http://www.sentimentrader.com

Wednesday, March 26, 2003  8:30 PM EST

Without a doubt, we'll see ample press over the coming days about the fact that many mutual and hedge funds will be closing their quarters (three-month period) at the end of this month.  Because quarterly reports are widely disseminated, and carry influence for gaining new assets and keeping existing ones, the theory is that these funds will goose the market in the final days of the month, looking to add a bit more to their performance and quite possibly generating more of a bonus for themselves.  I've never really bought into this theory, as it just doesn't prove consistent.  I took another look today, just to make sure.

The tables below look at various scenarios going into quarter-end, up to the third day prior to the end of the month, to see how the rest of the month (the final few days) played out.  For our current situation, this would mean the period covering tomorrow through Monday.  I looked at situations where the market was up going into tomorrow, where it was down, where it was above or below the 50-day and 200-day moving averages, and a combination of the above.  I also looked at ALL quarter-ends, as well as just MARCH quarter-ends.

The S&P 500 from 1950-2003 was used for purposes of the study.  The "Avg Return" and "% Pos" columns show how the market performed in the final few days of the month.  The yellow highlighted cells show instances that match our current situation.

Let's take a look at the ends of March first:

Doesn't look very impressive, does it?  No matter how the market was in relation to the previous month or its longer-term moving averages, the market went DOWN into the end of the month the majority of the time.  When we look at situations similar to now, where the market is higher than where it closed last month, and the S&P is above its 50-day moving average but below the 200-day, then the average return over the next few days is a lackluster minus 0.28%, and it closed higher only 43% of the time.  Since the bear market began, two out of last three March quarter-ends have ended badly.  What this suggests to us now is that taken in a vacuum, we should possibly expect lower prices, and not significantly higher ones, between tomorrow and Monday.

Perhaps it's just March.  Let's look at all quarter-ends (March, June,  September, December) put together:

The results here are more constructive, particularly when the market is up, and when it is above both its 50-day and 200-day moving averages.  Obviously, that suggests we were in uptrends, and the trend usually continued the last few days of the month during each quarter-end.  However, once again, situations similar to now do NOT generate positive results, as the average return and percentage of time the market was positive were less than random, and less than even.  So far during the bear market, quarter-ends have been about even, both in average return and percentage positive, although four out of the last five quarter-ends have ended badly.

The purpose of this exercise was not to suggest that we should short the market because it indicates that we are more likely to go down than up.  The purpose was to test the commonly-given excuse by market reporters that "window-dressing", or goosing the market higher during the last few days of the month, is the likely reason for any rise.  In fact, that would be an UNLIKELY reason for the market to rise, as it usually doesn't when we have situations such as now.  Of course, the war factor throws everything into question, but taken in a vacuum, you should not be buying because you think quarter-end means we have to go higher.

The Investor's Intelligence survey from Chartcraft came out today with another rise in bullishness.  I've been asked to comment on this a few times, as people have suggested that it shows that traders (or at least newsletter writers, which is the population of this survey) have been too eager to jump on this rally.  In fact, the rise in bullishness was on a par with what we usually see after a rise in the market of such magnitude, so I don't think it's any more bearish than usual.  Of course, the survey never really showed the amount of bearishness usually seen at intermediate-term lows in the first place, so this group of market pundits never truly shed their bullish bias.

The choppy action over the past two days has been enough to work off the extremes in most of our shortest-term measures, so we're fairly neutral in the short-term.  Longer-term, we became quite stretched in some of our breadth measurements, but they will be dropping positive readings for the next two weeks, so we likely won't become any more overbought as far as they are concerned, unless the market really rips higher.  The typical market reaction after such overbought readings are reached in these indicators is a choppy, two-sided market, so that argues strongly that we'll see volatility contract over the coming days, with more action like we've seen the past two days.  That should also work off some of the wild speculation we saw last week, as I've detailed in other commentaries this week.  We'll have spikes higher and spikes lower, but bereft of significant war news, we could be relatively unchanged a week or more from now.  It may be beneficial to pay more attention to oscillator-type indicators during a time like this, and back off on using breakout or breakdown kinds of strategies.  It's obvious to anyone who watches the market that we continue to be bound by the war - any and all indicators will be trumped by significant developments on that front.  So, I believe that the most successful strategy during these times will be short-term trades, small position sizes, and not chasing breakouts in either direction (unless we have extraordinary war news that changes the face of the conflict).  You can see from the model portfolios and disclosures at the ends of my commentaries that I have not been doing much trading in this war environment, as I've been practicing what I'm preaching and keeping things light and very short-term.  I don't see an end to that just yet.

