sentimenTrader.com

All rights reserved.

Thursday, November 14, 2002  8:25 PM EST

For the first time in two weeks, we got a solid trend day higher today.  Unfortunately, it looks as though it may not last.

Today's total and equity put/call ratios were very low, in fact the total p/c ratio pierced its lower standard deviation band while the equity ratio came close.  The total p/c ratio has pierced its lower trading band only five other times in the past year (the chart is posted to the site - click here), each of which lead to a slide in the market shortly thereafter.  I've stated several times that the put/call ratio during expiration weeks trends about 7% higher than non-expiration weeks.  The average expiration-week total put/cal ratio since 1995 has been .71, as opposed to .66 during a non-expiration week.  I also looked at markets that are trending up as opposed to down, to see if that affects the bias at all.  As a definition of trend, I simply used the slope of the 20-day moving average of the OEX (S&P 100).  If the 20-day average was sloping up, then I considered the market in an uptrend, and if it was sloping down then I considered it in a downtrend.  Interestingly, the average expiration-week p/c ratio in a downtrend was .80, as opposed to a random downtrend reading of .71.  The average expiration-week p/c ratio in an uptrend was .64, as opposed to a random uptrend of .63.  The standard deviation for each of the averages above ranged from .12-.16.  What's the point of all this?  My point was to see if we should "throw out" high or low volume-based put/call readings during expiration week.  My opinion is that we should not, that expiration-week put/call readings are as valid as any other reading.  In fact, there is no statistical difference between an expiration week and non-expiration week when the trend is up, as it is now.  Therefore, today's low put/call reading should be considered low and bearish.

We have a few other bearish signs tonight as well, particularly in the short-term.  The S&P and Nasdaq price oscillators are now extremely overbought.  Again, these indicators are made up of a score that's assigned to each price bar on a 30-minute time frame.  I then take a 13-period moving average (which equates to one full trading day).  Here's an example:

The percentages above each bar are the "score" that's assigned to that bar.  So the first candle, where opening price was the low price, and it closed near the high, gets a score of 87%.  The second candle, which is relatively neutral, gets a score of 45%.  And the last bar, which bearishly opened near the high and closed near the low, gets a score of only 3%.  When we have several bars in a row that get a high or low score, it pushes the moving average into an extreme territory, and that usually equates with a market that's either overbought or oversold.  Now that you know how the indicator is constructed, it should give more meaning to the fact that the current S&P price oscillator is at 66%, while the NDX is at 70%.  This means that the average bar over the past day in the NDX received a reading of 70%.  Even a reading of 60% can be considered overbought, so we're quite extreme here.  Only two other times in the past six months have we seen such a high reading, and both lead to a 20-point decline in the NDX the next day.

Another bearish indication is the intraday NYSE cumulative TICK indicator, which closed at +5091 today, just over the overbought threshold.  We've reached this level four other times in the past month, and each lead to short-term weakness (click here for the chart).

The VIX declined approximately 12% today (using the difference between today's low and yesterday's close).  This has happened 20 other times during this bear market.  13 of those occurrences then lead to a down market 5 days later, with an average return of -0.96%.  However, what's striking about this is that EVERY ONE of the 7 others that lead to a positive market occurred after a severe market decline.  None of the 7 occurred after the S&P, for example, had already rallied 17%.  Several of the 13 decliners had occurred after substantial rallies.  What this tells me is that there has been a sudden dwindling of the premium given to uncertainty, which usually occurs once we put in a reversal session near an intermediate-term low point.  If we get this kind of complacency (for lack of a better word) after we've already had a large rally, then it has uniformly lead to weakness within the next week.  Another bearish omen.

Each of the paragraphs above detail a bearish indication for the short-term, and each has been quite reliable when they occur.  Even though price action itself (always the most important) appears quite bullish, we must be aware that these indications may put something of a damper on the short-term action.  Most of these bearish signs could be alleviated within a matter of hours, so it's not like I'm suggesting we're heading to new lows here.  The only thing these are saying is that it is best to use caution right now, and in fact I would look at any large gaps up at the open tomorrow to be a high-odds selling (and shorting) opportunity for short-term traders.  For longer-term traders, I continue to stress that the risk/reward in any direction at the moment is unfavorable, as I have been saying for three weeks now.  I do not believe now is a good time to become aggressive with any positions with an intention to hold them greater than a couple of days.

