1/ 12 years of meandering performance and nearly 25% off the high water mark has many of us asking, "is value dead?" Or, perhaps more specifically, "is price-to-book dead?" 
2/ But what does "dead" really mean in the world of factors? Are we expecting some Ragnarok-esque cataclysm of performance?
3/ Much more likely is that a factor will just stop working and performance will meander randomly in a seemingly never ending frozen wasteland of performance.
4/ Consider that SMB has failed to make any meaningful new highs since it was originally published in 1981. It took nearly 30 years for practitioners and academics to become more vocal about its veracity.
5/ Part of the "problem" is that factor return variance is so large on a year-to-year basis that it will likely take decades before there is enough data to overturn the statistical significance.
6/ We explore this very idea in this week's commentary. We simulate out future returns for each factor assuming zero mean and use Bayesian inference to update our "prior" beliefs about the expected returns, and stop when returns are no longer statistically different from zero.
7/ 67 years.
That's the median length it will take for HML to stop being statistically significant assuming forward returns have zero mean, starting today.
That's the median length it will take for HML to stop being statistically significant assuming forward returns have zero mean, starting today.
8/ At the risk of being morbid, I'm far more likely to die before the industry finally sticks a fork in price-to-book, statistically speaking.
9/ As practitioners, we're left in a weird place. On the one hand, we are using the weight of empirical evidence to drive our expectations. But going forward, we won't know whether we are "right" until well after our careers are over.
10/ And even if we are "right," there is so much variance in factor returns that we can still appear very wrong for long periods of time.
11/ Factors will not simply stand up and declare themselves dead. Indeed, the death of a factor will be more Fimulwinter than Ragnarok: not so violent to be the end of days, but enough to cause pain and frustration among investors for a very long time.
FIN.
FIN.
