What you will learn: how to measure how well a passive fund follows its index using tracking difference and tracking error, where the gap comes from, and how to add up the full cost of owning an index fund or ETF over the years you will hold it.
Rohit and Sneha Kulkarni (an illustrative couple), both 37, in Pune, want ₹10 lakh of their long-term equity money in a passive fund on a broad Indian index. They have shortlisted two index funds on the same index, one with an expense ratio of 0.20% and one at 0.45%, plus an ETF on that index charging 0.05%. Rohit says the cheapest expense ratio wins. Sneha has read that "tracking error" matters more. Their question: which number tells them what they will actually keep?
The answer is that no single number does. The expense ratio is one input, tracking difference is the result, tracking error is the consistency of that result, and trading costs apply only to the ETF. This lesson puts them together. How expense ratios work in general, and direct versus regular plans, is covered in fund costs and distribution.
Three numbers, three different questions
SEBI's Master Circular for Mutual Funds (as on 20 March 2026) defines the two tracking measures precisely:
- Tracking difference (TD) is the annualised difference between the return of the index and the return of the fund's NAV. It answers: "How much did I fall behind (or, occasionally, get ahead of) the index?"
- Tracking error (TE) is the annualised standard deviation of the difference in daily returns between the index and the NAV, using the past one year of rolling data. It answers: "How steady was that gap from day to day?"
- The total expense ratio (TER) is the annual charge deducted inside the fund. It answers: "What does the fund house charge me?"
A fund can have a tiny tracking error and a large tracking difference: it lags by the same amount every day. Another can have a small average lag but a jumpy tracking error. For a long-term holder, tracking difference is closer to the money lost; tracking error tells you how much any single period could surprise you.
| Measure | Disclosure | Limit or target |
|---|---|---|
| Tracking error | Daily, on the fund house's website and AMFI | Not above 2% for equity and other non-debt index funds and ETFs, except in force-majeure situations reported to trustees |
| Tracking difference | Monthly, on the fund house's website and AMFI, for 1, 3, 5 and 10 years and since allotment | Debt index funds and ETFs: one-year average not above 1.25%. Equity passive funds on the indices covered by SEBI's "MF Lite" framework (broad domestic equity indices on AMFI's periodic list, and permitted overseas indices), whichever fund house runs them: targeted at 0.50 percentage point (50 bps) over and above the TER actually charged; if the target is not met, it must be reported to trustees with corrective action |
| Base expense ratio | In the scheme documents and factsheet | Open-ended index funds and ETFs: maximum 0.90% of daily net assets under the SEBI (Mutual Funds) Regulations, 2026, from 1 April 2026; brokerage, transaction costs and statutory levies are charged separately within regulatory limits |
Where the gap comes from
If the fund held exactly the index at zero cost, tracking difference would be zero. Real funds face several frictions:
- Expenses. The TER is deducted daily, so it shows up almost one-for-one in tracking difference.
- Cash drag. Funds keep some cash for redemptions, and new money takes time to invest. In a rising market, cash lags.
- Trading costs. Brokerage, impact cost and taxes on trades when the index is reconstituted, when inflows arrive and when redemptions are paid. Rebalancing within the permitted 7 days, rather than at the exact index closing price, can add or subtract a little.
- Dividend timing. The total return index assumes a dividend is reinvested on the ex-date; the fund receives the cash later.
- Offsets. Income from lending securities, where the scheme permits it, can narrow the gap. That is why a fund's tracking difference is occasionally smaller than its TER.
Two practical rules follow. First, compare tracking difference against the total return version of the index, because that is what the fund is trying to match. Second, compare the same plan (direct with direct, regular with regular) over the same dates, because the plans have different expense ratios.
Worked example: tracking difference and tracking error from a year of data
Assumptions (illustrative): over one year the index TRI rose in the months shown below. Fund A and Fund B track it. SEBI's method uses daily data; monthly figures are used here only to keep the arithmetic visible, so tracking error is annualised by multiplying the monthly standard deviation by √12.
| Month | Index TRI return | Fund A difference | Fund B difference |
|---|---|---|---|
| 1 | 2.10% | −0.02 | +0.10 |
| 2 | −1.40% | −0.03 | −0.25 |
| 3 | 3.05% | −0.01 | −0.05 |
| 4 | 0.80% | −0.02 | +0.08 |
| 5 | −2.60% | −0.03 | −0.30 |
| 6 | 1.95% | −0.02 | +0.12 |
| 7 | 0.40% | −0.02 | −0.15 |
| 8 | 2.70% | −0.01 | +0.05 |
| 9 | −0.90% | −0.03 | −0.20 |
| 10 | 1.20% | −0.02 | +0.09 |
| 11 | −3.10% | −0.02 | −0.12 |
| 12 | 2.25% | −0.02 | +0.18 |
- Compounding the twelve months, the index returned 6.39%, Fund A 6.12% and Fund B 5.88%.
