So on its first out of sample forecast, was the CS rule a success or failure?