http://www.sentimentrader.com

Wednesday, September 10, 2003  9:20 PM EST

Even though the S&P 500 index has lost only 21 points over the past two days, or a little over 2% of its value, a look at the components of the index shows that the selling has been more severe than the index itself would indicate.  Due to the way the index is calculated, a positive move by one or two large-capitalization components can mask negative moves by several smaller ones.   

The Down Pressure indicators for the S&P 500 and Nasdaq 100 that are posted to the site are computed using the actual points gained or lost (along with the volume flowing into those issues) for each of the 600 individual stocks in those indexes.  They’ve now worked off the extreme short-term overbought condition they entered in the first days of this month and are heading towards oversold.  In fact, even if we have only a flat to mild down day tomorrow, both indicators will become firmly oversold.  If we instead suffer another relatively large down day (say 10 points on the S&P and 20 on the NDX), then both indicators will almost certainly become the most oversold they have been in eight months.  That would argue strongly for at least a temporary bounce at that point. 

Yesterday, the stocks in the S&P 500 that rose on the day gained a total of 29 points, while those down on the day lost 193 points.  Today, 31 points were gained while 286 were lost.  So, over the past two days, a total of 419 points were lost in the component stocks.  This is the largest two-day loss in points since March 10^th/11^th.  The chart below shows instances over the past year where there was a two-day net loss of 400 points or more in the components of the S&P 500.  You can see clearly that this is something that we haven’t encountered since the Spring rally began. 

The headline put/call ratios rose significantly from where they have been over the past week, but once again it was due to QQQ options.  QQQ put options accounted for 43% of all equity put options, which shouldn’t be terribly surprising anymore.  The equity put/call ratio with QQQ options removed actually dropped from yesterday (from .56 to .51).   

The spread between the total put/call ratio and the equity put/call ratio with QQQ options removed is now the widest since 2001, when QQQ option volume became available from the CBOE.  The chart below shows a 21-day average of the two ratios plotted against each other, and it is clear just how wide of a difference we’re now seeing:

The total put/call ratio is well within its trading range; not particularly extreme in either direction.  The “pure” equity put/call ratio, however, is challenging the lowest reading in its history reached in June.  If you consider the “pure” ratio more reflective of individual investors (which is a big part of contrary analysis), then I would think this should be a cause for worry. 

As of yesterday’s close, the “score” for our intermediate-term indicators dropped to 62%, the second-lowest reading dating back to December 1998 (the lowest reading prior to yesterday occurred on 1/8/02, which kicked off a month-long decline in equities).  This translates to an average score for each of the 24 indicators of about 1.0.  The way the scores are determined, a reading of 1.0 would mean that the indicator is showing enough optimism that it is at least one standard deviation from its mean value.  To see such a low total score, we have to be seeing a large confluence of “too optimistic” readings across a broad array of measures.  That is easily seen in one quick glance on the site, as there is row after row of red check marks. 

On a short-term basis, some of our more responsive indicators have cycled down enough to push the STEM.MR model into positive (for the market) territory for the first time since August 4^th.  While the model isn’t particularly extreme at this point, it usually doesn’t have to be if we are still in the midst of a strong uptrend.  You usually only need to see a mildly oversold short-term condition before the uptrend resumes, if indeed the trend is still in place.  This is where watching measures such as this can provide a lot of value – it lets us know if the character of the market is changing.  If we cannot stage a decent rally by Friday at the latest, it will suggest that there is not enough demand to take us to new highs. 

I’ve been saying that the market still has a bullish overall structure as long as we are above 960 on the S&P 500, but a break of 1015 would serve as an initial warning sign that a more serious correction could be setting in.  The fact that we’ve already sunk back into the June-August trading range so soon after breaking out, combined with the terrible sentiment condition we have, increases the chances substantially that the breakout was “false” and we will be heading lower.  While in all probability we will see a rally attempt back to challenge Monday’s highs, be very watchful for a failure of that rally.  If the market shows signs of faltering near that level, or if we cannot even muster much of a rally at all, then I will be focusing on the short side once again.

- 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.

Monday, September 8, 2003  9:00 PM EST

Last week, I mentioned that due to the extremely low equity put/call ratios (minus QQQ options) that we were seeing, we would also likely see an extremely low R.O.B.O. (Retail-Only, Buy to Open) put/call ratio.  Recall that the ROBO put/call ratio is constructed using buy-to-open transactions only for trades under 10 contracts.  This is as pure a contrary indicator as you’re going to get. 

This did indeed prove to be the case, as the ROBO p/c ratio came in at 0.47, meaning these smallest of options traders bought more than twice as many calls as puts.  Perhaps most telling, however, is the fact that these traders paid an average of only $161 per contract for their puts, but they paid an average of $213 for their calls.  This is a 32% “demand premium” for calls over puts, and is the most egregious spread since the height of the bubble in August/September 2000. 

Also troubling is the priority these traders gave to their various strategies for opening transactions.  34% of the volume went to buying calls, 34% to selling calls, 16% to buying puts and 16% to selling puts.  Recall from last week that I said whenever call buying exceeded call selling, the market typically had trouble afterwards, and we’re close to that point now.  Overall, these traders distributed their priorities evenly among bullish and bearish strategies, which also is a warning sign that the traders who are usually most wrong on market direction are now quite comfortable with the upside here. 

