Sunday, 19 September 2010
The PE10 nears its 80 Percentile - S&P 500 cyclically adjusted PE (PE10 or CAPE) – September 2010 Update
Why in this world do people always continually assume that because you have earnings of x this time that next time earnings will be x + 10% or so? My company does the same. You have a great year this year so the expectation is that next year you will have a great year plus another 10%. They do the same thing with turnover targets. To me this just seems nigh on impossible as eventually you end up so far up the exponential growth curve with it compounding year on year that you are almost destined for failure. Also where is this additional growth coming from? The world is not growing at 10% so the assumption must be that you are taking market share from someone else. But with the S&P 500 you have on average 500 companies all growing earnings by 9%. That doesn’t seem sustainable. I guess a good example of this is Nokia and the rise in competition from Apple and also the Android Operating System phones. I wonder if Nokia’s board was up until a couple of years ago forecasting this never ending exponential growth also? Now the flavour of the month is Apple but for how long?
Of course the other option to increase earnings is to sack plenty of your staff (now that has happened in the US just look at the unemployment figures) and make the remaining people work that bit harder (that has happened in my industry so I’m sure is also happening elsewhere). That also doesn’t seem sustainable to me.
Anyway enough of a digression. Today I wanted to look at the S&P 500 PE10 (or S&P 500 CAPE) data that I monitor every month. As regular readers know I am use this ratio to try and squeeze some extra performance out of my portfolio. This method is used by Professor Robert Shiller however I modify it slightly by incorporating forecast earnings (even though as I showed above they can be very bullish in my opinion) up to the month of interest. For new readers some background information on the CAPE is available here and if you’d like some information on why I use the CAPE then that is available here.
My 1st chart plots the Shiller S&P 500 PE10 that now sits at 20.4 which is up from 19.8 last month. To put this in perspective the 80th percentile for this ratio is 20.6 (the 20th percentile is 11.0)! The chart shows it quite nicely. Prior to the silliness that took place with the dot com boom in 2000 a valuation this high was getting to the point of the peak of the market. What will happen this time around? My mechanical investment strategy forces me to take action no matter what. With the increase in this PE10 my UK Equities target asset allocation decreases to 18.4% (nominal based on buy & hold would be 21%) and my International Equities target asset allocation decreases to 13.2% (nominal based on buy & hold would be 15%).
The Shiller S&P 500 PE10 Average stays at 16.4. Assuming this can be considered “fair value” with a PE10 of 20.4 we are now over valued by 25%.
The reason why I use the Shiller S&P 500 PE10 is shown in my 2nd chart. The correlation with the Real (ie after inflation) S&P 500 Price is currently 0.78. This can also be seen on the chart, which while the R^2 is low, there appears to be a trend, particularly as the PE10’s get large (ie very over valued).
The 3rd chart plots Real (after inflation) Earnings and Real Dividends for the S&P 500. Eye balling the chart shows Real Dividends continue close to the long run trend line. Real Earnings though are now well above trend. This comes back to my first point today. As an analyst I probably wouldn’t be saying x + 9% but maybe working towards earnings returning to trend over the medium term with any forecasting that I was doing.
As always do your own research.
- Prices are month averages except September ‘10 which is the 17 September ’10 S&P 500 stock market closing price.
- April ’10 to September ’10 reported earnings are estimates from Standard & Poors. S&P has September 2010 annual earnings estimated at $69.71.
- Inflation data from the Bureau of Labor Statistics. September ‘10 inflation is extrapolated.
- July to September ‘10 dividends are estimated as the June ‘10 dividend.
- Historic data provided from Professor Shiller website.