both the S&P and NDX. Those will require at least one more day (and probably two) of an up market before entering overbought territory.

I've read several reports recently that suggest such overbought breadth is actually a positive sign - it shows that there is pent-up demand that will not go away easily. That could be the case (although I don't believe there are enough bull/bear cycles with this data in order to come to that conclusion), but history suggests that whatever type of new market environment we may be entering, it would be highly unusual for the market to continue higher unabated once this level of overbought is reached. Even in a confirmed bull market, we usually see consolidation of gains at the very least before a resumption of the upward trend. And considering that we are still in a weekly downtrend, I believe the suggestions that these readings are positive are extremely premature.

Bottom Line: Market breadth is overbought in nearly every time frame I study. This is not a positive indication no matter how you try to spin it - at the very best, it may be neutral if in fact we have entered the beginning stages of a super-bull market.

____________________________________

Indicator: STEM.MR MODEL

Status: BEARISH

Comment: With Friday's push higher, this model has gone back to negative territory for the first time since April 29th. We also have a rare divergence occurring. This is where price makes a lower high (so far) while the model makes a lower low. In essence, it shows us that traders are more optimistic on this rally than they were during the preceding one, even though price is lower now than it was then. This suggests that there is a high degree of confidence among market participants that the market will overcome that previous high and go on to new heights. While it is certainly possible, this is not the type of scenario that would lead me to believe it is a high-odds affair.

Bottom Line: This model is now at a low relative and absolute level. Also, it is diverging from price in a manner that is giving negative connotations for the market.

____________________________________

Indicator: COMPOSITE MODEL

Status: NEUTRAL

Comment: This model was pushed into negative territory in late April as some of our shorter-term measures made up for the lack of extremes in the longer-term ones. Now, some of the slower-moving indicators are entering extreme territory, while the shorter-term ones have worked off a bit of their extremes. If we get another push higher next week, the model would likely once again enter negative territory. If so, I would look to become aggressively short at that point.

Bottom Line: The fact that this model already entered negative territory suggests we saw a large amount of optimism during the latest rally. While the market has been able to shrug that off and work higher, another drop in the model would likely coincide with the last stage of this rally.

____________________________________

Indicator: SHORT SALES

Status: NEUTRAL

Comment: Two weeks ago, I said that my best guess for this week's Specialist Short Ratio was somewhere north of 36%. The actual number came in right at 36%, which is less than I thought it would be. While the market rallied close to 3.5% in those two weeks, the ratio only crept up by 3%. This is close to the historical average when the market rallies by that much, but I thought we would see even less public shorting than we have. We've already rallied about another 5% since the most recent figures were released, so I believe the report two weeks from now (which will show the activity for this past week) will probably show a Specialist Short Ratio closer to 40%. This level, or slightly above, is what was seen at the peaks of the rallies late last year.

If we take a longer-term view, the situation changes dramatically. The persistent public short selling we have seen over the past year has been very heavy from a historical perspective. The 52-week moving average of the Specialist Short Ratio is close to matching that seen at other major bear-market lows. Let's take a look at a very long-term view:

This is a "de-trended" look at the Specialist Short Ratio. Since the 1940's, this ratio has shown a secular downward trend, as I've discussed before. Mainly, it is due to the proliferation of short-side hedge funds, and the increasing availability of short strategies to retail investors. Because of that trend in the data, it distorts current readings compared to historical ones, so the indicator in the chart above has been modified to eliminate that trend. The higher the red line in the chart, the more the "smart money" specialists at the NYSE were shorting stock. The lower the red line, the more the "dumb money" public was shorting. Therefore, high readings are bearish for the market (since specialists are shorting a lot of stock, and they are typically correct on market direction), and low readings are bullish.

We can see that if one has a time horizon of several months, this indicator does a decent job of timing the long-term swings. It is one of the rare sentiment indicators to not give multiple false signals during the bull run in the mid-1990's. Currently, we are seeing one of the lowest readings in the 60-year history of the data, matched only by that seen in the late 1940's and early 1980's (the green vertical lines highlight those instances where the indicator spiked to a low level). As you can see from the chart, every time this indicator dipped to a low level, showing a large amount of persistent short selling by the public, the Dow Jones Industrial Average rallied. An argument could be made that since we are coming off the most bearish reading in history in late 1999, then there is a possibility that we will now reach the opposite extreme and see the lowest reading in history. We're close to that point now, but not quite there yet.

Bottom Line: The weekly reading here is not telling us much, as public shorting has been pulling back at a rate commensurate with the rise the market has seen. On a shorter-term detrended basis, the level of specialist shorting is neutral. If we look at a long-term detrended indicator of this data, it is apparent that the persistent public short selling of the past year has reached a level that has coincided with very nice upward advances in the DJIA since the 1940's. This gives some credence to the argument that "this time is different" as compared to other rallies during this bear market.

____________________________________

Indicator: SEASONALITY

Status: NEUTRAL

Comment: There is nothing particularly noteworthy about this coming week as far as seasonal tendencies. However, as I've stated previously, five of the next six months are historically the worst for the S&P 500. The graph below shows you the expectancy for each month (from 1950 - 2002). The red highlights show the worst six months and the green highlights show the best six months.

As you can see, the coming months are nearly all red.

Bottom Line: There is no special tendency about this coming week, although we have entered what is traditionally the weakest time of the year for equities.

____________________________________

Indicator: STEM MODEL

Status: NEUTRAL

Comment: This model has not been much help at all recently. It has been steady at around 19% for two weeks now while the market has climbed higher. When the model initially went into record low territory in late April, it suggested that any additional upside without more of a correction first would be more likely to fail than not. Obviously, that was wrong as the market has stair-stepped higher without missing a beat since then. As it stands now, there is a high probability that this model won't have anything useful to say for some time to come.

Bottom Line: The last signal from this model was pretty much useless. In its current state, I don't believe it will give us much insight for the next week at the very least.

Once again this week, we have only the positioning in the full S&P 500 futures contract to look forward to as a positive from a sentiment perspective. The larger (weekly) trend of the market is currently either neutral or down depending on your definition, so until a prior weekly high is taken out (say around 955 on the S&P), my preference remains selling rallies with a time frame out several weeks at the very least. Short positions taken earlier have taken some heat, and to be perfectly honest, I've been surprised at the resiliency the market has shown. Normally, when a market does not do what it "should", it's sending you a signal that things have changed. So I've been asking myself repeatedly whether I'm just being stubborn (which is a recipe for ultimate disaster when trading), or whether there is sufficient evidence to believe that what has worked for nearly three years cannot be relied upon any longer, and one should switch to a "buy the dips" mentality instead of "sell the rallies". From a sentiment perspective, there are a few things which bolster the "this time is different" argument - namely, the unprecedented behavior in the S&P futures, and the persistent level of public shorting over the past year. However, I continue to believe that until we see the market itself confirm a bull market, it is more prudent to "dance with who brung ya" and sell the rallies.

- Jason Goepfert

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.