An Alternative to Shiller CAPE

Even though value investors love to use the Shiller Cyclically Adjusted PE (CAPE) to justify the market’s overvaluation, the metric has some obvious shortcomings.   The most relevant one is that it is biased towards overstating the earnings multiple.  That’s because the Shiller PE takes a simple average of 10 years worth of earnings.  Assuming that the economy grows, this will give a depressed view of earnings power because the average relies heavily on old earnings.  CAPE does try to correct for this issue by adjusting for inflation.  But while this is better than nothing, it still under-represents growth and has a propensity to diagnose an overvalued market (especially in periods of weak inflation).

Below is an alternative long term valuation metric (I’ll call it the KAPE), which averages ROE rather than earnings.  I think that this does a better job of capturing growth because the ROE metric captures elements of growth on its own.  It measures how well a company managed a given pool of capital in each time period whether that pool is growing or shrinking.  By averaging the ROE, the company’s past performance can then be compared to current valuation using a price to book multiple.

Mathematically the equation is:

KAPE= (Price/Book)/Average(Earnings1/Book1+Earnings2/Book2+Earnings3/Book3+Earnings4/Book4+Earnings5/Book5)

You can see the reason why you use the current price to book ratio is that this cancels out the “book value” element of the final metric, leaving you with a more simplified Price to Earnings Power multiple, the Krisiloff PE, or KAPE.

The chart below compares the KAPE to the Shiller PE for the S&P 500.  It mostly tells the same story, but with a little less volatility.  Unfortunately I couldn’t find book value data pre-1990 to give a longer term comparison.  If anyone can find long term historical book value data on a major index, please let me know.

Krisiloff PE

One thought on “An Alternative to Shiller CAPE

  1. Pingback: Friday links: headline hunger | Abnormal Returns

Leave a Reply

Your email address will not be published. Required fields are marked *