Wednesday, November 13, 2002  8:45 PM EST

Today was essentially a repeat of yesterday, as equities made a couple of wide swings without going anywhere in the end.  From a longer-term perspective, I'm quite happy with the neutral stance our sentiment indicators and models have given for the past three and a half weeks.  Anyone trying to take position trades during that time has likely broken even at best, while taking on considerable risk, so neutral has been the correct position.  I see nothing tonight to change that stance.

On a lesser time frame, our shortest-term Rydex indicator, the three-day RSI bull/bear spread, did a good job of pinpointing a peak in optimism on 11/6 with a reading of +69, which coincided with the most recent market high.  As of yesterday's data (today's will not be released until around 11:00pm EST), the three-day RSI had cycled back down to oversold, with a reading of -73.  For those of you not familiar, an RSI (Relative Strength Index) reading is essentially a momentum reading that compares an index to itself over a given period of time.  The stronger the momentum, the higher the RSI reading and vise versa.  The Rydex RSI spread that I post to the site each day is the difference between the bullish fund RSI and the bearish fund RSI.  So, the higher the spread, the more momentum (i.e. assets) there is flowing into the bullish funds as opposed to the bearish funds, weighted by the leverage each fund employs.  When the spread is low, then there is more momentum in the bearish funds as opposed to the bullish funds.  The fact that the spread is so low now suggests that the Rydex timers have piled on during the decline over the past week, expecting further downside.  We are at a point now that often coincides with some short-term upside relief (just to make them nervous and get out of their bearish bets - the market truly is a cruel beast).

Seasonality-wise, we have two interesting days coming up.  The Thursday prior to option expiration has tended to be somewhat positive, with a 58% probability of ending up on the day (versus a random Thursday of 53%) over the past 7 years.  However, during this bear market, the probability has slipped to 52% versus a random Thursday of 57%, albeit with a slightly greater average return of 0.17% (versus a random Thursday return of 0.10%).  Not much of an edge there.  Come Friday, the probabilities become more decisive.  Over the past 7 years, expiration Friday has been up 53% of the days (versus a random Friday of 59%), with an average return of -0.20% (versus random of 0.18%).  But during the past two years, expiration Friday has been up only 32% of the time, with an average return of -0.83% (versus random of 50% and -0.05%).  Bottom line, there is a slight negative bias to the next two days.

About the only positive I can find for the short-term is a pessimistic overshoot by the Rydex timers.  While that indicator has performed very well, it's hard to hang one's hat on that alone.  Conversely, there's not much negative to write about either.  That leaves us twisting in no man's land, and I will continue to stress that when a solid edge is not to be had, it is best to keep trades light or nonexistent.  The risk/reward is just not justifiable in either direction, so my plan is to continue to trade lightly with a very short time frame.

ADDENDUM (Posted at 11:20 PM EST)

The Rydex ratios for tonight continued the short-term flow into the bearish funds, and out of the bullish, to an extent that has caused the three-day RSI spread I talked about in the commentary earlier tonight to reach an extreme of -91.  This is significant because during this bear market, there have only been six other days showing readings less than -90, consisting of five distinct occurrences (i.e. one instance was two days in succession).  The following charts show the performance of the S&P 500 and Nasdaq 100 after these occurrences:

As always, the percentages above or below each bar are the probability of the market being positive the given number of days later.  For example, five days after an RSI spread reading of -90 or less, the S&P showed an average gain of 3.61%, and was positive 83% of the time.  This is only one indicator, the sample size is small, and it's about the only one suggesting that upside is likely, but I thought I should mention this development so you can factor it into your trading.  While this should be considered a secondary indicator, and price itself is not what I would call oversold in the short-term, it makes me more willing to fade this group of traders and look for long setups over the next few days.

Tuesday, November 12, 2002  7:40 PM EST

In the intraday note this afternoon, I said that we didn't appear to have yet gone too far, too fast, so short entries would not yet be a high-odds trade (as I saw it).  We went on to make a higher high on the day in both the S&P and Nasdaq before the afternoon decline, which coincided with overbought price oscillators (and ominous cumulative TICKS), but not much else.

That decline, while steep, didn't take very much time, so many of our indicators are just beginning to roll over.  Once again, we are being left with no particular edge in either direction. 

