Many investors and even some advisors assume that two exchange traded funds (ETFs) tracking the same index should produce nearly identical returns. After all, if both funds follow the S&P 500 or the TSX 60, performance should be virtually interchangeable aside from minor fee differences.
But that is often not the case.
Small structural and operational differences between ETFs can compound meaningfully over time, creating noticeable performance gaps — even when funds track the same benchmark. Understanding these mechanics has become increasingly important for advisors as ETFs continue to attract investor dollars.
One of the most misunderstood concepts is the difference between tracking error and tracking difference.
Tracking error measures the consistency of an ETF’s returns relative to its benchmark. Tracking difference reflects the actual performance gap between the ETF and the index over time. An ETF can have low tracking error — meaning it closely follows the index day-to-day — while still underperforming the benchmark consistently due to fees, taxes or operational drag.
Management expense ratios (MERs) are the most obvious factor, but they are far from the only one.
A useful U.S.-based example is the long-term comparison between the State Street SPDR S&P 500 ETF Trust (SPY) and the Vanguard S&P 500 ETF (VOO), two ETFs that track the same index. While they provide virtually identical market exposure, VOO has historically outperformed SPY over time. The reasons are structural.
SPY operates as a unit investment trust, an older ETF structure that cannot reinvest dividends before distribution. As dividends accumulate, cash sits idle until quarterly payouts occur, creating what is commonly referred to as cash drag.
VOO operates under the more flexible ’40 Act ETF structure – which is regulated by the U.S. Securities and Exchange Commission under the Investment Company Act of 1940. It allows dividends to be reinvested immediately, reducing unproductive cash balances. Combined with VOO’s lower MER, the result is a small but consistent performance advantage that compounds over time.
Canada provides an equally compelling example of how ETF structure can materially impact investor outcomes.
The BMO S&P 500 Index Series Units ETF directly holds the underlying S&P 500 stocks, so BMO is responsible for keeping the portfolio aligned with the index as constituents are added or removed.
The Vanguard S&P 500 Index ETF (VFV), by contrast, gets its exposure by holding the U.S.-listed Vanguard S&P ETF (VOO). That fund-of-funds structure means Canadian investors are effectively relying on VOO to manage the underlying S&P 500 portfolio correctly, while Vanguard Canada’s role is mainly to package that exposure in a Canadian-listed wrapper.
That does not make VFV worse, but it does add one extra operational layer between the Canadian investor and the index. Direct replication can be cleaner because the Canadian ETF owns the index securities itself, while the fund-of-funds model depends on the underlying ETF’s portfolio management, trading, rebalancing and tracking. While direct replication may appear cleaner conceptually, investors should evaluate actual tracking outcomes over time rather than making decision based solely on structural differences.
Currency and methodology
Currency management can also create meaningful tracking differences. The CAD-hedged version of the Vanguard S&P 500 ETF (VSP) tracks the same underlying index as unhedged S&P 500 ETFs, but its returns can diverge materially over time due to the costs, implementation and effectiveness of maintaining a Canadian-dollar currency hedge.
As a result, sources of tracking difference in Canada often extend well beyond MERs and can include foreign withholding taxes, cash drag and currency hedging expenses.
Securities lending is another important, but often overlooked, driver of returns.
Many ETF providers lend portfolio securities to short sellers or institutional counterparties in exchange for a fee, generating incremental revenue that can help offset fund expenses and improve performance.
In some cases, securities lending revenue can partially or fully offset an ETF’s MER. However, securities lending programs differ significantly across issuers in terms of scale, collateral policies, revenue sharing and risk controls.
Although securities lending can enhance returns, investors should also understand the associated counterparty and operational risks, which are managed differently across ETF providers.
Portfolio construction methodology also matters.
Some ETFs fully replicate an index by holding every underlying security in its exact weighting, while others use sampling techniques, holding only a representative subset of securities.
Sampling can reduce transaction costs and improve operational efficiency, particularly in fixed income or international markets. But it can also increase tracking differences if the sample fails to perfectly mirror benchmark performance.
Trading costs such as bid-ask spreads can also affect investor outcomes, particularly for less liquid ETFs, even when long-term tracking characteristics are similar.
In an increasingly competitive ETF landscape, implementation details matter. Two ETFs may track the same index on paper, but deliver materially different investor outcomes over time. Understanding what sits beneath the ticker symbol is essential — because when it comes to ETFs, how exposure is delivered can be just as important as the exposure itself.
Eli Yufest is executive director of the Canadian Exchange Traded Funds Association (CETFA).