John Hussman of Hussman Funds writes:
"Among the simplest truths is that market risk tends to be
unusually rewarding when market valuations are low and interest rates are
falling. For example, since 1950, the S&P 500 has enjoyed total returns
averaging 33.18% annually during periods when the S&P 500 price/peak
earnings ratio was below 15 and both 3-month T-bill yields and 10-year Treasury
yields were below their levels of 6 months earlier. Needless to say, there are a
variety of ways to refine this result based on the quality of other market
internals, but it’s a very useful fact in itself.
The “canonical” market bottom typically features
below-average valuations, falling interest rates, new lows in some major indices
on diminished trading volume, coupled with a failure of other measures to
confirm the new lows, and finally, a quick high-volume reversal in breadth
(usually with an explosion of advances over declines very early into a new
That makes terrific sense to us. What about when the opppositie is true?
"Similarly, market risk tends to be poorly rewarded when
market valuations are rich and interest rates are rising. Since 1950, the
S&P 500 has achieved total returns averaging just 3.50% annually during
periods when the S&P 500 price/peak earnings ratio was above 15 and both
3-month T-bill yields and 10-year Treasury yields were above their levels of 6
months earlier. Again, there are a variety of ways to refine this result, but
note that anytime the total return on the S&P 500 is less than risk-free
interest rates, a hedged investment position increases overall returns (since
hedging instruments are priced to include implied interest).
The “canonical” market peak typically features rich
valuations, rising interest rates, often a reasonably extended and “flattish”
period where, despite marginal new highs, momentum has gradually faded while
internal divergences have widened, and finally, an abrupt reversal in
leadership, from a preponderance of new highs over new lows (both generally
large in number) to a preponderance of new lows over new highs, with the
reversal often occurring over a period of just a week or two."
Again, perfectly reasonable analysis borne out by the data. So where does that elave us at present? Consider the following:
"Though our investment position doesn’t by any means rely on
it, my impression is that recent market conditions fall very much into that
description of a canonical peak…
It doesn’t help the case for stocks to argue that, for
example, earnings growth is still positive, because it turns out that the
year-to-year correlation between stock returns and earnings growth is almost
exactly zero. It doesn’t help to argue that consumer confidence is still high,
because consumer confidence is actually a contrary indicator, as are capacity
utilization, the ISM figures, and other factors being used for bullish fodder.
It doesn’t help to argue that the Fed will stop tightening soon, because the end
of a tightening cycle has historically been followed by below-average returns
for about 18 months. It doesn’t help that 10-year bond yields are still lower
than the prospective operating earnings yield on the S&P 500 (the “Fed
Model”), not only because the model is built on an omitted variables bias (see
the August 22 2005 comment), but also because the
model statistically underperforms a simpler rule that says “get in when stock
yields are high and interest rates are falling, and get out when the reverse is
Good stuff, John.
via WSJ Marketbeat
John P. Hussman, Ph.D.
Hussman Funds, May 22, 2006