What is the tracking difference – and what is the tracking error?
The key figure indicates the extent to which the performance of an ETF deviates from the respective reference index in a period under consideration. This is particularly important for ETFs, because they have declared that they want to replicate an index as precisely as possible. The tracking difference shows whether and to what extent the ETF is achieving this goal, even overachieving it or staying below its target. The term tracking error is often used alongside the tracking difference. Sometimes it is used synonymously for tracking difference, sometimes it refers to the measured standard deviation. The tracking difference then indicates the (average) deviation in the performance of the fund and the index, while the tracking error describes their fluctuation over time. That is a subtle difference.
HOW IS THE TRACKING DIFFERENCE CALCULATED?
The simple formula is: Tracking Difference = benchmark index performance – ETF performance
EXAMPLE : On an annual basis, the return is the yardstick for calculating the performance of funds. If an ETF achieves an annual return of 7.5% and the relevant reference index achieves a performance of 8%, the tracking difference is 0.5%. If the ETF develops exactly like the reference index, the tracking difference is 0.
GOOD TO KNOW: Sometimes the order is reversed when calculating: Tracking Difference = ETF performance – Reference index performance. This is then associated with a change in sign. It is advisable to always check the underlying definition when specifying key figures in order to avoid misinterpretations.
How can the tracking difference be explained?
Why can there be deviations in performance at all when the ETF portfolio is a reflection of the respective reference index? Shouldn’t one always expect a tracking difference of 0?
The answer is: yes – in a world without transaction costs, without fees, with immediate reinvestment of all returns at any time and immediate follow-up of index adjustments. The reality is a little different. ETF fees are a major cause of tracking difference. Even if ETFs are very cheap – the annual management fee has a negative impact on ETF performance compared to index performance.
In addition, other influencing factors can influence the replication accuracy:
Transaction costs: Necessary portfolio shifts due to changes in the index cause additional transaction costs.
Dividends: Can cause deviations due to delays in reinvestment, sometimes also due to different tax considerations in the index than in the fund assets.
Liquidity: With the provision of liquidity, the liquid assets generate no or only very little income, but there are also no losses.
Securities lending: Additional income can be achieved by lending securities from the fund’s assets. Securities lending improves ETF returns compared to index performance.
What is the significance of the tracking difference?
The tracking difference is not static, even if the ETF fees as the main influencing factor are usually a strong constant. But other factors that are also relevant change over time. Transaction costs depend on the frequency and extent of reallocations, dividends vary, the conditions for securities lending are exposed to market influences, etc. The variability of the tracking difference is measured – as mentioned above – via the tracking error.
In principle, the tracking difference is always a so-called ex-post consideration. That means: the deviation can only be determined afterwards. An estimate for the future can be derived from this, but it is fraught with uncertainty. It is greater the more variable factors influence the tracking difference.
Positive / negative tracking difference – what does that mean?
If we start from the definition: Tracking Difference = performance of reference index – performance of ETF, then a value with a positive sign always means that the performance of the ETF has remained below the index performance by the relevant percentage. The higher the amount, the greater the deviation. A negative sign indicates that the ETF has outperformed its benchmark index. This does happen and is possible if the yields outside of the pure index replication more than offset the costs incurred. Another reason may be deviations in the index replication, which lead to the index fund “outperforming” its reference index. These statements also apply if the tracking difference is defined “the other way round”, then again with the opposite sign.
ETF construction and tracking difference
The type of ETF construction can also have an impact on the tracking difference. As a rule, the index replication of synthetically replicated ETFs is particularly precise. The tracking difference is then largely cost-related and relatively stable.
In the case of physically replicated ETFs, on the other hand, there may be more deviations. These tend to be higher for distributing funds than for accumulating funds. The reason: in the case of distributing funds, larger amounts must be held as liquidity for distribution purposes.
The tracking difference is built in almost automatically when sampling index funds. Sampling is a variation on physical replication. An index is not reproduced 1: 1, but – mostly for cost reasons – you limit yourself to a representative selection. Of course, this increases the likelihood of deviations – up or down.