I wanted to chime in about the use of such indicators as a possible timing tool by investors. Can the q ratio be used in this manner? Here is what Smithers himself has said ("Wall Street Revalued", p. 77):
Selling equities when they are overvalued can be sensible, but only if the prospective returns from the sanctuary asset, in which the funds previously invested in equities are temporarily lodged, will exceed those on equities over the uncertain time period in question. The uncertainty about the time when the investor will wish to return to equities makes cash (or its equivalent, such as money on short-term deposit), the most likely and generally sensible choice as a sanctuary asset. Although real cash returns vary, the asset can at least be realized in the short term at its original nominal value. Bonds have an even greater exposure to inflation and even if that is stable the nominal value at any time prior to maturity is uncertain.
How true. The return on a "safe" asset is very low. The return on equities, even when shown to be overvalued, can still be relatively high. To switch from equities to cash requires a fine and bold calculation that the sanctuary asset return, plus any avoided losses, will exceed the return given from equities in the interim. THat is a judgment of three uncertainties: the length of time it takes for "fair value" to reassert itself in the equity space; the return on equities until that time; and the return on the "safe" asset class chosen.
Smithers continues:
Historically, there have only been five peaks in the market's overvaluation since 1900... . The average time between peaks has been 24 years but the average is far from regular and each of the last two swings has taken over 30 years from peak to peak.
...
If, for example, the real return on equities is normally 6%, Treasury bills can be expected to give at least a 1% real return and the stock market has an equal chance of being over- or under-valued in 15 years' time, then the limits of overvaluation given a limited number of rational investors will be just over twice. (Ed note: log terms).
...
A glance at Chart 15 shows that, during the period for which we have adequate data, only twice has the market become so overvalued that it was worth while (sic) selling on either of these assumptions: the first time being prior to its 1929 peak and the next prior to its peak in 2000.
Is the q ratio then largely ineffective as a timing tool? Apparently Smithers believes- given his assumptions- its use as such to be quite rare: twice in nearly 100 years. That's hardly the holding time-frame for the average investor. (Endowment investing is different.)But the rub (there's always one isn't there?) is in Smithers' assumptions regarding time elapsed from observed over-valuation to realize fair value, the prospective return from equities as an asset class, and the return from the sanctuary asset. If any one of these assumptions is altered, the calculus changes. And as we have written on this blog, not every market expert agrees with these values. I would only point the reader to my posts about John Hussman's models or Jeremy Grantham's forecasts to see how tweaking the assumptions might work a bit differently for the creative investor. What if bubbles were observed to burst more frequently than granted by Smithers? What if the sanctuary asset class is not cash but has some, though limited, volatility and a much higher expected return? What if the observed over-valuation shows that the expected return for equities over a given timeframe is not 6% but two percent, one percent or even ZER0? The possibilities for thoughtful substitution by an investor become great.
No comments:
Post a Comment