The tracking difference measures the actual costs of exchange-traded index funds (ETF). For a meaningful ETF comparison, this key figure must be calculated accurately. However, this is not possible on the basis of the yield data from the legally required information sheets. An important key figure for the performance comparison of ETFs on the same index is the so-called tracking difference. This is the difference in return between the ETF and the index over a certain period of time (ETF return minus index return = tracking difference, can of course also be calculated the other way around).
The tracking difference ultimately contains the same information as the return. It just shows the performance of an ETF from the cost side. The smaller the tracking difference, the lower the actual total costs. Moreover the lower the cost, the better the ETF. The ETF with the lowest tracking difference over a certain period of time is always the ETF with the highest yield.
Unlike the tracking difference, the administration costs (Total Expense Ratio, TER) reported by ETF providers only indicate the fixed portion of the total costs of an ETF. The TER is therefore not meaningful and unsuitable for the selection of ETFs . In contrast, the tracking difference also includes the variable costs, which fluctuate over time.
The historical tracking difference is a good indicator of future returns
ETFs with the lowest tracking difference or the highest return over at least four years were among the best in their comparison group in six out of eight cases even after the following four years. According to the analysis, the long-term return and tracking difference are the only two metrics that provide useful information about the future performance of an ETF.
When comparing ETFs, some investors rely on the return data from the so-called key investor information, which some fund companies also call the Key Investor Information Document (KIID). The tracking differences can also be determined from the information from the ETF providers contained therein. For example, the ETF comparison website Trackingdifferences.com uses the KIIDs as a data source. The information sheets are required by law. Providers have to create them for each fund. In the case of ETFs, the fund companies indicate a maximum of the annual returns for the past five years compared to the index. Distributed dividends are included using a uniform method as if they had never been paid out.
How accurate is the tracking difference from the KIIDs?
Things are a little different when investors calculate the tracking difference from the KIID data. In contrast to the performance, the missing trading day does not play a decisive role in the tracking difference. Nevertheless, the tracking differences obtained from KIID data are not valid, suggests a fair value analysis. We have calculated the returns for a total of 26 ETFs on four different indices.
Tracking difference is also not meaningful because of different comparison indices
Problems when comparing ETFs with the help of the tracking difference also cause different index variants which the providers use for comparison. So-called total return net indices are common. These comparative indices assume that the dividends distributed are reinvested after deduction of withholding taxes, as is the case with the calculation of the returns for distributing ETFs in the KIIDs. Now, some providers do not use the net variants as comparison indices, but the so-called price indices. They show the performance of the stocks in the index without dividends. The return of price indices is therefore lower than that of net indices. However, a stock ETF maps the dividends. If it compares itself with a price index, it of course looks as if the ETF has easily outperformed the index. In this way, for example, the provider Deka puts sand in the eyes of its customers. Tracking differences calculated on the basis of different index variants are of course not comparable. The providers who use price indices will always lead the ranking.
The comparison period must be the same for all ETFs
The tracking difference over different time periods is also not comparable. Nevertheless, such questionable comparisons are made on the Trackingdifferences.com website and also on ExtraETF . This leads to rankings that are not meaningful. Returns, tracking differences and all other key figures must always be compared over the same period. Otherwise, the respective tracking differences came about under different framework conditions and the resulting rankings were purely random.