May 06, 2002

sentimenTrader.com

Friday, May 10, 2002 3:30 PM CST

Please check back on Sunday for the regular weekly commentary. Due to traveling plans over the weekend, it may be a bit late, but I will send it out via email as well. All daily charts (except for the CBOE put/call ratios) HAVE been updated as of today's close. Have a great weekend.

Thursday, May 9, 2002 3:45 PM CST

Pretty much a nothing day, sentiment-wise. There are some anecdotal reports of pessimism due to the "failure" of yesterday's rally, but that's nothing unusual. With short-term sentiment so overwhelmingly bullish (as defined by STEM.MR), it would have been difficult for the markets to have much more of a rally immediately after yesterday. We've now had the almost obligatory correction (which didn't even retrace 38% of yesterday's rally in either the S&P or NDX, and on much reduced volume), so where we go from here will be telling.

Unfortunately, I don't think we have an edge either way. I can just as easily see a resumption of the rally as I can see a test of the recent lows. However, if we get the rally, I don't believe it will last, as I said yesterday. I believe we WILL test those recent lows sometime in the very near future, and how sentiment appears if that happens could help determine the next few months of market performance.

There's just not much to say tonight. We're neutral across the board on a short-term basis, and will wait and see what the next few days bring.

Wednesday, May 8, 2002 3:55 PM CST

If you read financial publications or listen to very many "professionals" tonight, you will most likely be bombarded by statements like "this rally should not be believed" or "this was a short-covering spike, it didn't mean anything". That always bothers me to no end. Try telling the guy who was short 1000 shares of DELL yesterday that today's rally didn't mean anything or that it wasn't real. The margin clerk at his brokerage house sure knows it was real. Instead of DELL, how about CSCO or MSFT or 7 out of the top 10 stocks in the NDX that was up more than 10% today. The market is what the market is - it should ALWAYS be believed and respected. Sorry, I had to get that off my chest.

Today was a positive day for the markets, period. There's not much to argue about - the advance/decline ratio was good, volume was among the highest over the past two months, and it was a broad rally (every one of the 13 broad sectors I have on my quote screens is green today, and all were up over 1%). Up volume was 450% greater than down volume. But let's take a look at the past few really "great" days in the market, with up volume beating down volume as much as today. The last four times were 12/5/01 (preceding a 50 point drop in the S&P); 3/4/02 and 3/6/02 (preceding a 60 point drop) and 4/16/02 (preceding the most recent 60 point drop). So you can see the last few times things were popping to the upside, a violent reversal soon followed.

The STEM.MR model cannot get much more negative. At a low reading of around 5%, it is extremely overbought. All components were flashing danger signals soon after the opening, but like I said in the comments, we are due for a few trend days on the upside. I've often said that when we have a protracted trend, the first reaction move out of that trend will lead to extreme readings in the other direction. Those are the times sentiment readings (short-term anyway) are the least effective.

Now for some positive news. Today's range was more than twice the average daily range (over the past 10 days). This has happened only 23 times over the past four years. Almost every single time, this incredible volatility marked an intermediate top or bottom. Obviously with the recent action, we would be looking for a bottom, so this is yet another positive development for that argument.

The II and Consensus survey figures showed a mixed picture. The II group once again became more bullish, placing them in the same position they were in a month ago when the S&P was 80 points higher. That's definitely a negative for the bullish argument. The Consensus figures, however, dropped a point to 21% bulls. Over the past 10 years, this survey has been under 22% bulls for two weeks in a row only 9 times. Here are the returns following those occurrences:

                 1wk   2wk    3wk   4wk   5wk   6wk    8wk   12wk 16wk
random       0.2     0.5     0.7     0.9     1.2     1.4     1.8     2.9     3.7
<22 2wks    0.9     0.3     2.4     2.9     2.3     4.2     5.3     4.8     5.3
% profitable 44      56      89      89      78      78      89      89      78
avg loss     -0.6    -2.7    -7.6   -6.0    -4.6    -2.3    -5.6   -12.1 -2.9%
avg gain      2.7     2.8     3.6     4.0     4.3     6.1     6.7     6.9     7.6%
min           -1.7    -6.8    -7.6   -6.0    -8.6    -4.1    -5.6    -12.1 -5.1%
max           4.7      5.3     7.3    9.2     9.9     11.1   13.3    20.8   26.2%

If you take out a signal that occurred in December of 2000, that would erase the largest losses and make the results from 8 weeks on 100% winners. The average loss row would actually become positive without that one outlier (which is so negative because of the March/April 2001 meltdown). Of course, we cannot pick and choose our signals since we don't know if this time will be an outlier too, but I think it pays to look at data with AND without the outliers to see what a "typical" result is.

