But Arnott's statistical tests indicated that the natural tendency of good earnings to follow bad was a far less powerful predictor of ten-year earnings growth than the payout ratio itself.

FORBES: The great stock illusion