http://www.sentimentrader.com

Sunday, February 16, 2003

I have been stressing, especially during the past week, that although we had not yet seen the confluence of negativity normally seen at intermediate-term bear market lows, there was enough short activity that we should expect to see very quick and sharp spikes higher at any time.  Thursday and Friday provided evidence of the type of market action that very often results from these types of conditions.  This weekend, we are left with a little less short fuel, and we still have not seen the final selling phase that normally accompanies good, high-probability low points.

____________________________________

Indicator:  NYSE MEMBERS REPORT

Status:  BULLISH

Comment:  Specialist shorting activity picked up a bit during the latest reporting period, which is the week ended 1/31/03.  The Specialist Short Ratio increased to 35% from 31% last week.  This pushes the current ratio just above its lower trading band.  As I've pointed out several times recently, the four-week average of this ratio continues to decline, and at 34% it is the lowest since the heavy short-selling we saw prior to the major bull-market surge we saw in 1997.  Granted, prior to that upswing we saw nearly three years of the Specialist Short Ratio remaining consistently under 40%, whereas now we don't even have a year of such readings under our belt.  If we take a look at the very long-term 52-week moving average of this data, we can see that we are still recovering from the extended bout of public buying that marked the top of the bull market.  While this is of course a badly lagging indicator, we have not yet reached a level of persistent public shorting that has lead to major advances in the past 20 years.  Prior to 1980, there was significantly less public shorting of securities, so the Specialist Short Ratio had a much higher trading range than now.

The recent dip to 31% still counts as a bullish factor here, though, at least in the intermediate-term.  The spikes down in this ratio near the July and October lows occurred about three weeks before the ultimate low in the market, not at the exact low, which is one of the reasons I like this indicator.  Even though it has a built-in lag of several weeks, it tends to precede major market moves by a few weeks anyway.  So by the time we receive the data, it is not necessarily "too late". 

Bottom Line:  We have seen an extreme of public shorting, at least on an intermediate-term level.  Considering the built-in lag and forecasting record of the Specialist Short Ratio, we should expect a tradable low at any time.  Longer-term, the activity here suggests that we are in the process of building up the fuel necessary for a new, multi-year bull market, but more time - perhaps as long as another six months to a year - may be required to get comparable readings as past long-term bottoms. 

____________________________________

Indicator:  BREADTH RATIOS

Status:  BULLISH

Comment:  Last week, I profiled the 21-day moving averages of the advancing issues ratio and the up volume ratio.  The updated charts are below:

I said at the time that we were approaching a point that had coincided with rallies in the past.  We came much closer this week, almost matching the readings seen at the July and October lows on Thursday, but came up just a bit shy.  The readings we saw late last week could be considered extreme enough to be good for an intermediate-term low, so that is certainly a positive for the coming weeks.  While these longer-term measures have reached an extreme, the faster-moving 10-day averages never hit the extreme readings I look for to indicate that we have seen the type of indiscriminate selling that almost always marks good buying opportunities.  This simply highlights the obvious fact that we've seen persistent selling over the past month, but have not yet seen any type of swift wash-out activity that would indicate we were seeing the much sought-after capitulation.  The table below looks at the last three intermediate-term lows and compares them to our current situation:

We currently compare relatively favorably with these lows, although we're off by just a pinch.  We haven't seen that last push down to make our current readings truly extreme.  Could we have already seen the low based on these measures?  Absolutely.  Is it likely?  I'd give it about a 50-50 chance (on these measures alone).  We can see that we are on the upper end of the range for the 21-day measures, but close enough to consider them comparable.  As for the 10-day measures, we are not yet what I would consider comparable, but look at the pattern in the data - each low since September 2001 has been accompanied by lesser-oversold readings.  The panic after the terrorist attacks marked the most devastating two-week negative momentum we had seen in many years, and since then each subsequent low has been accompanied by less compacted selling, even though price reached a lower low each time.  This could be some type of very long-term positive divergence between price and internal selling momentum which could result in a long-term upswing in the market, but I'm not sure we wouldn't be reading too much into the indicators if we conclude that.  What it does suggest to me is that we may consider a lesser-oversold reading this time to be part of the recent pattern, and if it holds, then we may have already seen a tradable low and we should look for higher prices in the coming weeks.

