sentimenTrader.com

All rights reserved.

Thursday, October 24, 2002  8:00 PM EST

Towards midday today, we were getting extremely overbought - not only on an intermediate-term time frame as I mentioned last night, but also on a short-term basis.  Virtually every short-term sentiment indicator I follow was stating that optimism had reached an extreme, and hopefully the intraday note proved helpful to many of you and you were able to take advantage of the afternoon weakness.

The decline into the close served to alleviate the pressure on most of the indicators, so our condition tonight is not as critical as it was a few hours ago.  The STEM.MR model gained 12% by the close, recovering from a reading that was just on the edge of negative territory when I sent out the note.  The intraday cumulative TICKS, while not oversold by any means, have come off the extreme overbought condition generated this morning.  The price oscillators have also reversed from extremely overbought and are now actually oversold.

I talked about the overbought breadth measurements yesterday, and going forward the situation should become more constructive in a relatively short period of time, especially if we settle down here for a few days.  Over the next five trading sessions, we will be dropping three large positive advance/decline readings from the 10-day average of that indicator, which will serve to dampen it and possibly even swing it to oversold if we replace them with negative readings.  The same goes for the up volume ratio average, which as I pointed out yesterday is extremely overbought.

I haven't discussed the VIX for awhile, as there really hasn't been much going on there.  However, we're getting to a point now where the fear we saw near the October low has, of course, receded substantially.  I post the VIX fear premium to the site each day, and have included it below to show that the uncertainty premium built into the VIX has fallen dramatically over the past two weeks.  It is now closer to the lower standard deviation band than the upper.

While we are not near an area that would dictate caution, there is a good chance that this premium will shrink further over the coming days.  Remember, this indicator is constructed by comparing the historical volatility over the past 30 days to the implied volatility built into the VIX.  The historical volatility is increasing (which is no surprise considering recent action) and the VIX has been trending lower.  This is forcing them to converge, and they should quickly cross (meaning historical volatility would be greater than what the VIX is implying for the future).  So while this indicator is not saying anything in particular at the moment, the next few days may bring it into an area where caution is warranted.  I suggest you monitor it closely.

When the investigations into Wall Street analyst recommendations began coming to the fore, I had noticed the timing of their upgrades and downgrades did not seem to be random - there seemed to be something of a pattern.  When a stock had run up significantly, it appeared as though the analysts were more willing to downgrade a stock due to "valuation" or "technical" concerns.  Since these analysts supposedly know more about the company than anyone else - including many at the company itself - I had always thought that if they actually gave honest opinions (don't laugh), then monitoring the daily recommendation changes might prove to be a useful barometer of general analyst sentiment.  I've gone back and tracked the changes over the past 7 months, and the chart below shows what percentage of total analyst changes are downgrades (as opposed to upgrades):

The ratio on the chart is actually a three-day moving average of the downgrade ratio, to smooth out the daily fluctuations somewhat.  Although the number of upgrades and downgrades is affected by the quarterly release of earnings warnings and statements, and individual companies can have a big impact on the number if they disclose a surprise, the downgrade percentage seems to be following a pattern.  As analyst face more pressure to be "honest" with the public, they appear to be more willing to downgrade.  As you can see from the chart, their timing doesn't seem to be half bad, and is even improving.  This is by no means a perfect indicator, but I believe it bears monitoring going forward.  When the analysts become overwhelmingly bearish, it has tended to precede market weakness (and is perhaps even the catalyst for market weakness), while when they are relatively bullish as defined by a low downgrade percentage, the market seems to perform well in the following weeks.  The current reading is quite high, and would suggest that the analysts are seeing many of their stocks as "fully valued".

Our intermediate-term indicators are now in the process of working off their overbought condition.  If the market is as strong as it appears to be, then we should see a period of consolidation over the coming days instead of a cascading move lower.  Our short-term indicators are now mixed after today's decline, and not giving much of a hint as to future direction.  This has me in a defensive posture, as dips have recently been bought aggressively, causing risk to short positions, but the overbought condition should put a lid on any advances, limiting reward on longs.  From a sentiment perspective, I would have to say the current situation is neutral to mildly bearish.

