Why Index ETF Returns Differ from Their Index: Tracking Difference vs. Tracking Error
An index returns 10% while its ETF returns 9.6%. Before judging tracking quality, distinguish the return gap from the variability of periodic gaps, and match dividends, currency, dates, and valuation bases.
What tracking difference and tracking error measure
Tracking Difference is the difference between ETF total return and benchmark return over a specified period, while Tracking Error is the volatility of periodic return differences, not money the fund lost or its overall price volatility
Vanguard's description distinguishes the measures and warns that providers may define tracking error differently. This article uses Fund return minus index return and tracking error as the annualized standard deviation of periodic return differences
| numbers | Questions answered | How to interpret it | What it does not tell you |
|---|---|---|---|
| Tracking Difference | How far did the fund lead or lag the index during this period? | According to the formula in the article Negative values are behind, positive values are leading. | Is the difference consistent over time? |
| Tracking Error | How much periodic return differences fluctuate | A low value means the difference fluctuates little. according to the method and data range used | whether the fund offers higher returns or protection against losses |
Imagine two cars following the same route. One maintains a fairly constant gap; another moves closer and farther away. Similar average gaps need not have the same consistency. This is an analogy, not a market-price model.
Tracking Difference: How far ahead or behind?
The formula for the same period is Tracking Difference = ETF Total Return − Index Total Return Use comparable bases, including currencies and dividend treatment. When both returns are percentages, their difference is expressed in percentage points not a relative rate of change
Hypothetical example: The index has one year total return 10.00% and the fund has one year total return 9.60% will be 9.60% − 10.00% = −0.40 percentage points or -40 basis points (bps), meaning the fund lagged the index by 0.40 percentage points that year, not that the fund lost 0.40%.
Starting with 100,000 currency units, no interim cash flows, and exactly matching the total-return series, the fund value becomes 109,600 For the index, it is 110,000 different 400 currency units Round to an integer This example does not include trading fees, exchange fees, or investor taxes. and the index is the basis of comparison. It is not an asset that can be directly purchased.
Do not divide a three-year cumulative tracking difference by three and call it annual. First convert fund and index returns to compound annual rates over the same period, then subtract. Read more in Article: CAGR vs. Average Return
Tracking Error: How much does the difference fluctuate?
Start with each period's difference, such as the fund's monthly total return minus the index's. Then calculate the standard deviation of that series and annualize it, following the statistical definition in Vanguard explains
For the monthly example in this article, use the formula Annual Tracking Error = Sample standard deviation of monthly differences × √12 where sample standard deviation uses the denominator n − 1 where n is the number of periods. Daily data requires a defined number of trading days per year for the chosen method, rather than the monthly multiplier.
Square-root annualization relies on statistical assumptions. Serially correlated differences or changing patterns can make the result a poor guide to future risk. Before comparing sources, check their observation frequency, lookback period, and method.
Low tracking error does not mean no cost. Funds that trail the index the same every period may have tracking errors close to zero but still give lower returns than the index. As for funds that move with the market consistently, they can still lose when the index decreases. This value does not measure all market risk.
Two hypothetical funds with equal average tracking differences
This table is Four months of hypothetical data solely to demonstrate the formula not real ETF test results. Every return is assumed to be a total return in the same currency. Fund A lags by 0.02 percentage points monthly; fund B alternates between −0.12 and +0.08 percentage points.
| months | Index (%) | Fund A (%) | Difference A (percentage points) | Fund B (%) | Difference B (percentage points) |
|---|---|---|---|---|---|
| 1 | 1.00 | 0.98 | −0.02 | 0.88 | −0.12 |
| 2 | −2.00 | −2.02 | −0.02 | −1.92 | +0.08 |
| 3 | 3.00 | 2.98 | −0.02 | 2.88 | −0.12 |
| 4 | 1.00 | 0.98 | −0.02 | 1.08 | +0.08 |
- Average monthly difference: Both funds average −0.02 percentage points. Fund B: (−0.12 + 0.08 − 0.12 + 0.08) ÷ 4
- Monthly standard deviation: Fund A equals 0 because its difference is constant. Fund B is approximately 0.11547 percentage points, from √[(0.10² + 0.10² + 0.10² + 0.10²) ÷ 3].
- Annualized tracking error: Fund A is 0.00%; Fund B is approximately 0.40%, from 0.11547 × √12 rounded to two decimals. This is variability in return differences, not an expected annual loss of 0.40%.
What the example establishes is The same average gap does not imply the same consistency The average monthly difference is not the four-month cumulative tracking difference: cumulative returns must be compounded separately for each series. Four months are too few to judge an actual fund's quality or predict its future.
