[ThrustVectoring]

Being correct in 2/3rds of predictions isn't enough to get rich - he could very well have gotten more things wrong than the conventional wisdom that earns $0 on average.