Wednesday, October 23, 2002  10:17 PM EST

As I mentioned in the intraday update, the Investor's Intelligence survey showed a dramatic increase in bullishness during the reporting period ended Friday.  The bears retreated from 43.2% to 35.6% while the bulls jumped from 28.4% to 38.9%.  This is a swing in the bullish ratio of nearly +13%.  It's been nearly twelve years since this type of move has been recorded.  In fact, the next-largest jump in bullishness in the past decade was just two months ago, at the exact high in late August, when the bullish ratio jumped 11%.  The lowrisk.com survey gave an excellent heads-up that this type of thing may happen, as I stated on Monday.  Needless to say, such a huge shift in sentiment puts something of a hole in the bullish argument going forward.  As I usually do when we see a dramatic one-week shift in the sentiment surveys, I calculated the average price these investors may have gotten on their purchases during the week, which comes out to around 858 on the S&P and 917 on the NDX.  I do this because it's fairly safe to say that these surveys reflect the general attitude of investors, and if they become dramatically more bullish during a given week, then we can assume that they take actual positions in accordance with their attitudes about market direction.  So if we see such a large shift in bullishness, we can assume most of those traders actually bought equities.  Obviously, I'm making some broad generalizations and large assumptions here, but the past couple of times we've gone through this exercise, it has been a good guide.  So now what I am going to look for is the S&P and NDX to hold above the levels mentioned.  If they do, then those who bought will likely feel comfortable with their purchases and will not add to any selling pressure, and may in fact buy more.  If we go below 858 S&P and 917 NDX, then we will be below the average price at which these traders may have come into the market, and should expect the selling to increase even further as they try to get out at breakeven.  Again, I'm brushing with some pretty broad strokes here, but keep these levels in mind.

In addition to an increasing amount of bullish sentiment, our breadth measurements are beginning to enter nosebleed territory.  The chart below is the 10-day average of the NYSE up volume ratio (up volume / (up volume + down volume)), plotted against the S&P 500.

The red line is at 60%, a level that has coincided with overbought conditions during this bear market.  The green line is oversold.  You can see how important it is to keep these types of figures in context, as absolute levels on almost any sentiment indicator are dangerous.  Early in the bear market, a reading of 40% could be considered oversold, whereas now it takes a reading closer to 30% to have that status.  As we get further into the bear and the selling intensifies, it takes more and more to become "extreme" on the downside.  In any event, our current situation is the most overbought than at any time during this bear market.  The only way this can be justified is if we are at the beginning of a new, major leg up (possibly a new bull market).  While that is certainly possible, I don't believe it's prudent to think that way at this time.  I take much more of a "prove it to me" attitude when it comes to trend changes, and long-term, we have no indications that we are in a new bull market.  Therefore, this reading has to be considered dangerous if you're looking for further upside over the coming days and weeks.

Adding to the overbought breadth are the cumulative TICKS.  The dailies on the NYSE and Nasdaq are matching or exceeding their readings in late August, and the 30-minutes are beginning to push their upper boundaries.  Since I guess we can be considered to be in a 30-minute uptrend, those intraday overbought readings can be excused somewhat.  But we are still clearly in a daily downtrend, so the fact that the daily cumulative TICKS are becoming overbought is troublesome.

As each day goes by, the Rydex fund complex gets more bullish (bearish to us).  The bearish fund flow is now the most negative for the market since April, and while the flow into the bullish funds is not quite as bad, recent developments have pushed the 3-month stochastic of the bullish ratio to the overbought .70 level.  You can see from the chart on the site that the last few times this level was reached, the market soon weakened considerably.

As I write this, the S&P and Nasdaq futures are up and teasing the recent highs.  If we have a large gap up tomorrow morning, I expect it to get sold heavily and it may make for a good short trade for the nimble short-term traders among you, with the knowledge that at some point breakout traders will try to push it higher.  In the intermediate-term, we are not quite "maximum" overbought, but we're close.  Another day or two of significant upside will likely push us there, and at that time - for the first time since the 10/8 low - I would recommend that longer-term traders begin to pare out of their long positions, or at the very least tighten your trailing stops very closely.  Like the late August period, the risk/reward of holding long positions is getting skewed more towards risk, so please be careful.

