Welcome to a new quarter, back from the long holiday weekend, and to the start of another earnings season.
There are many cross currents going on, and today, we want to review one in particular. What happens after Markets suffer very large 1 month declines.
Several words of caution are due first: Traders can get into trouble when extrapolating. This is especially true when dealing with relatively small data sets, and fifteen examples over 2/3rds of a century certainly qualifies as small. Additionally, the causation/correlation issue arises especially in small samples (random or truly causative?)
Caveats aside, let’s have a look at the 15 prior biggest one month Dow selloffs. As the chart below shows, we often see a healthy snap-back after significant one month sell offs 6 and 12 months later. Except when we don’t, such as 1973 and 2001, where we see even larger losses one year later.
>
via Jake
>
What does this mean for traders? Well, if 13 out of 15 sounds like good odds to you, then you will be tempted to play this sell off for a bounce.
Perhaps adding another layer of questions to your analysis might be helpful. Is the current environment — high inflation and oil prices, modest growth, large write-downs and profits declining from a peak measure, and relatively low fear levels — more similar to 1973 and/or 2000, or is it closer to the other 13 periods?
Answer that question and you have your trade.
And speaking of fear: A recent Bloomberg survey of Wall Street strategists shows they have little or none. Instead, for the second half of 2008, they are forecasting the best S&P 500 since 1982.
Whether their analysis is similar to the chart we shown above, or is simply more of their institutional tendencies towards perma-bullishness, we cannot say.
Regardless, here is a quick excerpt (note the unusually skeptical tone of the Bloomberg piece).:
"Deutsche Bank AG, Lehman Brothers Holdings Inc. and UBS AG say the Standard & Poor’s 500 Index will gain the most in 26 years during this year’s second half. That isn’t going to happen, if history is any guide.
The S&P 500 will rise 18 percent by January, according to the consensus projection of 10 U.S. strategists surveyed by Bloomberg. The forecasts are based partly on estimates that profits will jump 50 percent in the fourth quarter after falling for the past year.
Even if that happens, it may not be enough. In 2001, the last time profits fell as much, they then had to climb for three straight quarters before stocks rebounded. Analysts’ earnings estimates for this year still represent a decline from 2006 levels, making the strategists’ optimism harder to justify, investors say."
There verywellmight be an 18% bounce out there somewhere, but I suspect it come from lower prices, more deeply oversold conditions, showing greater levels of fear and panic.
>
Source:
Deutsche, UBS Fight History Forecasting Best S&P 500 Since 1982
Eric Martin and Michael Tsang
Bloomberg, July 7 2008
http://www.bloomberg.com/apps/news?pid=20601087&sid=aTjWQbjpkpxk&
2 firms with the worst track records on the street have no chance at these silly forecasts
Catalyst?
For me it all comes back to basic math, consumer is 70% of the economy and unless the consumer catches a break somewhere, the market is still overvalued. Oil down about $20 should about do it.
But then there are those pesky financials, anybody besides me looking for a few banks to fail this week?
“The forecasts are based partly on estimates that profits will jump 50 percent in the fourth quarter after falling for the past year.”
Say What? What are these guys smoking??
Ah but the big piece of missing information is valuations in each event! I suspect this would be the richest normalized valuations after the 1-month drop, with the exception of 2001.
If we’re looking at extreme short-term momentum (large Dow drops in one month), the more pertinent question is how much of a bounce occurred in the NEXT month?
Just off the top of my head, the win-loss statistics probably would still be positive, but with more noise (typical of short-term windows of study). But if I were to buy this pig for a bounce, days and weeks would be the period of interest, not months.
I agree with Jim.
Most of the people who visit this website are probably short term traders. Forecasting 6-12 months into future on such limited statistics probably does not really help us out much.
I still think there is about a 50-50 chance that we will witness some sort of capitulation within the next 1-4 weeks which will produce a short term can’t miss trading situation on the long side.
Maybe it’s just wishful thinking, but that’s my gut feeling.
I question the choice of examples that were included. How comprehensive is this? What were the criteria?
For example, whoever compiled this left out this past Dec-Jan. The DJ dropped (from seat of the pants calc while still on my morning coffee) 11% from 12/27/07 – 1/24/08. Almost six months later, it’s down another 600 points, or 6 percent. Will we see a 600 point rally?
Sounds like postive thinking. I think the market will be lower by the end of the year, compared to current levels. We could have a rally but the path is down. Everyone I read is still bullish.
I always love the 100 Dow Industrial points rally in less than 30 minutes and fascinate a 300+ Dow Industrial points in one trading session. Be bullish!
A lot of ingredients are in place for a short term rally – oil pulling back, the 2 year was below 2.50, and Q3 fund flows gotta go somewhere. Earnings will probably be spotty – some companies are going to get crushed while others may surprise to the up side. I could see a rally up to SPX 1360, NAZ 2500, Shanghai 3000, but the next NFP number will sink the rally.
There was an insert in today’s New York Times for the 2008 Non-profit Excellence Awards which was sponsored by Citi.
I hope they got their own joke, because I’m sure they could use a chuckle over there.
This is a good chart for a starter, but it begs for more detailed follow-up. Such as:
How does the worst performing sector during the swoon perform in the following 6/12mo.
Ditto the best performing.
Does a particular sector tend to lead these rallies or is it related to the order of decline, etc.
Simply looking at the DJIA,like this piece, is of limited utility.
I would keep one eye on the dollar. This recent ECB rate increase might be the last one you see for a while. Which will put some pressure on oil..
“There … might be an 18% bounce out there somewhere, but I suspect it come from lower prices, more deeply oversold conditions, showing greater levels of fear and panic”.
Agree. I think we’ll get that 18% bounce during the October-to-January time period.
Pretty simple… in a REAL bear market (as opposed to something more cyclical in nature), buying dips or into downtrends is the formula for disaster unless one can guess the final low (not recommended). Is this the real thing, i.e. 1973-74, 2000-2002? Many signs point to the REAL thing being at hand as the market fails to see it (in denial), but in reality might well be approaching the lip of the Grand Canyon.
Here is a more expanded study including more asset classes:
http://worldbeta.blogspot.com/2008/07/mean-reversion-after-bad-months.html
and the original study:
http://worldbeta.blogspot.com/2008/01/time-to-put-money-to-work.html
Looking at the same chart for the S&P500, we see that June 08 (-8.55%) is between September 86 (-8.58%; 6M later +26.1%; 1Y later +39.13) and February 01 (-9.23; 6M later -8.58; 1Y later -10.74)