Apr 29, 2002

sentimenTrader.com

Friday, May 3, 2002 3:35 PM CST

Please check back Sunday for the regular weekly commentary. All daily charts HAVE been updated. Have a good weekend.

Thursday, May 2, 2002 4:55 PM CST

So far the scenario is playing out as history would suggest if we are indeed forming an intermediate-term bottom in this area. I suggested yesterday that we should see a day or two of consolidation before the rally resumes, and today did not do much to change that. The one kink is technology, but I'll talk more about that in a bit.

At this point, it looks as though we can still go higher without too much trouble from a sentiment perspective. Our breadth oscillators are now very close to overbought, which is worrisome, but some other developments are very positive, such as a continued high put/call ratio and low open interest put/call ratio. Confirming that is the TOPM model, which dropped again today to .43 (remember that below .40 is outright bullish) - recall that a low number means puts are being bid more aggressively than calls (as market makers push up the price to meet retail demand), which from a contrarian standpoint is bullish. Our other indicators are neutral to slightly oversold, so there's not much to be gleaned there.

I said above that the one kink is technology. It appears as though the semiconductor group within technology is leading the market at this point, and right now that group is at a critical juncture. That index is sitting at 500, which has served as some form of support no less than 7 times over the past three years. Looking at the component stocks, most are facing the same situation. If the support holds, it could finally release the broader indexes to make some decent gains. Technology makes up about 19.3% of the S&P 500 index, which is the largest group, followed by financials (at 19.2%) and healthcare (at 14%). Financials and healthcare are doing OK, but without technology the broader market will not make significant headway.

As of today the Nasdaq TRIN 10-day moving average is at 2.13. This is the highest number since the April 2000 and April 2001 bottoms, when that index subsequently rallied about 25% and 45% respectively. There has been some very serious selling in these issues, but it cannot last much longer - at this pace, the NDX will be at zero in just over 30 days. It would be best if we saw the VXN react a little more strongly to the decline, but as I suggested previously, I think that indicator may be suppressed for other reasons.

At this point, I am still net neutral. After seeing the COT and other sentiment surveys this weekend, I may change that, but I would like to see something of a washout session or two before I would be comfortable with anything more than a trade to the long side. If the SOX breaks its support, we may just get that. A washout in technology would probably be enough of a catalyst to get things going on the upside.

Late note (7:30 PM CST):
I just received the CBOE put/call and open interest put/call ratios. I've said before that when the put/call ratio is greater than the open interest put/call ratio, it is a significant sign of extreme pessimism. I haven't seen this mentioned anywhere before, so I did some research on the significance. Following are the percentage returns over given time frames when the put/call ratio exceeded the open interest ratio (this is for the S&P cash index only). There were 29 days where this event happened (out of 585 days). The data used for the study is daily data from January 3, 2000 through yesterday:

                    1day   2day   3day   5day 10day   15day   20day   30day
Random        .0      -.1       -.1      -.2      -.5        -.7        -1.0      -1.5
P/C>OI         -.1     .1        -.2       .6       1.9       2.7        2.8       3.5
%profitable    41      41       48      52       52        63         74        70

Many of these occurrences were repetitive. Meaning, the put/call ratio was greater than the OI ratio for several days in a row. That throws off the data, so I looked at the data again, this time using only the last day of the series as a data point. There were 12 such distinct occurrences:

                    1day   2day   3day   5day 10day   15day   20day   30day
Random        .0      -.1       -.1      -.2      -.5        -.7        -1.0      -1.5
P/C>OI         1.0     1.2      1.1     2.7      4.1       4.1        3.6       4.3
%profitable    67      58       67      83       75        75         83        67

What this tells us is that starting tomorrow, we should watch these two ratios very carefully. If and when the data correct - meaning, when the put/call ratio drops below the open interest ratio - history would suggest that there is a statistically significant chance we will have a decent rally over the coming month.

Wednesday, May 1, 2002 5:01 PM CST

An interesting day today to be sure, as the opening optimism was palpable, then changed to disappointment early (probably due to weaker than expected economic numbers), then changed once again mid-morning as the markets took off and never looked back.

We just missed a STEM.MR SELL signal yesterday afternoon and just missed a BUY signal right at the lows of the day today. We couldn't have been closer - just a percentage point or two, but the extremes just weren't there. We ended the day with STEM.MR right where is was at yesterday's close - at the cusp of a SELL indication.

The longer-term STEM model fell hard today, with all components (except one - see below) registering readings in the bottom 10% of all readings over the past three years. Over the next two days, we'll be dropping several high readings from the 50-period moving average of the index (which is our signal line), so provided we have more of the same tomorrow, we could conceivably get a SELL signal this week.

That one exception was the put/call ratio. It jumped right away this morning and actually went up as the market rallied, giving the suggestion that options traders don't believe this rally has any lasting power. It's impossible to know whether they were purchasing puts for outright speculation or portfolio protection, but either way it suggests they are sceptical of this rally. To confirm this, my TOPM model (explained in the weekend commentary) dropped a few points, which means puts were being bid more aggressively than calls - rather unusual with the magnitude of the rally off the lows we saw.

Our breadth oscillators have now come out of their oversold condition and are neutral to slightly overbought at this point, although there is quite a bit more room for them to go before we reach an extreme.

The VIX and VXN acted normally, and have had a pretty violent reversal from the poke out of their 10% envelope. Nothing really to comment on here until they get to the other end of their trading range.

