If you have ever picked an ETF by sorting a list by TER and buying the cheapest one, you have done what I did for my first few years of investing. It feels rigorous. It is also subtly wrong, because the TER is not what an index fund actually costs you. Your real cost is the tracking difference, which routinely differs from the TER by amounts that matter over a 20-year holding period.
The confusion deepens because a second, similar-sounding metric exists, tracking error, and half the articles online use the two terms interchangeably. They are not the same thing. Tracking difference tells you how much return you lost (or occasionally gained) versus the index. Tracking error tells you how consistently the fund hugs the index day to day. One is a cost. The other is a quality-control measure.
Here I want to nail down both definitions, show why a 0.20% TER fund can genuinely be cheaper than one charging 0.12%, and share the process I use to check these numbers before buying.
What tracking difference actually measures
Tracking difference is brutally simple: it is the fund’s total return minus the index’s total return over a given period, usually one calendar year.
If the MSCI World index returned 12.0% in a year and your MSCI World ETF returned 11.7%, the tracking difference is -0.3%. That 0.3% is what the fund really cost you that year, all in: management fees, transaction costs inside the fund, withholding-tax leakage on dividends, sampling decisions, and any income the fund earned on the side.
This is the key mental shift. The TER is one ingredient; tracking difference is the finished dish. A fund can charge 0.20% and still land within 0.10% of its index by clawing back cost through efficient dividend-tax treatment and securities-lending revenue. Another charging 0.12% can lag by 0.35% because it sits in a domicile with worse tax treaties or samples the index poorly. I compared exactly this dynamic in my Invesco MSCI World vs iShares Core breakdown, and the results surprised me.
One more nuance: tracking difference can be positive. Some synthetic S&P 500 UCITS ETFs have beaten their net-return index in certain years because of how swap structures handle US dividend withholding.
Tracking error is a different number entirely
Tracking error is the volatility of the gap between fund and index, usually the annualised standard deviation of daily or weekly return differences.
Read that again, because it is where most people trip. Tracking error says nothing about whether you lost money. It says how wobbly the fund’s path around the index was. A fund could lag its index by exactly 0.25% every year like clockwork: meaningful tracking difference, near-zero tracking error. Another could beat the index by 1% one year and lag by 1% the next: zero average tracking difference, large tracking error.
Why the two get confused
Partly sloppy writing, and partly the fact that fund documents lean on tracking error, because UCITS rules require funds to disclose their anticipated tracking error and explain divergences in the annual report. So tracking error is the number regulators talk about, while tracking difference is the number your wallet feels. ESMA’s guidelines for UCITS ETFs set those disclosure requirements; the framework is on the ESMA website if you enjoy that sort of reading.
For a buy-and-hold investor, tracking difference matters far more. Tracking error matters mainly if you trade frequently or need predictable short-window behaviour. I dug into a real-world example in my SXR8 vs CSPX tracking comparison, where two excellent S&P 500 funds show slightly different personalities despite near-identical long-run results.
Tracking difference vs tracking error at a glance
| Feature | Tracking difference | Tracking error |
|---|---|---|
| What it measures | Fund return minus index return over a period | Volatility of the fund-index gap (standard deviation) |
| What it tells you | The real all-in cost of owning the fund | How consistently the fund follows its index |
| Typical value (large index ETF) | Roughly -0.4% to +0.2% per year | Often under 0.5% annualised for physical funds |
| Can it be “good”? | Yes, a positive TD means the fund beat its index | Lower is better; there is no upside to wobble |
| Who should care most | Every long-term buy-and-hold investor | Short-horizon traders, hedgers, institutions |
| Where to find it | Comparison sites, annual reports, factsheet return tables | KID, annual report, provider fund page |
Why a 0.20% TER fund can beat a 0.12% TER fund
Three forces drive the gap between TER and tracking difference, and each can be worth more than the TER itself.
Dividend withholding tax leakage
When a fund receives dividends from foreign companies, the source country withholds tax before the money arrives, and how much depends on the treaty between the fund’s domicile and that country. Irish-domiciled UCITS funds pay roughly 15% withholding on US dividends thanks to the US-Ireland treaty, while a Luxembourg fund tracking the same index may lose closer to 30%. On an index yielding around 2%, that treaty gap alone is worth roughly 0.3% per year, which dwarfs most TER differences. It is the single biggest reason Irish domicile dominates US-heavy equity indices, and I cover the mechanics in my ETF tax efficiency guide.
Securities lending income
Many physical ETFs lend a slice of their holdings to short sellers for a fee, and most of that revenue flows back into the fund. For a large, liquid fund this can recover several basis points a year, quietly offsetting part of the TER. Providers disclose lending revenue and collateral policy in the annual report.
Replication method and sampling
A fund that fully replicates a 1,500-stock index carries real trading costs. A fund that samples holds a subset and accepts some drift, which can help or hurt in any given year. Synthetic funds sidestep both via a swap, often achieving superb tracking on US indices, in exchange for counterparty complexity some investors reasonably prefer to avoid.
