“Range-think” Still Pervasive
Thursday, July 22nd, 2004 7:30pm EST
Rumors of a Death Cross Have Been Greatly Exaggerated
Bottom Line: According to some, the Grim Reaper is about the visit the markets due to one moving average crossing under another. However, history tells us that as usual, such hyperbole does not present us with tradable information.
Due to a blurb on CNBC today, I received quite a few questions about the implications of a “death cross” on some of the major averages. This ominous-sounding signal goes by several names, but basically it occurs when a 50-day moving average crosses below a 200-day moving average on whatever stock or index you happen to be following. This is sometimes referred to as a “golden cross” when it occurs in the other direction (supposedly being a positive for the market). Whatever the name, it is becoming an issue now because the cross has already happened on the Nasdaq Composite and it is perilously close to happening on the Dow Jones Industrial Average. The S&P 500 looks safe for now.
I don’t watch TV during the day, so I did not see the interview, but my understanding is that the mention of this “death cross” was that it was a bad omen for the market and investors should be defensive because of it. As usual when something makes it on to CNBC, there is a grain of truth to this, but only a grain.
Usually when I present the results of studies such as this, for simplicity’s sake I just show the S&P 500. That index tends to be a pretty good proxy for “the market” and the DJIA and Nasdaq usually follow fairly closely. However, in looking at those three major indexes for this 50/200 day cross, the results are quite different between the three, so it’s a good idea to look at them separately. The series of charts below shows the future performance of each of the three indexes after the 50-day average crossed below the 200-day average. To further refine the results, I separated out those times the market was in an uptrend at the time versus those times it was in a downtrend (with “uptrend” defined as a rising 200-day average, and “downtrend” defined as a falling 200-day average – thanks to my friend Tony Dwyer at FTN Midwest Research for the refinement). There are two charts for each index, with the first one showing the average return the given number of days later, and the second showing the percentage of time the index was positive.
A few notes:
1. For the S&P, this cross was a decent short-term sell signal – provided the index was in a downtrend at the time. 20 days after the cross took place, the index was a little over 2% lower on average, with only 18% of the occurrences being positive. But when the index was in an uptrend, the cross actually resulted in a market that was positive more often than average up to 30 days later.
2. For the S&P, whether in an uptrend or downtrend, the cross resulted in longer-term performance that was sub-par. 60 days out and after, the index showed both an average return and percentage of time positive that was inferior to an average time during the study period.
3. Cross signals in the Dow, which has the most history behind it, were unusual in that the best performer in terms of average returns were those when the market was in a downtrend. There were a total of 25 of those signals, and the Dow handily outperformed an average period from 30 days to six months later. Cross signals when in an uptrend were generally poor performers, and the Dow often struggled to remain positive across all time frames.
4. The Nasdaq, which has the least history, was also the best performer (no shock there). In fact, these “death cross” signals were actually pretty good buy signals, especially when the index was in an uptrend. 30 days after a signal, the Nasdaq was up about 80% of the time, sporting an average return of about 4%. Across all time frames, performance after these signals (when in an uptrend) beat an average period. The most recent cross, which just occurred on June 22nd, has seen the worst performance out of all of them. Interestingly, the two other times the index was down 20 days after one of these 50/200 cross signals, the index was higher after 90 days, showing a gain of 17% once and 4% the other time.
These are contradictory and somewhat confusing results. For the S&P, a cross of the 50-day average under the 200-day average lead to underperformance going forward, particularly if the index was in a downtrend. For the Dow, the best return was had by buying these signals during just those times (downtrends), but it had difficulty moving ahead when the cross occurred during an uptrend. On the Nasdaq, the signals were actually pretty good buying opportunities, particularly if the index was in an uptrend.
