Closet indexing is a mismatch between an advertised active strategy and benchmark-like exposure, assessed through holdings, returns, costs, and disclosures.
Closet indexing is the practice of presenting a fund as actively managed while keeping its portfolio close to a benchmark, potentially charging investors for more independent investment selection than they receive. It is also called index hugging.
The issue is not that index investing is inherently undesirable. It is a possible mismatch between the service described, the exposures delivered, and the price paid. A fund that openly aims to track an index is not a closet indexer merely because it succeeds in doing so.
In its 2016 statement on potential closet index tracking, the European Securities and Markets Authority described concerns about funds whose claimed active approach did not match their benchmark-like portfolios.
ESMA combined numerical screening with a review of investor disclosures. It cautioned that statistical results alone were only an initial step toward establishing what a particular fund was doing. This historical statement explains the concern; it is not a universal legal test for funds in every jurisdiction.
A disclosed, tightly benchmark-constrained active mandate can legitimately have limited room to differ from its index. The question is whether the portfolio’s behavior is consistent with what investors were told. Do not infer a manager’s motives solely from portfolio similarity.
For a simple, fully invested, long-only equity portfolio, active share compares each security’s portfolio weight with its benchmark weight:
Here, the weights are decimals and the calculation includes every security in either the portfolio or the benchmark. A security absent from one side receives a zero weight on that side.
ESMA’s 2020 working paper, Annex 2, sets out this holdings-based measure and distinguishes it from return-based measures. The one-half factor avoids counting the same reallocated weight twice.
Suppose the following table contains all holdings of a hypothetical portfolio and its appropriate benchmark on the same date. Both are long-only, with no cash, derivatives, or leverage.
| Company | Benchmark weight | Portfolio weight | Absolute difference |
|---|---|---|---|
| A | 40% | 41% | 1 percentage point |
| B | 30% | 29% | 1 percentage point |
| C | 20% | 20% | 0 percentage points |
| D | 10% | 9% | 1 percentage point |
| E | 0% | 1% | 1 percentage point |
| Total | 100% | 100% | 4 percentage points |
The active share is:
The portfolio has moved one percentage point from B to A and another from D to E relative to the benchmark. Although it holds an extra company, its weighted difference is small.
This 2% is not an investment return, a turnover rate, or a probability of outperformance. It establishes close holdings overlap in this constructed snapshot. Determining whether the actual service was misrepresented would require the mandate, disclosures, costs, and a longer history.
Sector totals can hide security-level choices. Replacing one technology company with another can leave the technology allocation unchanged while materially changing the holdings. Conversely, owning many familiar benchmark companies does not reveal whether their portfolio weights are close to the index weights.
Different measurements answer different questions:
| Evidence | What it can show | What it cannot establish alone |
|---|---|---|
| Active share | How security weights differ at a measurement date | Return volatility, future skill, or the manager’s intent |
| Tracking Error | Variability of periodic portfolio-minus-benchmark returns | The exact holdings or whether disclosures were misleading |
| Portfolio-minus-benchmark return | Relative performance for the measured period | A pattern of close tracking from one observation |
| Prospectus and investor reports | The promised objective, discretion, restrictions, and reported costs | Whether every portfolio decision matched those descriptions |
A fund can trail its index by a similar amount each period and have low tracking error but persistent underperformance. Low relative-return variability also does not make its investments safe: the portfolio and benchmark can both fall sharply.
The basic active-share formula is useful, but its inputs need care. CFA Institute’s discussion of active-share measurement highlights complications involving benchmark choice, cash, different securities of the same issuer, and exposure held through other instruments.
For example, comparing an index fund’s ticker directly with an index’s underlying companies can make economically similar exposure look different. Fund holdings or derivatives may require a documented look-through method. Security-level and issuer-level calculations are not automatically interchangeable.
A review should also distinguish a single holdings date from a persistent pattern. Incomplete disclosures, unmatched reporting dates, or a recently changed mandate can undermine the comparison.
Do not turn a research cutoff into a universal definition. A higher active share can reflect different holdings without demonstrating better investment decisions.
If two portfolios deliver nearly the same exposure before expenses, different charges can still produce meaningfully different investor outcomes. The relevant comparison is the service and return after applicable costs, not the active or passive label alone.
For a simplified hypothetical comparison, assume two accounts each start with $10,000, earn the same 8% before an advisory fee, and have no other costs, cash flows, or taxes. Both charge their fee once on the $10,800 year-end value before the fee.
The difference is $97.20, entirely due to the assumed charges. These are illustrative account fees, not quoted fund expense ratios, typical market prices, or evidence that any real fund is a closet indexer.
Actual fund expenses usually follow different accrual arrangements. The SEC’s fund-fee bulletin explains operating expenses, share-class differences, and costs outside the expense ratio. Compare the correct share class and avoid deducting costs twice when reported returns already reflect them.
These steps help frame questions; they are not personalized investment advice or a legal determination about a manager. Losses remain possible in both active and index strategies.