in the tests we ran using Winner and Loser portfolios based on three-year “formation periods” and followed by three-year “test periods,” the average beta for the Winners was 1.37 and for the Losers was 1.03. So the Winners were actually riskier than the Losers.
