New Highs and Oversold?
Monday, December 15, 2003 9:15 PM EST
Down Pressure
Bottom Line: Heavy selling pressure most often leads to positive market performance (at least in the short-term).
On Thursday, I noted the lopsided buying pressure that we saw that day, and said that statistically, it is usually a better idea to use such strength to sell equities rather than buy, at least in the short-term. Despite all the hoopla over the past couple of days, the market is on track to keep those probabilities in line, as we closed today below where we did on Thursday. Now, however, we have the opposite scenario, at least on the NDX. Today’s daily Down Pressure reading on the NDX came in at 90% (the reading on the site is a 3-day moving average of the daily readings). You will recall that this measure is calculated by looking at how many points each of the 100 components of the Nasdaq 100 gained or lost on the day, as well as the amount of volume that flowed into each. A reading of 90% means that on average, 90% of the total points gained or lost today were lost, and 90% of the volume went into stocks that closed down on the day. The figure for the S&P 500 was not as severe, as it recorded a reading of 79%.
Still, I think it can be instructive to see what has happened in the past when selling pressure became so lopsided. Since the rally began in March, there have been 24 days with an NDX Down Pressure reading of 90% or greater. In 71% of the cases, the NDX closed higher the next day, with an average gain of 0.6%. The results were similar after 3 and 5 days, but with gains of 1.0% and 1.5% respectively. Probably least surprising of all, after 10 days the NDX was higher a whopping 91% of the time, or on 20 out of 22 occasions. The index was up an average of 3.3% ten days later, with a maximum gain of 9% and a relatively miniscule maximum loss of 2%.
This is the third 90%+ daily Down Pressure reading we’ve seen in the past 10 days. Since the rally began, such a concentration has only been seen three other times – the periods ending March 31st, June 23rd and September 26th. Obviously, the results 10 days later after those instances were very positive, as the NDX was higher each time, with an average gain of over 6%. It’s certainly tempting to suggest that since we’re seeing so much selling pressure now, so close to a new high, that a more substantial correction may be upon us. But that was a tempting suggesting after the three other occurrences, too, and it was not to be.
60-Day Highs and Momentum
Bottom Line: The lack of momentum we’re seeing now is extremely unusual.
Momentum can be defined a hundred different ways, from the simple to the complex. I prefer to use very simple measures that have stood the test of time, and that suit me fine enough. I use mostly what’s already posted to the site, like a 10-day average of the advance/decline line, or up volume, or cumulative TICKS. Those types of measures tend to give me a feel for how much pressure is being applied to the current move. What we’re seeing now is extremely unusual, in that we made a new 60-day high in the S&P 500 today, but the 10-day moving average of up volume as a percentage of total volume is below 50%. In fact, going back to 1997, such an occurrence has happened only six other times. Notably, our current reading is the lowest of them all.
Over the past six years, when the S&P has made a new 60-day high, the average reading of the 10-day average of up volume has been 57%. That has stayed relatively consistent – if we look at new 60-day highs for 2003 alone, then the average 10-day up volume reading has been 59%. As I said above, our current reading is the lowest one in over six years, at just over 48%. Of the six other occurrences where up volume has been under 50%, the average is 49%. So, the ones that have been under that 50% mark have been just barely so. Clearly, it is very unusual to hit a new 60-day high with this oscillator more close to oversold than overbought.
Historically, since there are so few instances, it’s difficult if not impossible to draw any meaningful conclusions from what happened going forward. One thing that I think may be instructive is that it took the market another 20 trading days on average before a new 60-day high was achieved. There are two real exceptions to this – June 1998 and December 1998, where new highs were set 1 day and 2 days later, respectively. The chart below highlights the six occurrences in orange.
Twice in 1998, we saw these “oversold new highs” prompt nothing but another squirt to more new highs over the next few days. Of course, the one in June of that year eventually lead to the crash that fall, but I think it would be spurious to correlate this discussion with that crash. Another instance in June 1999 pinpointed the exact top for that move, which lead to some quick and severe selling.
It is interesting to note that this type of behavior occurred in December of 1998 and December of 1999. I’m not sure if it has anything to do with “December” per se, but both instances lead to another move to new highs in January, then choppy to down action for the next several months.
Momentum has clearly been waning on the last few pushes to new highs, especially on the Nasdaq. By one measure of momentum, defined above, we’re now seeing an historically unusual occurrence, in that we’re nearly oversold at the same time we’re hitting new highs. A review of the precedents doesn’t yield much useful information, though it is perhaps notable that the two other December occurrences lead to a push to new highs in early January, then consolidation or decline for the next few months.
December Ends
Bottom Line: Shorts have to peddle into the wind in the coming weeks.
One of the most common arguments against betting against equities for the remainder of the month is the historically positive seasonality. I showed a few weeks ago that the weakest part of the month was from approximately the 8th through the 15th. So, now we’re through was has been the statistically weakest period of December, and can look forward to the consistent positive bias around the Christmas and New Years holidays.
I took a look at the last 10 trading days of December to see if there is any precedent to large declines during that period. If you go back long enough and have enough samples, you can usually find a precedent for just about anything, and this is no exception. Over the past 40 years, there have been 8 years that saw a decline of more than 3% during the last two weeks of the month. On balance, though, this is a tough battle to fight from the short side.
If we take just a static look at how the S&P 500 performed from 10 days prior to the end of December through the last day of the month, 31 out of the past 41 years have shown a positive return, with an average gain of 1.3%. The maximum loss seen during those two weeks (i.e. maximum drawdown if one was long) averaged 1.6%, with the largest loss being 6.5% in 2000. 24 out of the 41 years, or 59%, saw a drawdown of at least 1%. Of course, that means that there were 17 years where the S&P didn’t decline even 1% during those two weeks. On the other hand, 31 years, or 76%, saw a maximum rally of at least 1%, and the average was 2.5%. There were 21 years, or over 50%, that saw a rally of at least 2%.
I’ve said before that I like to think of seasonality as a gentle breeze blowing either with you or against you. It’s usually not enough to change your direction (except maybe on 3 or 4 days each year), but it helps to have it at your back. And for the next few weeks, the breeze will be a little stiffer than usual, and it will be squarely at the bulls’ backs.
Conclusion
For many weeks on end now, I have been longer-term neutral, as I just have not seen a high-odds opportunity on either side of the coin. While the S&P has been able to muster some small gains during that time, the Nasdaq has not (the Nasdaq looks like an especially dangerous place to me, and I would not want to be long any of those types of issues). Currently, I see no change to that outlook. I continue to believe that pushes to new highs will suffer a similar fate to what was seen today, while smallish declines of 3%-4% will find buyers, at least through the beginning of 2004.
Shorter-term, today’s reversal obviously was not encouraging from the long side, but it can’t be too much of a surprise to anyone given the hyperventilating done all morning in regards to the possible positive outcomes of Hussein’s capture. These types of reversal days have not had a good record at predicting future weakness, especially if looking out longer than a few days. In addition, many of our shorter-term indicators are now mildly oversold. In fact, if the Rydex traders threw in their towels today and reversed some of their bullish leanings, then our short-term indicator “score” could become quite positive (for the market). That would get me to look at any further weakness over the next day or two as a short-term buying opportunity. Right now, however, it’s not enough for me to want to be looking long, especially with the kind of short-term selling a nasty reversal day like today can create.
- 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.