Tracking error measures how much a fund’s returns fluctuate relative to a chosen benchmark. Analysts usually calculate the standard deviation of periodic active returns—the fund’s return minus the benchmark’s return—and annualize the result. A small figure indicates consistent benchmark-relative results, but it does not prove that a fund earned a strong return or avoided losses.
An index fund can lag its benchmark by a predictable amount every month yet report very low tracking error. Meanwhile, another fund may outperform over the year despite producing a much less consistent return pattern. The distinction matters when investors compare ETFs, review fund managers, or interpret statements about passive-investing accuracy.
The most useful approach is to read tracking error beside tracking difference, fund costs, benchmark methodology, and the period used to calculate the statistic. Each measure answers a different question.
What Is Tracking Error in Investing?
Tracking error is the variability of the difference between a portfolio’s periodic returns and those of its benchmark. That difference is known as the active return. For a fund intended to replicate an index, high variability indicates the fund’s relative performance changes considerably across periods. In active management, the same statistic is commonly called active risk.
For example, suppose a fund beats its index by 0.2 percentage points one month and trails by 0.2 points the next. Those swings contribute to tracking error. In contrast, a fund that trails by exactly 0.05 percentage points every month has no variation in monthly active returns in this simplified example, even though it persistently underperforms.
Importantly, tracking error does not measure the fund’s total volatility. If both a fund and its index fall by 15% in a month, their relative difference may still be close to zero. Nor does a low tracking error guarantee that a benchmark is suitable for an investor’s goals.
The Benchmark Must Match the Investment Objective
The comparison is meaningful only when the benchmark represents the fund’s intended opportunity set. An emerging-market equity portfolio compared with a developed-market bond index could display high active risk for reasons unrelated to implementation quality. Additionally, the currency, dividend assumptions, tax treatment, and observation dates should match.
For index funds, investors should distinguish a price-return index, which excludes reinvested dividends, from a total-return index, which incorporates them according to the index’s methodology. Comparing a distributing fund’s price-only changes with a total-return benchmark would exaggerate the apparent gap unless distributions were added back.
Tracking Error Formula: From Monthly Returns to an Annual Figure
A common ex-post formula uses the sample standard deviation of active returns:
Active return in period t = Fund return(t) − Benchmark return(t)
Tracking error = √[Σ(active return(t) − mean active return)² / (n − 1)] × √m
Here, n is the number of observations and m is the number of observation periods in a year. With monthly data, m equals 12; with weekly data, analysts may use approximately 52. Daily calculations often use a trading-day convention such as 252, subject to the reporting methodology.
The expression uses a sample standard deviation. Some reporting systems apply population standard deviation or other conventions, so results can differ slightly even when the underlying data match. Therefore, compare fund statistics only after checking frequency, time window, annualization method, and return definitions.
Why the Square Root of 12 Appears
Standard deviation is measured at the frequency of the observations. Under the customary square-root-of-time approximation, monthly tracking volatility becomes annualized by multiplying by √12. That is a scaling convention, not a prediction of next year’s relative outcome. Serial correlation, market closures, and changing exposures can make actual longer-horizon behavior differ from the estimate.
Tracking Error Calculation: A Twelve-Month Worked Example
Consider a fictional fund called the Meridian Index Portfolio. Assume the fund and its benchmark report monthly total returns in the same currency, using aligned month-end valuations. Every figure below is hypothetical, and the dataset is deliberately short to make the arithmetic transparent.
| Month | Benchmark return | Fund return | Active return |
|---|---|---|---|
| 1 | 2.00% | 1.86% | −0.14 pp |
| 2 | −1.50% | −1.53% | −0.03 pp |
| 3 | 1.20% | 1.27% | +0.07 pp |
| 4 | 0.80% | 0.64% | −0.16 pp |
| 5 | −2.00% | −1.96% | +0.04 pp |
| 6 | 3.00% | 2.95% | −0.05 pp |
| 7 | 1.50% | 1.40% | −0.10 pp |
| 8 | −0.70% | −0.65% | +0.05 pp |
| 9 | 1.80% | 1.67% | −0.13 pp |
| 10 | 0.60% | 0.61% | +0.01 pp |
| 11 | −1.20% | −1.28% | −0.08 pp |
| 12 | 2.40% | 2.38% | −0.02 pp |
Step 1: Calculate active returns. Subtract the benchmark’s monthly return from the fund’s return. For month one, 1.86% − 2.00% equals −0.14 percentage points. Continue for each month.
