Value and Regression
Lance on Dec 10 2008 at 7:48 pm | Filed under: Domestic Equities, Valuation
Over at dshort.com there is an interesting look at valuation and regression to the mean (click image to link to larger version)
The peak in 2000 marked an unprecedented 160% overshooting of the trend, which is double the overshoot in 1929. The index has been above the trend for 17 years. We also see that the major troughs saw declines in excess of 50% below the trend. If the S&P 500 were sitting squarely on the regression, it would be hovering around 820. If the index should decline over the next 12 months to a level comparable to previous major bottoms, it would fall to the vicinity of 400-425.
That is a teeny bit lower than how I see it, but it is about right. However, he provides a chart which adjusts for inflation using John Williams’ shadow government stats. That shows the markets at all time lows. You can see and read the rest here.
I have one problem with that analysis. While I do think some adjustment might make it look a little better, if John Williams’ numbers are correct then stocks should be as low as they are because GDP and earnings have been horrible in inflation adjusted terms for a very long time. We have been in a recession for two decades similar to Japan. I don’t think that is true, but if so then stocks deserve every bit of the undervaluation they have experienced.
I’ll stick with the first graph being slightly overstated. It should be noted that statisticians would likely have problems with the while idea of the long term regression line having any meaning. I disagree, there are fundamental economic reasons the line makes sense as an approximation. I won’t go into the details now, I just want it noted I am aware of the issue.
Hat tip: Barry Ritholtz.
