THURSDAY, MAY 1, 2008
Trend Day Stays Intact, But Can It Hold?
05/01/08 3:00 PM EST
 |
|
|
As of: SPX 1375 |
|
I mentioned earlier today that other than relatively weak breadth figures (which subsequently improved a bit), the indices were showing some classic signs of putting in a trend day to the upside. They've run with that so far, holding up extremely well going into the final hour.
I've been noting some nagging negatives, each of them minor but still troubling. There are a few more, such as what happens when the S&P 500 jumps 1% or more the day following an FOMC meeting (it showed a negative return 63% of the time over the next three days, averaging -1.1%) or when it is up 1% or more on the day prior to a payrolls report (it showed a negative return 70% of the time over the next week if the S&P gapped up the next morning, averaging -0.8%). If it gaps up the morning of the payroll report, it tends to carry a bit further into the close, but then give back the gains over the subsequent few sessions.
There are several of these statistical biases we've gone over, each of them pointing to at least a mild negative reaction to today's upside thrust. Given a couple of other signs of too much optimism we've discussed, I am not enamored with the idea of trying to chase this breakout (especially since over the past week, we've been discussing the possibility of a "false" breakout over the most-obvious-resistance-level-in-history at 1400 on the S&P).
If the S&P can continue its rally and then consolidate the gains while holding above 1400ish, then I'll look to the long side again, but not now. I've been trying very hard to stay objective and not find things specifically negative about today's rally even though I'm not participating, but the only things I've been able to come up with are at least moderately negative going forward.
As for shorting it, that would be a better probability if we continue higher into the close (increasingly likely given the trend-day qualities we've seen so far), and gap up tomorrow morning based at least in part off the payroll report. Unless that happens, I'll be staying mostly out and watching longingly from the sidelines.
An Unusual Day
05/01/08 11:25 AM EST
|
|
|
|
As of: SPX 1375 |
|
While we could point to a couple of potential reasons for some upside today in stocks (the first day of May has been up 9 of the last 13 years in the S&P 500 tracking fund, and stocks tend to reverse their post-FOMC moves), it's unusual in another respect.
Since the October top last year, stocks and bonds have moved together with a high degree of correlation, especially on days when either makes a big move. With bonds enjoying a nice rally today, the yield on 10-year Treasuries is down about 1% after rebounding a bit off its lows, so stocks "should" be lower.
Since October, I show that yields have dropped 1% or more 49 times. Of those 49 instances, the S&P 500 was negative on the day 38 times (78% of the time) by an average of -0.8%. There has been a clear pattern of money fleeing stocks and buying bonds, commonly referred to as a "flight to safety". This is also usually reflected in the Euro/Yen currency cross, which usually rallies on big days in U.S. stocks, but that is falling heavily today.
There were 8 times when the S&P closed higher by more than +0.5% and yields dropped by more than -1%. Over the next few days, the S&P showed a positive return 3 times but the overall average was -1.3%. Of course, part of that is simply a reflection of the downtrend we've been in since October. If we go back a few years, then the pattern was still to see weaker-than-average returns, as the S&P was positive three days later less than 50% of the time and showed an average return of -0.2%. Not a big edge there either way, but it's an unusual situation and I wanted to point it out.
It's also odd that the broader market is acting relatively limp while technology screams higher. I went back over the past eight years and looked for any time that the Nasdaq 100 was up by more than +1.5% by 11:00am EST, but the S&P 500 was up less than +0.5% and breadth on the NYSE was less than +750 (i.e. fewer than 750 more stocks up on the day than down). There were 19 occurrences, and through the next day's close, the S&P was positive only 7 times, with an overall average return of -0.4%.
There is also the "Fed reversal reversal" pattern. When the S&P closes in negative territory the day of an FOMC meeting, then rebounds the following day, buying that close and holding for a week resulted in only 9 winning trades out of 25 attempts since 1996, with an overall average return of -1.0%. This is something we've touched on quite a bit over the years, with a tendency to see reversals of reversals after Fed meetings.
All of these are just ways to try to quantify today's relatively unusual behavior. Surely there are other ways which may paint a more optimistic view (the indices are showing some initial signs of an upside trend day, but breadth is not nearly as positive as we typically see on those days), but I'm not finding much of anything supporting a sustained run higher. The only patterns I can find are pointing to neutral to at least slightly negative returns in the short-term.
When we combine that with the fact that we're still in a long-term downtrend, and the few signs of too much optimism and low volume we went over earlier today, I continue to think any upside here will more than likely be given back over the next several days/weeks...but I'm still not finding what I would consider a solid enough edge to put on aggressive short positions. Overall, this is confusing market action and I'm staying mostly out on a trading basis.
Lowest Volume in 28 Years?