For the purposes of a contrary indicator, I prefer to concentrate on the smallest of traders, as they are consistently those who are the most wrong the most often.  However, if we cast out our net a bit further and look at all options traders (those who trade under 10 contracts, 11-49 contracts and 50+ contracts), we get a buy-to-open put/call ratio of 0.60.  This is the highest number of calls bought relative to puts since January 2002, as all trader groups are showing put/call ratios on the lower edge of their trading range over the past couple of years.  This signals that there is broad-based optimism towards a continued price rise in equities. 

Today’s CBOE equity-only put/call ratio came in at an extremely low 0.38.  This is the lowest reading since June 20^th.  The only other reading in the past three years as low was 0.38 recorded on 8/16/02.  If we look at my preferred ratio by taking out QQQ options, then the put/call ratio drops 0.26, a new record low by a wide margin.  However, whenever I get such an extreme reading for anything, I like to double-check the information.  In checking with the CBOE, they state that they may have an issue with today’s data, and have pulled the option totals from their website.  When they are able to verify today’s numbers, they will post them again.  Due to this uncertainty, I would hesitate to read too much into any of the figures at this time. 

Also troubling is the AIM model.  At 35.2%, this combination of the four major sentiment surveys is the lowest since June 20^th, which in turn is the lowest since the Spring of 1998 (low model readings correspond to low levels of pessimism – the lower the model, the higher the confidence of investors).  It is not just one or two surveys that are skewing the model, as each of the four are showing levels of optimism that are somewhere between high and “nosebleed”.  The reason this is usually trouble for the market is because these surveys cover a broad swath of investor groups – from small individual investors to newsletter writers to professional futures traders to brokerage firm analysts.  When so many diverse traders are positive to an extreme on the market, it has proven to be the death knell for equities for four years without exception.   

Prior to 1999, when the S&P was in a steady uptrend, we saw extremely low model readings in 1996, 1997 and 1998.  None of those occurrences lead to any significant downside – if anything, the market paused for a few months while some of the excess receded.  In June of 1997 and February of 1998, when the model reached the lowest levels in its history (suggesting way too much optimism of the part of investors), it not only did NOT lead to further downside, it kicked off a multi-month rally instead.  That’s why it is so important to consider these readings in the context of the market environment, and why it is dangerous to rely too heavily on overbought readings in a strong uptrend.  Up until last week, I considered the bear market structure intact, and so I was willing to give low AIM model readings their full due.  With the breakout last week, that structure has changed and so overbought readings must be given less weight in my opinion.  However, there is no question that the extremely high level of investor optimism in 1998 helped exacerbate the plunge in September of that year, which is why I insist on focusing on “stop” levels here.  If we drop below 1015 on the S&P 500, then that is the first warning sign that something is wrong. 

We now have such a confluence of extreme optimism among such a broad group of traders and investors that our Composite model recorded a new all-time low reading today, dating back to January 2000.  Today’s single-day reading of 15% is the lowest ever recorded by the model, with second place going to 18% recorded on 3/12/02.  This has moved the 5-day average (which is posted to the site) down to 23%, which is tied for the lowest in the model’s existence.  The argument here, of course (and it’s a valid one), is that we are in a different market environment than what we have seen previously. 

Market breadth is now what I would call grossly overbought.  Over the past 10 days, advancing issues have averaged 60% of total issues traded (ignoring unchanged issues), and up volume has accounted for 64% of total volume.  If we combine these two readings, our current condition is in the top 4% of readings since 1965.  While conceptually this seems as though it should be bearish because it is unsustainable (the only other readings in the past few years that approach our current level occurred on 3/7/02 and early June 2003), historically that has not been the case.  Since 1965, the S&P 500 has actually tended to OUT-perform a random return after seeing such overbought breadth, which the table below outlines: 

KEY:

Avg Ret:  The percentage change in the S&P 500 from the date of the overbought reading to the given number of days later.

% Pos:  The percentage of time the S&P 500 closed higher the given number of days later.

Max:  The highest return seen the given number of days later.

Min:  The greatest loss seen the given number of days later.

For each time period observed, from 1 to 90 days out, the S&P showed a positive average return and closed positively greater than 50% of the time after such overbought breadth readings.  From 10 to 60 days forward, the S&P was higher greater than 70% of the time, with an average return typically much higher than that of a random period.  While overbought readings in a downtrend certainly beg to be sold, when we see overbought conditions in an uptrend it can actually be a better buy signal than sell signal. 

Without question, we are seeing some signs of speculation that border on mania.  When the Rydex asset flows are released tonight, I suspect that we’ll see more of the same.  I don’t think we’re seeing enough that it would make sense to short new highs for anything other than a day-trade (if that), so at this point I am still giving the uptrend the benefit of the doubt.  However, if we get more manic-type readings like we’re beginning to see now, I will be inching up the “stop” levels that would get me out of any long positions.  I’m not ignorant of the fact that more and more traders who were once bearish on the market are now turning at least cautiously bullish (such as myself), so we are probably closer to a meaningful top than we’ve been in some time.  Currently, I’m using the widely-watched 1015 on the S&P 500 as an initial warning that a more serious correction may be at hand if it’s broken to the downside.  If we break 960, the positive market structure that I have been mentioning will be invalidated.

- 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.