We reassess the performance of market-timing rules when controlling for data-snooping biases. For the first time, a comprehensive set of simple and complex market-timing rules is examined and tested for statistical significance, using the White (2000) "Reality Check," the Hansen (2005) SPA test,...