sentimenTrader.com

All rights reserved.

Sunday, July 7, 2002

From last week's commentary:

"The situation isn't as clear this week as it was last.  Last week, it was a good bet to me that any gaps down or cascading moves lower were buyable for a bounce, maybe even catching THE bottom on an intermediate-term basis.  This week, with the rally off of those lows, we're in a more precarious position.  We're getting overbought on a short-term basis, and when the market declines we will be facing a major test.  The sentiment studies I have done suggest that the test would likely hold, although another failure may generate the kind of fear so many seem to be looking for for a better bottom."

Those overbought conditions (in the context of a downtrend) did indeed lead to a decline which lead to the test of the recent lows.  The lows broke, only to lead to a reversal which prompted the holiday-shortened explosion on Friday.  It looks good on a chart, but like many others I think we need to be very cautious about reading too much into the action. 

This is why...

BULLISH

NYSE MEMBERS REPORT - Essentially no change here, as the specialist short ratio increased just a tad to 38%.  The entire short picture remained basically the same, as the level of activity didn't really change for any of the reportable groups.  As the market declines, the public continues shorting in an historically aggressive manner.  The most aggressive public activity turned out to be during the break of 1050 on the S&P, with relatively less shorting during the rally failures and subsequent testing of previous lows.  If that pattern holds, it's likely we won't see a lower reading than the recent 32% unless the broader market breaks the September lows on a weekly basis.  Longer-term, this indicator is still bullish.

COMPOSITE MODEL - Another week, another string of readings over the buy threshold.  Tuesday actually tied the highest absolute reading we've ever seen, as the model reached a high of 83% before closing Friday at 68%.  Even after Wednesday's reversal and Friday's large gain, the longer-term picture is still extremely oversold.  As we've seen repeatedly over the past couple of weeks, oversold can of course become more oversold, but I think we've reached a point now where it would take an out-and-out true panic day or two to get the model above the recent highs.

SENTIMENT SURVEYS - I've been hearing a lot of grumbling about the Investor's Intelligence survey and how bullishness is not nearly at a place where a decent rally can form.  That's ridiculous, as this is ONE survey covering ONE distinct population.  If you look at that survey and that survey only, you are doing yourself a great injustice.  When you look at the whole of investor surveys instead of only one, you begin to form a much different opinion of current sentiment conditions.  The Market Vane survey remained at 18% bulls, which is once again the 3rd lowest reading since 1982.  The AAII bullish ratio is the lowest since 1998 at 37%, as the population of this survey (individual investors) made a large move from neutral to bearish (so much for my "more patient investor" theory!).  The Consensus bullish percentage dropped a bit to 24% and the II bullish ratio increased a little to 56%.  If you construct the investor sentiment stochastic I've discussed several times (take a straight average of the one-year stochastics of each of the four major surveys), you will come up with a figure that has only been reached twice in the past 10 YEARS - the March 2001 bottom and the April 1994 bottom.  September 2001 comes close, but we're more negative than then.  October 1998 comes close, but we're more negative than then too.  And people are actually saying that sentiment is too OPTIMISTIC now?  That's incredible, unless you're looking for the mother of all rallies, and the end to the bear market.  I'm not...I'm just looking for an intermediate-term rally that gets us back to overbought. 

AIM - Obviously, with the large move from the AAII survey, this model has increased again, with this week's reading coming in at 93.7.  It's rare that the model increases in a week when the market did too, which it did during the most recent polling period.  That tells me there's not a whole lot of faith in whatever rallies the market manages to pull off.  Since this model is constructed in a different manner than the stochastics mentioned above, it is giving a slightly different picture of the sentiment surveys as a whole, although not by much.  The current reading has only been matched or exceeded a handful of times in the past 15 years, with the most recent occurrences being September 2001 and October 1999 before that.  The average S&P return coming off such a reading has been just over 8% within three months, with very limited downside after such an extreme has been recorded.

STEM - The extreme positive readings recorded during Friday's rise has begun to have an effect on this model, as the fast line (20-period moving average) has moved from extremely oversold to exactly neutral.  It will take another couple of days before the signal line will be deeply effected, although we will be dropping strings of very high readings during the early part of next week, so that line could drop quickly with a further rally as well.  For now, we're still in oversold territory.

TRIN - We're coming off of the 3rd highest 10-day average reading in the past 40 years, so even with two days of low readings, the average is still elevated (as it is on the Nasdaq).  One thing I want to touch on with the Nasdaq TRIN is that WorldCom is having a large impact on that reading.  I calculate the TRIN on the NDX by hand (well, by spreadsheet), and for Friday I show a reading of .38 (note:  this is different from the chart on the site because the one on the site is from my quote vendor, not from my personal calculations).  Since WorldCom was positive on the day and it had such tremendous volume, it skews the numbers greatly.  If that one stock had been negative on Friday (instead of up by $0.03), even by one penny, the TRIN would have been 40.71!  So a difference of four cents in one stock would have moved the TRIN from .38 to 40.71.  If we just take WorldCom out of the picture entirely, the TRIN would be .70.  That's low, but it's not absurdly so.  We're still oversold and bullish here.