Elsewhere in our breadth measurements, the new high ratio has been steadily declining and is finally approaching oversold.  This indicator gave a nearly perfect sell signal in mid-January by reaching the .70 level (meaning new highs on the NYSE accounted for 70% of all new highs and new lows...new high ratio = new highs / (new highs + new lows))).  Currently, it stands at .31, which is still high compared to recent history.  Every intermediate-term low during the bear market has been accompanied by readings under .20 except for the low in April 2001, which had a reading of .45.  I suggested last week that we should look for new lows to expand to around the 600 level, but the highest we reached was 219 on Thursday (your figure may vary slightly based on your source).  The new high ratio could continue to decline if we see either increased new lows or decreased new highs (or a combination of both of course).  I would certainly like to see further deterioration in this ratio in order to be more comfortable of a trading low having been reached.

Our Down Pressure indicators have been stuck in neutral for the past few weeks, as we have yet to see any real severe short-term push in selling pressure.   Whenever we begin to see one, it usually reverses intraday and we don't get a chance to record any real extreme readings on a closing basis.  All but one of the good lows over the past six months have occurred when the Down Pressure readings hit 80% or above in the S&P 500 and NDX, so if we have already seen the low, it would be relatively unusual based on this measure.

The daily cumulative TICKS for the NYSE has clawed out of oversold territory and is currently neutral, while the daily indicator for the Nasdaq never did enter extreme territory.  The intraday indicators have done pretty good jobs of identifying intraday selling extremes over the past few weeks, but they also are currently neutral.  You will notice from the charts on the site that every time the intraday cumulative TICKS have reached the upper end of neutral territory (approaching overbought) during the past month, it has coincided with market peaks.  In a severe downtrend, in which we remain, all it usually takes is an alleviation of extreme oversold conditions to continue the downtrend.  If that pattern holds, then short-term upside from here should be limited.

Bottom Line:  Our longer-term breadth measurements suggest that we have seen enough persistent selling that a good low could be here.  However, all shorter-term indicators show that we have not seen the final selling climax that so often accompanies intermediate-term bear market lows.  Overall, my opinion is that it is certainly possible that a low has been reached (reading these indicators only), but the confidence level is not nearly as high as it would be if we had seen a final push lower with heavy selling - enough to push the Down Pressure indicators above 80%, new lows above 600 and the intraday cumulative TICKS into truly extreme oversold territory.

____________________________________

Indicator:  RYDEX RATIOS

Status:  BULLISH

Comment:  The upside of the past two days has prompted quite a bit of money to flee the bearish Rydex funds, especially the leveraged ones.  Over the past two days, nearly $210 million has left the bearish S&P 500 and NDX index funds (65% of which was in the leveraged funds), but only $51 million has flowed into the bullish funds (64% of which went into the leveraged bullish funds).  Another $67 million went into the money market, so about $90 million has left the fund complex altogether (or went into one of the bond or sector funds).  This is a rather large shift of assets within the funds, but it does not appear that these timers have become especially bullish.  With so much money leaving the bearish funds, and much of that the leveraged ones, it seems more of attitude of removing some upside risk rather than betting on higher prices.  The fact that more money went into the money market than the bullish funds further supports that view.  All of the activity in this fund family - from the longer-term indicators to the activity over the past couple of days - suggests that we have seen a low, possibly even an important low, and remains a bullish indication for the coming weeks.

Bottom Line:  The large amount of short-side bets that have been made here will take time to unravel.  The recent upside has seemingly been met with a willingness to remove risk, but not bet on further upside, which from a contrarian point of view is positive.  These indications are positive for the market.

