Why there are labels at all

Until 2017 an Indian fund house could run three “large cap” funds with three different definitions of large, and a scheme called Opportunities could hold anything. Comparing funds meant first working out what each one actually was.

SEBI’s categorisation circular of October 2017 fixed that with two rules. Each fund house may run only one scheme per category (with a few exceptions such as index funds), and every category is defined by a hard portfolio constraint — a minimum the scheme must hold, checked continuously, not a description of intent.

That second rule is the useful part. A category name is a promise about the portfolio that the manager cannot break without SEBI noticing. Knowing exactly what each promise is — and, just as important, what it leaves open — is what lets you compare like with like, which the benchmarks post argued is most of the work in evaluating a fund.

Definitions below start from that circular (SEBI/HO/IMD/DF3/CIR/P/2017/114, 6 October 2017 (categorisation and rationalisation of mutual fund schemes), as amended in 2020), later consolidated in SEBI’s Master Circular for Mutual Funds, 27 June 2024 (clause 2.6). SEBI then replaced that clause with a revised categorisation circular (HO/24/13/15(2)2026-IMD-RAC4/I/5764/2026, 26 February 2026): value, contra, dividend yield and focused funds now need at least 80% in equity rather than 65%, a fund house may run both a value and a contra fund as long as their portfolios overlap by no more than 50%, sectoral and thematic funds face a similar overlap cap, and the “solution-oriented” category was discontinued. Existing schemes got six months to comply (three years for the sectoral/thematic overlap cap). From 1 April 2026 these rules sit in SEBI’s Master Circular for Mutual Funds, HO/24/13/11(1)2026-IMD-POD-1/I/7602/2026, 20 March 2026 (paras 3.7 and 3.8), which folds in the February 2026 circular; the equity thresholds below match its para 3.8.1. The earlier 65% figure for these categories comes from the 2017 framework and was not re-read for this post. SEBI revises these, so check the current text before relying on one.

Market cap bands: who counts as large, mid and small

Every equity category starts from one list. Twice a year AMFI (the Association of Mutual Funds in India) ranks all listed companies by average market capitalisation and publishes the cut-offs:

Large cap  =  ranks   1 to 100
Mid cap    =  ranks 101 to 250
Small cap  =  ranks 251 onwards

Two consequences people miss. The bands are defined by rank, not by a rupee size, so the 100th company today is a far bigger business than the 100th company in 2017 — “large cap” has quietly inflated. And because the list is republished every six months, a stock can move from mid to large (or back) without anything happening to the company. A mid-cap fund holding it then has one month to rebalance to the updated list.

Equity categories

Category The hard rule Core band
Large cap At least 80% of assets in large-cap stocks Stocks ranked 1st to 100th by market cap
Large & mid cap At least 35% in large caps AND at least 35% in mid caps Ranks 1-100 and 101-250
Mid cap At least 65% in mid-cap stocks Ranks 101st to 250th
Small cap At least 65% in small-cap stocks Rank 251st onwards
Multi cap At least 75% in equity, with at least 25% EACH in large, mid and small caps All three bands, forced
Flexi cap At least 65% in equity; no minimum in any band Anywhere the manager likes
Focused At most 30 stocks; at least 80% in equity Any band
Value At least 80% in equity, following a value strategy Any band
Contra At least 80% in equity, following a contrarian strategy; a fund house running both value and contra must keep their overlap at or below 50% Any band
Dividend yield At least 80% in equity, mostly dividend-paying stocks Any band
Sectoral / thematic At least 80% in one sector or one theme That sector or theme only
ELSS (equity-linked savings scheme) At least 80% in equity, 3-year lock-in, tax deduction under the old regime Any
Index fund / ETF (exchange-traded fund), filed under “Other schemes” At least 95% in the securities of the index it tracks Whatever the index holds

That’s the equity group only. SEBI also defines debt categories (below), hybrid categories, life-cycle funds, and an “other schemes” group that holds index funds, ETFs and funds of funds — hybrids and the rest are out of scope here.

Read the rules for what they don’t say.

Large cap must keep 80% in the top 100 — but the other 20% is free, and a “large cap” fund’s excess return over the Nifty can come from that 20%, which is not large-cap exposure at all.

Flexi cap has no band minimums whatsoever. Two flexi-cap funds can be 80% large cap and 60% small cap respectively and both be perfectly compliant. Comparing them to each other, or to one benchmark, tells you very little until you look at where the money actually sits.

