The Case for Market Timing

[doushioukanaa]

I know many of us put a lot of faith in a fixed allocation for a portfolio, but don't you think there is still a place for varying the allocation on the basis of P/E ratio or dividend yield or some other measure of whether stocks are overvalued?  I mean, any Japanese investor who had 80% in stocks in 1989 lost their shirt and it still has not recovered. 

http://finance.yahoo.com/q/bc?s=^N225&t=my

Is there something inherent to US equities that prevents us from anticipating and reacting to a possible bubble?

Isn't there a case that at some point an investment class can be just too expensive and therefore you should change your weighting in it?  And if so, what is the magic number?  I mean, the P/E ratio of the S&P 500 came close to 47 back in 2002 well above the mean of 24.  Everybody said "this time it's different" because of new technologies, etc.  But shouldn't the indicators have told us to become a bit more underweight in US stocks or at least retreat to value stocks? 

Fortunately, the P/E ratio of US stocks is not too out of whack, so it's not like we need to contemplate reducing our exposure at this moment, but aren't there conditions where you would really consider it?

 

[brewer12345]

Sure, great, just show us the magic indicator that tells us when to run for the hills or back up the truck.

I am a stock picker, by and large, but I have a hard time valuing entire markets and I know it. But I did buy some QQQQ puts and sold half of my remaining EFA exposure last week because I did not like the look of what was going on in the commodities market and I thought that the market had rallied on a Fed pause that I doubt is forthcoming. But can I put numbers to that? Nope, just gut instinct.

 

[samclem]

Ahhhh! The H**** welcome mat has been rolled out! Pull up the drawbridge!

 

[doushioukanaa]

Ahhhh!   The H**** welcome mat has been rolled out!  Pull up the drawbridge!

Huh?

 

[haha]

Huh?

Don't pay any attention to this. There are those here who believe more fully in the EMT and continuous high equity allocation than they do in the paternity of their children.

Both beliefs are likely misplaced.

HA

 

[samclem]

doushioukanaa,
A less glib reply to your post: Valuations do matter, but using the info is not easy. You could end up seling al your equities because you think the market is overvalued, then spend a decade on the sidelines missing out on good gains--it has happened.
What I'm doing is setting an asset allocation and sticking to it. If one asset class or market segment gets overvalued compared to others, then when you rebalance you'l automaticaly be selling the higher valued stuff to buy the things that have not run up so much.
Sure, I'd be tempted to significantly reduce holdings if something were obviously running up to unsustainable heights, but beyond the very occassional bubble this doesn't come up much for me.

 

[unclemick]

The case for market timing.

Benjamin Graham(edited by Zweig) - The Intelligent Investor. latest paperback edition or library version.

Geraldine Weiss - Dividends Don't Lie - or Goggle - Investment Quality Trends.

Most broadly based versions I've run across for mutual funds are mixtures of - heh heh heh heh heh - psst! Wellesley and Wellingon or Dodge and Cox setting the mix to get a benchmark dividend yield - ie raising/lowering the stock/bond mix to hit the benchmark.

Me - I let VG Target Retirement do the heavy lifting(age timing).

But male hormones never die - soooo - 15% in Norwegian widow dividend stocks.

heh heh heh heh heh - two kinds of timing - heh heh heh heh

 

[doushioukanaa]

Most broadly based versions I've run across for mutual funds are mixtures of - heh heh heh heh heh - psst! Wellesley and Wellingon or Dodge and Cox setting the mix to get a benchmark dividend yield - ie raising/lowering the stock/bond mix to hit the benchmark.

Thanks for the feedback! 

Just to confirm, you are recommending and endorsing those books, and you are indicating that some of the managed funds constantly adjust their portfolios in order to hit a certain yield, and in the process they end up having more or less weighting in stocks depending on the stock dividend yield, right?  I think Capital Income Builder probably falls into that category, and creates yet another reason to have some of one's portfolio in well managed funds as opposed to solely index funds.

 

[Cut-Throat]

No, I just saw a no money down real estate show on TV and we are all fools for investing in the Stock Market. These people were 'ordinary Joes' and had instant cash flows of mega bucks with no money down, bad credit, and no job. Today they are all driving Rolls Royces and living the good life.

We are fools I tell you - Fools!

 

[JohnP]

Here is the real case for Market Timing - from Fundalarm.com - HighLights and Commentary,  April 2006

According to MIT Professor Andrew Lo, $1 invested in the S&P 500 Index on January 1, 1926 would have grown to about $4,000 by the end of 2005 ($1 invested in the Lehman Brothers Aggregate Bond Index would have grown to about $18).....Over the same period of time, an investor who moved $1 between the indices at the precise time one was up and the other was down -- in other words, a market timer with perfect foresight -- would have accumulated a little more than $23 billion (yes, that's "billion" with a "b").....Now do we all understand why market timers will never give up their quest?
Pensions & Investments, March 6, 2006

Knowing when to go all-in to a market or to get out is the key to Market-Timing success.  I think we all hope that we know more about the market than the next guy, but... 

Keep your powder dry and limit your testosterone-based market timing to under 10% of your portfolio unless you really have 'the knack'.

