http://www.sentimentrader.com

Wednesday, July 2, 2003  7:50 PM EST

My plan is to have some type of comment this weekend, so tonight’s will be short and applicable pretty much to tomorrow’s shortened session only.  Since it is short and there are no charts, I am sending it via text instead of in the regular PDF format. 

We saw an extraordinarily persistent upward bias in the NYSE TICK today, as it spent a total of about one minute below zero all day.  Not only did it not go negative very much, but it also recorded an unusually large number of extremely high readings.  Our intraday NYSE cumulative TICK indicator, which sums the last 13 half-hourly readings of the TICK, hit a new three-year high this afternoon.  That’s quite remarkable considering the S&P was up barely 1% on mediocre volume, and isn’t even particularly close to a new high.  These types of true extremes are typically not seen coming off of major low points.  At those times, there is still some remnant of two-sided trade.  When we see these types of days where there is just one side (buying) all day, it has normally marked some type of at least a short-term exhaustion point.  It doesn’t always mark the exact high point, but some type of multi-day peak is usually formed within a day.  Frankly, if the S&P can shake off this reading and exceed the 1015 level without taking a rest, I would be amazed. 

This short-term overbought condition is being confirmed by our Down Pressure indicators, which have now entered overbought territory for the first time since June 18^th.  If we get another relatively large up day tomorrow, then the Down Pressure indicators for both the S&P 500 and Nasdaq 100 will be close to setting one-year records of overbought. 

With the STEM.MR model now in negative (for the market) territory for the first time since May 12th, an historic overbought level in the cumulative TICK,  and a plethora of other short- and intermediate-term indicators flashing caution, it appears as though a rally tomorrow would set up a decent shorting activity for the short-term at least.  I would be a little more inclined to look short intraday tomorrow if it weren’t for the positive pre-holiday bias I went over on Monday.  But with these types of overbought readings and the mildly negative seasonality after July 4^th, it looks like we’re setting up for a high-odds attempt to the short side for next week.  Unless the S&P can make it over the mid-June highs and hold there, my preference is firmly on the short side, especially if we have an up day tomorrow.  

Monday, June 30, 2003  10:05 PM EST

This week will be shortened by the July 4^th holiday on Friday.  Holiday trading strategies always pique the interest of traders, so I looked at the biases around this particular day.  It turns out there is not much to get excited about. 

I have written many times before that the typical trading pattern around holidays is that we see strength before the holiday and weakness afterward.  That is consistently true among the major national U.S. holidays, and July 4^th seems to follow that, although it is a weaker relationship than some of the others.   

The table below outlines the performance of the S&P 500 cash index during the four days surrounding Independence Day, from 1950 through 2002.  

KEY:

·          Average Return:  The average percentage change that day from the day prior, close-to-close.

·          % Positive:  The percentage of time the S&P 500 closed higher compared to the day prior.

·          Maximum:  The most positive return seen on that day over the study period.

·          Minimum:  The most negative return seen on that day over the study period.

·          Exceeded High:  The percentage of time that day exceeded the prior day’s high.  Also given is the average amount the S&P went above the prior day’s high if it did so.

·          Exceeded Low:  The percentage of time that day exceeded the prior day’s low.  Also given is the average amount the S&P dipped below the prior day’s low if it did so.

Here is a summary of the information in graphical form: 

From the chart, it is relatively plain to see that the average return and percentage of time the S&P is positive is greater before the holiday than after.  However, and this is a good lesson whenever you are looking at charts, be sure to look at the scales.  In this particular chart, the scales are very narrow, meaning there is not much distance between the maximum and minimum values.  So although at first glance the difference seems large, when you take into account that there is only a couple tenths of a percentage point between values, it doesn’t seem so impressive. 

Here are a few observations about this holiday bias: 

*  The most positive day is the day immediately prior to the holiday, which is Thursday this year.  This is the case in terms of each of the metrics studied, except maximum return.  This day had the highest average return; it was positive more often than any of the other days; when the prior day’s high was exceeded, it ran up more than any other day; and when the prior day’s low was violated, it dropped less than any other day.

*  The most negative day is the day two days after the holiday, which is Tuesday, July 8^th this year.  This day was negative more than any other day (although it was still positive more than 50% of the time); the prior day’s high was exceeded only 44% of the time, and it ran less than any other day once that high was exceeded; and it exceeded the prior day’s low more than 60% of the time, and ran quite a bit once it did so on average.

*  The day prior to the holiday was the least volatile of all the days, most easily seen by the fact that is has both the lowest maximum and lowest minimum returns.  Although it is not posted, this day also has the lowest volume of all the others, which makes sense because traders leave early to get a head-start on the holiday traffic.  If the forecast is for good weather on the East Coast, look for volume on Thursday afternoon to be almost non-existent.

*  While the days following the holiday have a negative tone overall, they also showed the highest maximum return (remember the trend day higher last July 5^th?). 

From this study, short-term traders may be just a little better served looking to the long side later this week, but then shift gears and look to the short side early next week.  This does NOT mean buy the S&P futures on Wednesday, sell them and go short at the close Thursday and hold short over the weekend.  While that may very well turn out to be a winning strategy, this data does not suggest enough of an edge to justify that risk, all else being equal. 

I was asked a couple of times today about the high level of the put/call ratio.  The CBOE equity put/call ratio ended the day at 0.83, which is quite high.  However, per usual, this was skewed to a great degree by a large institutional trade in QQQ (Nasdaq 100 tracking stock) options.  If we back out QQQ options, then the equity p/c ratio drops all the way down to 0.55, which is nothing more than average.  The 10-day average of the non-QQQ equity p/c ratio is still very low at 0.50 after having hit 0.45 on June 20^th.  This continues to suggest that pure equity options traders have not taken to betting against the market just yet, which is bearish from a contrarian point of view. 

Extraordinarily, the OEX put/call ratio hit 0.51 today.  This is the lowest in two years, and one of the lowest in the past 20 years.  There have been only 7 occurrences of this ratio being under 0.55 since 1992.  There have been 34 since 1982, but many of those were in the immediate aftermath of the 1987 market crash.  The following table shows how the S&P 500 performed after this ratio hit such a low level in the past decade: 

We can see that the market tended to have an immediately positive reaction, then gradually became more negative as time went on, finally becoming extremely negative after 20 days.  However, it MUST be noted that six of these occurrences happened after the year 2000, so that obviously has a large impact on market performance.  If we look at the instances prior to these, there are really just two distinct time periods in the 1980’s where there was a cluster of low OEX p/c ratios, which are circled in the chart below: 

In these cases, the market rallied strongly almost immediately after the readings were seen, then chopped around over the next few months.  There is really nothing to conclude here, with such a small sample size, but I wanted to make sure you had an historical perspective. 

As of today’s close, most of our shortest-term indicators are really a mush and I can’t find much of an edge in any of them at the moment.  Perhaps if we get some more downside over the next day or two, we will record some oversold readings to go along with the positive seasonality ahead of the July 4^th holiday, setting up a potentially high-odds long trade into later in the week.  However, any upside we do get continues to look like it will only set up longer-term trades (a time frame of several weeks at the least) on the short side, what with the intermediate-term sentiment condition I’ve been outlining in recent comments.

- Jason Goepfert

Disclosure: long OEX puts

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.