To me, this type of signal would have more meaning if there was some consistency among the historical examples and across the various indexes. The fact that performance varies so greatly among the three is probably due more to the how far they go back than that they behave so differently. Still, to have confidence in the data (and think that it has some kind of edge), we should see either generally bullish behavior after these signals or generally bearish behavior. The fact that we see neither – consistently – is a sign that these “death crosses” are overblown (probably because they look nice on a chart), and we best concentrate our worries on more worthy data.
A Bit of Pessimism in Options Market (but just a bit)
Bottom Line: A new index by one of the largest options exchanges shows traders exhibited a bit of pessimism yesterday, but it is nowhere near an extreme.
One of the side benefits of sentiment analysis gaining more attention is that more companies try to jump on the bandwagon by offering up their unique take on any data they may have. Ameritrade releases an index of their customers’ trades, UBS releases a monthly sentiment poll, Merrill Lynch creates a put/call ratio based on customer trades, etc. One of the newer entrants is the International Securities Exchange, or ISE, which over the past year has often surpassed the venerable Chicago Board Options Exchange (CBOE) in terms of the volume of trading in stock options. The ISE has created what it bills as a “better” put/call ratio, in that it includes only long, opening transactions in its volume figures, unlike the CBOE which includes everything. Click here for a more detailed discussion of opening and closing trades.
Despite its best intentions, the ISE really has not improved on the traditional CBOE volume figures, at least not from what I can see. The data only goes back to October 2002, but so far I am unimpressed – especially when the data is viewed on a moving average basis. However, the daily figures sometimes can give some insight, and like the figures from the CBOE, the data appears to be best interpreted in a contrary manner.
The way the ISE index is calculated, a reading over 100 means that customers have opened more long call positions than long put positions. Generally, this would mean they are more bullish than bearish. A reading under 100, of course, would then mean that customers have opened more long put positions than call positions, and would typically mean that on average they are more bearish than bullish.
Once the bull market got underway in earnest last year, it was rare to see a reading under 100. In fact, before mid-May of this year, the last reading under 100 occurred on August 14th. Just looking at the readings so far this year, there is a definite contrary tilt to the data. Any time the index closed the day at 125 or below (meaning traders were relatively pessimistic), the S&P was higher 10 days later 8 out of 10 times, for an average return of 1.4%. However, when the index closed at 200 or above (meaning they opened twice as many long calls as long puts – a very bullish statement), the S&P was higher only 12 out of 32 times, with an average return of -0.5%.
Today’s figure was neutral at 153, but after yesterday’s carnage the index closed at 125. That’s not extremely low, but it’s a start. We should continue to watch the index closely, and if we see a value of 100 or below, it may be a good sign that we are finally seeing the type of negative attitudes we need to form a better risk/reward proposition from the long side.
Conclusion
On Tuesday, I highlighted the grossly overbought nature of our shortest-term measures, and this morning I pointed out their opposite condition. The market has gone from one extreme to the other (and nearly back again) in two days. This type of volatility is good to see, because as we’ve seen many times before, turning points often come during times of extreme volatility. However, we still are not seeing any true scramble for protection or speculation on a falling market, unlike the previous two lows in March and May this year. It continues to look like the majority of traders believe in a trading range, and so we are not seeing an up-tick in bearish bets as we fall, unlike our prior lows. It would be healthier – and set up better risk/reward trades from the long side – if we could get some evidence that traders had exorcised their bullish bias to some degree. There are some minor positive signs out there (e.g. volume in the S&P 500 exchange-traded fund, SPY, was extremely high today – a hint of uncertainty), but not enough to make the long side particularly tempting.
In the short-term, most of our measures are at least back to neutral, and some back to overbought. In the context of a downtrend, that’s usually all that’s required for the selling to pick up again, so if we continue to decline tomorrow and into next week, it will be obvious that we have not yet seen a longer-term low. Initial reactions off intermediate-term lows typically continue rising right through short-term overbought conditions, and we have seen no evidence of that so far.
Jason Goepfert
President and CEO
Sundial Capital Research, Inc.
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.
© 2004 Sundial Capital Research, Inc. All Rights Reserved.