Step 2: Find the mean active return. The twelve active returns average −0.045 percentage points per month. This shows the average monthly difference, but does not yet measure variability.
Step 3: Calculate sample standard deviation. Subtract −0.045 from each active return, square each deviation, sum the results, divide by 11, and take the square root. The monthly sample standard deviation equals 0.07810 percentage points.
Step 4: Annualize the result. Multiplying 0.07810 by √12 gives 0.27055%, or approximately 0.271% annualized tracking error.
A spreadsheet can calculate this efficiently. When the 12 monthly active returns are in D2:D13 and expressed in percentage units, use =STDEV.S(D2:D13)*SQRT(12). The calculation should yield the same number, apart from rounding.
What Happened to Actual Fund Performance?
Compounding the monthly returns produces a 7.46% fund return and an 8.03% benchmark return for the illustrated year. The fund’s return was therefore 0.57 percentage points lower. That return difference is not the same as the 0.271% tracking error.
In short, tracking error describes variation around the average active return, while the annual return gap describes the eventual relative outcome. Both are useful, but neither substitutes for the other.
Tracking Error vs Tracking Difference
The two terms are often confused in fund marketing and casual discussion. Strictly speaking, a standardized volatility-based tracking-error measure and a holding-period return difference serve different purposes.
| Measure | Main question | Typical calculation | Example result |
|---|---|---|---|
| Tracking error | How much did periodic relative returns fluctuate? | Annualized standard deviation of active returns | 0.271% |
| Tracking difference | How far did the fund return deviate over the stated period? | Fund holding-period return minus benchmark holding-period return | −0.57 pp |
| Fund total return | What did the fund earn over the period? | Compounded periodic fund returns | +7.46% |
| Benchmark total return | What did the reference index earn? | Compounded periodic index returns | +8.03% |
A fund with a steady monthly drag from fees may post low tracking error even when the cumulative tracking difference is persistently negative. Meanwhile, a portfolio that outperforms its benchmark in volatile bursts may have a positive final difference and relatively high tracking error.
A Zero-Error Thought Experiment
Suppose a fund follows an index so closely that it trails by exactly 0.05 percentage points every month. Its active returns never vary, giving a sample standard deviation of zero. As a result, historical tracking error is zero in that artificial case. Nevertheless, the fund is still behind the index each month, and the compounded annual difference is generally negative.
This demonstrates why zero tracking error does not mean zero underperformance. Investors who focus exclusively on the volatility statistic may overlook a systematic performance shortfall.
What Real Fund Documents Reveal About Tracking Statistics
A useful example comes from the iShares PLC interim report for the financial period ended 31 August 2025. BlackRock explained that its realised tracking error was the annualized standard deviation of monthly fund-versus-index return differences, generally calculated over the preceding 36 months. The report also distinguished that historical measure from the anticipated tracking error stated in a prospectus.
For the iShares $ Corp Bond Interest Rate Hedged UCITS ETF, the report listed a 3.78% fund return and 3.48% benchmark return for the interim financial period. Separately, it reported 0.66% realised tracking error, versus an anticipated figure of up to 0.30%. Those statistics should not be compared as though they all measure the same interval: the performance figures cover the interim financial period, while the realized error uses its stated historical observation methodology.
This is a practical reporting lesson. A table can correctly display both return differences and tracking error, yet readers can reach the wrong conclusion if they overlook the different measurement windows. Moreover, exceeding an anticipated tracking-error estimate does not, by itself, establish that a fund violated a guaranteed performance limit; the prospectus terms and disclosure wording matter.
Another useful example appears in a 2026 U.S. ETF prospectus filed with the SEC: a fund described a goal of keeping tracking error below 5%, measured from monthly active returns over one year, while expressly noting that the target was not guaranteed. A stated tracking-error objective therefore needs interpretation alongside the fund’s investment strategy and risk disclosures.
Why Index Funds and ETFs Depart From Their Benchmarks
The benchmark is a calculation produced under a published methodology. A real fund must hold or replicate assets, process cash flows, pay operating expenses, and execute transactions. Those operational realities create differences.
Fees and Other Expenses
Management fees, custody expenses, and fund administration typically reduce investor returns relative to a frictionless index. However, predictable fees may create a relatively stable negative tracking difference rather than high tracking error. Unexpected costs or timing differences may increase variability as well.
Sampling Instead of Full Replication
Some index funds own most or all index constituents at similar weights. Others use representative sampling to reduce transaction costs or handle difficult-to-buy securities. The sample may broadly match sector and risk exposures while differing from the index when individual holdings behave unexpectedly.