05/01/08 10:20 AM EST
|
|
|
|
As of: SPX 1375 |
|
Good Thursday morning...We begin the day with a slight negative bias in the pre-market futures, as selling pressure came in after the release of a handful of economic reports earlier this morning. The ISM report due out at 9:00 could also be a market-mover, particularly if we see higher-than-expected inflationary pressures. Other than that, tomorrow morning's jobs report will take center stage.
Anyone who's read this site over the years knows that I consider volume to be a tertiary indicator (at best), particularly when considering it on a market-wide level. It's simply not as big a deal as it's made out to be, especially in terms of the dubious concept of "accumulation" and "distribution" days.
That said, I do pay attention when aggregate volume makes an extreme move, especially to the upside. Exceptionally high volume has historically been a good buy signal for equities, and that has not changed over the years.
We're dealing with the opposite situation now, as I've touched on in a couple of recent notes (see here and here). I was struck again last night when reviewing some month-end stats about just how poor volume was last month.
When looking at a simple chart of monthly volume, though, it's easy to be mislead because volume exhibits a definite seasonal pattern. Turnover tends to be higher-than-average during the early months of the year, and again in the fall. But it is often extremely low during August and December. That makes it more difficult to figure out what is "low" volume and what is "high" volume.
Therefore, when considering what "average" is, it's best to de-trend or seasonally adjust volume. Towards that end, I created a seasonally adjusted monthly volume figure for NYSE data, using a procedure similar to what the government does to adjust most of its data. I then checked to see how much each month's ending volume deviated from its seasonally-adjusted norm. The chart below shows this figure.
We can see that big spikes higher in the ratio have been bullish, no surprise there. The deviation reached +2% in August 1998, July 2002 and August 2007, all times of upheaval in the market.
But look at the past six months. Last month, along with October 2007, showed deviations of greater than -2%, the only two months in the past decade to do so. In fact, they're the only two months I could find going back to 1980. January 1988 came close, but didn't reach quite this kind of extreme (stocks chopped around, edging slightly higher, for seven months after Jan 1988).
High volume means that there is a lot of disagreement among traders about what value is. Uncertainty is rampant, and we often see large gyrations in prices because of it. As things settle down, prices tend to rise. Low-volume conditions are the opposite. A lack of turnover suggests that a relative amount of complacency has spread, and that tends to not work out so well going forward.
Last week, I mentioned that one of the reasons I was starting to tone down my intermediate-term expectations (besides the exceptionally low volume) was due to the latest reading in the American Association of Individual Investors sentiment survey. The percentage of respondents considering themselves negative on the market had dropped by a large enough degree to be concerning (relative to other readings over the past few years).
That didn't change during the latest week - in fact, it got a little worse. The bearish percentage fell to 26%, while the bulls jumped up to 53%. That bullish percentage can now be considered high on an absolute basis, and extreme on a relative basis. The last time we saw this many bulls was last October, which obviously wasn't a very good time to be among those bulls.
Yesterday afternoon's reversal was atypical of most post-FOMC reactions given the overall positive bias the indices had going into the announcement. I've tried to look for other precedents, but the sample sizes get so small there's really nothing we can read into them. Ignoring the fact that the FOMC met yesterday, instead I took a look at any time the S&P 500 hit at least a three-month high during the day, but then reversed enough to close below the low of the prior two days.
This, too, was relatively rare. Using data for the S&P 500 tracking fund (SPY), there were five other occurrences in that fund's history. The next day, SPY was lower four of the five times by an average of -1.2% (the lone up day was a meager +0.2%). That weakness tended to persist over the next few days, as the S&P was still lower three days later four of the five times. Any further out than that, and the returns reverted closer to random. For those curious, the dates were 1/3/00, 12/31/04, 1/3/05, 12/16/05 and 12/7/06. I have no idea why all the instances were in December or early January.
Using the S&P 500 cash index, we get 15 occurrences, and saw downside follow-through the following day 8 times with an average return of -0.3%. That again continued over the next couple of days, then returns reverted back more towards random.
Overall, there appears to be a moderate negative bias following reversals like yesterday, but that kind of goes against the norm of seeing a reversal of any extreme move following an FOMC decision. While the price losses yesterday weren't all that great, the reversal was fairly substantial. Bottom line, given some of the other modest negatives we've gone over during the past week and this morning, I'm more inclined to continue to side with the thought that it's still going to be tough for the indices to garner much sustained upside here.
All the best,
Jason Goepfert
President and CEO
Sundial Capital Research, Inc.
Forwarding or otherwise distributing this copyrighted material is a breach of your subscriber agreement. Violators are subject to termination of their subscription with any received subscription fees forfeited. Any references to historical performance are based on data we deem to be reliable, but are based upon feeds from third parties. We do not recommend subscribers take positions based on data presented here alone, but rather incorporate it into a comprehensive investment outlook.
© 2008 Sundial Capital Research, Inc. All Rights Reserved. www.sentimenTrader.com