CBOE RATIOS -  Not much has changed here, so I'm not going to repeat myself again about the historic level of bearish bets we've been seeing over the past couple of weeks.  Friday's put/call reading was a neutral .77, which was quite a bit higher than the .66 recorded during the previous Friday's not-nearly-as-large rally.  A lot of small players are apprehensive of Friday's low-volume surge, and were not positioning themselves for an upside continuation.  The OEX put/call ratio continues to hover under 1.0, meaning the "big money" is buying more calls than puts.  That's a positive sign over the short term.

BEARISH

STEM.MR - Friday's action pushed all component indicators into bearish territory, and we are now overbought within the context of a downtrend.  As I suggested in the signal commentary, the combination of light volume, a shortened session, oversold longer-term conditions and a holiday atmosphere could make for a strong trend day on the upside, which it did.  But come Monday, all of those "excuses" for rising prices will be eliminated.  This model closed Friday at 0.90, which is the most overbought reading we've seen since May 14th (the peak before the most recent decline).  It's not unheard of for the model to stay extreme for extended periods of time (as we saw with 25 straight readings above 95 not too long ago), but it's not typical.  When we consider that we're bumping up against resistance, it may be difficult for the market to make much headway before we see something of a correction.  Unless, that is, we're at the beginning of a new uptrend, in which case it wouldn't be unusual to see several explosive days out of the bottom.

NEUTRAL

BREADTH RATIOS - The advance/decline difference on Friday was an incredible 1,924 (meaning that 1,924 more issues closed up on the day than down on the day on the NYSE - this will vary some depending on your quote source).  The last time we saw such a high reading was...well, never.  At least not in the past 40 years.  Now, that's somewhat deceptive since as time goes on, more and more companies are traded on that exchange, so it's best to look at a ratio instead of absolute number.  In that case, we had a a/d ratio of 81% (meaning that 81% of all issues that were either up or down were up).  That's not the highest ratio in history, but it is in the top 1% over the past 40 years.  The last reading we had that was as high was on 5/2/97, which kicked off a tremendous 20% rally in the S&P over the next three months.  Interestingly, total volume was nothing spectacular on that day either, but it wasn't as bad as Friday, where volume was not only holiday-shortened, it was also lighter than it has been over the past week.  The average volume on the NYSE had been about 990,000,000 up to 1:00pm EST over the past 5 days, but Friday it was only 790,000,000.  Up volume accounted for an incredible 94% of all volume, which is in the top 0.1% (yes, 0.1%) of all readings over the past 40 years.  The last such reading was on 5/11/90 which lead to bit more of a rally before the ultimate lows were reached during that bottoming process.  Throughout the past major bottoms we've seen over the past few decades, it is typical to get extreme upside days such as this in the process of finding a bottom.  These days do NOT necessarily sound the all-clear sign.  In fact, quite frequently, the market will head back down and we will get the opposite day, where volume and the a/d line are heavily skewed to the downside.  That volatility is all part of the process.  But if we get another day such as Friday within the next few weeks, I will be more confident in suggesting that we may have seen a long-term bottom as opposed to just an intermediate-term one.  The breadth action we've been seeing, and continue to see, is very typical of bear market bottoms.  On a shorter-term basis, the a/d moving average and NH/NL indicator are still in neutral territory.  The cumulative TICK indicators, however, are becoming quite overbought, especially the NYSE intraday.  Although the theoretical maximum for that indicator is around 30,000, the practical maximum is around 8,000.  The highest reading I have recorded is 7,545 which we saw in early January of this year.  At 6,632, we are now the most overbought on this indicator since May 14th, which was the peak of the most recent leg down.  I've said that these indicators are excellent at pointing out overbought conditions within downtrend (or oversold conditions within uptrend), and since we are now overbought within a clear downtrend, it makes me leery of the very short-term prospects for a continued rally.  Granted, it's only one indicator and it does give false signals like any other, but I'm paying close attention to it in here.

VIX and VXN - The VIX made another perfect reversal, making a spike high before closing down and continuing down the next day.  This type of action usually bodes well over the next two to three days.  If that happens, we will almost certainly become overbought in the volatility indices, as they will most likely pierce their lower 10% bands and have a RSI reading below 30.  What we've seen over the past couple of weeks is that rallies tend to deflate these indicators fairly quickly, so any upside continuation early next week could push these into the short-term bearish column.

COT - Due to the holiday schedule, the COT information will not be released until Monday, July 8th.  I will discuss it in the daily commentary.

Although Friday looks positive on the surface, and it may in fact bode well for the longer-term health of the market, I don't believe it was a rocket that will serve to get us out of the morass we've been wallowing in for the past several weeks.  Considering the fact that we're now overbought (from a sentiment perspective) on a very short-term basis within a clear downtrend, it's likely that we will once again head lower next week.  Looking at the intermediate-term picture, however, there's nothing to suggest that the recent lows will not hold and serve as a base from which to judge forthcoming action.  The odds do suggest, once again, that the lows will hold, and that's the assumption that I will work with going forward.

 - Jason Goepfert