What the two terms mean

An index fund has one job: deliver the index’s return. Two numbers measure how well it does that, and they answer different questions.

Tracking difference is the gap between the fund’s return and the index’s return over a period. It answers: how much did I actually give up (or gain) by holding the fund instead of the index? This is the number that costs you money.

Tracking error is the variability of that gap — the standard deviation of the fund’s return minus the index’s return, measured at daily or monthly frequency and annualised. It answers: how tightly does the fund hug the index day to day? A fund can have a large tracking difference with tiny tracking error (it lags the index by exactly its expense ratio, like clockwork), or a small difference with large error (it wobbles around the index and happens to end up close).

Both numbers are now published. SEBI’s passive funds circular of 23 May 2022 (SEBI/HO/IMD/DOF2/P/CIR/2022/69) caps tracking error for equity index funds and ETFs (exchange-traded funds) at 2% (measured on one year of daily data), requires it to be disclosed daily, and requires tracking difference to be disclosed monthly over 1, 3, 5 and 10 years. Tracking error is still the one that tends to get quoted, because it’s almost always a small, reassuring number. Tracking difference is the one to look at first.

The formula

Tracking difference  =  R_fund − R_index               (over a stated period)

Tracking error       =  σ( r_fund,t − r_index,t ) × √N

  where r_t are period returns (daily: N = 252; monthly: N = 12)

Worked example: UTI Nifty 50 Index Fund against the Nifty 50

Fund NAV from AMFI via mfapi.in (regular and direct plans); index from Yahoo Finance, price index. Data to 31 March 2026, so the most recent full calendar year is 2025. Historical data, for illustration only.

Tracking error, regular plan, 1 Apr 2023 – 31 Mar 2026:

Measured on Tracking error (annualised)
Daily return differences 0.27%
Monthly return differences 0.33%

Small, as it should be for a fifty-stock index fund — a fraction of one percent. The fund hugs the index closely.

Tracking difference, regular plan, same three years: fund 9.61% a year, index 8.71% a year, difference +0.9 pp a year.

And by calendar year, both plans:

Year Index (price) Regular plan Difference Direct plan Difference
2016 3.01% 4.0% +0.99 4.1% +1.09
2017 28.65% 29.68% +1.03 29.79% +1.14
2018 3.15% 4.26% +1.11 4.33% +1.18
2019 12.02% 13.25% +1.22 13.33% +1.3
2020 14.9% 15.5% +0.6 15.56% +0.66
2021 24.12% 25.2% +1.08 25.3% +1.18
2022 4.33% 5.33% +1.0 5.44% +1.11
2023 20.03% 20.89% +0.86 21.04% +1.01
2024 8.8% 9.64% +0.84 9.8% +0.99
2025 10.51% 11.57% +1.06 11.68% +1.17
Mean     +0.98   +1.08

Reading the plus sign correctly

An index fund that beats its index by a point a year, every year, for ten years is not a good index fund. It’s a fund being measured against the wrong index.

The benchmark here is the price index, which excludes dividends. A positive tracking difference against it is mostly the roughly 1–1.5% dividend yield of the fifty stocks, minus the expense ratio — not outperformance.

The alpha post made the same point from the other direction. Against the total return index (TRI) — which is what SEBI has required funds to benchmark against since February 2018 — both columns would flip to small negative numbers: roughly minus the expense ratio, minus a little friction from cash held for redemptions and the timing of dividend reinvestment. That negative number is the true cost of indexing, and it’s what you should be comparing across index funds.

To be clear about this post’s own numbers: tracking difference should be measured against the TRI, and every figure above is against the price index, because a TRI series isn’t in this post’s data. The table shows the method and the size of the dividend distortion — not the fund’s true cost of indexing. For that, use the TRI-based tracking difference the fund house publishes each month.

One thing the table does show correctly, because the benchmark error is the same for both plans: the direct plan’s difference is 0.1 pp a year better than the regular plan’s. That gap is the distributor commission, and it is the entire subject of the expense ratio post.

🧒 Explain it like I'm 10 (optional — skip if this is already clear)

You’re trying to walk exactly in your older brother’s footsteps in the sand.

Tracking error is how wobbly your line is compared to his — do you weave left and right, or stay right on top of his prints?

Tracking difference is where you end up: if he walks 100 steps and you’ve only managed 98 by the time he stops, your difference is 2 steps, however straight your line was.

Common mistakes

  • Judging an index fund by tracking error alone. It measures wobble, not cost. Two funds with identical tracking error can differ by half a point a year in tracking difference, and that half point compounds.
  • Reading a positive tracking difference as skill. For an index fund against a price index, it’s dividends. Check whether the benchmark is the PRI (price return index, no dividends) or the TRI (total return index, dividends reinvested) before reading any sign.
  • Comparing tracking errors computed on different frequencies. Daily and monthly figures differ (0.27% vs 0.33% here) and neither is “wrong” — but they aren’t interchangeable.
  • Expecting zero. Even a perfect index fund can’t return the index: it has costs, holds a little cash, and receives dividends a few days after the index books them. A small, steady negative difference against the TRI is what good looks like.

Takeaway: tracking error is how much an index fund wobbles around its index; tracking difference is how far it ends up from it — and only the second one costs you money. When the difference is positive year after year, as it is here against the price index, the fund isn’t winning; the benchmark is missing its dividends.