Tuesday, March 10, 2009

Stocks Still Well Above Levels Typical For Deep Slumps

Bloomberg news has a story which makes a point that I have made several times before (for example here), namely that while stock valuation (by the measure of 10-year moving average of profits) at this point in time has finally fallen below the historical average, that doesn't mean that stocks can't fall further, because during really deep slumps like the Great Depression and the deep recessions of 1973-75 and 1981-82, valuation levels have been a lot lower than now.

Indeed, the reporters Alexis Xydias and Michael Tsang, underestimates just how much stocks will have to continue to fall to reach levels typical of deep economic downturns. At one point in the article, they say its 27% and at another point they claim it is 21%. The real number is much higher. If you look at the actual data from Robert Schiller, you can see that the current 10-year P/E ratio is 12.03. During the Great Depression, it was as low as 5.57, during the 1973-75 recession it fell to 8.29 and during the 1981-82 recession it fell to 6.64. Thus, even if you use the most favorable reference point, the 1973-75 recession, stocks have to fall another 31%. If you use the number from the 1981-82, stocks have to fall nearly 45% more and if you use the number from the Great Depression, stocks will have to fall another 54%.

While this slump will probably not be as deep as the Great Depression (though that can't be ruled out), it will definitely be worse than both the 1973-75 and the 1981-82 recessions. Compared to those periods, stocks are still far too expensive.
That is why I don't think we've seen the end of this bear market. While it is possible and even likely with some kind of short-term rally soon, that will only be a temporary one, and stocks will after that fall to even lower levels than today.


Blogger Wille said...

I heard someone say "Today is the best buying opportunity, until tomorrow and the day after that".

The thing with looking at P/E values today is that you have no visibility whatsoever of what they will be tomorrow - when earnings fall of a cliff, looking at last years profits is no good, the real P/E value might quadruple overnight when the next numbers are made available.

10:03 AM  
Blogger stefankarlsson said...

P/E ratios are of course no perfect method, but the 10-year average earnings method (which was the basis for the numbers mentioned in this post) actually addresses to a large extent the shortcoming you mention, as it smoothes out the usual cyclical fluctuations in earnings.

10:43 AM  
Blogger fds said...


4:33 PM  

Post a Comment

<< Home