- Tracking difference: Fund A −0.27 percentage point; Fund B −0.51 percentage point.
- Tracking error: Fund A about 0.02%; Fund B about 0.56%.
Fund A lagged a little, every month, very predictably. Fund B beat the index in six months and still ended the year further behind, and its monthly gaps swung widely. With gaps that size, a different twelve-month window could show Fund B looking much better or much worse, which is why one period proves little. Look at tracking difference over several periods and read tracking error alongside it.
The total cost of ownership
For an index fund, total cost is mostly tracking difference plus any exit load if you redeem early, and tax when you sell. For an ETF, add the costs of trading on an exchange: brokerage, the bid-ask spread, any premium paid or discount accepted against the ETF's indicative value, and demat charges if the account exists only for this. Tax rules for these funds are covered in mutual-fund taxation and are left out of the comparison below because they would apply equally to all three routes.
Worked example: ₹10 lakh for ten years
Assumptions (illustrative): ₹10 lakh invested once and held for ten years; the index TRI compounds at a constant 11% a year (an assumption, not a forecast), giving ₹28,39,421 if it could be owned at zero cost. Index fund X has a tracking difference of 0.30% a year; index fund Y, 0.75%. The ETF has a tracking difference of 0.15%, but each trade costs 0.20% (half the spread plus a small premium or discount) and ₹20 brokerage, and the demat account costs ₹300 a year. No exit load applies at ten years.
| Route | Value after 10 years | Shortfall versus index |
|---|---|---|
| Index fund X (TD 0.30%) | ₹27,63,607 | ₹75,814 |
| Index fund Y (TD 0.75%) | ₹26,53,298 | ₹1,86,123 |
| ETF, normal trading (TD 0.15%) | ₹27,87,013 | ₹52,408 |
| ETF, adverse trading (1.5% lost on entry, 1.0% on exit) | ₹27,28,596 | ₹1,10,825 |
Three lessons sit in this table. A 0.45 percentage-point difference in tracking difference between two funds on the same index costs about ₹1.1 lakh on ₹10 lakh over ten years. The ETF's lower running cost can be cancelled by poor execution, and the ETF's adverse case still beats fund Y but now trails fund X. And trading costs are a one-time hit per trade, so they matter more for someone who buys in small monthly lots or trades often than for someone who invests once and holds. The trading ETFs lesson shows how to keep those costs small.
Reading the published tracking data
Fund houses and AMFI publish both measures. When you compare funds:
- Use the same index, the same plan and the same end date. A fund on a different index is a different product, not a cheaper version.
- Prefer the 3-year and 5-year tracking difference to the 1-year figure where they exist; a newer fund simply has less evidence.
- Check that tracking difference is broadly in line with the TER. A gap much wider than the expense ratio, repeated over time, points to weak execution.
- Read tracking error as a stability check. A low figure over a full year is reassuring; a sudden rise deserves a look at the factsheet or notices.
- Note fund size and age. Very small funds can find it harder to replicate an index and to absorb flows smoothly, and the regulation's targets do not guarantee any figure.
Try this: a one-page passive cost check
- Write down the index name and confirm it appears in each fund's scheme name.
- For each candidate, record the TER of the plan you would buy (direct or regular).
- From the fund house or AMFI tracking data page, record tracking difference for 1, 3 and 5 years and the latest tracking error.
- Subtract TER from tracking difference. A small, stable remainder suggests efficient management.
- For an ETF, add an estimate of trading cost per purchase and sale, and multiply by how many times you expect to trade.
- Note any exit load and your expected holding period, then compare the totals, not any single figure.
Common questions
Is lower tracking error always better?
Lower is usually better, but tracking difference is the bigger cost for long-term investors. A fund that lags by the same amount every day has very low tracking error and can still trail the index by more than a rival with a slightly higher tracking error.
Why can tracking difference be smaller than the expense ratio?
Income such as securities lending, where the scheme allows it, and favourable trade timing can offset part of the expense ratio. Over long periods, the expense ratio is usually the floor, so a tracking difference consistently far below it deserves a check that you are comparing against the right index version.
Does a 0.90% expense ratio limit mean index funds charge 0.90%?
No. That is the ceiling for the base expense ratio of open-ended index funds and ETFs from 1 April 2026. Many charge well below it. Check the actual figure in the factsheet.
Should I compare an index fund's return with the Nifty 50 figure on the news?
Not directly. News usually quotes the price index, which excludes dividends. Fund tracking is measured against the total return index, which is higher by roughly the dividend yield.
Where to go next
- How an index is built and maintained if you skipped the rules behind the index itself.
- Buying ETFs well for spreads, iNAV and premiums.
- Choosing and reviewing funds for a repeatable review routine.
- Further reading: direct vs regular mutual funds.