Sunday, April 27, 2003
Last weekend, I suggested that short-term weakness was more likely than
strength. That wasn't much help, as the choppy action on Monday just
served to provide time for another run higher into mid-week. I
thought that a move over 905 would lend enough support to technical-type
traders that we would probably push into the next obvious layer of
resistance, which I believed to be around 930/940, but any such move over
that 905 level would just give us a better opportunity to sell or hedge
long positions and/or initiate longer-term shorts. We only made it
to 920, but I think there's more likely than not another push higher in
front of us before we ultimately roll over.
____________________________________
Indicator: COMMITMENTS OF TRADERS
Status: BULLISH (int-term), BEARISH (short-term)
Comment: Once again, nothing much has changed here from what was reported two weeks ago. In the large S&P 500 futures contract, commercials remain relatively long (although they pared those positions to the largest degree they have in over two months). The small speculators continue to shrink their net long position, marking the sixth week in a row that they have gotten less long. Our Futures Balance Matrix, which compares long-side momentum from the commercial traders versus the small specs, has continued its record string of 100% readings, notching its sixth consecutive week. This is the most bullish configuration possible for the market, and I don't think its message can be ignored.
Also once again, I must temper this bullishness with the readings from the e-mini S&P 500 contract and the 30-year bond. In the e-mini, the commercials have added to their record net short position at the same time the small specs have increased their record net longs. This is essentially the eighth week in a row we've seen this pattern. With the commercials now short about 314,000 contracts, the small specs long 355,000 contracts, and average daily volume in the e-mini's around 630,000 contracts, these positions are beginning to get relatively sizable. There is a chance that if the market continues higher, the commercial traders will begin covering their shorts, lending something of a bid to the futures. However, I think that's much less likely than them just continuing to add to their short position. On the other hand, I think there is a distinct possibility that the small specs will begin to unload their record net long position should the S&P start to roll over and a downtrend begin in force. Small specs tend to have a less steady hand with market gyrations, and are more likely to bail on their position once it goes against them. While their net long position is only about 1/2 of the average daily volume, it's not the positions themselves that I'm worried about - it's the probability that those positions represent the sentiment of these small traders, and they have bet on further upside in other ways that will be just as quickly abandoned if it appears we're not headed to new highs.
In the 30-year Treasury bonds, our variance between commercial traders and small speculators remains positive, as it has for five weeks in a row now. This is the longest stretch of a positive variance since January 2000, which coincided with a major low in bonds (and high in yields). As I've stated several times, bonds and stocks have had a very high negative correlation for the past three years, so this positive development in bonds would have negative connotations for equities. Again, however, please be aware that this negative correlation is not perfect, and goes against their historical relationship, so at some point it will break down. We've seen some evidence of this over the past two weeks, and I think it bears watching.
Bottom Line: The conclusion here is the same as two weeks ago - short-term weakness is more likely than strength, but the intermediate-term outlook is positive.
____________________________________
Indicator: RYDEX RATIOS
Status: BEARISH
Comment: The longer-term bullish and bearish asset flows in this mutual fund complex are coming down off of their extreme levels from a couple of weeks ago. Since these flows are measured by the deviation of the 10-day average from the 50-day, this type of action is not surprising considering that the 50-day average has been dropping the heavy asset moves into the Rydex bear funds seen in mid-February. As this longer-term average now takes into account fewer bearish readings and more bullish ones, it takes more and more extreme readings from the 10-day average in order to keep these indicators steady. In any event, this past week has indicated a lessening of the momentum into the bull funds, which while not quite a positive for the market, is at least less of a negative.
Our shorter-term measurements, such as the 3-month stochastic of the bull ratio, and the RSI Spread, are still rather elevated and in bearish (for the market) territory. This tells us that in the very short-term, meaning the past few days, the market timers who trade the Rydex funds have shown a pretty heavy predilection towards those funds that bet on further upside. This is confirmed by our Beta Chase Index, which has risen for three days in a row for the first time in over a month. This demonstrates that the Rydex fund timers have been moving assets rather aggressively into those funds that have a high beta, or correlation to the broader market. We see this type of activity when the timers want to speculate on higher prices with the most aggressive vehicles that Rydex offers. The Index has not yet become extreme, but it bears close watching - if we see such fervent, frothy bets being placed that the Index reaches 3 or higher, then we have a pretty good clue that the party is about to end, and cheaper equities will be the likely result.
Bottom Line: In what has proved to be a fairly rare occurrence, all of our indicators in this fund complex entered extreme territory by last Thursday's close. It would be expected to see (at least) short-term weakness after such a confluence, and how these timers react from here will be telling. Should we get more upside, and their bets remain aggressive to the long side, then the chances for a longer-term top being put in place are high.
