Tracking difference vs tracking error

The clean distinction in one paragraph

Tracking difference (TD) = the cumulative gap between fund return and index return over a period. A positive TD means the fund beat the index; a negative TD means the fund trailed. Usually quoted as % per year for the trailing 1, 3, or 5 years.

Tracking error (TE) = the volatility (standard deviation) of the differences between fund return and index return, sampled at regular intervals (typically monthly). A high TE means the gap is noisy; a low TE means the gap is consistent. It says nothing about whether the gap is positive or negative.

Said differently: tracking difference is "did the fund keep up?"; tracking error is "did the fund keep up smoothly?".

The mathematics, with numbers

Imagine a fund vs its index over four quarters. Fund returns: +5%, +3%, −4%, +2%. Index returns: +5%, +3%, −4%, +2%. Tracking difference = 0%; tracking error = 0%. Perfect replication.

Now imagine fund returns +4%, +4%, −5%, +3%; index returns +5%, +3%, −4%, +2%. Per-quarter differences: −1%, +1%, −1%, +1%. Cumulative fund return ≈ +5.9%; cumulative index return ≈ +5.9%. Tracking difference ≈ 0% — the fund kept up perfectly over the year. But the per-quarter swings give a standard deviation of about 1% — tracking error ≈ 4% annualised. So this fund kept up but did so noisily.

Finally, imagine fund returns +4.7%, +2.7%, −4.3%, +1.7%; index returns +5%, +3%, −4%, +2%. Per-quarter differences are nearly constant at −0.3%. Cumulative fund return ≈ +4.7%; cumulative index ≈ +5.9%. Tracking difference ≈ −1.2% (fund trailed by 1.2%). Tracking error ≈ 0% (the gap is highly consistent). This fund missed the index by 1.2% per year, very smoothly.

Which fund would you rather hold for 30 years? The second one (zero TD, high TE) — over the long horizon the cumulative outcome is the same as the index, even though the monthly numbers wobbled. The third one (negative TD, zero TE) is the worst — it's quietly and consistently missing.

What drives tracking difference

For a UK-listed UCITS ETF, tracking difference is roughly the sum of:

• OCF — the explicit annual cost, taken out daily. This is usually the largest single contributor.
• Withholding tax drag — tax on dividends from underlying companies that the fund can't reclaim
• Sampling vs full replication — if the fund samples (e.g. an EM fund holding 800 of 1,400 index members), the sample will deviate from the index by some amount
• Cash drag — uninvested cash held to meet redemptions can underperform the index in a rising market
• Securities lending revenue (a POSITIVE driver) — many ETFs lend out portfolio holdings overnight and earn ~0.02–0.15% per year, which offsets some of the negative drag
• Index licensing fees, transaction costs and FX hedging costs — smaller but real

The "expected" TD for a Reporting Fund S&P 500 tracker is approximately: OCF (0.07%) + WHT drag (0.20%) − securities lending (0.02%) ≈ −0.25% per year. The actual TD varies year to year depending on dividend timing, market volatility and the fund manager's operational performance.

What drives tracking error

Tracking error is about noise, not direction:

• Sampling — the more the fund samples, the noisier the fund-vs-index relationship
• Rebalancing timing — index rebalances on, say, the 3rd Friday of the quarter; the fund might rebalance one or two days off
• Cash drag and inflows — large fund flows can cause short-term tracking noise as the fund equitises the cash
• Currency hedging mismatches for hedged share classes
• Securities lending can add noise on top of its positive drift
• For synthetic ETFs: the swap counterparty's daily rebalancing also feeds in

A well-run physical-replication ETF on a deeply liquid index (e.g. S&P 500 ETF) typically has tracking error of 0.02–0.10%. On thinly-traded indices (small-cap, single-country EM, niche themes) tracking error of 0.30–1.0% is normal.

Real numbers: trailing 3-year TD and TE for popular UK ETFs

Indicative figures based on published fund-vs-benchmark data:

Indicative numbers as of recent annual reports. Exact figures vary by year; always check the current fund factsheet.

What "good" looks like

For a large-cap developed market tracker, good is:

• TD better than −OCF — the fund made up some of its OCF through securities lending or efficient operations
• TD worse than −OCF by less than ~0.10% — the WHT drag is the dominant residual
• TE under 0.10% — the relationship is smooth

For an emerging market or small-cap tracker, good is:

• TD within ~0.30% of −OCF — bigger WHT drag and rebalancing costs are inevitable
• TE under 0.50% — some sampling noise is unavoidable

When each metric matters more

Tracking difference matters more for long-term holders. If you're holding 20+ years, what you care about is cumulative return. The relevant question is: did this fund keep up with its index?

Tracking error matters more for traders and benchmarked institutions. If you're holding for a few months, or you're running a portfolio measured against a benchmark, you care about whether the fund deviated from the benchmark on any given day — not just over the period.

For UK retail investors building long-term portfolios, focus on tracking difference. The lowest-TD fund in your shortlist usually wins, regardless of how the trailing tracking error compares.

How to find these numbers

1. Fund factsheet — usually shows trailing 1, 3, 5 year fund return vs index return. The gap is the tracking difference.
2. Fund annual report — usually quotes tracking error in the "fund performance" section. UCITS rules require this disclosure.
3. justETF.com — an independent ETF database, lets you sort by tracking difference across funds tracking the same index. Useful for comparison.
4. The fund manager's institutional reporting — sometimes more detailed than retail factsheets

Frequently asked questions

Related guides

• Tracking error explained — the metric in depth
• Why your tracker isn't matching the index
• ETF cost comparison calculator
• Synthetic vs physical ETF replication
• ETF metrics glossary