At its essence, due diligence on any investment strategy is aimed at answering whether the existing record is repeatable. Better said, when will the strategy be likely to perform well, however you define ‘well’.
Decades of empirical evidence suggest that all investment styles episodically have their day, so after a while you begin to dismiss recent good/(bad) performance out of hand as being only one data point. (Explaining that dismissal to clients is a different issue.) Is the good/(bad) performance merely reflecting a climate that is unlikely to repeat any time soon?
No matter the investment climate there’s always a strategy that performs well. Did the manager correctly identify the climate and pick the right strategy? Or is the manager always pursuing the same strategy – mindfully or mindlessly – and this just happened to be his time?
You can always identify styles that are in favor at different periods of time. Global consumer stocks in the mid-1990s gave way to the Internet/Tech bubble. In both cases a narrow set of stocks was in favor, and the market was strong so remaining fully invested seemed crucial. The bubble was followed by a long bust – a couple of years of a severe bear market, when it was better not to reinvest cash in a hurry. 2003 hosted a “junk rally,” a rebound in volatile stocks – the more volatile, the better they performed. Capitalization cycles have been longer – large caps reigned in the 1990s, small and mid-caps in the 2000s. Et cetera.
Aware that the climate is always shifting, it’s natural to be skeptical when viewing someone’s backtested simulation results. Data snooping is always a likely suspect.