P/E Ratios and Inflation

One of the most basic rules for valuing the stock market is that the overall price to earnings ratio should be equal to 20 minus the inflation rate. I was curious to see how good a measure of the market this is. Given how simple it is, it’s not too bad.

Here’s a look at where the rule would have valued the S&P 500 compared with its actual value, since 1900:

The trouble spots are when there’s serious deflation. Still, I wouldn’t recommend using this as a market timing device. The average error is about 10%.

According to the latest numbers, the S&P 500 should be at 1,092.

Comments

[Michael Zhuang:]

The chart shows that P/E ratio is a simple but effective tool. My own research supports this conclusion

2008 Aug 03 08:01 AM

[EE:]

What are you using for inflation? The false gvm't less than 3%, or the real double digit?

2008 Aug 03 10:36 AM

[mixter:]

Don't worry about statistics. Obama and his $5.95 bicycle tire pump will do more for our energy crisis and economy than all of the multi million dollar off shore drilling platforms could ever do. Also, he will tune up engines to save fuel. How many people do you know that have to tune up their modern day car engines?

2008 Aug 03 12:37 PM

[jakedeez:]

Obama is neither an idiot, nor is he a moron. He is in fact very intelligent. What he is, is a very good politician who has far, far more money to deploy in his effort than does McCain. I am not saying he is the right choice. I am just saying, all reasonable conclusions at this point suggest he will be president. There is no point complaining about it, you can't trade the market you want, you have to trade the market in front of you.

2008 Aug 03 03:09 PM

[Owen:]

This formula is another variant of the famous equity premium model. Inflation is used as a proxy for the risk-free interest rate, and 20-x is just 1/x in disguise, when x is between 7 and 15%. For a more comprehensive review, look up the works of Cliff Asness, Jeremy Siegel, and Robert Shiller. An article by Siegel in the November 2005 issue of the Financial Analysis Journal summarizes the issue nicely, for those interested in more theoretical background.

2008 Aug 03 03:28 PM

[juggerMan:]

Owen, thanks for the info.

2008 Aug 03 04:19 PM

[Larry Belle...:]

Eddy,

Interesting article. I also believe that inflation has not yet made its way fully through the system. And, I also agree with EE that the Bernanke version of inflation is much lower than the Volker one.....by Volker's standards, we'd already be in the upper single / lower double digits range.

I think we are in for a further correction, then hopefully, another beginning to a long sustained bull market, probably in late 2009.

2008 Aug 03 05:59 PM

[bick2:]

If the equity premium model is based on 1/(interest rate) instead of inflation the calculation would predict that the S&P 500 is undervalued by a factor of two.

Now I don't necessarily believe that, but I do think that basing this calculation on inflation rate rather than interest rate is likely to be wrong when the two are different by a factor of two or so like they are now. Equities are valued relative to interest bearing vehicles, after all.

So it looks to me that it would be more correct to state that the S&P 500 looks undervalued relative to cash.

2008 Aug 03 07:13 PM

[nonsumdignus:]

The mates at CXO advisory think we're in for a rebound relative to P/E's somewhere in the beginning of next year, just after a bottom in September:

www.cxoadvisory.com/st.../

But, even their model doesn't have us falling below 1200.

2008 Aug 03 07:30 PM

[Pipo:]

Nice article. Probably its time to look to some asian value plays like Aluminum Corp of China and some others.

investinchinastocks.bl...

2008 Aug 03 07:51 PM

[Alex Sebastian:]

What happens when the inflation rate exceeds 20%?

2008 Aug 03 09:18 PM

[BrianZach:]

This is nonsense

A true value investor will compare the future discounted cash flows to risk free rate of return. What you have to do is compare the S&P 500 P/E yield (assuming it is an accurate measure of DCF) to the treasury rates.

The ten year treasury bond trading at about 4% trades at a P/E of 25 to put things in perspective. A value investor would be comparing their stock's or index's performance to this number.

2008 Aug 03 10:26 PM

[Mr. Yukos:]

According to "your" chart, there were roughly 3 decades where the market was over or under priced according to the rule. Conclusion=f* that rule. The fundamentals have changed.

2008 Aug 03 10:33 PM

[Mr. Yukos:]

Also, you are combining the P/E ratios for all firms in the S&P and then saying "they" are overpriced? Some firms trade at high forward earnings realtive to others for a reason, to say that all 500 firms are still overpriced is also naive.

2008 Aug 03 10:35 PM

[bearfund:]

User 237528, think about what you're actually saying. I should compare stocks with Treasuries instead of inflation? Why exactly? If interest rates are negative, am I not losing purchasing power in either investment? In fact one should judge equity returns against the higher of the risk-free rate of return or the rate at which cash loses its purchasing power. Sometimes both stocks and bonds look awful. Stupid investors buy whichever looks less so. The rest retreat to gold and wait.

2008 Aug 03 10:42 PM

[Kunst:]

mixter is well know to be an idiot and a moron.

2008 Aug 03 10:55 PM

[yea ok:]

flawed analysis. Totally useless.

2008 Aug 03 11:26 PM

[Steven Hansen:]

the analysis is not flawed. it is an extrapolation of historical tools. the question is whether this rule works today. my personal opinion is that while price to earnings ratios are somewhat effected by inflation and in the past have paralleled inflation, inflation is not the primary driver of price to earnings ratios, and this is the reason for the lack of precision when plotting price to earnings ratio versus inflation.

2008 Aug 03 11:51 PM

[bbzz24:]

paraphrased: because snakes and monkeys have coexisted in the jungle for centuries, there must be causation rather than correlation.
Cum hoc ergo propter hoc

2008 Aug 04 02:48 AM

[John Lounsbury:]

I have to disagree with those who feel this analysis is useless. Using my eyeball to estimate, it looks like careful analysis would reveal that, on average this comparison has shown too high an estimate more of the time than too low. However, there appear to be more extreme deviations for estimates that are too low, the most recent being in 2002.

Over the time of the graph, estimates that have been much too low have quickly returned to close agreement, although the 2002 extreme took longer (2-3 years to return) than the others. I find that useful.

Also, some estimates that are much too high have taken years to return to agreement: the 1920's, the 1950's and mid-1970's to 1980 and 1980 to 1990.

Useful information but, as the auther has stated, not a timing tool.

2008 Aug 04 08:32 AM

[truthinvesting:]

All Pes are not created equal.The homebuilders were the WORST investment when PE was so low. I do agree PE shold be under 20 and there are plenty like MO KO MSFT BA etc that will do fine with a combo on safety and growth.My free website documents this fact

2008 Aug 05 09:05 AM