John Hussman has been beating the drum (‘till his hands bleed…) about the risks of current market valuations in his weekly commentary.
I think, supporting this message, it is worthwhile paying attention to historical relationships; at the moment, while the water is pushed to one end of the bathtub, and the line is indeed high in this respect, it tells you little about how much water the bathtub actually holds and the relative stability of the dynamics.
The current P/E on the S&P 500 is around 18 times historic earnings, which, based on “rules of thumbisms” has tended, historically, to mean dangerous territory. I know some point to low interest rates and significant capacity and high levels of unemployment to suggest that valuations are not rich and that as the economy continues to recover there is much earnings growth to come. But this cannot be the whole story….
One of the perspectives I want to develop is the historically high level of corporate profits as a percentage of national income.
Between 1960 and 1989 corporate profits either matched (till 1979) or lagged the growth in National Income. But since then, the growth rate of corporate profits has well exceeded that of National Income.
And if we look at the longer term relationship between profits growth and growth in GDP, we find similar outcomes:
And the following showing the same data but dividing the differential (growth in profits less growth in GDP) by the GDP index (base Q3 1952, 100%):
A P/E of 18 in the 1980s meant a totally different thing to a P/E of 18 in 2012. They are not comparable and likewise incomparable are the risks to return.
The Shiller P/E based on inflation adjusted earnings gives us a P/E of close to 24 times earnings (my approximation based on current valuations and earnings).
If we were to assume that after tax profits as a % of National income – currently 10.9% as of Q4 2012 – should be closer to 8%, then valuations would need to be some 24% lower, and the current adjusted P/E, if after tax profits were 8% of NI, a touch more than 24 times earnings.
If in fact corporate profits after tax should have a lower average equilibrium level, of say 7% of NI, then the market is some 36% over valued, assuming a P/E of 18 is appropriate. Obviously, if a P/E of 18 does not reflect the actual earnings growth profile going forward, we would be even more overvalued. A risk event combined with a normalisation of earnings could take the market (S&P 500) down to 900 or below.
In fact, based on the real earnings data from the Shiller/Yale dataset, real earnings growth looks to have peaked for the moment:
You just hope the three peaks in earnings growth do not yield the same graphical outcome for the main market indices from this point on!
And of course, profits growth to date has been well outside historical averages, which post 2000 and the lower GDP/National Income growth profile, should be a concern for valuations and risks to return going forward.