____________________________________
Indicator: VOLATILITY MEASUREMENTS
Status: BEARISH
Comment: As you can see from the VIX Fear Premium posted to the site, options traders are not expecting future volatility to be much different from the recent past. This type of certainty, where the Fear Premium is so low, is rare. While we're not there yet (but likely will be this coming week), the indicator has dropped below zero on only 10 distinct occasions since 1987, and 7 of those have occurred in the past five years. Whenever options traders get confident of ANYTHING, it makes me nervous, so this low future premium has put me on guard.
I've touched on the VIX Transform indicator several times in the past week, so I'm not going to go into depth again today. However, I do want to mention that the indicator is now the forth-most extreme in its history (dating back to 1987), and is at the highest level since June 1991. This is extreme in a statistical sense, and it gives strong indications that equities are not favored here from an intermediate-term perspective.
Bottom Line: Everything that we monitor volatility-wise is more indicative of a topping process than a bottom. Indicators from the VIX and VXN to average volatility to relative volume suggests that higher prices are more likely a selling opportunity than the beginning of a new, sustained leg higher.
____________________________________
Indicator: BREADTH RATIOS
Status: BEARISH
Comment: I pointed out the current (very overbought) levels of the 21-day advance/decline line, up volume ratio and cumulative TICK late last week, so I suggest you check Thursday's commentary for a view of those indicators. As a recap, they are overbought to a degree not seen since the bear market began, and while we remain in a downtrend, that has to be considered a major negative for our prospects going forward.
The shorter-term 10-day moving averages of the a/d line and up volume ratio have likely peaked for the time being. For six out of the next seven days, we will be dropping positive figures from these averages. Incredibly, four of these days will see a/d readings over +1000 disappear. This means that it will be extremely difficult for the averages to become any more overbought than they were this past week, and it sets up the probability that if the S&P squirts to new highs, it will be accompanied by negative breadth divergences (signaling a loss of momentum).
I haven't mentioned the new high/new low 10-day average for a long time, because it really hasn't been telling us much. However, the indicator is now essentially overbought and is at the highest level since mid-January. I'm always confused when analysts trumpet the expansion of new highs on the NYSE, suggesting that it's a sign of strength and is a good signal to buy. The market has had great difficulty retaining upside momentum when new highs have trumped new lows to such a great degree and for such a sustained period, especially during the past few years. There is some room for this indicator to become more extreme, but again it is telling us that the rally is getting long in the tooth and further upside will be exhaustive.
Our Down Pressure indicators did a good job at hinting that Tuesday's rally was a bit "too much, too soon". After reaching extreme overbought territory at the close Tuesday, they have worked off that condition and are now closer to oversold due to Friday's ugly session. In fact, if we have a single-day Down Pressure reading above 60% on Monday, then the indicators as posted to the site will enter oversold (since they are based on a 3-day moving average). In order to get a reading over 60%, we'd likely have to see the S&P and NDX down by more than 0.5% tomorrow.
The daily cumulative TICK indicators remain overbought on the NYSE and NDX. The intraday indicators have relieved their overbought condition and are currently neutral. This "neutral" position in the intraday indicators have been all that was necessary to give the market a breather for another run higher during the past month. However, the daily indicators are suggesting that if we do get another upside attempt, it will be exhaustive and likely not have the legs to become sustained.
I'm really not a big fan of the TRIN since I believe that it has some inherent weaknesses that severely limit its usefulness, but during the bear market, its sell signals have been a thing of absolute beauty (if you can call declining prices "beautiful"). On Thursday, both the NYSE and NDX 10-day TRIN gave overbought signals for the first time since mid-January. When these indicators have become overbought during the past few years, it has been an excellent sell signal in the short- to intermediate-term. Since the TRIN is constructed using the advance/decline and up volume/down volume figures, it will be extremely difficult for the TRIN on either index to become any more overbought than it was this past week (for reasons described above).
Bottom Line: Other than in the very short-term, our breadth measurements are telling us that the market is not only extended, it is OVER-extended in an historical way. I cannot see how this is anything other than a major negative for the market, unless we have entered a new bull market and this time really is different from every other over the past three years.
____________________________________
Indicator: COMPOSITE MODEL
Status: BEARISH
Comment: As I stated on Thursday, this model has entered strong negative territory for only the third time in three years. The first being August 2000 and the second March 2002. There is an extremely high probability that the model will not become any more overbought next week, as many of the components are already "maximum" overbought and will almost certainly be relieving at least some of that condition next week. Regardless, the fact that the model has slipped this far, while some of the longer-term components are still neutral or even bullish, suggests that we saw a truly massive amount of optimism and overbought breadth in our shorter-term measures. The model tends to be most effective out several weeks, so the suggestion here is that additional upside attempts will be met with weakening momentum and are more likely to fail than not.