I still feel that we will not go substantially higher from this point. The most likely scenario would be a short rally, then a retest of yesterday's lows. At that point, I would hope to see a pickup in bearish sentiment that would give us enough of a push to get BUY signals across our models. It doesn't HAVE to happen that way of course, and we will take market action as it comes, but historically, good intermediate bottoms have not taken off from the condition we are in.

Tuesday, May 7, 2002 3:50 PM CST

Like I said yesterday, these types of days can play havoc with sentiment indicators. The rally this morning was accompanied by high put/call ratios but low TRIN readings. That's actually quite positive, but it didn't last long as the markets sold off after the Fed announcement. It didn't affect the models much as the opposing readings canceled each other out. As I said I feared, we got a gap up opening which immediately took the STEM.MR model into neutral before we could even generate a signal. Oh well, that stuff happens. That's why we don't rely on sentiment exclusively to trade (or at least I hope you don't).

We're still at the same place we have been for some time - following the trend down until there is reason to believe it may change. All models are currently at the edge of BUY signals, but they are not there - which means I follow the trend until they get there, then I begin to anticipate a trend change. I cannot emphasize enough that price action always comes first.

The VXN is still putting in an impressive performance above the 10% band, and the longer it stays up there, the better our chances for a sustainable rally. It will only be able to stretch so far before being pulled back to the mean, so we need to keep a close watch on a possible reversal there sometime soon.

The breadth oscillators I touched on this weekend have now been relieved of their overbought condition. I would consider them neutral at this point but it will not take much to make them oversold and bullish. That's another peg that's falling into place.

We have now undercut the October lows in the S&P. The NDX did that awhile ago of course and is within a whisper of the September lows. All the technical analysis texts suggest that the NDX will at the very least challenge those September lows. If that happens in the next few days, I believe that's all it will take to push all of our models into BUY signals. Like I've said ad nauseum, that's a rare and significant occurrence. If that happens (when it happens?), I will become very aggressive on the long side.

Monday, May 6, 2002 9:41 PM CST

I noted in the weekend commentary that I thought there had to be further downside before I could justifiably turn bullish, and I hold that same opinion today. Today did just not do enough damage to make our indicators and models turn bullish. As much as I want to say that we have enough sentiment pieces in place for an intermediate bottom, we are just not there yet.

You may be wondering why there was no STEM.MR model BUY signal today, and the reason is because there was no reversal in the index to trigger it. We finished the day with the model at a high reading of 92.7 and that high print was at the close of trading. The worst case scenario for this model would be a gap up opening tomorrow, which would more than likely spike this model down to neutral before a signal has enough time to be generated. Hopefully there was enough pessimism going into the close that it will prompt emotional selling tomorrow morning. Unfortunately (or fortunately depending on how it plays out), we have some economic numbers before and right after the opening, then of course the FOMC meeting in the afternoon. That stuff is hard to game, and could play havoc with our indicators.

The Composite model I mentioned in the weekend commentary is still in neutral, although it is getting close to BUY territory. Just for your information, this model has given seven signals in the past year (and ONLY seven). Here they are: BUY on 4/10/01; SELL on 5/23/01; BUY on 9/19/01; BUY on 10/31/01; SELL on 12/31/01; BUY on 2/8/02; and SELL on 3/13/02. If you pull those points up on a chart, you can see that they have come within a few days (before or after) of an intermediate market turn. Unfortunately, with the COT figures where they are, it will be difficult to get a BUY signal this week unless we get some low a/d readings and/or high VIX readings and/or high put/call readings. The other components are in place - these three need to step up in order to get a signal out of this model.

Unfortunately, there really isn't much to add today that wasn't discussed in the weekend commentary. The VXN has popped significantly out of its 10% band, the TRIN (both NYSE and Nasdaq) is wildly oversold, the CBOE spread is still bullish, the put/call ratios are bullish (but not extreme), and the breadth oscillators are overbought (but that did actually correct quite a bit today). Volume was NOT what is typically seen at bottoms. I wish I had more to say, and when I sat down to write today's commentary, I really thought there would be more to add. Simply put, today's action just gave us more of the same, and there is nothing actionable at this point. Now, if we sell off heavily tomorrow due to the economic numbers or the FOMC release or whatever, that will change the story. If we rally (which would be the worst possible scenario), we will most likely be left behind.

Someone asked me not too long ago what it means when the models are neutral, and why weren't they on SELL signals for this decline. What it means to me is that whatever trend is in place is the dominant force when a model is in neutral. For example, even though the STEM model has been neutral since late March, the last signal was SELL and that is the trend I stay with even though the model components are neutral. Price action ALWAYS comes first.