http://www.sentimentrader.com

Wednesday, March 12, 2003  9:35 PM EST

I've seen several reports about how positive today's action was.  We were down a fair amount by mid-morning, then came back and closed at the highs on the greatest volume in over a month.  For those of you who follow candlestick charts, you've no doubt noticed that today's action formed a bullish "hammer" formation on the daily charts.  For those who are not familiar with this charting method, a "hammer" is formed when we have action like today's - an open and close near the highs and a long lower body, showing the market hit a low and subsequently recovered.  When coming at the end of a decline, it can often signal a bottom.  The problem is, these patterns can give frequent false signals, especially in such emotionally charged times as we have now. 

I went back over the past few years and looked at several breadth measurements at major lows, then at other "hammers" preceding those lows and compared both readings to our current situation.  Here is the comparison:

We can see here that we compare much more favorably to "Other Hammers" (all of which ultimately failed) than we do "Major Lows".  We're not near ANY of the average readings seen at the other major lows.  In fact, we're not even as extreme as the average failed hammer except in the number of new lows.

I've received a few requests to reprise the "bottom spotter" table I last highlighted on October 9th of last year.  This table looks at (mainly) breadth readings seen at an average intermediate-term low formed over the past 40 years.  Here it is:

We're a substantial amount away from these past readings.  So, we're nowhere close to breadth readings seen at almost any other major low, yet compare relatively well to several failed hammer formations seen during the past few years.  Doesn't sound like we just hit the bottom to me.

There are always two sides to the story, so let me present something that is giving off bullish implications.  The VIX Fear Premium is an indicator we track which shows how the VIX (CBOE Volatility Index, or implied volatility on OEX options) compares to recent historical volatility.  When the Fear Premium is high, it shows there is greater uncertainty priced into OEX options than its historical volatility would warrant.  This type of fear is usually a precursor to a good intermediate-term market low.  Conversely, when the Fear Premium is low (or especially, negative), then it shows that options traders are comfortable with current price action and do not believe the market will make a major move anytime soon.  This complacency is often seen near market peaks.  Currently, the Fear Premium is the highest it has been since July, as today's high in the VIX was more than 20 points greater than the average volatility the OEX has experienced during the past 30 days.  The indicator posted to the site is actually a 10-day moving average of the daily readings, so today's high reading is watered down a bit.  The following two graphs show every other occurrence of the daily reading exceeding the 20 point level since 1986.  The vertical green bars represent the days of the high Fear Premium:

We can see that such spikes in uncertainty almost invariably lead to some sort of market rebound soon thereafter.  Of course, we are in the midst of one of the most vicious bear markets ever, and that must be taken into account, but you can see that the two other times the Fear Premium reached such an extreme during this bear was in September 2001 and July 2002 - major low points.

Undoubtedly, if you pay even casual attention to sentiment, you know that the widely-watched VIX has not shown the type of readings we often see near low points.  That's true, as the VIX hit at least 50 during the lows in 9/01, 7/02 and 10/02.  However, the historical volatility near those lows was at least 30%, and over 40% on two of the three occasions.  This means that the daily changes we were seeing at the time projected that the OEX would rise or fall by at least 30% over the course of a year if it continued.  Our current historical volatility is barely even 20% - a low figure for current times and the lowest we've seen since the market high in March/April 2002.  So even though the VIX is relatively subdued at 40%, it is nearly double the historical volatility of 20%.  This tells us that although the market itself has been lacking volatility, options traders are pricing into their trades a future volatility premium - which is something they do during uncertain times.  This type of activity is usually resolved by the implied volatility "coming in", or declining, as market prices rise.

Our current situation volatility-wise is very unusual.  Volume and volatility are at levels normally seen after significant market advances, certainly not declines as severe as we've seen.  Obviously, the geopolitical and economic question marks are having a huge impact on trading activity, and that is suppressing our volatility as traders do not want to take positions ahead of a situation that is impossible to game.

Consider what happened during Desert Storm:

The tan line is the VIX Fear Premium, the green line is actual historical volatility, and the black line is the OEX (S&P 100).  We can see that after Iraq invaded Kuwait, historical volatility rose dramatically as the market sold off.  However, as the deadline for war approached, volatility in the market dropped off drastically, as you can see from the sharp decline in the green line. 

What we're going through now certainly has precedent.  Traders are afraid to take committed positions - long or short - ahead of an uncertain event.  When there is some clarity as to what the outcome will be (or even what the path to that outcome is going to be), I suspect volatility will rise and volume will come back into the market.  Can we bottom before that happens?  Sure.  Is it likely?  Nope.  Considering the tables above comparing current breadth to what has been seen previously, the most clear precedents to our recent action lead to further downside before a meaningful rise took place.  I think that's the most likely scenario this time around as well.  Therefore, I would not chase the upside here, as I do not believe the risk/reward favors it for anything longer than a short-term trade.

