Backtested a trading system that trades an index only three ways, 100% long, 0% (out of market) or 100% short from 2004-2011. This system returned 80% annualized with a 60% standard deviation over that period.
Are those GOOD results? Would you could consider this just data-fitting or...