sentimenTrader.com

All rights reserved.

Thursday, July 18, 2002  10:50 PM EST

Tonight's comment will be abbreviated.  All charts have now been posted to the site as of today's close.

And the beat goes on.  Sentiment has just not been working well over the past two months, as you can plainly see from the wretched signals the longer-term models have given.  I guess the string of several years of consistently profitable signals has to end sometime.  This will happen at times, of course.  1997 was the most recent example of the continuous failure of most sentiment indicators.  I always follow price and never plunge in just because the indicators say it's time to buy, but watching this day after day is becoming disheartening.

There is really nothing new to go over tonight.  We're in pretty much the same situation as yesterday, although the STEM.MR model is now flashing "extreme oversold" with a closing reading of 98.  As I've said ad naseum, when all four models are flashing buy signals, I begin to look very aggressively for long-side trades.  It has served me well over the past month, and I will continue to take the low-risk, high-probability trades when they're presented.

Technically, we're still holding the lows, although the S&P futures have broken the previous lows in the Globex session.  These night sessions are so illiquid that I don't read much into them, but if we open below the recent lows there could be a severe selling wave.  Like I said a couple of days ago, if that happens all bets are off on my part until we either reverse back through Monday's lows or we see another round of extreme sentiment measures (I've laid out the thresholds I'm looking for several times on the site).

Wednesday, July 17, 2002  8:14 PM EST

Although I said yesterday that there were signs we could see another upside surge in the next couple of days, I was not expecting the bulk of it to be between the hours of 5:00 - 9:00am EST.  I have grown to detest these gaps up and down because it wastes so much energy that otherwise could have some real meaning.  Fading gaps in either direction has been a consistently profitable trade for a long time, and traders aren't about to stop their money machine now.  The only good to come out of situations like this is the increased volatility.  I have been suggesting that we should see an increase in volatility if this is truly the bottom it could be, and so far it is not disappointing.

The OEX put/call open interest ratio continues to drop, now to a reading of 0.57.  This is the second-lowest reading since 1995, with the lowest being 0.56 on September 20, 2001.  You can see from yesterday's scatter plot that the return after that reading was 10%+ after 30 days.  Lest you think this is an expiration-dominated event, the open interest ratio has steadily RISEN in 56% of the expiration weeks over the past two years.

I have received several requests for an update on the OEX put/call ratio - total put/call ratio spread.  This spread tracks the difference between the 10-day moving averages of the OEX put/call ratio (ostensibly the "big money") and the total put/call ratio (including the "dumb money").  When the spread is low, then that means the OEX ratio is relatively low while the total ratio is relatively high - the big money is betting on the upside while the dumb money is betting on the downside.  We are currently at one of the lowest ratios we've seen over the past 7 years.

The Investor's Intelligence numbers came out and we have finally seen the inversion of bulls and bears that has been widely watched in the past.  Maybe I'm reading the wrong material, but I only saw it mentioned once today.  Last week everyone was all over the large drop in bulls but this week it's going by with nary a mention.  The bullish ratio of 47 this week is the lowest since 44 in September 2001 and September 1998.  This inversion has usually lasted at least a couple of weeks and sometimes for months on end when things are looking really dismal.  In fact, the bullish ratio was lower than the current reading during the entire period from April of 1994 through February of 1995, with not even a peak above 48.  The difference that time is that the other sentiment surveys were much more bullish during that period.  Now we have a confluence of negativity that's rarely seen and always meaningful.

I still see nothing in the recent action to change my working theory of increased volatility with an upward bias.  I will be looking to buy the dips and most likely lighten up a bit if we rally.  I am not thrilled with pressing my bets in either direction right now, but will become more aggressive on the long side should we have a decline of any significance.  I will keep doing that until we can say with some confidence that we have put in an intermediate-term low.  We are not at that point yet, although it looks increasingly likely that we have.

Tuesday, July 16, 2002  8:09 PM EST

There's not much to read into today's session.  You should be expecting the volatility, and today's was not even anything particularly special.  Whether it was because of Greenspan or Intel's pending earnings, I don't particularly care, but it is normal and healthy.  You shouldn't expect the market to go up every day, even after a reversal session.  We need a follow-through day, but it doesn't have to be the next day.

Yesterday, I mentioned some research I did on OEX reversals.  Today's action does NOT negate the stats I mentioned then.  Three of the 11 reversals in the past 20 years had next-day negative changes greater than the -1.8% we saw today - and the market went on to be higher than the reversal day by an average of 1.2% 10 days later.  The study is still valid and suggests higher prices ahead.  These are only odds and not certainties, but the results were significant.

The OEX put/call open interest ratio hit a very low 0.61 today.  Following is a scatter plot of the open interest ratio versus the percentage return on the OEX 30 days later.  If you're not familiar with scatter plots, they can be a bit confusing until you realize what you're looking at.  All it shows is each data point plotted at the correct juncture - for example, if you look at the dot to the furthest right on the graph, that represents an instance where the open interest ratio was 2.74, and the OEX was 0.2% lower 30 days hence.  These are excellent graphs for showing correlations.  If there was a perfect correlation between the open interest ratio and percentage return, you would see the dots line up in a straight line going from the upper left quadrant to the lower right quadrant (meaning a low open interest ratio is bullish and a high open interest ratio is bearish).  This graph shows all data from 1995 through 30 days ago.