Bottom Line: We don't have enough history with this model to be staking an entire outlook on its readings. However, the model covers a very broad range of sentiment readings and when we get such a confluence of extremes, it should put us on caution.
____________________________________
Indicator: SHORT SALES
Status: NEUTRAL
Comment: During the reporting week ending on April 11th, public investors picked up their short-selling activities in relation to specialists on the NYSE. The difference was enough to push the Specialist Short Ratio down to 33% (meaning specialists accounted for 33% of all short selling activity during that week), which is historically low. And when public demand for selling short is so great that the specialists make up such a low proportion of total short-selling activity, it is usually a sign that the wrong-way public is too pessimistic and we are due for a rebound. Unfortunately, I have to once again temper this bullish outlook by just a tad. First, the market had an overall negative tone during the reporting week, so an increase in public shorting should be expected. Second, the NYSE Composite has rallied nearly 3% since the time those figures were collected, so I can (almost) guarantee you that public shorting has dropped off during the past two weeks. Based on historical precedent, I would say that there's a greater than 80% chance the Specialist Short Ratio will be higher two weeks from now (when the figures for this past week are released). My best "guestimate" is that the ratio will be somewhere north of 36%.
One thing I haven't touched on very often, since it so rarely gives signals, is the monthly short interest figures published by the NYSE and Nasdaq. Short interest is simply the total number of shares currently held short by investors. These investors have borrowed shares from their broker and sold them, hoping to buy them back later at a reduced price, and keeping the profit. Selling short is the exact opposite of a typical stock trade - you SELL first, then BUY later.
As you can see from the chart below, short interest on the NYSE has been growing steadily over the past decade:
You can also see that the chart doesn't do you much good, mainly because total volume has grown exponentially during this time. Typically, you take the short interest figure and divide it by that month's average volume to adjust for the secular upward trend in volume. This is the "short interest ratio" you may have hear about.
The charts of short interest that I post to the site, in contrast, are modified to take into account volume patterns and seasonality. Since both volume and short selling itself are very seasonal, taking the figures at face value can give misleading signals. To correct for this, I use the past 12 months' average volume, and normalize the short interest based on its seasonal tendencies. If we see a high short interest ratio, it means that there has been a large amount of shares sold short in relation to average volume, and indicates investors are betting aggressively on declining prices. From a contrarian perspective, this is a positive sign for the market. Conversely, a low short interest ratio is a sign of confidence in rising prices (after all, who would short stock when they think prices are going to rise?), and is a bearish omen for the market when it reaches an extreme.
The chart below shows this modified short interest ratio, with highlights showing periods that had a high short interest ratio (green shading) compared to those that showed a low short interest ratio (red shading).
This chart makes a lot more sense in terms of giving us some kind of clue as to what is extreme and what is not. And from a long-term perspective, it does a pretty good job at identifying the extremes. Most significantly, this short interest ratio is one of the only sentiment measures I know of that accurately stayed bullish from 1995 through 1997. Most others were suggesting that optimism was too high and gave multiple false signals. Currently, the ratio is suggesting that even though total short interest is close to an all-time record, on a modified basis that adjusts for volume and seasonality, it is at a low (but not extreme) level that has been more closely related to falling markets than rising ones.
On the Nasdaq, we actually have a drastically different scenario, as the modified short interest there is the highest since 1998. This should be a positive for technology shares, although we have much less data on which to rely for this information.
Bottom Line: The relatively heavy public short selling seen a couple of weeks ago could be considered a positive, however much of that enthusiasm should be tempered since we've already rallied 3% since those figures were recorded. Also, on a "de-trended" basis, the ratio of specialist shorting is neutral at best. From a (much) longer-term perspective, we have not seen the type of large short interest ratios that corresponded to the big bull moves in the 1990's.
____________________________________
Indicator: SENTIMENT SURVEYS
Status: NEUTRAL
Comment: I went over the most recent readings from Investor's Intelligence (positive for the market) and AAII (negative) on Thursday, so I'm not going to repeat myself. The other surveys came in pretty much as expected, with a slight increase in bullishness.
Bottom Line: While most of the surveys have not recorded what I would call truly extreme levels of optimism (except the most recent AAII reading), most of them have reached a point that has coincided well with past peaks during the bear market.
____________________________________
Indicator: AIM MODEL
Status: NEUTRAL
Comment: The drop in bullishness in the II survey was cancelled out by the increase in AAII, with the other surveys not adding much to the mix. The end result was basically no change in this model, which remains on the lower end of its trading range - certainly not positive for the market, but not yet exactly negative, either.
