Understanding what is an ETF is the first step toward understanding why an ETF does not always produce exactly the same return as the index it is designed to follow. An exchange-traded fund, or ETF, typically seeks to track an index, commodity, bond market, or another investment strategy. Ideally, its performance should closely match its benchmark. In reality, small differences can appear.
These differences are known as tracking error. Tracking error is important because it shows how closely an ETF follows its intended benchmark over time. While some level of difference is normal, larger or persistent tracking errors can affect an investor's actual returns.
This comprehensive guide explains the main causes of ETF tracking error, how it is measured, why it happens, and what investors should consider when comparing ETFs.
What Is ETF Tracking Error?
ETF tracking error refers to the difference between the performance of an ETF and the performance of its benchmark.
For example, imagine an ETF is designed to track an index that rises by 10% during a particular year. If the ETF rises by 9.7%, the ETF has underperformed the index by 0.3 percentage points.
Tracking error can be discussed in two related ways.
The first is the tracking difference, which is the actual gap between an ETF's return and its benchmark's return over a specific period.
The second is tracking error in the statistical sense, which measures the volatility of the difference between the ETF's returns and benchmark returns over time.
This distinction matters. An ETF might consistently earn slightly less than its benchmark because of fees, but its performance could still be very stable. Another ETF might have returns that move above and below its benchmark considerably, producing a larger statistical tracking error.
Understanding what is an ETF therefore involves more than knowing that it trades on an exchange. Investors also need to understand how effectively the fund follows its underlying benchmark.
Why Does Tracking Error Matter?
Tracking error matters because investors usually purchase an index ETF expecting it to provide exposure similar to its benchmark.
If an ETF consistently differs significantly from its benchmark, the investor may receive a different result from the one expected.
For long-term investors, even small differences can become meaningful when they compound over many years.
Suppose two ETFs follow the same index. One ETF closely follows its benchmark, while the other regularly experiences larger performance differences. Even if both ETFs have similar investment objectives, their actual results may differ over time.
Tracking error can therefore provide useful information when evaluating an ETF alongside other factors such as expense ratios, liquidity, assets under management, and portfolio construction.
Major Causes of ETF Tracking Error
Several factors can cause an ETF to perform differently from its benchmark. Some are related to the fund's expenses, while others result from market conditions, portfolio management, taxes, or the way the index itself is constructed.
Expense Ratios and Fund Expenses
One of the most common reasons for tracking differences is the cost of running the ETF.
An ETF charges investors an expense ratio to cover costs associated with managing and operating the fund. These costs can include administration, portfolio management, custody, legal services, accounting, and other expenses.
Because the benchmark index itself does not pay these fund expenses, an ETF will generally lag behind the index by approximately the amount of its ongoing costs, all else being equal.
For example, if an index gains 8% and an ETF has annual expenses of 0.20%, the ETF may have a return somewhat below the index.
The actual difference will not always equal the expense ratio because other factors can either increase or reduce the gap.
Transaction Costs
An index may be changed periodically as companies are added, removed, or given different weights.
When an ETF changes its portfolio to reflect those adjustments, it may have to buy or sell securities. These transactions create costs.
Brokerage commissions, bid-ask spreads, market impact, and other trading expenses can reduce the ETF's return.
The benchmark index generally assumes that securities can be adjusted according to its rules without reflecting the same real-world trading costs.
This creates a natural difference between theoretical index performance and actual fund performance.
Sampling Instead of Full Replication
Not every ETF holds every security in its benchmark.
Some indexes contain hundreds or even thousands of securities. Buying every component may be expensive or operationally complicated. Instead, an ETF may use a sampling strategy.
With representative sampling, the fund holds a selection of securities designed to behave similarly to the overall index.
This approach can reduce costs, but it can also introduce tracking differences.
If the selected portfolio does not perfectly replicate the characteristics of the benchmark, its return may differ from the index.
Cash Holdings
ETFs may maintain small amounts of cash for operational purposes, including meeting redemptions, handling dividends, or managing portfolio transactions.
The benchmark may assume that all capital is invested in its constituent securities.
Even a relatively small cash position can produce a performance difference when markets move significantly.
For example, if an index rises sharply while part of an ETF's portfolio remains in cash, the ETF may gain slightly less than the benchmark.
