A large cap fund beat its peers by 2.4% a year for three years, but its quarterly holdings show 26% to 31% in stocks ranked outside the top 100 by market value. Is the outperformance skill, and what do you do about a fund whose label no longer describes it?

Most of the outperformance is not skill but a different fund: about 2.0 of the 2.4 points came from holding mid and small caps in years they beat large caps. The remaining 0.45 points is too small to tell apart from luck over three years. Investors bought a large cap fund and got a large and mid cap fund. Move it off any large cap recommendation, raise the breach with the AMC, and judge it against the right benchmark.

Suppose a vegetarian restaurant wins a taste award, and you then discover a third of its dishes contained meat. The food may still be good, but the award says nothing about vegetarian cooking. A fund's return only tells you about the manager's skill within its category if the fund actually stayed in that category. Meghavan held 28.5% outside the top 100 on average, every quarter above the 20% the rules leave for such stocks. This is style driftWhen a fund’s holdings move away from the style or market segment its category and name describe, so it no longer behaves like the fund investors chose., and it ran for three years.

Compare it with the peers, since that is the claim being made. The fund held about 19.5 points more than peers outside the top 100. Those stocks beat large caps by about 10 points a year. 19.5% of the portfolio earning 10 points more adds about 1.95 points a year, so the tilt alone explains about four fifths of the lead. This is a simple attribution; a fuller one would compare each segment's holdings with its own index. The direction would not change.

Not provably. Suppose the fund's returns differ from peers' by about 4 points a year in a typical year. Over three years the uncertainty around its average lead is about 4 divided by the square root of 3, roughly 2.3 points. A 0.45 point residual against 2.3 points of noise is a t-statistic of about 0.2, nowhere near evidence of skill. The tilt also cuts the other way. In a year when mid and small caps lag large caps by 15 points, the same positioning would cost about 2.9 points, and investors who bought it as their large cap core would get a fall they did not sign up for.

Three steps. First, check the innocent explanation: stocks move in and out of the top 100 when the official ranking list is updated, and funds get a window to rebalance. A breach that persists across twelve quarters is not that. Second, take the fund off any large cap recommendation and judge it against a large and mid cap benchmark, which describes what investors actually own. Third, raise it with the AMC in writing and, if you are a distributor, tell clients who hold it as their large cap fund that it carries more mid cap risk than its label suggests. Reporting suspected category breaches to the regulator is the AMC's and trustees' job in the first instance, but a research house that sees a persistent one should not stay silent.