On this page
The short answer#
The TER is what the fund says it charges. The tracking difference is what holding it actually cost you. Only one of those is a measurement.
The tracking difference already contains the TER, and then adds everything the TER leaves out. The fund's own trading costs, dividends sitting in cash, withholding tax that stayed abroad. From that it subtracts anything the fund earned back by lending its shares out. It is a single number that settles the question the fee only gestures at.
The practical consequence is that the cheaper fund on paper is not always the cheaper fund. Two trackers on the same index can rank one way on fee and the other way on what they kept for you. It holds for index funds of every kind, exchange traded and the older mutual sort alike, whether they hold shares or bonds, because the fee and the outcome are simply two different quantities.
You will not find it on the KID. You take it from the provider's own annual figures, over three to five years, and you compare like with like. Same index, same currency, net asset value (NAV) and not market price.
Then hold the whole exercise at its proper size. Five minutes once, before you buy, is the right amount of attention. Switching funds to chase it is not.
What the TER does and does not cover#
The total expense ratio is the fund's stated running cost. It covers the management fees, plus administration, custody, audit and the licence the provider pays to use the index's name. You may see it called the ongoing charges figure instead. On a plain index ETF the two labels describe the same thing.
The mechanism matters more than the label. Nobody bills you and the money never leaves your account. The charge is an annualised figure that accrues daily against the fund's net assets, so the price you see already has it taken out. No invoice arrives, no payment is approved, and no line appears on a statement, which is precisely why a charge is the easiest cost in investing to stop thinking about.
What it leaves out is where the trouble starts. The TER is a defined figure with a defined scope, since the prospectus of the investment company or unit trust behind the fund sets out what may be charged against it. Several genuine costs of owning the fund sit outside that scope by construction.
Transaction costs inside the fund. Every time the index adds or drops a company, the fund trades to match it, paying spreads and commissions. How often that happens is the index's turnover, set by the index rules and not by the fund. A stable large-cap index costs less to follow than one that reshuffles constantly, so smart beta and other rules-based indexes tend to carry more of this cost than a plain market capitalisation one. Those come out of the fund's assets and out of your return, and they are not in the TER.
Cash drag. Dividends arrive on one date and are reinvested on another. In between, that money is cash sitting in a rising market, earning you no return.
Withholding tax on dividends. When a company in the index pays a dividend to the fund, its home country usually withholds tax first. The fund is the shareholder here, not you, so this happens before any return is reported to you. How much comes back depends on treaties, and whatever does not come back is a cost that appears on no document you will read.
The spread you pay to buy. That one is yours, not the fund's, but it belongs in the same mental column. A fund with a wide bid-ask spread costs you real money on the way in and out. It is set by market makers, who can have units created and redeemed against the fund's holdings whenever the two drift apart. So an ETF's spread follows the liquidity of what it holds, not how busy its own screen looks, and the redemption mechanism, invisible as it is, ends up in your brokerage account as a cost or the absence of one.
Tracking difference: the number that includes everything#
Tracking difference is the fund's return relative to the index's return, over a stated period: one subtracted from the other. If the index returned 10.00% over a year and the fund returned 9.82%, the tracking difference is −0.18%.
That single figure captures everything at once. The TER, the trading costs, the cash drag, the dividend tax lost, and anything the fund managed to earn back. It measures what the fund did, and does not estimate what it costs.
One condition applies to the arithmetic. Both halves have to be the index's total return, the version that counts dividends as reinvested. Compared against a price index, which leaves them out, every competent tracker on earth appears to be failing by roughly the dividend yield. Providers publish against the right one. A table you find elsewhere may not.
It can also be positive. A fund that recovers more dividend tax than the index assumes, or earns enough from lending its shares out, can outperform the thing it is tracking over a year. The fund manager has not been clever. The effect is structural, and one of the reasons a headline fee on its own tells you so little.
The fee is what the fund charges you. The tracking difference is what the fund cost you. Only one of those is a promise.
What it is
- TER / ongoing charge
What the fund states it charges to run itself.
- Tracking difference
What holding the fund cost you, measured after the fact.
Where it comes from
- TER / ongoing charge
The provider, in the KID and on every fund page.
- Tracking difference
You, from the provider's annual return table: fund return minus index return.
The management fee
- TER / ongoing charge
Included — it is most of the figure.
- Tracking difference
Included.
Trading costs inside the fund
- TER / ongoing charge
Not included.
- Tracking difference
Included.
Dividends waiting in cash
- TER / ongoing charge
Not included.
- Tracking difference
Included.
Dividend tax that never came back
- TER / ongoing charge
Not included.
- Tracking difference
Included — and this is often the largest of the omissions.
