How do I know if my fund tracks a capped variant?

The index's full name on the factsheet (the "25/50," "35/20," "Capped" suffixes) and the methodology link. If the exposure is concentrated (sector, single country), assume a cap exists and find it.

Is capping good or bad for returns?

Neither inherently: it trims winners (drag in momentum eras) and redistributes (cushion in reversals), while making the fund legally possible at all. It's a behavior difference to know, priced into the capped index your fund is honestly tracking.

Why did a huge IPO take months to enter my index fund?

Seasoning rules, float thresholds (early float is often small), and review calendars: plus, for the S&P family, committee discretion. The stock's full float-adjusted weight arrives on schedule, not on excitement.

Do float factors change often?

On defined review dates and corporate events (lock-up expiries, buybacks, secondary offerings). Each change is a mechanical weight adjustment trackers implement: part of the ordinary turnover Chapter 41 prices.

Whose buffer rules are "best"?

Wrong frame: buffers trade turnover against definitional crispness, and every major provider's are published and defensible. The comparative question that matters is Chapter 35's: realized TD of funds on each manual.

Index Construction: Float, Caps and Buffers

Free float: counting only the tradable shares

Raw market cap counts every share; float-adjusted cap counts only those actually available to public investors, excluding founder and family blocks, strategic corporate stakes, government holdings and shares under lock-up. All major equity indices have been float-adjusted for two decades (S&P completed the shift in 2005), for a reason that is pure Chapter 31 mechanics: an index weighting shares that funds cannot buy forces trackers to squeeze into the tradable remainder, distorting prices and tracking alike.

Consequences you can observe: a company can be enormous by raw cap yet modest in your index because insiders hold half of it (float factors are published per constituent); privatizations and lock-up expiries produce scheduled weight increases with zero price opinion attached; and state-heavy markets (parts of EM) punch below their raw-cap weight in global indices by design. When a headline says "Company X is Y% of the market" and your factsheet disagrees, float adjustment is usually the referee.

Screens: the bouncer at the door

Before weighting comes eligibility: minimum float percentage, minimum tradable size, liquidity thresholds (median traded value), listing standards, and in some indices profitability or seasoning requirements (the S&P 500's positive-earnings rule; new-listing waiting periods). Screens are why "the 500 largest US companies" is approximately true rather than exactly true, and why some famous names linger outside famous indices longer than their size suggests. For you, screens are mostly protective: they keep untradeable paper out of the basket your fund must actually buy: Chapter 37's cost logic, enforced at the manual level.

From census to constituents. Four filters, each preventing a specific failure.

Caps: ceilings with legal teeth

Concentrated markets meet fund law here. UCITS diversification rules (the "5/10/40" regime: positions above 5% may together not exceed 40% of the fund, no single position above 10%, with index-tracker allowances up to 20%, exceptionally 35%) mean an uncapped index dominated by one giant is literally untrackable by a European fund. The industry's answer: capped index variants: "MSCI World 35/20 Capped," sector indices with 25%/50% aggregate rules, country indices with issuer ceilings: rebalanced back under the caps at each review.

Where this bites in real portfolios:

Sector ETFs (Chapter 42 next): a technology index without caps would be a two-stock fund in mega-cap eras; the capped variant you can actually buy behaves measurably differently from the raw sector: check which your fund tracks.
Concentrated national markets: a single dominant company (a Nestlé, a Samsung, a TSMC in its home index) gets capped, redistributing weight to the rest: your "country exposure" is the capped country, a feature to know, not a flaw.
Cap-induced turnover: every review that trims the winner back to its ceiling is a small systematic sell-high (and buy-back-after-declines): modest costs, modest rebalancing character, visible in TD.

The US analogue: RIC diversification rules push the same engineering into US sector and thematic products. Same physics, different statute.

Buffers: hysteresis against churn

An index ranking cutoff applied naively would eject and readmit borderline companies every review as they wobble around rank 100 or the size threshold: pure turnover, pure cost, zero information. Buffers fix it with asymmetric gates, Chapter 5's FTSE 100 example generalized: enter at rank 90 or better, exit only at 111 or worse; size-segment indices use band overlaps the same way (a stock must fall well into mid-cap territory before large-cap indices release it). Russell's annual reconstitution, MSCI's semi-annual reviews, all run buffered logic. Effect on you: meaningfully lower turnover (and cost) than the index's crisp definition implies, at the price of slight definitional fuzziness at the edges: an engineering trade this book endorses without reservation.

Between reviews: fast entry, deletions and corporate actions

Manuals contain event pages: fast-entry rules admit qualifying IPOs or spin-offs before the next review (with size thresholds; Tesla's famously deliberated 2020 S&P entry showed committee discretion layered on rules); deletion rules handle acquisitions, delistings and bankruptcies immediately (replacement from the bench, Chapter 5's FTSE promotion mechanics); corporate actions (splits, rights issues, special dividends) adjust weights mechanically by published formulas; and the emergency pages, sanctions and market closures, got their grim demonstration in Chapter 39's Russia case. Trackers execute all of it on effective dates: the crowd-trade dynamics of which are exactly Chapter 41's subject.

Buffers and caps in action. The engineering that makes an index trackable — and slightly different from its own name.

A worked example

Selin holds a Taiwan equity ETF and reads one October factsheet line she'd always skipped: "Index: MSCI Taiwan 25/50 Capped."

Fifteen minutes with the methodology: TSMC's uncapped weight would be ~45%+ of the raw index; the 25/50 rules hold the largest issuer near 22–24% with quarterly trims; the trims mean her fund systematically sold slivers of TSMC into its great run (a drag versus the raw index) and would buy them back in drawdowns (a cushion). Her fund's tracking versus the capped index: excellent; versus the raw country: structurally different, by design and by law. The generalizable lesson she writes down: the index name's suffix was the risk disclosure: capped/float/buffer vocabulary is where a fund's real behavior is specified, one suffix at a time.