____________________________________

Indicator:  COMMITMENTS OF TRADERS

Status:  BULLISH

Comment:  On Monday, I touched on the current dichotomy between activity in the large S&P 500 futures contract and the e-mini.  I said that the recent record of the two contracts in terms of forecasting accuracy were at opposite ends of the spectrum - the large contract was performing poorly while the e-mini was doing a good job.  This is with the knowledge, however, that the long-term track record of the large contract is very good, while the record of the e-mini is no better than random, so that must temper whatever indications the e-mini is giving.  The most recent activity in the two contracts continued the recent pattern.  Once again, during a week where the S&P lost over 2% of its value, commercial traders in the large contract reduced long positions and increased their shorts, for an addition to their net short position of around 8,700 contracts.  Meanwhile, the small specs in that same contract increased their long holdings more than their shorts, for an addition to their net long position of about 7,800 contracts.  While neither change is particularly large, the fact that it occurred during a down market is unusual, and continues the odd pattern that has existed for the past few months.  Normally, commercials pare long positions and increase shorts as the market rises, and increase longs while covering shorts as the market drops.  Small speculators do the opposite.  This runs counter to what has happened over the past few months, as I have pointed out numerous times.  While this has me on guard about reading too much into the data, I am not ready to throw the large contract out as a forecasting tool.

On the other hand, the action in the e-mini is conforming more to what should be expected.  Commercial traders, after becoming net short to a large degree at the January peak, are now the net longest they have been in the history of this contract (in terms of the absolute number of contracts).  Conversely, the small specs are currently holding their fourth-largest net short position since inception, after having one of their largest net long positions at the January peak.  Considering the fact that we consider commercial traders to be the "smart money", and the small speculators to be the "hot money" (that chases momentum and usually become extreme at the exact wrong moments), this activity has to be interpreted as positive.  Once again, however, I need to stress that the longer-term record of this data is questionable, and that should put a cap on reading too much into this.

Bottom Line:  The recent record of the large S&P contract is poor, while the e-mini is coming on strong.  If we give greater weight to the indicator that has performed better recently, then overall this complex is bullish.  In the e-mini, commercial traders are the net longest they have ever been, while the momentum-seeking small speculators are holding their fourth-largest net short position.

____________________________________

Indicator:  SENTIMENT SURVEYS

Status:  BULLISH

Comment:  I went over a bigger-picture review of the Investor's Intelligence survey on Wednesday, so I recommend checking that out if you haven't already.  In a nutshell, it suggests that that survey has recently given readings normally associated with market peaks and certainly not long-term lows.  Otherwise, each of the other major surveys came in this week with a reduction in bullishness, and each of them are close to readings seen at other lows during the bear market.  We are still not seeing any extreme weekly readings, however.  One possible exception to that is the four-week average of the bull ratio of the AAII survey (American Association of Individual Investors).  With a current reading of 39%, this indicator is matched only by the lows in September 1998, July 2002 and October 2002.  Going back over the past decade, the indicator has dropped to the low 30%'s and lower, so there is certainly a precedent for it going even lower than it is now.  In the Dow Jones Industrial Average-oriented lowrisk.com survey, the four-week moving average of the bull ratio has just dropped to one of its lowest levels ever.

Bottom Line:  The more erratic surveys (AAII and lowrisk.com) are currently in extreme territory, particularly on a moving average basis.  The more well-known and slower-moving surveys are showing increased bearishness that compare relatively well with past lows during this bear market, but none of them are particularly extreme.  The granddaddy of them all, Investor's Intelligence, is showing the most - and most persistent - bullishness and needs to show more negative readings before it could be considered a bullish market indicator.