Check the cumulative results from the same four months.
Use (1 + month 1 return) × … × (1 + month 4 return) − 1, converting percentages to decimals first. The index yields 1.01 × 0.98 × 1.03 × 1.01 − 1 = 2.968894%; fund A is approximately 2.887131% and fund B approximately 2.892026%.
Subtracting cumulative index returns gives four-month tracking differences of approximately -0.081763 percentage points for A and -0.076868 for B, calculated from unrounded figures. These are not -0.02 percentage points, the monthly average, or annual rates. Small differences in hypothetical data are insufficient to rank actual funds.
Why do ETF returns not match indexes?
Index is a set of rules and calculation base. As for the fund, it must actually operate. Vanguard breaks down the main factors are fees, management ability, index-replication methods, and taxes, affecting both the size and variability of differences
- Fund expenses and trading costs: Operating expenses create drag on returns And trading costs within the fund are not included in all types of expense rates.
- Cash management and composition adjustment: Funds manage flows, index constituent changes, and portfolio construction rather than merely copying index return figures
- Index tracking method: Full replication and representative sampling or optimization may produce different tracking patterns. Read the fund's actual policy rather than inferring from the ETF name.
- Internal taxes: Withholding tax on dividends from a fund's holdings may affect NAV and returns, depending on its structure and domicile. This is not the reader's personal tax rate.
Therefore, do not conclude that Tracking Difference must always be equal to the fund's fees and should not deduct the Expense Ratio again from the return figure after deducting that expense. Read about the costs of each level in Expense Ratio and ETF Expenses
Match bases before calculating differences
Before applying the formula, record what each return series measures, following this KrobClick data check. If the series use different bases, the calculated gap may come from how the data are presented and cannot yet establish how well the fund tracks its index.
- Index and return variant: Use the actual fund's benchmark. Distinguish price return excluding dividends from total return including them, and gross from net total return with different tax assumptions. Check the index methodology.
- Dividends: Do not compare a dividend-paying fund's price chart with a total-return index. Use total-return data and read how dividend reinvestment is modeled.
- Currency and hedging: Do not directly compare a baht portfolio with a dollar index, or interpret a currency-hedged share class versus an unhedged index as poor tracking
- Dates and time periods: Use matching start dates, end dates, and periods. Do not compare year-to-date with trailing-year returns. Check valuation times when assets trade in different markets.
- NAV and market price: To study portfolio management, start with the NAV-return series the manager uses against the index. Market-price returns are a separate perspective. Do not mix these two bases.
- Cumulative versus annual returns: Compare annualized figures with annualized figures over the same period, not multiyear cumulative returns with annual rates
Thai investors should still separate The fund’s index-tracking performance from own net return because trading timing, exchange fees, commissions, and personal taxes belong to separate calculation layers. Read more in DCA US stocks and ETFs and currency risk
Checklist for reading actual fund information
Start with the manager's website, prospectus, factsheet, and annual report rather than undated comparisons without methods. Record the following when reviewing performance tables.
- Do fund names, share classes, benchmarks, currencies, and hedging policies match?
- Are fund returns NAV Total Return or Market Price Return? Does the index use Price, Gross Total, or Net Total Return?
- What date does the return cover? Is it cumulative or annual? And what kind of dividends are included?
- Does tracking difference use fund return minus index return, as this article does, or the opposite sign convention?
- Is tracking error historical or forecast? Does it use daily, weekly, or monthly observations, how many periods, and has it been annualized?
- Do changes in the gap over different periods coincide with policy, fee, or index-tracking changes? If information is insufficient, say the cause is unknown rather than guessing.
If the document does not show Tracking Error but only has one year's returns Tracking Error cannot be calculated from a single pair of numbers. There must be a consistent set of periodic returns. Finding the difference in one year can be done when the two bases match. For an overview of how to read the document, see Fund factsheet reading checklist
Do not judge a fund by one figure
Tracking difference measures the return gap over a period; tracking error measures how variable periodic gaps are. A fund can lag consistently or have a small average gap with greater variation, so read the measures together.
Before choosing the lowest figure, match bases, check methods, and examine multiple periods while considering policies, costs, and other risks. Past tracking performance does not guarantee future results. This article does not rank ETFs or recommend purchases.
Sources
Vanguard explanations reviewed October 12, 2026 (Thailand time). Fund A and B figures are hypothetical KrobClick calculations, not actual returns. The demonstration uses sample standard deviation and assumes 12 periods per year. Providers' methods may differ. This is general education, not guaranteed returns or personalized investment advice.