I have circled the concentration of highest returns.  You can see that they ALL occurred when the open interest ratio was less than 1.0 30 days prior.  I highlighted with a square all occurrences of an open interest ratio of less than 0.70.  You can see that ALL returns were positive, with none less than 4.5%.  The green vertical line on the graph represents our current reading of 0.61.  You can see that it has been matched or exceeded three times, all of which were positive for the OEX 30 days later.

The OEX put/call ratio registered a low, bullish reading of .54 today, which places it in the bottom 0.5% of readings over the past 17 years.  Such heavy call buying by the "big money" in relation to puts has typically led to very short-term bounces in the past.  We also saw a drastic reduction in put purchases on the S&P 500 index, which could be considered even more of a "big money" haven than the OEX.  That put/call reading dropped below 1.0 for the first time in the past week.  We go through this every month, but expiration could be having an effect on these numbers.  I don't believe it renders them useless, however.

Although sentiment is not particularly oversold on a short-term basis (the STEM.MR model closed at 81%), there are some other signs popping up that we could see another upside surge sometime in the next couple of days.  I'm considering any declines at this point to be buying opportunities, but these are for short-term trades only, and with tight stops.  I still believe that we are in the midst of a bottoming process, and the volatility should continue.  If we break yesterday's lows, all bets are off and I wouldn't consider the long side again until 1) price reversed back through Monday's low or 2) we once again recorded extremely low TICK and A/D readings and saw a spike high in the VIX.

Monday, July 15, 2002  9:00 PM EST

I said over the weekend that my working theory going forward would be that we would likely see increased volatility with an upward bias.  Today's session didn't do anything to refute that theory, so I'm still going with it.  I also said I didn't see anything that would indicate "CRASH AHEAD", although with the DOW down 400 points, it was coming close.  But with the TICKS flashing -1200 and the A/D line at -2300, I was looking for a reversal to go long.  You might recall from the commentary two weeks ago that a TICK reading under -1000 and an A/D reading under -2000 were two of the indicators I was looking for to determine if we were seeing an arbitrary dumping of positions.

Today's reversal was impressive, with a 5% decline from the previous close and a 5% rally from the low.  Going back over the past 20 years in the OEX (S&P 100), I looked at reversals where we dropped over 3% from the previous close and reversed up in the same session more than 3% from the low.  Such occurrences are of course rare, with only 11 instances being recorded in those 20 years.  It boded well for the short-term, as the table of future returns below illustrates:

The VIX also posted a nice spike reversal, and the stats I posted on Thursday once again apply.  The combination of the VIX reversal and the OEX reversal (the same would apply to the S&P 500) bode well for the short-term.  There were only three occurrences of both reversals happening on the same day.  Here are the results:

We can't infer much from only three samples, but the evidence we do have strongly suggests that this reversal will have some legs.  It should be enough to bring us back to overbought on our shortest-term indicators, where once again we will face a major test.

The 10-day advance/decline line is at -644, the lowest reading since September 2001.  This is one of my favorite overbought/oversold indicators, and it is very oversold although not extreme at this point.  We would have to see something closer to -1000 to call it historically washed out.  Many of the better bottoms over the past 40 years have been formed when the 10-day A/D line shows a negative number about 30% of total issues traded on the NYSE.  The current reading is 19% of total issues.  September showed 35% and March 2001 showed 24%.  We reached -2320 intraday which was close to the lowest in recent history, next to three -2350 readings during the September 2001 bottom.

The NYSE intraday cumulative TICK indicator hit -5200 this afternoon, which is the third-lowest I have on record (going back about four years).  We were seeing some major negative TICK readings for a sustained period as the S&P hit new low after new low.  It's hard to tell when it's going to turn, but sustained TICK readings like that are a good clue that the trend will not continue for long.  We remain oversold on that indicator even after the +1000 readings we saw during the reversal.

I mentioned several days ago that most tradable bottoms that have occurred over the past four years saw average 10-day range readings in the S&P of around 30 points.  We're at 27 now.  On the Nasdaq Composite, I stated that the average range at past bottoms was around 70 points (or 5% of price) and now we are at 52 points (3.8% of price).

The STEM.MR model, after issuing the shortest signal possible today (since I use 30-minute data), is now back in neutral territory.  Because the moving averages are so elevated, it won't take much to push this into sell territory.  I wouldn't be terribly surprised if we got a sell signal tomorrow, in fact.

The lowrisk.com sentiment survey for last week showed a persistent lack of bullishness, as the bullish ratio stayed at the same low level as last week.  These two weekly readings have begun to push the moving average toward the lower end of the range (which is positive for the market).

In all, it was another good reversal opportunity for short-term traders open to taking low-risk, high-probability counter-trend trades.  For those with a longer horizon, it was another gut-wrenching puke session that means nothing without a follow-through day within the next couple of weeks.  How the market reacts to the next overbought situation will tell us most of what we need to know to make an intelligent investment decision.

 - Jason Goepfert