I've really been struggling tonight to find some useful to report sentiment-wise, but I simply can't find anything.  A great majority of our indicators and all models continue to thrash around in neutral territory, and there is just no edge to be had.  Most of you pay me to find something that helps you make money, and I take that very seriously.  But when I can't find anything, I feel it's my obligation to tell you so instead of trying to come up with confusing filler.  I continue to believe that traders on all time frames except the shortest should keep it light until a high-odds opportunity presents itself.  Normally in times like this I suggest following whatever trend is in place, but at this time we really don't even have a trend, unless you look longer-term and consider the daily and weekly to be down.  Please wait for a good pitch before swinging...

Monday, November 11, 2002  8:20 PM EST

In today's intraday note, I suggested that it may pay to look for an upside pop in the Nasdaq, and we ended up getting the largest rally of the day immediately after.  Of course, that "largest rally" was only 10 points, and would have likely resulted in minimal profit for anyone attempting a trade.  I also said that for the S&P, we would have to see more selling in order to achieve a level of short-term oversold.  We did get that further selling, and we are approaching oversold on a good number of our indicators.

The STEM.MR model is oh-so-close to switching to a positive bias, as it nearly pierced its upper standard deviation band at the close.  If we have another bout of selling in the first hour or so tomorrow, then it's very likely that we will see a positive or strong positive indication here.  If so, I would be looking to take short-term long positions at the first sign of a price reversal.  The price oscillators on both the S&P and Nasdaq are now bullish, and the intraday cumulative TICKS are essentially also oversold.  For me, it's difficult to tell if we are in a 30-minute uptrend or downtrend, so these oversold readings have a little less impact on me than they would if we were in a clear uptrend.

The lowrisk.com sentiment survey, released today, came in with the recent bullishness receding a bit.  The bullish ratio (bulls / (bulls + bears)) is now 45% as opposed to 51% last week.  This is the first hint that the major surveys released later in the week may also show a bit less bullishness.  However, as you can see from the chart below, the 4-week average of the bullish ratio on this survey upticked to 48%, which is a level that has previously urged caution during this bear market.

The selling we have seen over the past week has been incredibly intense.  The 10-day TRIN has ratcheted quite a bit higher, and is actually beginning to flirt with oversold.  But even more telling is our proprietary Down Pressure indicator.  I have not added this to the site since I am still in the process of gathering historical data.  This indicator is computed by taking a three-day average of the volume going into advancing issues versus declining issues, and the percentage of points gained versus lost for all of the stocks in the S&P 500.  For example, today there were 58 stocks in the S&P 500 which ended positive on the day, and there were 431 which declined (this may vary slightly depending on your quote vendor).  Those 58 advancing stocks gained a total of 21 points on volume of 124 million shares, while the 431 decliners lost a total of 297 points on volume of 1.3 billion shares.  This resulted in a Down Pressure reading of 93%.  Over the past three days, the average has been 88%, making it one of the most intense selling climaxes in the past four months.  In fact, it is eclipsed only by late July, which reached a three-day average of 90%.  I keep the same indicator for the Nasdaq 100, and it is currently giving a similar reading as this one.

On the chart, I have circled every instance in the past four months where this indicator approached the 80% level.  As you can see, it coincided with each important low point.  There was one "signal" in mid-July that was very deceiving, since it kicked off the slide into the end of the month.  There was another in mid-September that preceded a two-week trading range.  The other three marked a good low.  Considering that we have such an elevated reading today, this indicator is suggesting that we may be near a tradable low point, although I don't have enough history to be entirely confident of its significance.

Traditional technical analysis suggests that if we have a "false breakout" from a trading range, then the signal in the opposite direction is usually quite strong.  Since it appears as though we had a false breakout on the upside of the recent trading range, that would mean that the move down could be significant.  However, we have now built up some support levels over the past month, and considering the intense selling we've seen the past few days, I think we will see a tradable bounce in the next day or two.  My plan for tomorrow is to look for long setups if we get either a large gap down or an hour or so of selling pressure in the morning (especially if it's enough to give us a positive indication from the STEM.MR model).  Of course, this wouldn't mean we HAVE to rally, but at that point the risk/reward would be favorable for a short-term long trade.

- Jason Goepfert