Bottom Line: We have not seen the type of rapid shift to the bull camp in the sentiment surveys that would get this model into extreme territory. Unless (or until) that happens, the model will be suggesting that while it's not time to sell just yet, one should be on guard.
____________________________________
Indicator: SEASONALITY
Status: NEUTRAL
Comment: The days around Easter performed pretty much as expected this year, and it should have helped short-term traders who were looking for a bias to trade around those days. This is one of the more consistent seasonality biases extant, so put it on your calendar for next year as well.
This coming week, we will have the end-of-month, beginning-of-new-month pattern working for higher prices, so that could lend some support to the market, particularly later in the week. Unfortunately, we are about to enter the worst time for the market, historically. Five out of the next six months are the worst for the market in terms of expectancy, with May being fourth-worst.
Bottom Line: The positive bias around the beginning of a new month may lend some support to equities later in the week, but the historical record for equity returns over the next six months is negative.
____________________________________
Indicator: PUT/CALL RATIOS
Status: NEUTRAL
Comment: The put/call ratios were generally mixed this week, although the moving averages have become a bit more negative than they were at this time last week. The OEX p/c ratio, which I view in a non-contrarian manner (meaning low ratios are positive for the market while high ratios are negative) has been steadily ticking higher, in conjunction with the open interest in those contracts. I pointed out a couple of times during the past week that such an open interest configuration tend to be a negative for the market, and that continues to be the case.
Meanwhile, the equity p/c ratio has been generally declining, especially if QQQ options are backed out (see Tuesday's daily comment for some background on this). The 10-day moving average of the equity p/c ratio minus QQQ options dropped to 0.56 on Friday as opposed to 0.62 the week before. This is a level that has coincided with excessive retail optimism in the past, with declining market prices soon to follow.
Bottom Line: While not yet at extremes, the equity put/call ratio has entered a territory that should raise caution flags. This is especially so if QQQ options are backed out of the equation, as they tend to inflate the typical equity ratio. At the same time, the OEX put/call ratio has been rising along with the open interest ratio, which has signaled trouble with consistency in the past. If these ratios become more stretched than they already are, they will become a firm negative for the market.
____________________________________
Indicator: STEM.MR MODEL
Status: NEUTRAL
Comment: The decline on Thursday and Friday, particularly Friday, served to push our shortest-term model to the highest level in over a month. It is still nowhere near suggesting extreme short-term pessimism, but it does show that the ebullient optimism of early last week has worn off. Each time the model has reached this level in the past month, the S&P has been able to stage a rally of at least 10 points, so it may be instructive to see how this one is treated. If we cannot muster much of a bounce, it may suggest that the upward momentum has left and we are in store for a steeper retracement.
Bottom Line: The short-term optimism which was so bothersome last week has worn off to a point that has allowed the market to continue its advance, at least for a short-term pop, during the recent past. If we cannot do the same this time, then we may have more of a decline to look forward to.
____________________________________
Indicator: STEM MODEL
Status: NEUTRAL
Comment: We saw a true extreme here this week, as the model recorded a new record low reading. That has begun to wear off with the market action seen late in the week, but we remain extremely low here on an absolute basis. In fact, even after rising 6% in the past couple of days, it's still below the previous record low reading. The model will gradually be rising this week with all probability, but it's message has already been given - we've seen a momentum peak, and any subsequent rally will likely be accompanied by less pressure and thus is more likely to fail.
Bottom Line: The extreme seen in this model suggests that additional upside attempts from here are more likely to fail than not.
The short-term trend of the market is up, and most of our short-term sentiment measures are back to neutral. If we see continued weakness on Monday, then a few of them will enter oversold territory and it would be a good indication that the selling is a bit "overdone". This suggests that further selling pressure early this coming week will likely lead to something of a reaction back to test the recent highs. If we do get such a bounce, I would view it as yet another opportunity to pare or hedge long positions and/or establish intermediate-term shorts. There are some positive developments out there from a sentiment perspective (e.g. futures positioning in the full S&P contract), but they are outweighed by the many negatives as outlined above. We have reached a historic level of overbought in many respects, and it will take time to wear that off. If we continue higher from here, I would become very concerned that we will fall back just as hard and just as fast. The most positive thing the market could do is churn around in a trading range, which is a possibility. However, the types of readings that we have seen are not normally worked off in that manner - they are usually resolved by a declining market, particularly during the last three years. From an intermediate-term standpoint, out several weeks at least, the risk/reward firmly favors the short side.
- 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.
© 2003 Sundial Capital Research, Inc. All Rights Reserved.