Stack those three forces together and the TER looks like what it is: a list price, not the till receipt.
How to actually check tracking difference before you buy
Here is my process, which takes about ten minutes per shortlist.
First, pull the fund’s own return table. Every provider factsheet shows fund performance next to index performance by calendar year. Subtract one from the other for the past five years; that row of small negative (or positive) numbers is the fund’s tracking-difference history. Note that funds track the net-return version of an index, which already assumes some withholding tax; MSCI publishes its methodology and net-return conventions on the MSCI website.
Learning ETFs the slow way?
The free 12-page Starter Kit gets you to a first, sensible portfolio faster: the 7-filter fund shortcut, the 3-question broker matrix, and the 7 costliest beginner mistakes.
Second, use a comparison site that computes this for you. justETF shows fund-versus-index charts, and trackingdifferences.com exists purely to rank UCITS funds by multi-year tracking difference. I treat any single year as noise and use the three-to-five-year average.
Third, compare rivals tracking the same index. This is where the exercise pays off, because funds with identical benchmarks and near-identical TERs can sit 0.1-0.2% apart on realised tracking difference year after year. My UCITS ETF comparison tool lets you line up candidates side by side, and I keep a running ranking in the cheapest MSCI World ETF guide, where “cheapest” means cheapest by tracking difference, not by TER.
A caveat on precision: all figures above are approximate and move every year, so verify current numbers on the provider’s official factsheet before committing money.
How much tracking error should worry you
For big mainstream index funds, honestly, not much. Physical funds on liquid developed-market indices typically show annualised tracking error well under half a percent, and as a monthly buyer I cannot feel that in practice. It deserves attention at the edges: small funds, small-cap or frontier indices, ESG-screened indices with sampling, and anything thinly traded. A tracking error drifting upwards year over year is telling you something, and the annual report has to explain divergences beyond the anticipated level. If the explanation reads like an excuse, I move on; there is almost always a close substitute.
My verdict: how I’d use both numbers
If you hold for decades, rank your shortlist by multi-year tracking difference and treat TER as a tiebreaker, not a decision-maker. My working rule: take the funds tracking your chosen index, discard anything small or new, then pick from the two or three with the best three-to-five-year average tracking difference. If they sit within a few basis points of each other, choose on practical grounds: your broker’s fees, fund size, accumulating versus distributing.
If you trade or rebalance frequently, glance at tracking error too, and prefer the steadier fund when returns are otherwise equal. And if a fund’s TER looks temptingly low but there is no tracking-difference history because it launched last year, I simply wait. A low list price is a marketing decision. A low tracking difference has to be earned, one year at a time.
Frequently asked questions
Q: What is the difference between tracking difference and tracking error?
A: Tracking difference is the fund’s return minus the index’s return over a period, so it measures your real cost of owning the ETF. Tracking error is the standard deviation of that gap, so it measures how consistently the fund follows the index. A fund can have a meaningful tracking difference with almost no tracking error, and vice versa.
Q: Is tracking difference more important than TER?
A: For long-term investors, yes. TER is one component of cost, while tracking difference captures everything: fees, internal trading costs, withholding-tax drag, and lending income. Funds with the same TER can differ by 0.1-0.2% per year in realised tracking difference, which compounds into real money over decades.
Q: Can an ETF have a positive tracking difference?
A: Yes. If securities-lending revenue plus favourable withholding-tax treatment outweigh the fund’s costs, it can beat its net-return index. Some synthetic S&P 500 UCITS ETFs have done this in certain years because swap structures handle US dividend withholding efficiently.
Q: Where can I check an ETF’s tracking difference?
A: Subtract index return from fund return in the calendar-year performance table on the provider’s factsheet, or use comparison sites such as justETF or trackingdifferences.com. Look at three to five years rather than a single year, because annual figures are noisy.
Q: Why do Irish-domiciled ETFs track US indices better?
A: The US-Ireland tax treaty cuts withholding on US dividends to roughly 15% for Irish funds, while some other domiciles lose up to 30%. On an index yielding around 2%, that advantage is worth roughly 0.3% per year, often more than the entire TER of a cheap index fund.
Q: What is an acceptable tracking error for a UCITS ETF?
A: Large physical ETFs on liquid developed-market indices commonly show annualised tracking error well below 0.5%, which is fine for buy-and-hold investors. Higher figures are normal for sampled, small-cap, or emerging-market funds. A rising tracking error, or one the annual report cannot explain convincingly, is a reason to pick a rival fund.
This guide is part of our hub on sector and thematic UCITS ETFs — browse the full cluster for related situations.
European UCITS/ETF investor with 11+ years building a global dividend and index portfolio, and founder of Financial Expert Class. I write every guide here from first-hand experience investing across EU brokers and tax regimes — not theory. More about me →