One very interesting development today was the release of two of the sentiment surveys. The II survey didn't show much change once again (bulls dipped a bit but bears did too, same pattern as the past three months), but the Consensus Inc survey showed a sharp drop in bullish percentage to 22 (from 29). Over the past 10 years, we have seen this level or lower only 33 times.
Following is a breakdown of the percentage returns following an occurrence of 22% or less bullishness, compared to random returns over 1 - 16 weeks.

For the S&P...
                    1wk   2wk   3wk   4wk 5wk    6wk    8wk    12wk   16wk
Random        .2      .5      .7       .9      1.2     1.4      1.8      2.9      3.7
< 22%           .8     1.2    1.6      2.6    2.9     3.2      3.8      3.2      4.6

For the Nasdaq...
                     1wk   2wk   3wk   4wk 5wk    6wk    8wk    12wk   16wk
Random        .3      .7       1.0      1.3    1.7     2.0     2.7      4.3      5.6
< 22%           .2      .8       2.2      4.0    4.0     3.9     5.3      5.4      9.1

You can see that generally, this level has preceded a time of outperformance for both the market in general and for technology shares. Of course, this is over the past decade, which we all know was one of the greatest bulls markets in history. But there were times, like in 1991, 1994 and the past two years, where these figures held up, so I think it may be safe to say that they are significant. One other consideration is that only about half the time, a low % of bulls in this survey was confirmed by a low % in the II survey. Quite a few times, the II % bulls was still over 50%, like we have today. However, the biggest gains did come when both surveys showed a low level of bullishness. Also, many times this survey double-dipped - meaning it went below 22, then back above, then below once again as the market tested or exceeded the previous lows.

If we use history as a guide, it would give us a target of about 1110 on the S&P and about 1730 on the Nasdaq Composite (NOT the Nasdaq 100!) in the next four weeks. However, I am not a fan of using projections or predictions, because we never know if this time will be different. It does give us a heads-up, though, as to what has happened before when conditions were similar.

Using history and our current situation as a guide, it leads me to believe that the most likely scenario over the next few weeks would be a small rally to continue, then a restest or slight overrrun of the lows just achieved, then a fairly significant rally from there. Like I said, I'm no fan of predictions, but it gives us a framework to work with.

Tuesday, April 30, 2002 4:55 PM CST

We got the rally but the pattern remains the same. The only thing that was different today was the magnitude of the rally off of the STEM.MR BUY indication. During the previous days, the rally was for a little over 5 points and today it was about twice that. But when the model went to neutral, there was not a whole lot of punch left in the bowl. I truly thought after the morning rally that we would have a trend day to the upside, with a slow, constant upward pressure, like the reverse of the past month. Instead we stalled mid-morning then sold off a bit into the close. Actually, that could be a positive as it leaves more powder to be burned later in the week.

At this point, we are neutral across the board. We still have a mildly oversold longer-term condition, but vey short-term we are already overbought. However, that will happen when we come out of a prolonged trend - after the heavy selling of the past month, it doesn't take much to make our indicators look overbought. So, at this point, I would not put a whole lot of weight on the fact that the TRIN and put/call numbers are so low today. The VIX also dropped hard, forming a perfect reversal from a peek outside the 10% band. What I would like to see (and what I think is the most likely scenario) is a day or two of consolidation, then another rally attempt. We will have to wait until later in the week to see if this has the potential to be an intermediate-term rally.

If you've been short during this most recent decline, I hope you see the value in apprising sentiment before major economic numbers or news events are scheduled to be released. With sentiment as negative as it was going into this morning, it should not take much more than mediocre numbers to ignite a short-covering rally (or even real buying). You have to look at the risk-reward of staying in a position that has potentially limited downside (due to highly negative sentiment) but good chances of reversing to the upside.

Monday, April 29, 2002 4:10 PM CST

And the pattern repeats. Seemingly every day now, we've been getting a STEM.MR BUY signal which is good for a few points, and as soon as it goes to neutral, the markets head south. Like I said this weekend, though, it cannot and will not last much longer. The STEM.MR model closed at 95.7, the high of the day - the past two days have seen readings not seen since February 22, right before the last significant rally.

The VIX is now about 18% above its 10-day moving average, which places it in the top 4% of all readings over the past three years. That's significant, although as we saw in September, it can go much higher. Watch for a reversal, as that can tip us off to at least a short-term rally in the market.

The TRIN is also extremely oversold, and like I said this weekend, we are now very close to having the 10-day moving average hit 1.50 once again. If we get a daily reading of 1.50 or more tomorrow, the 10-day MA will indeed exceed 1.50. To have two 10-day readings over 1.50 within two weeks of each other is a notable event.

The Advance/Decline and NH/NL breadth ratios can now be considered oversold and bullish. We will be dropping a very large positive reading tomorrow on the 10-day moving average of the A/D line, which will drop that average quite a bit into oversold territory (unless we get a large positive reading to cancel it out tomorrow of course). In fact, if we get an A/D reading of zero tomorrow, the 10-day moving average will be in the bottom 15% of all readings over the past two years - quite oversold.

One aberration is the put/call ratio. It did not get to a high level today like one would expect, and in fact call buying was rampant this morning during the tiny rally we had off the open. However, as the day wore on and the market dropped, the put/call ratio rose. Just no extremes here. The TOPM that I mentioned this weekend bears that out, as the index rose back above .40 today (to .44), suggesting increased demand for call options. By the way, that May 570 put I talked about on Sunday is now bid at $44.00.

I feel pretty comfortable saying that I think we're very close to a short-term bottom (of a few days to a week). Longer-term, I have to say I'm still neutral to slightly bullish, as we just don't have the confluence of extremes I prefer to see at intermediate bottoms.