0.05% cost sounds better than 0.20%. For an investment volume of 10,000 euros, that’s just a 15 euro difference per year. Nevertheless, exactly this number is often enough to drive investors into new comparison tables, forum discussions and thoughts about changing ETFs. The costs of investing should not be underestimated and anything but irrelevant. But many investors look at the most obvious metrics when it comes to ETFs and confuse them with the most important one. The TER is large in the product information sheet, is easy to sort and looks objective.
However, it does not tell the whole truth, because what matters is not just how much an ETF costs. What is also crucial is how much its actual return deviates from the respective index, and this is where the Tracking Difference comes into play, which we would like to examine in this article.
The TER is just the price tag
The TER, i.e. the total expense ratio, shows the running costs of a fund per year and is usually taken directly from the fund assets once a year. An ETF with a 0.20% TER charges the fund assets with 20 euros per 10,000 euros of investment volume per year. At 0.05% it is only 5 euros. This is transparent and easy to understand, but is largely just a price tag of an ETF and does not show the actual result in returns.
An ETF may seem like a no-brainer for the owner, but in the background, securities must be constantly bought and sold, dividends collected and taxes taken into account. Changes to the index must also be replicated and the liquidity of thousands of investors must be managed efficiently. All of this can cost returns, and some of it can even bring returns back.
In practice, this means: An ETF with a low TER can replicate its index worse than an ETF with a slightly higher TER. Conversely, a supposedly more expensive ETF may in practice be closer to the index. That is why it is too short-sighted to sort ETFs only by the cheapest expense ratio.
What the Tracking Difference measures
The Tracking Difference measures the difference in returns between an ETF and its benchmark index over a certain period of time. So it shows what really came out after costs, taxes, implementation and additional income. A simple example:
The index increases by 10% in a year. The ETF is up 9.80%. Then the ETF missed its index by 0.20 percentage points. However, if the index increases by 10% and the ETF increases by 10.05%, the ETF has actually outperformed the index by 0.05 percentage points. The pure return is more important for investors than the mere TER. The Tracking Difference not only shows what was on the price tag, but also what actually arrived as a result.
Examples from practice
A few numbers make the difference more tangible, but are not a product recommendation, but rather simple examples to illustrate the mechanics. The examples are from trackingdifferences.com, where a variety of products can be compared accordingly. Among other things, the Xtrackers MSCI World UCITS ETF 1C is listed there with a TER of 0.19%. The average annual Tracking Difference since 2015 has been minus 0.05% per year. This means: On average, the ETF was not only cheaper than the TER suggests, but was even slightly ahead of the index. The iShares Core MSCI World is also listed there with a TER of 0.20%. The average Tracking Difference is -0.08% per year. A Tracking Difference of 0.0% is stated for 2024. This also shows that the visible TER does not automatically explain what happens in the end compared to the index.
An ETF with a 0.19 or 0.20 percent TER is not automatically worse than the index by exactly this value. The real deviation can be higher, lower or even positive for the investor. Implementation is crucial.
The sign trap
There is a practical commonality with the Tracking Difference: not every source calculates the same way. Some portals calculate it by taking the index return minus the ETF return, which results in a positive value if the ETF performed worse than the actual index. A negative value, on the other hand, means that the ETF performed better. Other sources calculate the ETF return minus the index return, as we did above. Then the meaning of the sign is reversed. A positive value shows a better ETF return, a negative value shows a worse one. If you only look at plus or minus, you can misunderstand the key figure.
Why ETFs differ from the index
An index is first and foremost a theoretical construct, without a custodian or trading costs that have an impact on actual performance. The pure index also does not have to control the liquidity of its buyers or has operational costs. An ETF, on the other hand, is the actual and genuine product that changes this theoretical construct as a basis. For this reason alone, an index can never be exactly matched by an ETF. Inevitably they always diverge somewhat.
The main reasons for deviations are ongoing costs, transaction costs, taxes on dividends, rebalancing, replication method, securities lending, spreads and the question of which index variant is used as a comparison. This is particularly visible in broad world indices. An index with more than a thousand titles can theoretically be completely recreated. In practice, some providers use optimized sampling and then do not buy every single stock with exactly the same weighting, but rather a selection that should reflect the index as closely as possible. This saves costs and effort, but can create deviations.
Why an ETF can beat the index
Many investors expect that an ETF must always perform slightly worse than its index because of the costs. This is obvious, but not always correct. An ETF can outperform its index under certain circumstances. The most important parameters here are withholding taxes on dividends. Many international indices rely on certain tax assumptions. Depending on the fund domicile, double taxation agreement and structure, a real ETF can sometimes perform more favorably than the index assumption. Irish fund domiciles are discussed particularly frequently in Europe because Ireland can be tax advantageous for many fund structures when it comes to US dividends.
There is also securities lending. An ETF can lend shares held for a fee. Part of this income flows into the fund and can partially offset the ongoing costs. This is not a free gift, as securities lending involves counterparty risks. Reputable providers secure such transactions, but it remains a component that investors should understand.
The replication method can also help. A synthetic ETF does not replicate the index by directly purchasing all shares, but rather via an exchange transaction, a so-called swap. This can bring tax or operational advantages for certain indices. But it also brings with it a different risk structure.
In short: the TER deducts returns. Other effects can cost or recover returns. The Tracking Difference shows the sum of these deviations.
Tracking Difference is not Tracking Error
Tracking Difference and Tracking Error are often confused. However, they measure different things. The Tracking Difference is the result calculation and shows how far the ETF and index differed over a period of time. The Tracking Error, on the other hand, measures the fluctuation of this deviation. It shows how calmly or restlessly an ETF follows its index.
An ETF can have a small Tracking Difference, but can fluctuate more around the index at times. Another may follow very calmly, but end up permanently losing some return. For long-term investors, the Tracking Difference is usually more tangible because it shows the result. The Tracking Error is also useful when it comes to evaluating the stability of the index mapping.
A simple reminder is enough: Tracking Difference shows what was missing or added at the end. Tracking Error shows how much the ETF deviated from the index along the way.
A year is not enough
In order to be truly meaningful, the longest possible observation period should be used. Perhaps the sampling worked particularly well in one year or individual dividend dates were particularly favorable for the calculation. But perhaps withholding tax effects also helped or the rebalancing was unusually complex and therefore much more expensive than usual. That’s why a single year is not enough. It makes more sense to look over several years. Three to five years give a more meaningful picture; comparing several ETFs on the same index is even better.
If an ETF consistently meets or even slightly outperforms its index over many years, that’s a good sign. If it consistently performs worse than comparable products despite a low TER, the low cost ratio should not be blinding.
Where investors find the Tracking Difference
The Tracking Difference is rarely as prominent in the fact sheet as the TER. Nevertheless, it is easy to find. Practical points of contact are ETF portals such as extraETF and justETF or specialized sites like trackingdifferences.com. It is often easier to compare ETFs based on the same index there than on the providers’ advertising websites. The providers’ annual reports and documents are more official but more complex to research. There you can find return data from funds and benchmark indexes. If you want to check it exactly, you have to pay attention to which index variant is used and which calculation logic is behind the deviation.
For investors, the bottom line is that the Tracking Difference is the most important key figure, because in the end it is not the best expense ratio in the fact sheet that counts, but rather the actual return after deducting all costs, taxes and transaction fees. Anyone comparing ETFs should therefore not simply sort by the lowest TER, but always look at several key figures and above all make the comparison period as long as possible.
Most importantly, an ETF is always better than no investment. The most detailed research should never fail because of simple implementation!
