Tracking Error in Index-Linked Products: Decompose the Gap
Products linked to the same index rarely deliver identical returns. Fees, tax, cash, sampling, corporate actions and valuation times create a gap. Separate the average tracking difference from the variability called tracking error.
Who this guide is for: Readers comparing ETFs or index funds, analysts looking beyond the expense ratio, and investors separating NAV tracking from traded-price experience.
Key points to understand first
- Tracking difference is commonly the average product-minus-benchmark return; tracking error is the variability of that active return.
- Match benchmark variant, currency, NAV or market price, frequency and period.
- Fees, withholding tax, cash, sampling, rebalances and securities lending can widen or offset the gap.
- Low historical tracking error does not guarantee future tracking, liquidity, credit quality or safety.
Main steps from benchmark to product return
- Benchmark return
Select the exact index variant
Starting point - Fees & cost
Management, trading and operating expenses
Persistent drag - Tax & cash
Withholding, dividend timing and cash balance
Timing gap - Replication
Full, sample, derivative and rebalance
Variable gap - Product return
NAV or reinvested realised result
Comparison output
Tracking difference versus tracking error
Active return_t = product return_t − benchmark return_tTracking difference = average active returnTracking error = standard deviation of active return, optionally annualisedDocuments can use different terminology. State frequency, sample convention and annualisation factor.A product that lags by roughly 0.2% every year can have a −0.2% tracking difference and low tracking error because the gap is stable. Another can average near zero but vary widely day to day, creating high tracking error. One measures level; the other measures consistency.
Before comparing published figures, match price, gross or net total-return benchmark, currency and hedge, return frequency, period, and NAV or market price. A benchmark mismatch is not manager skill.
Seven sources of the gap
| Driver | Transmission | Evidence |
|---|---|---|
| Expense | ongoing management and operating charge | prospectus and report |
| Tax | withholding versus index assumption | benchmark and fund domicile |
| Cash | dividend, subscription and redemption timing | holdings and cash policy |
| Sampling | subset approximates the index | policy and holdings |
| Rebalance | execution differs from index effective price | notice and turnover |
| Corporate action | mergers, rights and fractional shares | fund report |
| Lending/derivative | revenue or counterparty and roll cost | annual report |
Equal expense ratios can produce different outcomes because tax, lending, execution, cash and sampling add or offset drag. Study stress periods, rebalance months and dividend seasons, not only the long-run mean.
Full replication, sampling and derivatives
Full replication aims to hold every constituent near index weight. Sampling uses a subset when liquidity, cost or ownership limits make full replication inefficient. Synthetic or derivative implementation introduces swap, futures, collateral, counterparty, roll and basis considerations.
In a broad small-cap or emerging-market index, sampling can reduce trading cost and improve net tracking, but can deviate in company, sector or factor exposure. Inspect holdings count, country and sector differences and derivative exposure.
A reproducible product comparison
Save formal product and share-class identifier, exact benchmark, return type, currency, hedge, NAV frequency, expenses, replication, lending, assets and spread. Build active returns from matching daily or monthly data and calculate mean, standard deviation, worst gap and rolling windows.
Align return type, then inspect event gaps with index reviews. Add spread, commission and funding separately in the Trade Cost Calculator.
- Use reinvested total returns for distribution-paying products.
- Address survivorship when products closed or merged.
- Do not present low past error as a guarantee.
Frequently asked questions
Is the lowest tracking error always the best product?
It is one tracking measure. Also review cost, liquidity, spread, tax, credit and structure.
Is expense ratio the same as tracking difference?
No. Tax, cash, sampling, trading and lending revenue also matter.
Should I compare an ETF index with NAV or market price?
Use NAV for portfolio tracking and market price for the investor’s traded experience; align timestamps.
Can a product perfectly match its index?
Perfect identity is not guaranteed because cost and implementation frictions remain.
Primary sources and verification links
- Investor.gov — Index FundsOfficial explanation of fees and tracking error.
- U.S. SEC — Investor Bulletin: Exchange-Traded FundsOfficial ETF, NAV, market-price and cost explanation.
- S&P Dow Jones Indices — Index Mathematics MethodologyBenchmark return types and index calculation.
- IOSCO — Principles for Financial BenchmarksBenchmark governance and quality principles.
Editorial approach: We prioritize primary materials from central banks, regulators and international institutions. Rules, product terms and release times can change, so verify current information at the linked source and with your provider before acting.
Important notice: This article is general education about index mechanics, not a recommendation, signal or promise of return. An index cannot be held directly. Index-linked products introduce their own market, currency, liquidity, credit, leverage, fee and tax risks. Verify the latest index methodology, product documents and regulator information before acting.