JohnP

 

[modlair]

According to MIT Professor Andrew Lo, $1 invested in the S&P 500 Index on January 1, 1926 would have grown to about $4,000 by the end of 2005 ($1 invested in the Lehman Brothers Aggregate Bond Index would have grown to about $18).....Over the same period of time, an investor who moved $1 between the indices at the precise time one was up and the other was down -- in other words, a market timer with perfect foresight -- would have accumulated a little more than $23 billion (yes, that's "billion" with a "b").....Now do we all understand why market timers will never give up their quest?
Pensions & Investments, March 6, 2006

Knowing when to go all-in to a market or to get out is the key to Market-Timing success. I think we all hope that we know more about the market than the next guy, but...

Keep your powder dry and limit your testosterone-based market timing to under 10% of your portfolio unless you really have 'the knack'.

It's really hard to address this softly. Here's reality.

There is no knack. There is no system. There is no compelling evidence that anything other than earnings growth predicts the price of a stock (albeit loosely) -- and there is no compelling evidence that anyone accurately predicts earnings growth.

Try to wrap your minds around this next concept. The markets' variance is explained by a mixture of rational behavior and by a process which is indistinguishable mathematically from randomness. The mixture of those two influences varies with time, and the schedule of the mixture is itself indistinguishable from a random process.

The calculations quoted above did not include taxes or brokerage fees, but certainly perfect knowledge would outperform buy and hold. The DFA folks are measured as the best, apparently, and they only outperform the market by a few %.

When you play with timing and picking, you are engaged in a random process where the house advantage is brokerage fees, assymetric taxation of gains vs deductions for losses and time out of your life studying a largely random process. It's not particularly different from Las Vegas, where everyone knows you can win once in a while and be nicely entertained trying.

But next time you're there look around at the skyline and ask yourself if those hotels got built by depending on a process where players win.

 

[doushioukanaa]

When you play with timing and picking, you are engaged in a random process where the house advantage is brokerage fees, assymetric taxation of gains vs deductions for losses and time out of your life studying a largely random process.  It's not particularly different from Las Vegas, where everyone knows you can win once in a while and be nicely entertained trying. 

Ok, but be honest. If the S&P 500 P/E ratio went to 75 and the dividend yield went to 0.5%, don't you think you would lighten up on equities even a little bit?

 

[bpp]

I mean, any Japanese investor who had 80% in stocks in 1989 lost their shirt and it still has not recovered.

If they had been properly diversified (with at least international stocks, and a more meaningful bond allocation), then they would probably have lost only a sock, not a whole shirt. Might even have come out all right by now.

Bpp

 

[wabmester]

Ok, but be honest. If the S&P 500 P/E ratio went to 75 and the dividend yield went to 0.5%, don't you think you would lighten up on equities even a little bit? 

The P/E is meaningless by itself.   What is the "right" price to pay for earnings?    In your P/E=75 scenario, what if bonds were only yielding 1%?   Would you move from "expensive" stocks to even more "expensive" bonds?

A lot of people judge the market in terms of historical P/E values.   They compare to historical values because there is no intrinsically right value.   A P/E of 10 might have been right in the 1970's.    After the introduction of discount brokers, IRAs, and 401-K's in the 1980's, maybe a P/E of 20 became the right value.

In this decade, with the historically unprecedented influx of foreign capital, maybe a P/E of 30 is the right value.

One simple way of looking at P/E is to invert the ratio and treat it as a "yield" on the stock market.    In today's market, an E/P of 1/20 corresponds to a yield of 5%.    5% is about what bonds yield right now, so a lot of people (including me) believe it is becoming a good time to lighten up on stocks and move to a lower risk portfolio.

I've recently posted this on another thread, but using the spread between market E/P and bond yields is a successfully back-tested timing stragegy:

(Simple) Market-Timing Strategies That Worked

 

[doushioukanaa]

I only read the first few paragraphs of the document you recommended, but it looks like a very thought provoking read. Thanks!

 

[justin]

Wab,

I sort of rationalize equity investing like you do. 1/PE = earnings yield. At PE=20, I'm paying $100 for $5 of earnings each year (a 5% yield), plus getting a dividend of ~2%/yr, for a total yield of ~7%. I also expect the earnings in the future to grow.

A bond yield of 5% won't grow in the future.

For these reasons, I don't feel too bad about a PE of 20. If prices double, and I'm getting a 3.5% yield on stocks (with growth potential) versus a 5% yield on bonds, I'm starting to think bonds are the better deal, given the lack of risk. I wouldn't go 100% bonds, but I might up my allocation for bonds/cash from 0% today to 20-40% if stock PE's go to 40.

Given the recent runup in PE's, I'm starting to consider a tactical allocation to cash/bonds at this point. If the market moves up 30% in the next few months or year or so from where it is now, I'll be putting some money in bonds/cash.

 

[brewer12345]

justin, the dividend yield comes from the earnings yield, so it isn't additive. Total yield is still 5%, you are just getting part of it in cash and letting the rest ride.

An aalternative to pulling out of the market is to look for cheap stuff. I start getting interested at a PE of around 12 (8.3% yield), I get real interested at 10, and below that I just about wet myself if there isn't a really bad problem with the business. Obviously it requires some time to find the cheap stuff in today's market, but there sure is a lot of it.

 

[wabmester]

I sort of rationalize equity investing like you do.

That's the nice thing about that spread-based market timing strategy -- as long as enough people believe like we do, it works!