Guide · Portfolio construction · Updated 2026-07-05

Mutual fund portfolio overlap: why more funds is not more diversification

Adding a fund feels like adding safety. It usually is not. Diversification is a property of how holdings move relative to each other, not of how many line items sit in a portfolio. Because Indian equity funds are sorted into categories with a shared, regulator-defined universe, two funds with the same label tend to own most of the same names. This guide sets out why that convergence happens, how overlap is measured from published holdings, why overlapping funds offer no cushion in a fall, and what genuinely lowers risk. The value here is the part the fund-count instinct hides: past a point, another fund in the same category is cost without diversification.

Why same-category funds converge on the same stocks

Overlap is not an accident of manager taste. It is structural, and three forces drive it. The first is the investable universe. Under SEBI's 2017 categorisation framework, a large-cap fund is confined to the top 100 listed companies by full market capitalisation. That single rule means every large-cap fund in the country is choosing from the same hundred names. Two managers can disagree on weights, but they cannot escape the pool.

The second force is the benchmark. Same-category funds are measured against the same broad index, which is itself built from those same large companies. A manager's performance is judged as a deviation from that index, so the index is the gravitational centre of the portfolio. The third force is closet indexing: a manager who strays far from the benchmark risks a stretch of underperformance and the loss of the mandate, so many quietly stay close to it. Researchers measure this with active share, the fraction of a portfolio that differs from its benchmark. A genuinely active large-cap manager runs a high active share; readings that drift toward the benchmark signal a fund that charges active fees for near-index behaviour. Put the three forces together, a shared universe, a shared benchmark, and a career incentive to hug it, and convergence is the natural outcome.

How overlap is measured

Overlap has a precise definition, and it is simpler than it sounds. For every stock held by both funds, you take the lower of the two weights, because only the smaller amount is genuinely held in common. You then add those minimums across all shared holdings. The total is the percentage of one portfolio reproduced inside the other. A stock held only by one fund contributes nothing, since there is nothing to duplicate.

The intuition behind taking the minimum is worth pausing on. If a name is 8 percent of Fund A and 6 percent of Fund B, the two portfolios coincide on that stock only up to 6 percent; the extra 2 percent in Fund A is exposure that Fund B does not share. Summing the smaller weight across every common name captures exactly the part of the two portfolios that is the same position. Free portfolio-overlap tools from several Indian platforms perform this on published fund holdings, so the calculation itself is not something you have to do by hand.

Extended across the full list of shared names, the same sum gives the overlap between the two funds as a whole. For a portfolio of several funds, the more revealing exercise is to collapse every fund into a single combined book, weight each stock by the rupees behind it, and rank the result. That is where a holder of six or eight funds often discovers that a handful of names carries the bulk of the equity exposure, regardless of how many fund labels sit above them.

Why overlap matters: correlated drawdowns and duplicated cost

The reason overlap is not a harmless quirk is that diversification only works when holdings are imperfectly correlated. The whole point of spreading capital is that when one part falls, another holds up or falls less, so the portfolio's swing is gentler than any single holding's. That cushioning depends entirely on the holdings not moving in lockstep. Two funds that share most of their names are, for practical purposes, the same position held twice, and two copies of the same position are perfectly correlated. In a decline they fall together, at the same time, by the same magnitude.

So the investor who spread capital across several same-category funds to feel safer has bought no cushion at all. In a drawdown the "diversified" book falls like the concentrated book it actually is, and the surprise is sharpest precisely because the fund count implied protection that was never there. Meanwhile the duplication is not free. Each fund charges its own expense ratio, so on the overlapping portion the holder pays two or three times over for what is, underneath, a single exposure. A second cost is quieter and can matter more: because the same top names recur across funds, the combined single-stock exposure to a given company can climb well past what the holder would ever choose deliberately in one place. The concentration is real; only the appearance of spread is manufactured.

Overlap by combination: where it is high and where it is low

Overlap is not uniform. It tracks how much two funds' universes intersect, which is why the combination matters more than the count. Two funds in the same category sit at the high end, because they are constrained to the same pool of names. Move across a boundary, a large-cap fund paired with a mid or small-cap fund, and the overlap drops sharply, because the SEBI universes barely touch: the top 100 and the 101-to-250 band are different companies by definition. Cross an asset-class boundary, equity against debt or gold, and overlap approaches nothing, because the instruments are not even the same kind of thing.

Read the table as a rule of thumb, not a measurement: it says where to expect duplication before you ever run the numbers. If two funds you hold fall in the top rows, one of them is probably doing little the other does not, and the overlap tool will usually confirm it.

What actually diversifies

If the number of funds is the wrong lever, the right one is the number of genuinely different sources of return. Diversification is bought by combining holdings that respond to different conditions, so they do not all fall on the same day. Four kinds of difference do real work. Different market-cap segments, large against mid against small, lead and lag through different phases of a cycle. Different sectors carry different drivers, so a shock to one need not be a shock to another. Different factors or strategies, value against momentum against low volatility, take turns outperforming and are imperfectly correlated with each other. And different asset classes, equity against debt against gold, are the strongest separators of all, because their prices answer to different forces entirely.

This is where the idea of diworsification earns its place. Peter Lynch coined the word in One Up On Wall Street for the act of adding holdings that dilute a portfolio without improving it, and it captures the overlap problem exactly. The marginal diversification benefit of each new holding falls as the portfolio grows; once the genuinely different exposures are in place, further additions of the same kind contribute cost and complexity while barely touching risk. More funds, past that point, is not more safety. It is more of the same bet, at a higher total fee, dressed as prudence. Knowing which additions still cut risk and which only cut into returns is a judgement about correlation and structure, and that upstream judgement is exactly what the method we teach is built around.

Where this fits, and what to do with it

Overlap analysis is a diagnostic, not a trade. Its job is to replace the comforting fund count with the real picture: the combined, rupee-weighted list of what you actually own, and the degree to which your funds duplicate each other. Once that picture is in front of you, the questions become concrete. Do two of these funds cover the same ground? Is a single company quietly larger, across funds, than you would ever hold on purpose? Is the spread you are paying several fees for actually there, or is it four labels over one core?

None of that is a recommendation about any particular fund, and none of it is advice to buy or sell. It is the structural literacy that lets you read a portfolio honestly before deciding anything: understanding categories as constrained universes, correlation as the thing diversification actually manages, and cost as a certainty where diversification is only a hope. Build that lens first, and the fund count stops being the story. What the funds hold, together, becomes the story.