There's a thread at Old School Value that discusses Michael Burry's VIC write-ups. His write-ups all recommended statistically cheap micro-caps, and one commenter, "somrh," notices that most of them did poorly over time. This leads him to wonder if the stocks that Burry recommended were truly undervalued:
"Part of what I'm interested in is whether or not value strategies really are 'value' strategies. Some of them seem to work more like short-term mean reversion strategies that may be really lousy over the long-haul."
My hunch is that somrh is right: many statistically cheap stocks are
cheap because there's some structural handicap that depresses their
ROIC, and this dooms them to low long-term returns regardless of valuation.
A lot of academic studies claim that statistically cheap stocks outperform the market, but they typically look performance over a single year. The ones that look at performance over longer periods typically re-balance annually. Rather than measuring outperformance, these studies may be measuring cheap stocks' ability to
stage a dead cat bounce.