sentimenTrader.com

All rights reserved.

Sunday, July 21, 2002

The action over the past week has made it abundantly clear to me that we should not prepare for an intermediate-term bottom, but instead a long-term one lasting several months or longer.  The readings we have seen and continue to see do not occur at intermediate-term bottoms with great regularity - instead they occur at long-term market bottoms.  Most of the specific indicator discussions below will be short - not because they're not significant, but because they are at historical extremes and there's just not much more to say than that.

Take a look at the following  indicators since 1943 (I understand this does not include the 1920s-1930s market condition, but then again we are not in a depression).  I looked at occurrences where the market was down 20% or more from the past high and subsequently rallied at least 15% from the low over the next three months.  That is my definition of a long-term market bottom.  Some will argue that a 20% drop is not enough and some will argue that a 15% rally in three months is too much or not enough time, but I'm not trying to find a consensus.  These readings are as of the day (or the week in the case of the weekly readings) of the bottom.

We are now at a point that mirrors those of past major bottoms.  The advance/decline ratio is already extremely oversold on a historical basis, and the 10-day average will be dropping a positive number on Monday, so another large negative a/d reading will give us one of the most oversold 10-day averages in 60 years.  We have also seen down volume overwhelm up volume over the most recent two weeks, in a scale that is historically meaningful.  The number of new lows we are seeing, although extremely high, does not quite meet the historical average.  We would have to see something on the scale of 500+ new lows on the NYSE to be "average" as a percentage of issues traded.  With 403 recorded on Friday, 500+ is not far off.  I am sure there are some that will argue that because of the amount of money that has been pumped into the market and the influence of mutual fund redemption and hedge fund liquidations, that historical numbers are irrelevant.  That could be.  But trying to quantify that and determine the impact on the market is impossible, so I don't see the sense in trying.  My opinion is that I will let history be the guide, and not try to guess at why the market is doing what it's doing.

Indicator Summary:

BULLISH

NYSE MEMBERS REPORT - A bullish drop here as the specialist short ratio dropped 5% to 34%.  An argument could be made that as the public becomes more comfortable with the shorting strategy, this ratio will show a secular decline.  Even so, this is still a significant reading, as the five-year average of the ratio is 45% with a standard deviation of 4%.  That means that we are close to being 3 standard deviations from the average.  For those of you not familiar with standard deviations, the essence is that 99% of all readings should be contained within 3 deviations if the distribution shows a normal bell curve (which this data does although with a very slight bias to higher numbers).  Anyway, you can forget the statistical mumbo-jumbo - the brass tacks is that the public is shorting this decline extremely heavily, and they have been uncharacteristically accurate so far.  This is an anomaly, and it will not last much longer.

COMPOSITE MODEL - There is nothing new to say here.  The model continues to set new highs as the market sells off.  In the past five years (which seems quaint now) we have never seen the confluence of negativity we are seeing now, both short and long-term.

SENTIMENT SURVEYS - The confluence of negativity extends to the popular sentiment surveys as well, as the four majors are displaying an historical level of pessimism.  In the fifteen years that I have data for all surveys, there has NEVER been a period of greater pessimism than now, in the way that I measure it.  Sure, you can take out an individual survey and say that it's not the most bearish it's ever been, but when you look at the whole picture, we are in unprecedented territory. 

AIM - We would have the highest reading in the fifteen-year history of this model if it wasn't for one reading in April 1994.  The only reason it's a bit higher than this week's reading is because the II survey has more weight than the other surveys and it was showing more bearishness than it is now.

STEM - We are of course extreme here as well as the 20-period moving average is now sitting at 98 while the 50-period is at 90.  These are both flirting with all-time highs and again suggests that the selling we are seeing is unsustainable.

STEM.MR  - Sitting at 99, short-term pessimism is about at the same level as the longer-term periods.

BREADTH RATIOS - Most of what needs to be said here was already discussed above.  The cumulative TICK indicators are also oversold, with the intraday NYSE and Nasdaq indicators coming close to historically oversold during the early part of trading on Friday.  It goes without saying that the selling has been ferocious - the positive side of that is that it is occurring at an unsustainable pace.

CBOE RATIOS -  The OEX open interest ratio set a new low on Friday, hitting .52.  This is one of the most accurate indicators I follow, and the fact that it has made a new low is significant to me.  I can pretty much guarantee that expiration will have a large effect on the numbers tomorrow, and it will likely shoot back above 1.0 for a day or two.  That's typical and doesn't diminish the significance of the reading.  The open interest - put/call spread reached -.62 on Friday, which is just a tad below September's crash reading.   The OEX p/c - total p/c spread is currently tied for the lowest reading in nearly 20 years, as the "big money" OEX traders continue to bet on the upside while the "small money" equity traders bet on further downside.

VIX and VXN - Bullish again here, as the VIX made another high above the 10% threshold I prefer to measure it by.  It did close at the highs, so we would have to wait for something of a reversal before considering this all-out bullish, but it is now in the right position for that reversal to occur.  The VXN made a puzzling decline on Friday, which could certainly be due to expiration.  Although technology shares have been outperforming (in the sense of holding the recent lows), it is still unusual for the VXN to make such a drop.

BEARISH

COT - This is not what we would want to see near a market bottom, as the commercials added significantly to their net short position while the small specs added a similar amount to their net long positions.  We should see the exact OPPOSITE of that as a reflection of optimism from the smart money and pessimism from the dumb money.  The numbers only reflect positions through Tuesday, so they could be changed drastically from then, but forecasting these positions are virtually impossible.  Although the stochastics are still supportive of something of a rally in here, the longer-term picture remains very bearish.  The market just doesn't make it easy by giving us a complete picture one way or the other, so we'll just have to balance this indicator with the others.

NEUTRAL

TRIN - Although not historically extreme, the 10-day TRIN remains elevated at 1.38.  You can see from the table above that past bottoms have averaged 10-day readings of 1.37, so we're right in line here.  We will be dropping a string of 3 high readings early next week, so it's unlikely the 10-day average will continue to increase beyond the popular 1.50 threshold.

Cleary I have been wrong by turning bullish as early as I did.  Although the short-term moves have been profitable when short-term extremes were reached in conjunction with the longer-term situation, the overall mood of the market has reached a point not easily forecast using sentiment alone.  The odds of this decline were small from a sentiment perspective, unless one was looking for a bottom of historic proportions.  Sentiment now suggests that we may have, indeed, reached that point.  A large gap down opening would be very welcome tomorrow, although once again the market usually doesn't make things so pat and easy.  I fully expect the volatility to continue (the 10-day average range on the S&P has now reached the 30 level seen at past recent bottoms), and I again will be expecting an upside bias.  That has been wrong, but as we go along it becomes more and more likely (by the way, if, once the market rallies, I suggest even once that I have been right by forecasting the rally, someone please take away my computer).

 - Jason Goepfert