Securities lending revenue
- TER / ongoing charge
Not reflected. The charge cannot go down.
- Tracking difference
Reflected, as a credit. It can pull the gap toward zero or past it.
Can it ever favour you
- TER / ongoing charge
No. It is always a cost.
- Tracking difference
Yes. A fund can finish a year ahead of the index it tracks.
How to read it
- TER / ongoing charge
One current figure.
- Tracking difference
Three to five calendar years, same source, same index, same currency.
It is not the same as tracking error#
These two get used interchangeably and they measure different things. Tracking difference is the size of the gap, meaning how far behind the index the fund finished. Tracking error is the volatility of that gap, meaning how much it fluctuated from period to period.
A fund can be consistently 0.30% behind every single year — a meaningful tracking difference with almost no tracking error. It is reliably expensive. Another can average 0.05% behind while swinging wildly either side of the index, which is a small tracking difference with a large tracking error. It is cheap on average and unpredictable in any given year.
For a long-term buy-and-hold investor, tracking difference is the one that decides what you keep. Tracking error tells you how closely the fund does the job it advertised, which matters more if you are using it as a precise building block or trading around it. If a comparison table gives you only one of the two, check which.
Why a dearer fund sometimes tracks better#
Say it plainly first. Indexing does not remove fund management, it changes the job. There is no judgement to admire here the way there is in active management, because nobody at the asset management firm is deciding what to own. The passively run fund has an operational task instead. Hold the list, deal with the changes cheaply, collect what is collectable. Tracking difference is the mark for that task. It is a tracking number and not a performance one, and managed funds and index funds are not judged on the same thing at all.
Three structural things then move the number around, and none of them appears in the fee.
How the fund replicates the index. Full replication buys every constituent, weighted by market capitalisation, exactly as the index is. Sampling buys a representative subset instead. The large-cap and mid-cap names can be bought in size, while the small cap tail on a very broad index costs more to hold and rebalance than the tracking it buys you, so sampling is cheaper to run and introduces a gap of its own. Synthetic replication holds a basket of collateral and enters a swap with a bank that pays the index return, which on some US indexes avoids dividend withholding tax in a way a physical fund cannot and has historically produced the tightest tracking on those indexes. The cost is a counterparty. You now depend on a bank honouring the swap, collateralised but not free of risk.
Where the fund is domiciled. An Irish-domiciled fund holding US shares recovers more of the US dividend withholding than a fund domiciled somewhere without the same treaty. The difference is deducted inside the fund before any return is reported, so it never appears on a statement, in the TER, or anywhere a comparison table would show it — but it lands squarely in the tracking difference. Why domicile does this has nothing to do with where you live.
Securities lending. Many funds lend their shares out to short sellers and other institutional investors for a fee, and pass some of that revenue back into the fund. It can offset a meaningful part of the running cost. It also introduces a real, if usually small, risk. The borrower may fail, and you are relying on collateral. A provider that lends aggressively and keeps a large cut of the revenue is making a different trade on your behalf than one that does not lend at all. Providers do not all decide this the same way, and their prospectuses say so. That document, and the annual report, is where an investment fund has to state the policy in full.
What that looks like in money#
“A dearer fund can win” stays a debating point until it has numbers on it, so here is a pair. Two funds, same index. The first charges 0,20% a year and has finished 0,12% behind the index over the last five years. The second charges 0,07% and has finished 0,15% behind. On the headline fee the second is nearly three times cheaper. On what it actually cost to hold, the first is ahead.
Put that through a real plan — €300 a month, 7% before costs, thirty years — and the fund with the higher fee leaves you about €1.900 better off, on a balance of roughly €343.000.
Hold that result at its proper size, because it cuts both ways. €1.900 over thirty years is not a windfall, and no argument for auditing your funds every quarter. What it does argue for is checking the number once, before you buy, instead of ranking two funds on a figure that leaves out most of what they cost.
Where the gap stops being small#
Everything above assumes an equity fund on a mainstream index, with a deep stock market underneath it, where the whole argument plays out in hundredths of a percent. It does not stay that small everywhere.
The asset class decides how large the gap can get, through one variable. How easily the thing underneath can be traded. In fixed income the range is wide. A bond ETF holding government debt, meaning Treasury issues and their European equivalents, can track closely, because that part of the bond market changes hands constantly and in size. A high yield fund has no such luxury. A good many of the bonds in its index trade rarely, so the fund samples instead of replicating, and both the sampling and the cost of dealing in an illiquid market land in the tracking difference. Small cap and emerging-market equity funds sit in the same position for the same reason.