Dividend Differences
Dividends can also contribute to tracking differences.
An index may use specific assumptions regarding how dividends are treated. Depending on the index methodology, it may be a price-return index, total-return index, or another form of benchmark.
An ETF receives actual dividends from its holdings. The timing of dividend payments, reinvestment, withholding taxes, and other factors can affect performance.
Investors should therefore compare an ETF with the correct version of its benchmark.
Comparing an ETF's total return with a price-only index, for example, can create a misleading impression of tracking performance.
Taxes and Withholding Taxes
Taxes can be another source of tracking difference, particularly for international ETFs.
When companies in foreign countries pay dividends, withholding taxes may apply before the ETF receives the income.
The benchmark may reflect a particular tax assumption, while the ETF experiences actual taxes based on its structure and the countries where its investments are held.
Tax treatment can therefore create differences even when the ETF accurately follows its portfolio methodology.
For investors researching what is an ETF, this is an important point because an ETF's legal structure and investment location can influence its after-tax performance.
Index Rebalancing
Indexes are not always static.
Index providers regularly rebalance their benchmarks. They may change the weights of companies, add new securities, remove others, or make adjustments following corporate events.
An ETF needs to respond to these changes.
The timing of the ETF's trades may not exactly match the theoretical changes used to calculate the index. As a result, temporary tracking differences can occur.
Large rebalancing events can be particularly important when many funds are attempting to trade the same securities at approximately the same time.
Corporate Actions
Corporate actions can create additional complexity.
Examples include stock splits, mergers, acquisitions, rights offerings, spin-offs, and special dividends.
An index provider has rules for dealing with these events. The ETF must implement those rules through actual portfolio transactions.
The difference between the theoretical index calculation and the practical execution of trades can create tracking differences.
These differences are often temporary, but they can contribute to tracking error measurements.
Market Liquidity
Liquidity refers to how easily a security can be bought or sold without significantly affecting its price.
Highly liquid securities generally have narrow bid-ask spreads and large trading volumes. Less liquid securities can be more expensive or difficult to trade.
If an ETF tracks an index containing less liquid assets, such as certain small-cap stocks, emerging-market securities, or bonds, the fund may experience greater trading costs.
This can increase the difference between ETF performance and benchmark performance.
Bid-Ask Spreads
Investors should distinguish between an ETF's tracking error and the trading costs they personally experience.
The ETF itself has a market price that changes throughout the trading day. Investors buying or selling the ETF may pay the ask price or receive the bid price.
The difference between these prices is the bid-ask spread.
A wide spread can increase the cost of entering or exiting an investment.
However, this investor-level trading cost is not necessarily the same thing as the ETF's internal tracking error.
A fund can track its benchmark very closely while having a relatively wide trading spread, particularly if it has low trading volume.
Premiums and Discounts to Net Asset Value
Another factor investors should understand is the relationship between an ETF's market price and its net asset value, or NAV.
NAV represents the value of the securities held by the fund, calculated according to the fund's valuation methodology.
An ETF's market price can trade slightly above or below its NAV.
When an ETF trades at a premium, its market price is higher than its NAV. When it trades at a discount, its market price is lower.
These differences can occur because ETF shares trade continuously while the underlying securities may trade at different times or have different levels of liquidity.
This is especially relevant for international ETFs and bond ETFs.
Currency Movements
Currency fluctuations can create tracking differences for international ETFs.
Suppose an ETF holds Japanese stocks but is priced in U.S. dollars. The ETF's return for a U.S. investor depends not only on the performance of Japanese stocks but also on movements in the yen relative to the dollar.
If the benchmark is calculated using a particular currency methodology, differences can arise from currency conversion and timing.
Currency-hedged ETFs introduce another layer of complexity because the hedging process itself has costs and can produce differences from the unhedged benchmark.
Securities Lending
Some ETFs lend securities to other market participants.
The ETF may receive lending revenue in return. This revenue can help offset fund expenses and may reduce the ETF's tracking difference.
However, securities lending involves operational and counterparty considerations.
The impact depends on how much revenue the fund generates and how that revenue is handled.
Therefore, two ETFs with similar expense ratios may still produce different results because their securities-lending arrangements differ.
