Measure how far an ETF's return fell short of its index, or beat it.
Farhan has just finished Lesson 3.1, "Expense ratio and the costs it leaves out", and is choosing between two ETFs on the same all-world index. Fund A charges a TER of 0.15% and Fund B charges 0.22% (invented examples). Farhan takes it for granted that the lower fee means the better result. Then both annual reports show that over five years, Fund B stayed nearer the index.
The fee table Farhan relied on was accurate, but it didn't show everything. A TER tells you something true about what a fund costs, and it leaves part of the cost out. With an index fund, what you need to know is how much of the index's return ended up with you.
Take the fund's return for a period and subtract the index's return for the same period. The result is the tracking difference, and it is normally measured per calendar year. Say the index made 10.3% and the fund made 10.1% (example figures). The tracking difference is minus 0.2 percentage points. That is the amount you gave up by owning the fund compared with owning the index directly at zero cost.
In most years the figure is negative, because the fund carries costs that an index never pays. Now and then it comes out positive, which means the fund did better than its own index that year. A positive figure is real, so don't treat it as a mistake in the data.
Make sure both numbers measure the same thing. An index provider usually publishes more than one version of an index's return. In the net return version, a standard rate of withholding tax on dividends has been taken off, and that is the version most funds measure themselves against. Use whichever version the fund's own documents name.
The TER pulls the fund's return below the index. Several other things also affect it, and some of them push the return up.
Trading costs pull it down. Each time the index changes its members, and each time investors put money in or take it out, the fund has to buy or sell shares and pay commissions and spreads. The index pays none of these.
Cash drag also works against the fund. Funds normally keep a small cash balance, and in a rising market that cash earns less than shares would.
Tax can help or hurt. If a fund pays more dividend withholding tax than the index assumes, it falls behind. If it pays less, it gains. A fund may pay less because of the country it is set up in or the tax treaties it can use, which Lesson 3.3 covers.
Securities lending tends to help. Many funds lend some of their shares to other market participants for a fee, and part of that income goes back into the fund. The fund's documents tell you whether it lends shares, how the lending income is split and what small risks come with lending.
Taken together, these factors mean a fund with a higher TER can trail its index by less than a fund with a lower TER. That is what happened with Farhan's two funds.
Tracking error is a separate measure. It describes how much the yearly gap varies, usually as a standard deviation, and it says nothing about how big the gap is. Take a fund that trails its index by exactly 0.2 points every year (an illustration). Its tracking error is close to zero, yet it still loses 0.2 points each year.
Many factsheets show tracking error prominently, so check which measure you are reading before you draw conclusions. If you hold for the long term, the average gap is what you pay, and that is the tracking difference. The wobble matters less to you, because good and bad years tend to cancel out over a long stretch. A fund with a noisier record, meaning a higher tracking error, can still give you more of the index's return on average.
Farhan copies each year's returns from the two annual reports and works out each fund's gap (all figures invented):
Year 1: index 10.3%. Fund A 10.1%, gap minus 0.2. Fund B 10.4%, gap plus 0.1. Year 2: index minus 18.2%. Fund A minus 18.4%, gap minus 0.2. Fund B minus 18.9%, gap minus 0.7. Year 3: index 22.3%. Fund A 22.0%, gap minus 0.3. Fund B 22.6%, gap plus 0.3. Year 4: index 7.7%. Fund A 7.6%, gap minus 0.1. Fund B 7.1%, gap minus 0.6. Year 5: index 15.1%. Fund A 14.9%, gap minus 0.2. Fund B 15.2%, gap plus 0.1.
Fund A's five gaps add up to minus 1.0. Dividing by five gives an average of minus 0.20 points a year. Fund B's gaps add up to minus 0.8, an average of minus 0.16 points a year. Fund B's gap changes much more from year to year, so its tracking error is much higher. Even so, it kept more of the index's return over the five years.
If Farhan had checked only one year, the answer could have been wrong in either direction. In year 2, Fund B looks much weaker, at minus 0.7 against Fund A's minus 0.2. In year 3, Fund B beat its index outright, at plus 0.3. Any single year can mislead like this. When you compare funds that track the same index, use the average tracking difference over several years. Don't rely on one year or on the TER alone.
First, find the version of the index return that the fund's documents name. This is usually the net return version. Then collect the fund's yearly returns and the matching index returns for every year the fund has published. You'll find both in the annual report or in the performance table on the factsheet.
Work out each year's tracking difference by subtracting the index return from the fund return. Add up the yearly figures and divide by the number of years to get the average. Note any year when something unusual happened, such as a change of index, so you know to read that year with care. Whenever the funds you are comparing track the same index, use this average to compare them, not one year or the TER alone.
A fund less than three or four years old doesn't have a long enough record for a fair average. For a new fund, go by the TER until more years of returns are published.
Now take two real funds through the same steps: list the yearly returns, subtract, add up and divide. Choose two funds you would seriously consider buying, so the result can go into your comparison sheet in Lesson 3.5.
Pick two ETFs that track the same index. Write down the fund and index return for every year published, then calculate each fund's average tracking difference.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).