The Fed Model in 1981: Responding to a Surprising Question

How did the Fed Model do in 1981?  Commenting on our site here, Barry Ritholtz states:

…according to the Fed Model, in 1981 stocks with depressed earnings
and sky high interest rates were crazy expensive — just as the
greatest bull market in US history began.

That is a VERY significant flaw in the Fed Model. Missing what may
very well be the greatest buying opportunity in a generation raises
some very serious questions about any model that proclaims to be able
to determine valuation.

Listed below are the monthly data we have for the time in question:

Date            FV                       S&P

1-Jan-81 140.9 133
1-Feb-81 128.9 128.4
1-Mar-81 135.7 133.2
1-Apr-81 130.6 134.4
1-May-81 128.1 131.7
1-Jun-81 135.3 132.3
1-Jul-81 128 129.1
1-Aug-81 123.6 129.6
1-Sep-81 121.9 118.3
1-Oct-81 123.6 119.8
1-Nov-81 139.1 122.9
1-Dec-81 134.8 123.8
1-Jan-82 127.9 117.3
1-Feb-82 120.5 114.5
1-Mar-82 128.2 110.8
1-Apr-82 126.7 116.3
1-May-82 127.2 116.4
1-Jun-82 120.3 109.7
1-Jul-82 121.7 109.4
1-Aug-82 127.6 109.7
1-Sep-82 134.8 122.4
1-Oct-82 151.7 132.7
1-Nov-82 156.4 138.1
1-Dec-82 157.2 139.4
1-Jan-83 158.6 144.3

While the comparison shows the market as slightly overvalued in April and May of 1981, and again in August of 1981, those are the only such months.  We have repeatedly said that the Fed Model is not a short-term timing method, but it actually would have worked in 1981.  An investor selling in the cited months would have had ample opportunity to enter the market at better prices.

There is nothing to suggest that an investor would have missed out on a major move.  We covered the history of valuation deviations in this article, including the chart below.Sp_over_fv
As we noted in our review, many researchers do not actually use forward earnings.  They attempt to "simulate" expected earnings through some arithmetic method.  We wonder if Barry is using data from such a source.  Perhaps he will enlighten us with an actual data series or chart illustrating his point.

You may also like


  • Barry Ritholtz April 5, 2007  

    Jeff —
    How do you derive fair valuein your measures? I can look at interest rates and earnings (actual and forecast) — but I don’t see how you made the leap to over/under valuation.
    Also, I get the same weird typepad glitch with tables — you get a huge gap.
    It looks like you have quite a few “” tags in your table. If you go to the post and click on the Edit HTML tab, try removing every instance you see from to , this will resolve that white gap (you may want to remove the extra tags as well.)

  • Barry Ritholtz April 5, 2007  

    what shows up as a double quotation mark should be the letters “Ampersand-N-B-S-P-semicolon”
    the extra tags I mentioned were:
    open caret, b, forward slash close caret.

  • RB April 5, 2007  

    I don’t know what model Jeff uses, but here’s a suggestion:

  • oldprof April 5, 2007  

    Hi Barry,
    Thanks for the HTML tip! I think we have it fixed now. It did not look as bad in FireFox as it did on IE.
    The “fair value” calculation is just the forward earnings divided by the ten-year rate. It is what the S&P would be if the expected earnings yield was the same as the ten year.
    The original data was from Thomson and the Fed, compiled on Dr. Ed Yardeni’s site when it was public. We got our own data to extend the series.
    Thanks again.