Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Shilling for Shiller or Ranting by Reynolds

|Includes: Vanguard Total Bond Market Index Fund Inv (VBMFX)

In a March 12th editorial in Investors Business Daily, Alan Reynolds criticizes Robert Shiller and his use of cyclically adjusted earnings (essentially a ten year average) as a valuation tool for the S&P 500.  Mr. Reynolds dismisses the idea that a cyclically adjusted earnings price earnings(NYSEARCA:CAPE) ratio over 20 has any value as a market timing tool since using it as an entry/exit point would have kept you out of the market from 1992 to 2008. His opinion is that an investor following that advice  "would have forgone many lucrative investment opportunities". I thought it would be an interesting experiment  to test Mr. Reynolds opinion.

Professor Shiller makes his data publicly available here. His monthly price for the S&P 500 appears to be the average price for the month and he interpolates earnings for the months between quarters. To simulate how you might use his method in real time, I used the monthly average price but only updated S&P 500 earnings two months after each quarter ends (so the CAPE 20 for the fourth quarter in any year would not be updated until February).

I decided to use two Vanguard funds that have been in existence since 1993, the Total Bond Market Index Fund (MUTF:VBMFX) and the 500 Index Fund (MUTF:VFINX). Using the CAPE 20 rule, Mr. Shiller would own the bond fund from January 1993 through October 2008, while Mr. Reynolds would own the stock fund during that time.

I first compared their total return on a $1,000 invested at the beginning of the test period. I used Yahoo's adjusted close which gives returns that assume no taxes, no transaction costs and dividend reinvestment.

Fund End Value Total Return
Stock  $     2,449.12 145%
Bond  $     2,465.43 147%

The total return for the period are very close but for most of the time the stock fund greatly outperformed the bond fund as you can see on the chart below.

Total Retrn from Jan 1993 through Oct 2008


Next I compared if they had invested $1,000 into each of their funds on the first trading day of the month for the test period.

Fund End Value Total Return
Stock  $  206,160.32 8%
Bond  $  296,897.91 55%

$1000 invested per month

The bond fund significantly outperformed over the whole period but, again, the stock fund was actually the better performing asset most of the time. To be sure, there are tax advantages to owning a stock versus a bond fund. It is difficult to model what these returns would be since tax rates vary between individuals and over time periods. The results below show what the returns would have been with a 35% tax levied on dividends and interests (I know, not necessarily realistic).

For $1000 invested only at the beginning of test period.
Fund End Value Total Return
Stock  $        194,078 108.0%
Bond  $        251,006 78.0%


And for $1,000 invested every month.
Fund End Value Total Return
Stock  $        194,078 1.6%
Bond  $        251,006 31.4%



If you were investing regularly over this period you obviously would have been better off in the investment grade bond fund. The stock index fund has rebounded (up 44%) but using the CAPE 20 entry/exit point, both investors would have enjoyed most of that gain. Alan Reynolds is correct that there were many periods in those 16 years where it would have been lucrative to invest in stocks but he gives no guide to when those periods might be. I do not propose that anyone use the CAPE 20 metric as a hard and fast rule but I think that Mr.. Reynolds underestimates its utility in gauging the risk of the market. Is the market less risk at a 19 CAPE than at 21? Probably not much, but does your risk increase as the CAPE goes to 25, 30 and higher? I think investors would be wise to consider this valuation when determining the amount of exposure they want to the more volatile equity markets.




Disclosure: No positions in stocks mentioned