Replication Methods and Their Impact
ETF providers generally use different methods to replicate their benchmarks.
Full Replication
With full replication, the ETF attempts to hold all or nearly all of the securities in the index in their appropriate weights.
This can provide close exposure to the benchmark, particularly when the index contains liquid securities.
However, full replication may become expensive when an index contains a very large number of securities or securities that are difficult to trade.
Representative Sampling
Sampling involves holding a smaller selection of securities that are designed to represent the broader benchmark.
It can reduce trading costs and simplify portfolio management.
However, the portfolio's performance may differ from the full index because it does not hold every component.
Synthetic Replication
Some ETFs use derivatives, such as swaps, to obtain exposure to an index.
Instead of directly holding every underlying asset, the fund may enter into an agreement designed to deliver the benchmark's performance.
Synthetic structures can reduce certain forms of tracking difference, but they introduce other considerations, including counterparty and structural risks.
Investors should review the fund's documentation to understand how the ETF obtains its exposure.
How Is Tracking Error Measured?
Tracking error is commonly calculated using the variability of the difference between the ETF's returns and benchmark returns.
In simplified terms, analysts examine the ETF's return for each period, subtract the benchmark's return, and evaluate how much those differences fluctuate.
A lower statistical tracking error generally means the ETF's returns have stayed closer to the benchmark over the measured period.
However, investors should not examine the number in isolation.
The measurement period matters. An ETF could have low tracking error during a calm market but experience greater differences during a period of extreme volatility.
The benchmark itself also matters.
Tracking Difference vs. Tracking Error
These terms are often confused.
Tracking difference describes the gap between the ETF's performance and the benchmark's performance.
Tracking error describes the variability of that gap.
For example, an ETF might underperform its benchmark by approximately 0.25% every year. Its tracking difference is consistently negative, but the statistical tracking error could be relatively small because the difference is stable.
Another ETF might outperform the benchmark by 0.5% one year and underperform by 0.7% the next. Its average difference might appear modest, but its tracking error could be larger because the differences fluctuate more.
Both measurements can provide useful information.
How Investors Can Evaluate Tracking Performance
When researching an ETF, investors can examine its historical performance against the appropriate benchmark.
The fund's prospectus and official documentation may provide information about its investment strategy, benchmark, expenses, and replication method.
Investors can also review historical returns over different periods rather than relying on one short-term measurement.
It is useful to consider whether differences are caused by normal expenses or by unusually large deviations.
For example, a modest and consistent gap may be explained primarily by the fund's expense ratio. A sudden large deviation may require further investigation.
Other Factors to Consider When Comparing ETFs
Tracking performance is important, but it is not the only factor that matters.
Investors may also examine the ETF's expense ratio, liquidity, bid-ask spread, assets under management, portfolio composition, benchmark methodology, tax considerations, and fund structure.
An ETF with a slightly different expense ratio may serve a different purpose or use a different replication method.
The goal is to understand how the fund operates rather than relying on one statistic.
Investors should also remember that past tracking performance does not guarantee future results.
Conclusion
Understanding what is an ETF requires looking beyond the simple idea of a fund that follows an index. An ETF operates in the real world, where trading costs, taxes, portfolio adjustments, liquidity, cash balances, dividends, currency movements, and market conditions can all affect performance.
Tracking error occurs because an ETF cannot always reproduce the theoretical performance of its benchmark perfectly. Expense ratios and transaction costs are among the most common causes, while sampling strategies, rebalancing, corporate actions, taxes, currency movements, and liquidity can create additional differences.
The distinction between tracking difference and statistical tracking error is also important. A fund may consistently lag its benchmark by a small amount without experiencing large fluctuations in that difference. Another fund may have a similar average gap but much greater variation.
When evaluating an ETF, investors can review its benchmark, historical performance, expenses, replication strategy, liquidity, and fund structure. Looking at several factors together provides a clearer picture than focusing on one number.
Ultimately, knowing what is an ETF and understanding why tracking differences occur can help investors interpret ETF performance more accurately. A small difference from a benchmark does not automatically indicate a problem. Instead, investors should consider the reason for the difference, how consistently it occurs, and whether it is reasonable given the ETF's strategy and operating costs.