There are two common measures used to evaluate the performance of an ETF – "tracking error" and "tracking difference". While both are useful indicators of deviation from benchmark performance, they measure different characteristics. Understanding both performance metrics is therefore very important.
Tracking difference is simply the return difference between the ETF and its underlying index over a given time period.
Tracking error is the volatility (as measured by standard deviation) of that return difference.
So while tracking difference tells us how much the performance of an ETF can deviate from its underlying index over a given time period, tracking error shows the consistency of returns relative to the index throughout this entire time period.
Generally, for long-term investors, tracking difference is their primary concern, whereas for short-term investors, tracking error is much more significant, since volatility relative to the benchmark may have significant impact on returns over shorter time periods.
Looking at both tracking difference and tracking error in tandem will give a fuller picture of how well an ETF tracks its index over time.
Below is an example to further illustrate the difference of tracking difference and tracking error:
If one assumes that the blue line is the index, which ETF has the higher tracking error?
Investors might be inclined to assume that the green line (ETF 2) has the higher tracking error, since it seems to track far away from the index. However, the opposite is actually true – the orange line (ETF 1) has significant tracking error, but almost zero tracking difference. ETF 2 on the other hand has almost zero tracking error, but it "consistently" underperforms the benchmark every month. Although this is an exaggerated example, it demonstrates that the distinction between tracking difference and tracking error is an important one.