Multi cap is the opposite: it forces at least a quarter each into large, mid and small, so it is structurally more mid-and-small-heavy than most “flexi” funds. The two names sound interchangeable. The portfolios are not. (The flexi-cap category was created in November 2020 precisely because the multi-cap rule was tightened in September 2020, and many funds that had called themselves multi cap moved over rather than be forced into small caps.)

Index fund / ETF is the one category where the constraint is total: at least 95% in the index’s securities. That’s what makes the arithmetic in this series clean — there’s no discretionary residue to explain.

🧒 Explain it like I'm 10 (optional — skip if this is already clear)

Think of a school sports day with age groups. “Under-10” means every runner must be under 10 — the rule is checked, so you know a race is fair.

But imagine one race called “Open” where anyone can run, and another called “Mixed” where the organiser must pick a third of the team from each age group. Both sound like “everyone welcome”. They’re completely different races. If you compared an Open team full of 15-year-olds against a Mixed team, the Open team would win, and it would mean nothing.

Flexi cap is the Open race. Multi cap is the Mixed one.

Debt categories: duration by name, credit by rule

Debt categories are sliced two ways at once, and the debt funds post explained why both matter: duration decides how much the NAV moves when rates move; credit quality decides whether a borrower might not pay.

Most debt categories are defined by Macaulay duration — the weighted-average time to the portfolio’s cash flows, a close cousin of the modified duration used to estimate price sensitivity. A few are defined by what they hold instead.

Category The hard rule What it implies
Overnight Securities maturing in 1 day Rate: ~none. Credit: ~none.
Liquid Debt and money-market paper maturing within 91 days Rate: very low. Credit: low, mostly top-rated.
Ultra short duration Macaulay duration 3 to 6 months Rate: low.
Low duration Macaulay duration 6 to 12 months Rate: low.
Money market Money-market instruments maturing within 1 year Rate: low. Credit: depends on paper.
Short duration Macaulay duration 1 to 3 years Rate: moderate.
Medium duration Macaulay duration 3 to 4 years Rate: moderate to high.
Medium to long duration Macaulay duration 4 to 7 years Rate: high.
Long duration Macaulay duration above 7 years Rate: high.
Dynamic bond Any duration; manager decides Rate: whatever the manager is betting on.
Corporate bond At least 80% in AA+ and above rated corporate bonds Credit: low to moderate.
Credit risk At least 65% in AA and below rated corporate bonds Credit: HIGH by design.
Banking & PSU At least 80% in debt of banks, PSUs (public sector undertakings) and public financial institutions Credit: low.
Gilt At least 80% in government securities, any maturity Rate: usually high. Credit: sovereign.
Gilt with 10-year constant duration At least 80% in G-secs (government securities), Macaulay duration held at 10 years Rate: high and constant.
Floater At least 65% in floating-rate instruments Rate: low by construction.

The row to stare at is credit risk: the category is defined as holding at least 65% in bonds rated AA and below. That is not a warning label someone forgot to remove — it is the mandate. The higher yield is the market’s price for exactly that.

Two more things the debt table reveals. Dynamic bond funds have no duration constraint at all, which means their rate risk is whatever the manager currently believes; two dynamic bond funds can be as different as an overnight fund and a gilt fund. And liquid funds are constrained by maturity (91 days) rather than duration, which is why their NAVs look like straight lines.

Why the label matters for comparison

Every performance comparison — returns, rolling returns, drawdowns, Sharpe ratios, rankings, star ratings — is only meaningful within a category, because the category constraint is what makes the funds comparable. A small-cap fund beating a large-cap fund over five years is not information about either manager; it is information about small caps over those five years.

That is also why category boundaries are where the mischief happens. A fund at the loose end of its category (a flexi cap that is really a small-cap fund; a “large cap” leaning hard on its free 20%) will top its category’s table in the years its tilt works, and look like a genius against peers who took the label literally.

Common mistakes

  • Treating flexi cap and multi cap as synonyms. One has no band minimums; the other forces 25% into each band. They are structurally different products.
  • Assuming “large cap” means 100% large cap. It means 80%. Check the factsheet for where the other 20% sits — the factsheet post shows where.
  • Comparing across categories. A fund winning against a different category’s funds has told you about the category, not the fund.
  • Reading “dynamic” as “safe”. It means unconstrained duration, which is the opposite of a promise.
  • Not noticing a category change. Funds do get reclassified; a scheme’s five-year record may belong to a different mandate.

Sources

Takeaway: SEBI’s categories are hard portfolio constraints, not marketing adjectives — large cap means 80% in the top 100 companies, credit risk means 65% in bonds rated AA or below, flexi cap means no band minimums at all. Know the constraint and you know what a fund’s record is actually a record of, which is the only basis on which comparing two funds means anything.