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.

Monday, March 24, 2003  7:31 PM EST

As I mentioned this weekend, we finally reached a daily overbought status late last week.  All it took was less-than-spectacular developments on the war front to trigger the lopsided selling we saw today.

Speaking of lopsided selling, we once again saw an abnormally high TRIN reading today.  Today's closing reading of 4.64 was the sixth-highest in 40 years, but four of those higher readings were directly related to the 1987 crash.  The other higher reading was the 5.50 we saw two weeks ago.  Today's action compared to that day highlights a peculiar aspect of the TRIN, and I think it's important to understand this if you ever use the indicator.  The TRIN is calculated by the following formula:

(advancing issues / declining issues) / (up volume / down volume)

Two weeks ago, we had almost the exact same up/down volume pattern that we had today, yet the TRIN was almost a full point higher.  Therefore, the only possible explanation is that the advancing/declining issues component was different than today's which is indeed the case.  Today's a/d reading was -1825 (meaning 1825 more issues declined than rose) compared to -1644 two weeks ago.  So a difference of 181 issues, less than 5% of total issues traded on the NYSE, made quite a large impact on the TRIN.  This isn't exactly correct, because the up/down volume figures are slightly different, but you get the idea.  Therefore, if the selling pressure is more broad-based, and more issues decline on the day, it can actually cause the TRIN to DECREASE.  Typically, when traders see a high TRIN reading they automatically think the selling was severe and indiscriminate.  In fact, that's not the case - many high TRIN readings are caused by heavy selling in just a few issues.  That's why so many panic-type lows do not have extraordinarily high TRIN readings.  The selling is indiscriminate, thus causing the a/d to widen, which does not normally allow the TRIN to reach extreme levels.

Today's high TRIN does not make me more bullish from a contrarian standpoint.  In fact, it's the exact opposite situation of one week ago, where we recorded a very low TRIN reading.  At that time, I said such a reading did NOT indicate an overbought situation unless we saw such low readings over several days.  Conversely, today's reading does not indicate an oversold situation - it simply shows that we saw heavy selling pressure, which is bearish if anything, and it would take several more days of such action before it could be considered bullish from a contrarian point of view.

Adding to the (inverse) comparison to last Monday's buying spurt, the Down Pressure reading today on the S&P 500 was 98%, compared to the 17th's reading of 2%.  Today, the total points gained and lost by the components of that index were almost exactly opposite:

While last week's low reading lead to further upside, today's high reading doesn't necessarily mean the opposite, though it seems likely.  By the time we got that low reading last week, the 3-day average was already overbought.  This time around, we got the high reading while the average was still overbought, suggesting that there's more room to go before we would have seen the type of selling pressure that is usually unsustainable.

Today's decline did serve to work off the overbought condition of our shortest-term indicators.  The intraday cumulative TICKs on the NYSE and NDX have gone from overbought to close to oversold, and the price oscillators have done the same.  When combined with a rise in the VIX and increased put volume, it was enough to push the STEM.MR model from overbought territory on Friday all the way to oversold as of the close today.  This is only the second time in two years the model has traveled such a great distance in such a short period of time.  The other instance was 8/28/02, which lead to a small rebound over the next two days before the downtrend resumed in force.

The rampant speculation seen last week (see this weekend's note) will take some time to unwind.  While today seemed vicious, it wasn't particularly notable in terms of volume, and traders' reactions were relatively ho-hum from what I can gather.  Everyone knew the party had to take a breather at some point.  I do believe we are more likely to see further downside over the coming days, although I expect to see something of a bounce after today's large point decline (especially the closer we get to the widely-watched 850 level on the S&P).  Any bounce we do get may serve as a good shorting opportunity for short-term traders.  As I said this weekend, I am finding it more and more likely that the recent low may serve as a low of some import, so I would look for a further retracement to possibly establish some longer-term long positions.  It depends on how things look when (if) we get there, and war developments can change things on a dime, but that's what I'm going with for now.

- 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.