There is a quiet trap built into the mutual fund industry. A small-cap fund finds an edge, genuine stock-picking skill in an under-researched corner of the market. It delivers 12-15% alpha in its early years. Retail investors pour money in. SIP mandates stack up. The fund’s AUM doubles, then doubles again.
And then something changes. Not the manager. Not the mandate. The math.
THE AUM CURSE As a small-cap or mid-cap fund grows beyond a certain size, it is forced to hold more stocks simply to deploy capital, until it effectively owns the entire investable universe of its category. At that point, it is the benchmark. Alpha becomes structurally impossible.
We tested this theory using 15 years of portfolio data from one of India’s largest and longest-running small-cap funds.
The AUM tidal wave
CHART
Annual AUM (₹ Crore)
AUM grew from ₹339 Cr in 2010 to nearly ₹49,600 Cr at peak, a 146x increase. The fund went from boutique to behemoth in 15 years.
The portfolio bloat
As AUM grew, the fund needed to deploy capital into progressively more stocks. India’s small-cap universe has roughly 200-300 investable names. At ₹49,600 Cr, the fund was systematically forced to own most of them.
CHART
Number of stocks held in portfolio
Stock count grew from ~50 in 2012 to 251 today, nearly the entire investable small-cap universe. At this point, concentration bets become impossible.
The alpha collapse
As the portfolio spread thin across the entire small-cap universe, the fund’s ability to outperform its benchmark eroded, and then reversed.
CHART
Rolling 3-year alpha vs. benchmark (%)
Peak alpha of ~9.5% in early 2023 collapsed to -1.1% by late 2026. The fund now underperforms its own benchmark on a 3-year rolling basis.
THE SEQUENCE MATTERS First comes AUM growth, then comes portfolio bloat, then comes alpha decay, in that order, with a lag of roughly 18-24 months between each stage.
Is this specific to one fund?
The stock-count part, no. We ran the same analysis across active small-cap and mid-cap funds, grouping them by peak AUM, and stock count rises consistently across every bucket. Alpha is messier: it does not decline cleanly with bucket size, and the largest bucket posts the highest average alpha of the four.
CHART
Average stocks held by AUM bucket
Stock count rises with AUM across the four buckets, 72 to 107. Alpha does not fall with it - it dips to +1.2% in the middle bracket and then recovers, and the largest bucket posts the highest average alpha of the four.
| AUM Bucket | Funds | Avg Stocks Held | Avg 3Y Alpha |
|---|---|---|---|
| Below ₹3,000 Cr | 44 | 72 | +3.4% |
| ₹3,000-8,000 Cr | 14 | 74 | +1.8% |
| ₹8,000-20,000 Cr | 7 | 105 | +3.3% |
| Above ₹20,000 Cr | 4 | 107 | +4.0% |
Stock count behaves as the mechanism predicts: it rises with AUM, 72 to 107 across the four buckets. Alpha does not follow it down. The weakest bucket is the middle one, ₹3,000-8,000 Cr, where 6 of 14 funds sit below their benchmark, and the largest bucket is the strongest, with all four funds positive in a tight +2.8% to +5.5% band. No size story predicts that shape.
Two reasons this table cannot settle the question, and neither is a defect in the single-fund chart above.
It measures a different thing. The fund chart is one portfolio’s own 15-year path: same mandate, same house, alpha falling as its own stock count climbs. The table is a snapshot of different funds against each other in one trailing 3-year window. A fund that already bloated and lost its edge does not stay in the top bucket to be counted, it shrinks out of it, or is wound up and leaves the sample entirely. The mechanism moves funds between the buckets that are supposed to detect it.
Diffusion is a choice, not a ceiling. The clearest evidence is inside the largest bucket. Motilal Oswal Midcap runs ₹22,220 Cr in 24 stocks. Nippon India Small Cap runs ₹22,912 Cr in 230. Both beat their benchmark. Bandhan Small Cap holds 198 stocks at ₹11,916 Cr and posts the highest alpha in the entire sample, +11.6%. Scale forces a manager to deploy more capital; it does not force them to spread it across the whole universe, and the funds that refuse are not the ones underperforming.
Read the two together, not as one claim. The single-fund history is the mechanism in isolation and it is real. The bucket table shows the mechanism is not the only force acting on returns at the category level, and that a manager willing to stay concentrated can carry size that the mechanism says should sink them.
The mechanism
A fund managing ₹45,000 Cr in small caps cannot build a meaningful position in a stock with a ₹2,000 Cr market cap - it would move the price on entry and be unable to exit. So it is forced to spread across 200+ stocks. That is not active management. That is index replication with fees.
A few reference numbers behind the mechanism: roughly 250 investable small-cap stocks exist in India (liquid, regulated, above ₹1,000 Cr market cap); funds typically cap any single position at 2-3% of the portfolio before market impact on entry/exit becomes a problem; at ₹5,000 Cr AUM, that caps a typical position at ₹100-150 Cr; and the lag between an AUM milestone being crossed and alpha decay becoming visible is roughly 18-24 months.
What this means for you
AUM isn’t a quality signal, it can be the opposite one. The same fund that delivered 12% alpha at ₹2,000 Cr may structurally be unable to repeat that at ₹25,000 Cr, regardless of manager skill.
A few practical implications:
- AUM caps matter. SEBI has been pushing AMCs to soft-close small and mid-cap funds past certain thresholds. That pressure is data-backed.
- Popularity forces a structurally different portfolio. The funds with the most SIP mandates and highest AUM are the most constrained on stock count, even if a particular large fund still manages to post decent alpha. Constrained does not automatically mean underperforming, it means the manager has less room to make concentrated bets.
- Watch the stock count. If a mid-cap fund crosses 80-100 stocks, ask what it is actually doing differently from the benchmark index.
- Consider newer entrants. A mid-cap or small-cap fund with ₹1,000-4,000 Cr AUM has far more room to run concentrated, high-conviction portfolios.
THE IRONY The best time to invest in a star fund was before it became a star. By the time it appears in top-quartile lists and attracts inflows, the structural edge may already be eroding.
METHODOLOGY Analysis uses publicly disclosed monthly portfolio data across 15+ years and quarterly AUM disclosures filed with the regulator. Rolling 3-year alpha is measured against each fund’s declared benchmark index. Stock counts reflect distinct equity positions per monthly portfolio snapshot. Cross-sectional bucket analysis covers active, direct-plan, growth-option equity funds in small-cap and mid-cap categories with AUM data available from 2024 onward. The two largest buckets hold only 7 and 4 funds. In the smallest bucket, 14 of the 44 funds have no 3-year alpha yet, too recently launched, so its average is computed on the 30 that do, which tilts it toward funds that have already survived three years. Alpha is never imputed or set to zero. Funds with incomplete holdings data are excluded from the average stock count calculation but included in the alpha computation. The specific fund charted at the top of this story is not named deliberately - the pattern, not the fund, is the point.
Data: AMFI public disclosures. Analysis: Punji Research. Not investment advice.