Tuesday, October 22, 2002  7:10 PM EST

Tonight's comment will be short because there is just not much that has changed from a sentiment perspective.

The Rydex funds continue to give bearish short-term implications, and now the intermediate-term picture is starting to give cause for concern.  Each day, I post the bullish and bearish flows to the site.  To construct the flow, I use the deviation of the 10-day moving average from the 50-day moving average of the bullish ratio and bearish ratio, weighted by the leverage each fund employs.  The following chart, which is also posted to the site daily, superimposes the bullish fund flows on the bearish flows.  It is fairly obvious that over the past year and a half, when the flows reach an extreme, it has typically coincided with an exhaustion of the trend.  What is not so obvious is that when the bullish fund flows first cross from negative to positive while the bearish fund flows cross from positive to negative, it has preceded a market high within a couple of weeks every time.

The three instances (except today's) have been circled on the chart.  We can see that the momentum out of the bearish funds and into the bullish funds has not been positive for the market during this bear market.  Considering that we have this same situation today, it should add a note of caution to our outlook.

About the only indicators which made notable movement today were the price oscillators, which were incredibly overbought yesterday but are now back to neutral tonight, and the S&P 500 put/call bid/ask bias ratio, which was also very bearish yesterday but is back to neutral.  Our shorter-term indicators are just not reaching extremes long enough to be able to give decent odds of a reversal.  The constant flip-flop of gaps up and down are keeping them in flux, and it will take something of a sustained trend in one direction or another for them to be able to reach an extreme.  In the midst of such mixed signals as we have tonight, the only suggestion I have is to stay cautious and wait for a better edge before becoming aggressive in either direction.  Sentiment is best used when extremes are reached, and I want to avoid trying to read something into them when nothing is there. 

Monday, October 21, 2002  9:06 PM EST

The short-term bearish arguments I outlined in the weekly commentary manifested themselves for about an hour before the steady march higher into the close.  As I stated in the intraday note, when the market isn't doing what it "should" do - in this case go down - then it's best to respect that dynamic and step aside or trade lighter than normal.

Our "sneak peak" sentiment survey from Lowrisk.com showed a drastic increase in bullish sentiment in the latest survey period, which ended yesterday.  The percentage of respondents who were bullish rose from 33% last week to 48% this week, and those bearish dropped from 57% to 41%.  This makes the most current reading the largest bullish ratio in over 4 months, and suggests that we will likely see an increase in bullishness in the major surveys released later this week.

I mentioned yesterday that the 10-day TRIN was extremely overbought, and it continued in that direction today.  At a current reading of .76, the 10-day is the lowest since March of 2000 - the peak of the bull market.  Before that, there were only three other such low occurrences in the previous decade - and that was during one of the greatest bull runs in history.  I showed this weekend how overbought TRIN readings in the context of a bear market tend to precede a continuation of the downtrend, so this lends an air of caution.

The 10-day average of the up volume ratio (up volume / (up volume + down volume)) is now at 60%, which is close to the practical maximum for this indicator.  It is extremely rare for the market to make any headway without a pause when this indicator reaches such an extreme, especially during a bear market.  Adding to the overbought breadth argument, the daily cumulative TICKS on the NYSE and Nasdaq have cycled back up to where they were at the peak in late August.

In the S&P 500 options, today's rally was used to create some apparently bearish positions, as our put/call bid/ask bias ratio dropped to 0.18.  As a reminder, the bias ratio is created by comparing trades going off at or below bid (sales) to those going off at or above ask (purchases) for both puts and calls.  Today's ratio is the lowest (most bearish) since the mini-peak on October 1st, and tells us that for the first time since this latest rally began, the institutions are beginning to aggressively use the options markets to either hedge or speculate on some near-term downside.

While it's entirely possible for equities to continue to march higher, each day brings us closer to "maximum" overbought.  I realize there is no true "maximum", but there is a practical maximum where the odds are heavily favored for a sustained move down.  We're not there yet, but two more consecutive days of positive breadth should bring us close.  In the meantime, I suggest short-term traders stay cautious but look closely for potential short candidates.  For longer-term traders, just keep trailing those stops.

- Jason Goepfert