After a Down Month, Wait a Month to Buy an Asset Class

Maybe the market is trying to tell you something - if you just listened?

I have written about this subject a number of times. For background you can check out some earlier posts:

Time to Put Money to Work
Mean Reversion After Bad Months
When is the Time to Buy Homebuilders?
When to Buy Japan?
Idiocracy and Mean Reversion

In one study I examined asset class performance after a really bad month.

The take-aways from this study were:

• It does not pay to buy an asset class after a really bad month for the following 1 month.


• 12 Months later the return is not much different than average


• 3 and 6 month returns, however, are stronger.


• You pick up on average about 3-4% abnormal returns buying after a terrible month.

A simple strategy would be:

After an asset class has a terrible month (ie MSCI EAFE in January), wait a month, then take a 2 month position. I.e. buy March 1 with a two month hold.

It looks like a good "trigger" for equity such as asset classes is around -9%, and for bonds around -3%.

Below is a summary chart of this strategy for the five asset classes I mentioned in my paper. The returns are simply the excess returns (nets out the average monthly return over the entire time period) to a strategy of buying an asset class a month after a really bad month, with a two-month hold.

Note that this does not work for the commodity index, and one could speculate that is due to differing risk premiums and sources of return to that asset class. This system would have triggered for REITs after a horrible June (-11.25%). To illustrate, one would wait a month, then buy the index August 1.

I imagine if you broke out the asset classes into further subdivisions (sectors, countries, etc.), the same principles would apply but the triggers would widen due to the increases in volatility.

Comments

[dlaw:]

"Mean Reversion" is based fundamentally on a fallacy.

The fallacy is that markets are best described by a Gaussian statistics. Mandelbrot and others proved this was not the case decades ago. If you want to argue with Mandelbrot on the math, be my guest, but you're wrong.

"Mean Reversion" only works *most* of the time and when it doesn't work, you lose all your money - as any mean reversion strategy contains within it an implicit martingale. When that martingale meets a liquidity limit, the strategy explodes.

It's just that simple.

You're not that smart.

2008 Sep 12 05:54 PM

[Twest:]

Thank you for an excellent article. For people who love mathematical theory, they should stay in the mathematical world and apply their theories to the physical world. Financial markets are made of human emotional cycles and patterns, many of which are difficult to detect and that change constantly. Hence, mathematical theories do not apply to human emotions because they are variable and given the same conditions their will not be the same response. Chaos theorists will argue differently about that, but psychologists would agree. What the author so creatively has accomplished is to give people facing a difficult choice, buying a market after a large losing month some guidelines as to what has happened in the past. Either, buy now or wait a month. What changes, of course, is what defines a large, down month over time. What also hasn't changed, however, is that investors tend to look at the market after receiving their monthly statements from their brokerage firm. Their will always be change to this analysis because the source of the sellers AFTER down months will change as the mix of investors in the stock market changes as the population curve shifts over time. Young people make very different investment decisions than older people.

Thank you for doing the study of this human emotion pattern to see 'what has worked' in the past. I appreciated it very much and hoped that I have addressed the criticism of the first responder. Isn't it amazing too that an article can generate such anger in someone! I guess that person shouldn't listen to a weather forecast either or they might get angry at the weatherman. Cheers to you.

2008 Sep 13 09:09 AM

[doctordoom:]

Would TA assist in buy/sell decisions?
As to answer 1, I call on Ms Manners: Dear Gentlepersons.

2008 Sep 13 09:56 AM

[etf_investor:]

I did similar statistics. I think your method is not accurate enough. The problem is that you are doing "average," without looking at the different result at different time period. For example, statistics in a "bear market" shows a very different result from "bull market." Time periods from 1929 to 1932, 2000 to 2002 are so different from 1994 to 1999 that if you apply "average" to a bear market, you may lose your shirt. Please try what you recommend to the 2000-2002 bear market and do some back test, you will understand what I mean.

2008 Sep 13 10:30 PM

[ACEMAN:]

Y'all have to understand that "mean reversion" is a probability construct and Gaussian applications are not. But the real issue to which DLaw is on the trail of is this: "median" and "average" could be pure noise or just bad numbers that look good for analysis sake. The fact is the mean reversion construct merely approaches or in truth approximates a mean, it can 't really get there, i.e., to any real mean, and that was proven by Goedel's hypothesis. So what y'all have is something that approaches your "mean" but never really gets there. Moreover, a mean reversion cannot, because it is impossible, provide all the data that is necessary to carry out a quantum analysis. So to make this applicable to trading, y'all need to forget about mean reversion because y'all can't trade on it as it is not predictive in any way, shape or form. In fact, it probably gets in the way of trading intuitions and complex individual understanding. Reversion to the mean is a fiction and cannot be proven to be effective in making you a better and more successful trader. Probability theory should not be used as a predictive tool, cause it ain't!

2008 Sep 14 12:05 AM

[AFJ Garner:]

What an unpleasant and unnecessary first comment above. I have much enjoyed Mr Faber’s comments on various websites and very much enjoyed his paper on asset allocation. Mr Faber somewhere described himself as a “quant lite” and has made some valuable and interesting comments on stock market behaviour based on his back testing and observation. I do not happen to favour mean reversion as I have never been able to prove its efficacy in my own back testing but that does not lead me on to insult Mr Faber.

How sad that can humanity work itself up into such unjustified and inappropriate rage over such small issues. Is it any wonder that peace, in every sense of the word, eludes us?

2008 Sep 29 02:08 PM