Two kinds of fund produce a gap that is not a cost at all, and confusing them with one leads you badly astray. A commodity fund usually holds futures instead of the physical thing, and rolling those contracts forward adds or subtracts a return the spot price never had. Leveraged and inverse products reset their exposure daily, so over any period longer than a day their result is not the index's return multiplied by the headline number. In both cases the gap is how the product works, not a measure of how well it is run.
None of that argues against holding any of them. It argues for checking the number where it is large and not only where it is small, and for leaving a rule of thumb formed on a world tracker out of a corner of the market that does not behave like one.
What it charges
Ongoing charge (TER)
Fund B looks almost three times cheaper.
What it cost
Tracking difference, five-year average
Fund A was the cheaper one to hold.
The order reverses. On the number every comparison site prints, Fund B wins by a distance. On what holding it actually cost, it loses. Both panels are drawn to the same scale.
Where to actually find the number#
This is the part most explanations skip. Tracking difference is not on the KID and usually not on the fund's summary page.
Where it does live is the issuer's own factsheet and annual report, which print the fund's return and the index's return side by side for each of the last several calendar years. Subtract one from the other and you have the tracking difference for each year, from the source, with nobody's methodology in between. Independent data sites compile these tables for you, with Morningstar and the ETF screeners the usual starting points. That is faster, and you should cross-check it against the provider once.
Three traps when you read them. Take the calendar years and not the since-inception figure, because inception dates differ between two funds, so that number compares different stretches of market history and quietly rewards whichever fund launched at the better moment. Use the return calculated on net asset value, not on market price. Market price includes whatever the spread happened to be on the closing day and tells you about the exchange, not the fund. And check the currency. A fund reporting in dollars against an index quoted in euros will produce a gap that is entirely exchange-rate movement and has little to do with cost. A currency-hedged share class is a third case again. Measured against the unhedged index it will show a gap that is the cost of running the hedge, not a failure to track.
Questions this raises the first time#
Four that come up whenever someone goes looking for this number for the first time.
“Is the OCF the same thing as the TER?” On a plain index ETF, treat them as the same figure under two labels. They come from different rulebooks and can diverge on more complicated products, but on a mainstream tracker the distinction is not where your money is going.
“The tracking difference is negative. Is that bad?” It is the ordinary case, and the sign is only a convention. A tracking difference of −0,15% means the fund finished 0,15 percentage points behind its index, which is what you would expect from a fund that has costs. A positive figure means it finished ahead. Some sources flip the sign and quote the same thing as a positive cost, so check which convention a table is using before you compare across two of them.
“What counts as a good tracking difference?” There is no published threshold, and this site will not invent one. A rule of thumb here would be exactly the kind of unsourced number that goes stale without telling anyone. Compare the funds you are actually choosing between, on the same index, over the same years. That comparison answers the question you have. A benchmark figure would only answer a question you do not.
“My broker's fund page does not show it.” It usually will not, and neither will the KID. It lives in the provider's own factsheet and annual report, which print the fund's return and the index's return side by side. Independent sites compile those tables, which is faster. Cross-check one against the provider once, then trust it.
How to use it without fooling yourself#
One year of tracking difference is noise. An index reconstitution that happened to fall well, or a single dividend timing quirk, moves a year enough to reverse a ranking. Look at three to five, from the same source, and look at whether the number is steady, not only at its average.
Check you are comparing the same index. MSCI World and FTSE Developed are not interchangeable, neither is the same as MSCI World ex-USA, and none of them is the Nasdaq-100. Two funds with different tracking differences against different indexes are not being compared at all, and a ticker will not tell you, since the same fund carries different ones on different exchanges. Investment strategies are chosen at the index level. Tracking difference only says who implements one more cheaply.
Then sanity-check the fund's size — its assets under management, printed on every factsheet. A very small fund with excellent tracking is still a fund that can be merged away or closed, and that forces a sale at a time you did not choose. That is a taxable disposal, with whatever capital gains consequence it carries where you live, unless the holding sits inside one of the tax-advantaged retirement accounts your country offers, where that half of the problem falls away and the cost half does not.
Finally, hold the whole exercise at its proper size. Tracking difference is more persistent than performance, because its drivers are structural instead of lucky. But past tracking is still not a promise, and the gap between two mainstream trackers on the same index is usually measured in hundredths of a percent. Five minutes and one comparison is the right amount of attention. Switching funds every year is not, and realising a disposal to chase it certainly is not.
It is also the smaller half of the decision. Which index you decided to invest in, the diversification that buys you, and the allocation you hold it at will move your investment portfolio by more than the gap between two funds tracking the same thing ever will. Tracking difference settles which fund. It says nothing about the investment strategy you are running, and remembering which of the two you are actually choosing is the useful thing to do when you find yourself an hour into a spreadsheet.