Rebalancing and Corporate Actions
When an index adds or removes securities, its calculation changes according to index rules. The fund may trade on a different schedule or at different market prices. Share splits, mergers, cash dividends, and index reconstitutions can also create temporary mismatches.
Cash Drag, Taxes, and Securities Lending
Cash held for redemptions or distributions can influence relative returns during market moves. Foreign withholding tax, tax reclaim timing, or securities-lending revenue may widen or narrow differences. Furthermore, different treatment of gross versus net dividends in an index can materially affect comparisons.
Trading Costs and Market Conditions
Funds pay spreads and may incur market impact when trading underlying securities. Their results can also depend on whether instruments are liquid during an index rebalance. The mechanics of trading slippage explain why an intended portfolio change and its executed prices can diverge, especially during volatility or thin trading.
Importantly, the ETF’s exchange-traded share price may differ from its net asset value. A premium or discount arising in secondary trading is a separate consideration from the fund’s NAV-based benchmark tracking. Investors should avoid mixing ETF market-price returns with NAV-based index comparisons unless that is the intended methodology.
Why Tracking Error Can Rise During Market Stress
Tracking efficiency may weaken when prices move rapidly or investors withdraw capital. For example, an index assumes particular valuation times, while portfolio securities may trade on markets that are closed or less liquid. In addition, a fund may need to hold more cash or adjust positions at unfavorable prices.
Suppose a fund owns a representative basket rather than every benchmark constituent. On an ordinary day, the holdings may move almost identically to the index. During a sudden sector-specific decline, an omitted security can contribute disproportionately to the index’s return. As a result, active returns can swing sharply even if the fund’s strategy has not changed.
However, a short-lived tracking-error spike does not automatically identify poor management. Investors should ask whether the fund’s objectives, holdings, benchmark rules, liquidity conditions, or corporate actions explain the episode. A persistent unexplained mismatch is more concerning than a one-off fluctuation with a documented cause.
Is Lower Tracking Error Always Better?
For a fund marketed as a close index tracker, low and stable error is generally desirable, provided the index is appropriate and investor costs remain reasonable. However, it is not a universal quality ranking.
An actively managed fund may intentionally take positions that differ from its benchmark. Such a strategy can generate greater tracking error because it accepts active risk in pursuit of a different outcome. Consequently, comparing its tracking error directly with that of a full-replication index fund without considering their objectives would be misleading.
Also, two low-error funds may have different fee levels and long-run return gaps. The one with the lowest tracking error might not produce the best after-cost investor outcome. The investment decision therefore requires both consistency and economic results.
Tracking Error Is Not Maximum Drawdown
Consider a fund and index that both lose 25% in a crisis. If their returns match period by period, tracking error can remain near zero. Nevertheless, the investor has suffered a substantial portfolio decline. Maximum drawdown measures the deepest peak-to-trough fall over a specified history and addresses a fundamentally different risk question.
Likewise, a fund can keep tracking error low while holding concentrated or volatile securities because the benchmark itself has those characteristics. Benchmark-relative consistency should never be mistaken for capital protection.
Taxes and Turnover Can Complicate the Comparison
Funds may rebalance, realize gains and losses, distribute income, or employ tax-aware trading. Those actions can change portfolio weights and transaction costs. If the investor also adjusts holdings for tax reasons, the investor’s personal after-tax return can diverge further from the fund’s reported NAV return.
For instance, a taxable investor practicing tax loss harvesting might temporarily replace a holding with a different instrument. That replacement may change benchmark exposure and risk, while local rules determine whether the realized loss produces any tax benefit. The performance effect should be evaluated separately from the possible tax result.
Tax treatment also varies between funds domiciled in different jurisdictions. Therefore, an ETF tracking a global equity index may have dividend withholding or reclaim effects not replicated perfectly by a benchmark’s assumed tax rate.
How to Compare Two Funds Without Misreading Tracking Error
A sound comparison uses a consistent checklist rather than choosing the smallest published percentage.
- Confirm the benchmark. Compare the exact index version, not merely a familiar index name. Check total-return conventions and currency.
- Align measurement dates. Use the same end date and sufficiently long periods, particularly around major market disruptions.
- Review the methodology. Determine whether the figure uses daily or monthly data, sample or population standard deviation, and what annualization factor applies.
- Check tracking difference. Evaluate the fund-versus-index return gap separately from variability, and prefer returns after published fund fees when assessing investors’ experience.
