Unless you manage to purchase an index fund that produces zero tracking error, at some point the investments you own will deviate from their associated benchmark. Whether it's the tracking error of an index fund, or the relative under- or out-performance of an active fund manager, returns can vary over time. And while most of us don't fret over small deviations from a benchmark - nor do we mind when the deviation is due to outperformance! - but at some point, a fund may underperform its relevant benchmark by enough, and for a long enough period of time, that you have to question whether it's time for a change. But the caveat is... what IS a big enough underperformance deviation, and how long must it persist, before you actually do decide to make a change? What is the best practice for deciding when a fund has lost its Mojo?