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 10, 2003  9:53 PM EST

I have received a lot of requests to discuss the NYSE TRIN tonight, so here's my two cents.  Today's closing reading of 5.50 was certainly high, in fact it was the highest reading in 40 years that was not related to the crash in 1987.  The reason the TRIN was so abnormally high today was the concentrated volume going into issues down on the day.  The up volume ratio was one of the lowest we've seen in 40 years, while the up issues ratio was not even close - this greatly skews the TRIN.  The primary reason for this skew was the extremely heavy volume in issues such as FNM and CE.  If we play "what-if" and switch only those two stocks to the plus column for today (meaning they would be counted as advancing issues instead of declining issues), then today's TRIN would have been almost half of what it actually was.  Alternatively, we could adjust the up issues ratio (i.e. advance/decline) to be more in line with the down volume by increasing the number of issues that closed lower on the day.  By doing that, the TRIN drops to 1.5 - again, quite a difference from what we actually saw.  The purpose of this exercise is not to suggest we adjust today's data to reflect possible "outliers".  It is simply to make you cognizant of the fact that today's reading was skewed because of very heavy volume in a few issues.  Ironically, if the selling had been more broad-based, the TRIN would have been significantly lower.

Regardless of the reason, today's daily reading has pushed the 10-day average close to 1.70, also one of the highest readings in 40 years.  This type of 10-day reading has often lead to at least a short-term (2-3 day) upside correction, even during this most vicious bear market.  The table below shows the percentage of time the S&P 500 closed higher the given number of days after a 10-day TRIN reading of 1.60 or higher since the bear market began:

We can see that the short-term positive impact of this type of lopsided selling evaporates after 3 days, which is when the downtrend most often resumes its force.  5 days after such readings, the market actually underperforms a random return.

The TRIN on the S&P 500 today was a low 0.60 and on the Nasdaq 100 it was 0.54.  However, for such a limited universe of stocks, I think the TRIN can be a misleading indicator.  I much prefer our Down Pressure gauge, which measures the MAGNITUDE of down issues, instead of just whether the issues were up or down like the TRIN.  For example, if MSFT was down 3 points today, it would have a greater effect on the Down Pressure gauge than it would the TRIN, which I believe is appropriate.  The Down Pressure readings were quite extreme in both indexes today, as the S&P recorded a reading of 96% while the NDX registered a reading of 90%.  The Down Pressure indicators posted to the site are a three-day average of these daily readings, so they did not move up as much as the daily readings would suggest.  Another day of selling should push them firmly into oversold territory.

Today's advance/decline reading of -1644 issues (meaning 1,644 more stocks declined than rose) is the most extreme reading since January 27th, and October 9th before that.  As a percentage of total stocks traded on the NYSE, it was 48%, which is shy of most other major lows seen during the bear market.  However, it was large enough to expect at least a slight pause in the selling over the short-term, as it has been a rare occurrence for the market to continue down unabated after seeing such broad selling pressure.  Each time the market DID NOT stage at least a small rebound after such a reading, we cascaded down into a panic selling climax.  If we DO rally a bit from here, there is a chance that it could carry further and develop into a tradable low.  Therefore, even though I think this term is way overused, the coming days should prove critical to our intermediate-term prospects.  This may be something to watch for over the coming days - if we don't relieve this oversold condition even a small amount over the next few days, there is a strong precedence that it could lead to a waterfall-type decline.  In order to reasonably expect a stop to the selling pressure, we should watch for an a/d reading of -2000 or more, or over 60% of total issues.  If you see a figure greater than 60%, there's a very good chance we are near a tradable low and longer-term traders and investors should consider building long positions.

The lowrisk.com sentiment survey came out with another drop in bullishness, to one of the lowest overall bullish levels seen in the history of the survey (about 6 years).  Each past reading comparable to the current one lead to a market low within two weeks.  I expect the more well-known surveys released later in the week to also show a drop in bullishness, although I seriously doubt it will be to a similar degree.

Our shortest-term model, STEM.MR, closed at 63% today.  This extreme of a reading has only been seen four times in the past two years - the low in September 2001, the low in July 2002, the low in October 2002 and at the end of January of this year.  In 9/01, the model hit this level on the day of the low.  In July, it hit this level on the 15th, which lead to a small bounce, then again on the 22nd-24th, which was the ultimate low.  In October, it reached 64% on the 4th which lead to a choppy move lower until the final low on the 10th.  The January reading of this year was unusual in that it did not ultimately lead to a significant move higher, instead just a week of choppy up and down trading.  If history is a guide, then we should look at lower prices from this point forward as opportunities to possibly establish long positions, at least for short-term traders.

We are at a crossroads moment in the market.  If a bounce from here catches hold, it has a chance to develop into something stronger, though a tight leash should be kept on any long positions taken.  If we instead keep moving lower from this point, without a significant bounce, then there is a good chance that it will transform into a panic-selling atmosphere.  Of course, personal preference will dictate your trading style, and if buying weakness here does not seem especially high-odds to you, then please don't do it.  We have not yet reached a point where I would suggest you take long positions even if it makes you uncomfortable.  I think there's a good shot we could reach that point sometime in the next two weeks or so, but we're not there yet.  For short-term traders, I think focusing on the long side here (good setups only) presents a better risk/reward scenario than the short side, with the understanding that if we continue lower, it could get ugly as I said above.  We have not yet reached a point where aggression is warranted from either 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.