Tracking Error — The Only Number That Judges a Passive Scheme

They get used interchangeably and they answer different questions.

Tracking difference is the gap between what the scheme did and what the index did over a period. It is usually negative, because costs come out of the scheme and not out of the index. This is the one that affects your money directly.

Tracking error measures how much that gap wobbled from period to period. A scheme can lag the index consistently, which shows up as a small tracking error and a steady difference, or lag erratically, which shows a larger error.

For most investors the difference matters more, because it is the amount actually given up. The error matters because a scheme that wanders unpredictably is not doing its job reliably, whatever the average looks like.

Several reasons, and the first accounts for most of it.

The expense ratio, deducted from the scheme every year while the index has no costs at all. This is the largest and most predictable component, which our page on the expense ratio covers.

Cash sitting in the scheme. Money arriving from investors is not invested the same instant, and a small uninvested balance drags in a rising market.

Trading costs when the index changes its constituents and the scheme has to buy and sell to follow.

Redemptions, which force selling at times the index does not experience.

We will not publish a threshold, because it varies by what is being tracked and quoting one would be pretending to a precision that does not exist.

What is meaningful is the comparison. Take two or three schemes tracking the same index, and compare their tracking difference over the same period. That is a fair comparison, because they were attempting an identical task under identical conditions.

Comparing a scheme tracking one index with a scheme tracking another is not that comparison, and our page on the benchmark explains why the reference has to match before any number means anything.

Not all indices are equally easy to follow, and it is worth knowing which are harder.

An index of large, heavily traded companies is straightforward to replicate, because the scheme can buy and sell what it needs without difficulty. Gaps there are small and mostly explained by cost.

Indices of smaller companies are harder, since building and adjusting positions takes longer and moves prices, which our page on fund size and AUM discusses in the active context. The same pressure applies to a tracker.

Rules-based indices that rebalance frequently are harder again, because every rebalance means trading, and our page on factor and smart beta funds covers that family. Overseas exposures add their own complications on top.

For an exchange traded fund there is a second gap that has nothing to do with the manager.

An ETF trades on the exchange, so the price you pay can differ from the value of what the fund holds at that moment. In a thinly traded ETF that gap can be meaningful, and it lands on you rather than on the scheme.

So judging an ETF means looking at tracking difference and at how actively the units themselves trade. Our page on ETF versus index fund covers the choice, and for many ordinary investors the index fund version avoids this second problem entirely.

Tracking error becomes the deciding number in one situation above all others: choosing how to hold the largest companies in the country.

In that segment the case for a manager is narrowest, so a good many investors end up comparing an active scheme with a tracker, and then comparing two trackers against each other. Our page on large cap fund versus index fund works through the first of those.

Once you have decided on the passive route, the second comparison is entirely this page. Two schemes following the same widely held index are doing an identical job, and the only things separating them are what they charged and how closely they followed. Past return between them is noise.

That is unusually clean as investment decisions go. Very few choices in this business come down to two knowable numbers, and it is worth using that clarity when it is available.

Four steps, and it takes about ten minutes.

• Confirm the schemes track the same index. Similar names are not enough.
• Compare tracking difference over the same period, from the scheme documents or fact sheets.
• Compare the expense ratio, since it explains most of the gap and is knowable in advance.
• Check the scheme size, since a very small tracker can find replication harder.

One caution on the numbers themselves. Tracking measures are reported over a stated period, and a scheme can look tidy over one window and less so over another, particularly if the index changed its constituents during it. Look at more than one period before concluding anything, and prefer the longer one where both are available.

What we would not do is choose between two trackers on which showed a marginally higher return last year, since over the same index that difference is mostly noise. Our page on index funds covers the category and active versus passive covers the wider argument. If you want help comparing what you hold, get in touch.