Database · 1,972 ETFs with TD from primary sources · growing every week as new issuers join

Tracking difference: an ETF's real cost

The TER tells you what an ETF should cost. Tracking difference tells you what it actually cost: the gap between the fund's return and its index's, year by year. Look up your ETF below — or start from the explanation.

ETFs without an honestly computable TD appear with a dedicated label (see methodology). The full page shows TD year by year.

What it is, in plain words

A passive ETF makes one promise: to copy its index. Tracking difference measures how well that promise was kept:

TD = fund return − index return
Negative = drag (the fund returned less than its index) · positive = extra return.

If the index gains 10% and the ETF 9.8%, the TD is −0.2%: holding it that year cost 0.2%, whatever the TER said. Everything is inside this number: the TER, the fund's internal trading costs, taxes on dividends received, and securities-lending revenue pulling the other way.

Which is why it can even be positive: iShares Core S&P 500, TER 0.07%, shows a 12-month TD of +0.14% — it returned more than its index.

Medians across our census

CategoryMedian TD (12 months)ETF
Equity-0.07%1,242
Bonds-0.10%617

Median of the 12-month fund − index gap, where the index is the one each issuer declares (normally a total-return variant; inconsistent variants are discarded by the sanity net). Refreshed with the monthly data run.

Our methodology, declared

  • PRIMARY-SOURCE index series: the one declared by the issuer itself, never rebuilt from third parties.
  • Fund side: total return (dividends reinvested) where published, otherwise NAV but only for accumulating classes — never a distributing class's NAV, which would drop dividends and distort the TD.
  • Sanity net: if the number doesn't add up (e.g. a price-return index variant instead of total return), the figure is discarded and never published.
  • Hedged share classes: TD only makes sense against the index hedged to the same currency. Many issuers publish, next to the hedged class, only the unhedged index: the gap would measure the exchange rate, not replication. Declared rule: if a class is hedged and the computed TD exceeds ±2 points in any year, the figure is not published — the fund page explains why and points, where one exists, to the unhedged sister class. Hedged classes with TD below that threshold (a truly hedged index) remain visible.
  • External check: the yearly fund/index returns we compute are periodically reconciled with independent sources; where a discrepancy emerges, a case is opened and the figure stays suspended until it is explained — never published as an uncertain number.

Where tracking difference comes from

TD is not a random number: it is the sum of a few well-known effects that financial literature and regulators have described for years. Knowing them helps read the figure without surprises.

  • Ongoing charges (TER). The most predictable component: charged to the fund every day, they push TD down by roughly the TER, year after year.
  • Withholding tax on foreign dividends. Net-total-return indices assume a standard withholding rate; an Irish- or Luxembourg-domiciled fund, thanks to tax treaties, often bears a lower one and gains ground on the index. It is the main reason many US-equity ETFs show a TD better than their TER — and swap-based synthetic funds even more so.
  • Securities lending. Many physical funds lend part of their holdings against collateral and earn a fee: revenue the index does not have, which improves TD. Our fund pages flag it whenever the issuer discloses it.
  • Replication frictions. Internal trading costs, sampling on very broad indices, liquidity in small markets, dividend reinvestment lags: small frictions that, added up, widen the gap — especially in emerging markets and small caps.

The framework is also regulatory: ESMA’s Guidelines on ETFs and other UCITS issues (ESMA/2012/474, consolidated text ESMA/2014/937, paras 9-11) define the annual tracking difference as “the difference between the annual return of the index-tracking UCITS and the annual return of the tracked index” and require the prospectus, KIID and annual report to disclose anticipated and realised tracking error, the factors affecting replication (transaction costs, illiquid components, dividend reinvestment) and the year’s tracking difference. It is the very number we compute here — from the declared series, for every fund, with one formula.

Frequently asked questions

What is an ETF's tracking difference?

It is the difference between the ETF's return and its index's return over the same period (TD = fund return − index return). If the index gains 10% and the ETF 9.8%, the TD is −0.2%: the effective cost of holding it that year.

How is tracking difference different from the TER?

The TER is the cost declared upfront in the documents; tracking difference is what actually happened. It includes the TER, internal trading costs, tax frictions on dividends, and securities-lending revenue — which is why it can beat the TER, and sometimes even be positive.

Why can tracking difference be positive?

Because some ETFs recover more costs than they pay: securities lending generates revenue and dividend-tax handling can be more efficient than the index assumes. A real example from our database: iShares Core S&P 500 has a 0.07% TER yet returned more than its index over the last 12 months.

Why do some ETFs have no tracking difference?

We compute it only where it is honest: it requires the issuer-declared index series and the fund's total return (or an accumulating class). Where these are missing, we prefer showing nothing to publishing an imprecise number.

Why do other sites show a TD for some hedged ETFs while you don't?

Because for those share classes the issuer publishes, next to the fund — hedged to euro — only the unhedged index in foreign currency. Subtracting one from the other does not measure replication quality: it measures the exchange rate. In a year when the currency moves 5%, “the TD” would come out near 5 points: a plausible-looking number, with decimals, and wrong. We would rather show one figure less than one figure wrong, and we say so on the fund page. To see how well that fund tracks its index, look at the unhedged sister class, where fund and index share the same currency: the management is the same.

Is tracking difference comparable across websites?

With caution: different methodologies, windows and index variants produce different numbers. Our methodology is declared on this page, and the yearly returns we compute match the sector's international references.

Want to compare more ETFs?

The screener has every census ETF with real costs (TER + transaction), white-list tax share, securities lending and charts — and every row links to the full page with year-by-year TD.

Open the ETF Screener
Informational figure computed by Rebalix on issuer-declared series; it describes the past and is not indicative of future results. Not an investment recommendation.