In traditional asset allocation the goal is not to find the "best" asset class, it is to find the best combination of asset classes that gives the client the best chance of achieving the desired result. Each asset class, be it large cap growth, value, small cap, etc, goes through periods where it appears to be the best. Imagine you were doing due diligence on a large cap growth manager in March 2000 and looking at how they had done over the past five years. It would be easy to fool yourself into believing you found the Holy Grail. The same would go for any asset class, that is why you are better off combining asset classes than you are trying to be in just one.
The same logic works in Tactical Asset Allocation . There are a number of tactical methodologies (sets of tactical rules) that work over time. Things like buying the S&P 500 when it is above the 200 day moving average and selling below or using a 2 period RSI to determine overbought and oversold points will have periods where they will do extremely well. It is also not a rare occurrence to do a backtest on a set of rules and find that it does extremely well over the entire backtest period. Tactical methodologies be even more misleading than asset classes as they can go through much longer periods where they are doing well than an asset class can. However, just like there is no one best asset class, there is no one best tactical methodology. Every methodology will fall on its face at some point. Instead of expending effort to find the one methodology that does well in every market, the goal should be to have a mix of different methodologies and expend effort finding the combination that produces the desired results.