Prominent portfolio managers, Don Hays and Jeremy Grantham, look at market history and come to wildly different conclusions about whether the US stock market is currently overvalued or undervalued. Hays says it is undervalued and Grantham says it is overvalued. Mark Hulbert of MarketWatch stated the difference as follows:
Even if you believe the markets follow a regression-to-the-mean process, you still can conclude that stocks are either as much as 40% overvalued or as much as 40% undervalued.
Although both Hays and Grantham came to their conclusions using regression-to-the mean analysis, Hays used four decades of market history and Grantham went back much further. Which is right?
Since none of us know whether the market will adhere to mean-reversion norms in place over the past four decades, ten decades, or twenty decades, doesn’t it make sense to adhere to a process that can adapt to paradigm shifts?
Posted by Andy Hyer 