- Inspect transaction and fund costs. Consider ongoing charges, turnover, spreads, potential market impact, and any platform costs.
- Review replication and securities lending. Understand whether the fund samples holdings, uses derivatives, or earns lending income.
- Examine outliers. Determine whether one unusual month dominates the result or whether deviations recur systematically.
- Separate NAV tracking from trading price. An ETF’s market premium or discount can affect an investor independently of its NAV tracking statistic.
For example, Fund A may report 0.10% tracking error with a persistent annual return shortfall of 0.45 percentage points. Fund B might show 0.20% tracking error and a smaller shortfall of 0.20 points. Fund A tracked more consistently, but Fund B delivered a closer final return over the stated period. The numbers are illustrative; neither fund can be declared superior without costs, risks, and benchmark suitability.
Common Tracking Error Mistakes
Mistake: Treating a Negative Tracking Difference as Negative Tracking Error
Standard deviation is nonnegative. A fund’s return difference can be positive or negative, while its tracking-error statistic measures variability. Report both with their units instead of using the terms interchangeably.
Mistake: Comparing Different Observation Frequencies
A daily tracking-error calculation may capture pricing noise that a monthly calculation smooths out. Even after annualization, the results need not match. Fund A’s daily result should not be ranked against Fund B’s monthly result without checking the methodological difference.
Mistake: Using Prices Instead of Total Returns
A distributing fund can appear to underperform if the calculation ignores dividends paid out to investors. Likewise, a benchmark using gross dividends can differ from one using net dividends. Align the return series before calculating active returns.
Mistake: Ignoring the Number of Observations
A calculation from three monthly returns can produce a numerical standard deviation, but it is a fragile estimate. The 12-point example in this article teaches the method; it does not provide the confidence of a longer operational track record. BlackRock’s 2025 fund report, for example, commonly uses a 36-month observation period for realized error.
Mistake: Assuming Tracking Error Is a Guaranteed Limit
A prospectus objective or anticipated estimate is generally not a promise of a future maximum. Liquidity shocks, index changes, and different market conditions may cause realized results to exceed the estimate. Always read the actual disclosure.
Frequently Asked Questions
What Is Tracking Error in Simple Terms?
Tracking error shows how consistently an investment fund follows a benchmark from period to period. Analysts usually calculate the standard deviation of the fund’s periodic return differences against the index. A smaller number indicates less variation in relative returns, but it does not guarantee an attractive absolute return.
What Is a Good Tracking Error for an ETF?
There is no universal threshold. A broad, liquid, fully replicated index fund may be expected to track closely, while a sampled, hedged, emerging-market, or complex strategy may naturally deviate more. Compare the fund with similar products, its stated objective, and its actual tracking difference.
Can Tracking Error Be Zero When a Fund Underperforms?
Yes. If a fund trails the benchmark by the same amount every observation period, the standard deviation of those active returns is zero. The persistent shortfall still appears in tracking difference. Real-world outcomes rarely match this exact thought experiment, but the distinction is essential.
How Do You Calculate Tracking Error in Excel?
Subtract each benchmark return from the corresponding fund return to form a series of active returns. Then use STDEV.S on that series and multiply by SQRT(12) for monthly data, assuming the chosen sample-standard-deviation convention. Check that fund and index returns use the same dates and currency.
Is Tracking Error the Same as Volatility or Beta?
No. Total volatility measures variation in the fund’s own returns. Beta estimates market sensitivity under a model, while tracking error measures variation in returns relative to a specific benchmark. A fund can have high market volatility and low tracking error when its benchmark is similarly volatile.
Does Tracking Error Include ETF Trading Spreads?
NAV-based tracking-error reports reflect fund-level return differences and implementation effects, but not necessarily the bid-ask spread a particular investor pays to buy or sell ETF shares. Market-price premiums, discounts, and brokerage costs can therefore affect investor returns beyond a reported NAV-based tracking-error figure.
Final Perspective
Tracking error is a valuable measure of benchmark-relative consistency, not a complete score for investment quality. Calculate it from aligned periodic active returns, verify the observation frequency, and compare it with tracking difference and the fund’s costs. Real fund reports show why anticipated error, realized error, and holding-period return gaps can sit in one table while describing different things.
For passive products, investigate persistent unexplained departures from the stated benchmark. For active strategies, consider whether benchmark deviations reflect deliberate investment decisions and whether their results justify the accompanying risk. Above all, assess the suitability of the underlying benchmark before treating closeness to it as an advantage.
This article provides general educational information and does not constitute personalized investment advice.