When selecting an ETF, most investors ignore tracking error, or downplay its importance. But does it matter? Here are my thoughts.
Tracking errors and how they arise
Before we talk go further into the analysis, it is important to understand how ETFs work.
When an ETF tracks an index, the fund manager deploys strategies to track the target index in real time – by buying and selling securities or their derivatives. Trades must be executed among hundreds and thousands of securities in proportion to their weight in the index when the market capitalisation of the index changes.
Managing these complex processes, implementation friction, liquidity of the underlying, incurring operating expenses, cash drag and trading fees, all come with costs and creates a drag on returns for the fund.
The difference between the fund returns and the target index returns is known as the tracking difference. It is usually expressed as a percentage. The standard deviation (i.e. variability) of those differences is known as the tracking error.
Is it really bad?
Though the word error sounds like it might be a bad thing, it’s actually quite normal to have tracking errors in some cases. As managers of actively managed ETFs deploy certain strategies to outperform the index or benchmark, they tend to increase the tracking error (perhaps favourably).
For passive ETFs, when the sole purpose of the ETF is to track the index, tracking errors tends to be low. That’s because market participants expect the ETF to deliver market returns as the benchmark.
Ultimately… it depends on the investment objectives
As the tracking error measures the deviation of the fund returns from the benchmark, for actively managed portfolios, you’d want a consistently high tracking error for a fund that delivers outperformance from the target index.
However, if you’re a passive portfolio relying on broad market ETFs in your portfolio, then you’d want to own them as cheap as possible, with the lowest tracking error from the benchmark they are tracking.