A reader left a comment that gave a reasonable "I hear ya but..." about the 200 DMA because three out of four timing indicators that Phil DeMuth and Ben Stein use say the S&P 500 is now undervalued.Let me be clear I don't know anything about what Stein and DeMuth are saying, if anything, I am relaying the reader's comment and choose to believe the reader has his facts straight.
I have written about this sort of thing before, the Fed Model is an example of the type of market valuation process the reader is talking about.
All indicators of all types have flaws. If you are going to to rely on an indicator you have to know its flaws in order to make an intelligent decision about it.
First let's stipulate that Stein and DeMuth's indicators are correct and the market is cheap. OK, when will it go up? Will it get cheaper? What about being cheap will lead to a move in the market? Is cheap important right now? Do the risks that made it cheap in the first place outweigh the cheapness?
Every model along these lines I have ever seen can stay cheap or expensive for years as the market continues to move up while expensive or down when cheap.
My take on the 200 DMA is that the dynamics of the market are communicating the health of demand. Quite simply if demand is healthy the market stands a better chance of doing well regardless of how cheap it is. The drawback is that occasionally it will breach the 200 DMA in one direction for only a couple of days and then go right back--which is a reason why I don't make big bets all at once but just start with a tweak.
The way I look at it, valuation models/indicators rely on too many "shoulds." I don't find anything forward looking when armed with the knowledge that stocks are cheap. Additionally, demand healthy yes or no is a much simpler concept and where possible I think simpler is better.
12 comments:
With apologies to your female readers, a couple of sites have pointed out today that the Sports Illustrated swimsuit cover is a bullish indicator for US stocks. :)
My point being that people forget the market is an auction place. When demand outstrips supply at a price, the price will rise until the demand is met. I don't know if the 50, 100, or 200 dma is the best indicator of healthy demand, but it stands to reason that the longer the time frame, the fewer false signals an investor will see.
Hi Roger. A couple of comments on the Stein/DeMuth model.
I think your reader is referring to the case made in book "Yes, you can time the market." I read it a number of years back and it started me on journey of completely rejecting general case of Efficient Market Hypothesis. It was a good, quick read and made a strong case for questioning certain buy-and-hold assumptions that constitute conventional wisdom.
But to be clear, my recollection is that they didn't make a particularly strong case that anyone should actually follow the simplistic model used in the book. It presupposes a 15 year holding period for any purchases. Idea is just to refrain from buying when a range of moving average indicators are clearly showing market valuation is above average based on past nominal values. So e.g. if P/B, P/S, P/E, etc. are all above long term average, don't buy. But if they are below average, buy and hold for 15 years. A backtest of this algorithm significantly outperforms blindly buy and holding since you avoid buying in most egregiously overvalued periods.
Of course downside is that you could well be sitting on cash for a decade or more at a time, and there's a big difference between saying market is not clearly overvalued, and saying it is cheap.
If your reader is committed to following this formula, he should go ahead and do so. But I would suggest that if he is in that category that even reading your blog may be counterproductive. He would be better off completely tuning out financial markets so not tempted to stray. 15 years is an incredibly long time.
For most of us mere mortals that feel the need to be a bit more active than 15 minutes/year to make a buy/no-buy decision, Stein/DeMuth is impractical. But the thought process I think is quite valuable.
Just my $0.02.
Michael
Nice picture. A while ago, you wrote that you used an 8% trailing stop on some of your portfolio. Does that change in such a volatile climate? Thanks for giving so much of your time to this blog. I know it has helped me. Tom in Indy
Michael,
If one is going to use fundamental metrics to time the market I would suggest "The little Book that Beats the Market" by Joel Greenblatt. It outlines a case for buying "cheap" companies on valuation basis who are performing well on ROE basis.
The holding period is one year and back tests smoke the buy-and-hope results.
Buy-and-hope has so many problems because it is based on a set of lies, bad hypothesis, misunderstandings and for most people a broken benchmark... but that's a rant for another day.
Tom, i think that either you mis read or I constructed my thoughts poorly. I have never been much of a stop user and i have never been an 8% person.
i wrote an article for TSCM at one point that echoed some sentiment written here where I said 8% might make sense for some stocks, but that some other stocks should have tighter stops and some others looser stops for those who even use stops. For me, no I only use them occasionally. I have written numerous time about many stocks not lending themselves to stop orders and why. I have no stop orders in currently.
sorry that was not clear.
Michael 7:19--I read the book as well and I think your take is quite accurate. I've given my copy away so I can't double-check.
"cheap" and "expensive" must be viewed in the context of overall risk tolerance. The government couldn't give away 10-year Treasuries last summer - you could get 5.25% on July 12th. Of course, they've been on a tear ever since.
"All indicators of all types have flaws." - How true. That's why it's best to go with models.
I also agree with your view of valuation indicators (add monetary indicators as well). The signals are too imprecise imo. However, I believe trend and sentiment indicators, although not perfect, are very helpful especially when used in combination.
I highly recommend the Research Driven investor by Tim Hayes of Ned Davis Research. It's pretty much the "Stock Market Logic" for the 2000s. I hope he comes out with a revised edition soon.
I would reword your approach in the following way:
cheap market - an estimation of current market relative to the past. thus, contains just little insight about future developments.
200 DMA - tries to reveal any possible change in future trend, i.e. projects into the future.
I agree that your approach is more powerful and provides a better extrapolation. Problem is that you do not know well natural frequency of the process. may be 183 DMA or 220 DMA is better? As you understand 200 DMA is a bit artificial.
From my point of view, 200 DMA will provide an excellent result during the next two years since the period of natural oscillations is around 11 months.
Ivan, interesting point.
Over the years I have read about people using a slightly different number than 200 with the belief that their number was not random.
What I mean is that the 220 DMA (this is an example only as I do not recall exactly) was deemed to be more useful than 200 DMA.
I will say that I am trying to keep things as simple as possible but as I (hopefully) learn more over time going with a different number, in an attempt to be less random, is certainly possible.
Roger,
I guess that you will be a bit reluctant to trust my approach to the prediction of market stock prices. I have introduced a linear link between SP 500 returns and the change rate of the number of 9-year-olds. (I have mentioned
relevant paper in your blog.)
I presume that the DMA 200 is a very good moving average because it is a bit longer that half a year. So, it provides a clear view on annual cycles, i.e. very (if not the most) sensitive to seasonal swings.
On the other hand, using only the 200 DMA one can not avoid problems with longer periods and higher oscillation amplitudes in natural economic process like between 1999and 2003.
The next two years ( from my point of view as related to my own model of SP 500 returns) the 200 DMA should be an excellent tool.
Do not miss the feel good rally in August-November, although.
With severe contraction in credit the markets health has to be iffy.The dropping of interest rates may not lead to material improvement in lending as the major banks are UNDERCAPITALISED. Just look at the Freeze in Auction Rate Bonds (Roger your comments on this and how it affects closed end mini funds will be appreciated)
Post a Comment