____________________________________

Indicator:  AIM MODEL

Status:  BULLISH

Comment:  Two weeks ago, I said that every time this model has formed a peak above 60% over the past five years, it has corresponded to an intermediate-term market low.  I also said that we should see a reading closer to 65% to match the lows seen during the bear market.  We have accomplished that this week, as the model currently stands at 64.9%.  Every time this model has formed a peak near 65% over the past decade, it has corresponded to a very tradable market low.  The model has not yet peaked, but I think it's a pretty good bet that next week's sentiment survey readings will show more optimism than this week's, meaning the model will likely peak right here.

Bottom Line:  This model has reached a point that has marked a good buying opportunity every time during the past decade, even during the bear market the past few years.  This suggests that any further weakness could be bought with a reasonable chance for a successful trade if held for at least a few weeks.

____________________________________

Indicator:  PUT/CALL RATIOS

Status:  BULLISH

Comment:  I've discussed the put/call ratios, particularly the equity put/call ratio, several times over the past two weeks, and nothing has really changed other than that they have become more extreme, for the most part.  The 10-day equity put/call ratio has set a new 5-year record high as of Friday's close at .92, which is quite an accomplishment considering all that we've been through in the past five years.  I should acknowledge the fact that this data is being largely influenced by the apparent single large trader that had been operating in QQQ options during this time, and that may decrease the weight this new record high should be given in our analysis.  Regardless, the indicator was rising before this trader entered the market in such an influential manner (apparently), and the indicator is now at a point obviously that has coincided  with past intermediate-term lows, so I believe it is another indication that pessimism has reached an extreme level.  The OEX put/call ratio has been neutral during the past week, so I don't believe there's much to read into there.  In the SPX options, our put/call bid/ask bias ratio was elevated on Tuesday, Wednesday and Thursday of last week (suggesting there was a heavy bias towards selling puts and/or buying calls by institutional traders - a bullish development), but that changed on Friday as the ratio dropped back to the lower end of its range.  For the coming week, this indicator will be heavily influenced by expiration activities, so those of you who follow this ratio should keep that in mind this week

Bottom Line:  The equity put/call ratio, even taking into consideration the skewed data from the large QQQ trader, has reached a level that indicates a trading low is likely very near.  Nothing in the other options complexes would argue with that indication.

____________________________________

Indicator:  VOLATILITY MEASUREMENTS

Status:  NEUTRAL

Comment:  There is not much going on here.  The VIX and VXN remain relatively subdued (which could be good or bad depending on how you look at it - personally, I consider it neutral).  Our other volatility measurements, however, remain a bit troublesome if looking for a low in here.  The 10-day average daily range of the S&P 500, currently at 1.9% of price, is about half of where it should be.  Volume has not picked up at all, and the TICKS still do not have the range they "should" have.  All of this suggests that we have not seen ANY of the volatility signs that normally accompany an intermediate-term bear market low.  While not bearish in and of themselves, the current readings certainly are not bullish.

Bottom Line:  We need to see a pick-up in volatility to be more confident that we are close to a bear-market low.  Current readings are lukewarm, and that is not usually what precedes good long-side bets in a downtrend.

____________________________________

Indicator:  TRIN

Status:  NEUTRAL

Comment:  I'm going to cheat here and repeat what I said two weeks ago, since things have not changed one bit.  The 10-day TRIN for both the NYSE and Nasdaq continue to recover from the near-perfect sell signals they gave in mid-January.  While the selling pressure has not been as heavy in the Nasdaq, it has been enough to bring the NYSE TRIN to an elevated level, though not yet extreme.  The 10-day NYSE TRIN did come close to 1.50 during the week, but I would need to see a reading closer to 1.60 or, preferably, 1.70 before reading too much into the data here.  One troubling development is that the Nasdaq TRIN is actually significantly closer to its lower (bearish) trading band than it is its upper band, suggesting a complete lack of intense selling pressure.  This is something normally found near market peaks, not troughs.

Bottom Line:  These indicators suggest that we have not yet seen the type of short-term intense selling pressure that normally accompanies bear-market lows.

____________________________________

Indicator:  STEM.MR MODEL

Status:  NEUTRAL

Comment:  This model continues to do a decent job of identifying the intraday selling extremes that have lead to some very short-term upside over the past couple of weeks.  However, the current reading of 29% is comparable to readings that have coincided with the peak of those recent moves.  Since mid-January, the model has not dropped appreciably below its current reading before the downtrend resumed.  If the pattern holds, then the upside prospects for the market early next week should be limited.

Bottom Line:  This model suggests that short-term sentiment is neutral, and that has been all it has taken recently for the downtrend to resume its natural course of action.

____________________________________

Indicator:  SEASONALITY

Status:  NEUTRAL

Comment:  There is nothing particularly outstanding this coming week, seasonality-wise.  We will have a slight positive expiration-week bias working for us, but that's about all.  If that pattern holds, any strength seen should be early in the week with weakness on Friday.

Bottom Line:  Not much going for or against us here.

____________________________________

Indicator:  COMPOSITE MODEL

Status:  NEUTRAL

Comment:  Two weeks ago, I said that although this model had reached a high level, it likely meant that we were entering the highest-risk time for the market, and we did indeed drop about 6% from that time.  Even with that drop, though, the model failed to become more extreme, instead staying relatively steady around the 66% area.  This is mainly due to the fact that the 10-day advancing issues ratio has not become extreme, as noted above, and the VIX has been rather subdued.  If both of those measures would have become extreme sometime during the past week, this model likely would have matched the readings seen at the other intermediate-term lows seen during the bear market. 

Bottom Line:  Two weeks ago, I mentioned that once the model hit the 70% level, on average the S&P bottomed about 8 days later and 9% lighter.  Last Wednesday marked the 8th day this time around, and as I said we dropped about 6%.  Both compare somewhat favorably with past instances and suggest that we may have already seen the low for this move, though the fact that the model did not continue to increase throws some doubt on that conclusion.

____________________________________

Indicator:  STEM MODEL

Status:  NEUTRAL

Comment:  This model reached its peak at a very high level on February 10th, which indicated that we would likely be about to go through a bottoming process.  Not that we had seen the exact low, but that the following week(s) would be laying the groundwork for a move higher.  Since then, we've seen what we often see after this model reaches an extreme - choppy action that ultimately leads to a more extreme price level before a good, multi-week rally is launched.  Whether or not we are about to embark on a multi-week rally remains to be seen of course, but so far the market is following the typical pattern.

Bottom Line:  While the current level of the model is neutral (but still quite high), the fact that it reached such an extreme and price action is conforming to the typical post-extreme pattern is positive.  It suggests that lower prices should be used to become more aggressive in long-side entries for a multi-week holding period.

Two weeks ago, I suggested that we were at a point where things looked bad enough for traders to sell, but not yet good enough for other traders to buy.  Now that we have dropped about 6% from that level, things have changed a bit.  Now, I believe that any violation of this week's low will be looked at as an opportunity for a large number of traders - and a large amount of money - to go back into the market on the long side.  I believe the negativity is high enough, and the short side crowded enough, that large traders will begin to see value in holding equities for at least an intermediate-term trade.  It is possible that that point has already been reached, although I believe that the rally late last week is more reminiscent of the fits and starts we saw towards the end of the decline in July and not the stuff that is normally seen at intermediate-term lows.  This type of volatility is good, and we need to see more of it.  A steady rise from here is unlikely and unhealthy, and will most likely fail outside of some important positive fundamental catalyst.  While the overall sentiment picture is negative, we still do not have the confluence or extremes that are almost always seen at important low points.  That tells me that the risk is greater that rallies will fail, and I would look to possibly short much more upside.  If, however, we drop back down (and preferably exceed the October lows), I would look to be an aggressive buyer with a multi-week timeframe.

 - 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.