How Index Funds Are Built: Replication, Tracking Error, and Securities Lending

Index funds appear simple from the outside: buy the index, hold it, collect market returns. But the mechanics of how a passive fund achieves this — and the genuine differences between funds tracking the same index — are more complex and more important than most investors realise. Choosing between two S&P 500 trackers with TERs of 0.07% and 0.20% is straightforward; understanding why a "cheap" fund might actually underperform a "more expensive" fund, or why one fund lends its holdings to short-sellers, requires a deeper look.

Full Replication

Full physical replication is the most intuitive approach: the fund buys every constituent of the index, weighted according to the index methodology. A fund tracking the FTSE 100 via full replication would hold all 100 constituent shares in their correct proportions.

This method minimises tracking error (the deviation between fund return and index return) because the fund holds exactly what the index specifies. It is most appropriate for indices with a manageable number of liquid constituents — FTSE 100, S&P 500, Nikkei 225 — where every constituent can be bought and sold efficiently at index weights.

Full replication becomes problematic for indices with hundreds or thousands of constituents (MSCI ACWI contains ~3,000 stocks), or indices with illiquid small-cap constituents. Trading costs accumulate, and the fund's market impact when rebalancing can be significant for less liquid names.

Optimised Sampling

Optimised sampling (also called stratified sampling or representative sampling) holds a subset of the index constituents, selected to replicate the index's return characteristics without buying every constituent.

The optimisation process selects holdings based on: sector and country exposures matching the index, factor characteristics (beta, volatility) matching the index, sufficient liquidity in selected holdings, and overall portfolio tracking error minimisation. A fund using optimised sampling to track the MSCI World might hold 800–1,200 of the ~1,500 index constituents.

The trade-off is slightly higher tracking error — the fund will not exactly replicate index returns — but lower trading costs, better liquidity management, and ability to track indices with illiquid constituents. Most large, well-managed passive funds using optimised sampling achieve tracking errors below 0.1–0.2% annualised, which is acceptable for most investors.

Synthetic Replication

Synthetic replication uses financial derivatives — specifically Total Return Swaps (TRS) — to deliver the index return without holding the underlying shares. In a typical unfunded swap structure:

1. The fund deposits collateral (high-quality bonds, government securities) with a swap counterparty (typically a major investment bank).
2. The swap counterparty agrees to pay the fund the total return of the index (dividends plus capital gains).
3. In exchange, the fund pays the counterparty a fee and the return on the collateral.

The fund never holds the actual index constituent shares. The index return is delivered synthetically through the swap.

Advantages of synthetic replication: lower tracking error (the swap is designed to exactly replicate the index), lower or zero withholding tax drag on dividends for certain indices (particularly important for US stocks, where physical funds incur US withholding tax on dividends while synthetic funds may avoid it), and ability to track difficult-to-replicate markets.

Disadvantages: counterparty risk (if the swap counterparty defaults, the fund may not receive promised index returns), limited transparency (investors own derivatives, not shares), and regulatory complexity. UCITS regulations limit counterparty exposure to 10% of fund NAV, providing some protection. In practice this caps the maximum loss from a single counterparty default at around a tenth of the fund's value in a worst case where posted collateral proved entirely unrecoverable — a scenario mitigated further by most providers using multiple swap counterparties simultaneously and posting collateral that is itself marked to market daily, rather than relying on the regulatory limit alone as the sole safeguard.

Synthetic funds became controversial post-2008 due to opacity concerns, and many fund providers switched to physical replication. Synthetic funds remain preferred for indices where dividend tax efficiency materially improves returns — particularly for non-US investors tracking US equities.

Securities Lending

Many physical index funds — including some of the world's largest — lend their holdings to short sellers and other borrowers in exchange for fees. The borrower pays a fee, posts collateral (usually sovereign bonds or cash), and returns the securities on demand.

Securities lending revenue can be meaningful: for an S&P 500 ETF, lending income is modest (US large-cap stocks are easy to borrow, so fees are low — typically 0.01–0.05% annually). For ETFs holding harder-to-borrow securities (small-cap, emerging market stocks, or stocks heavily shorted by hedge funds), lending income can be substantially higher.

This income is typically credited to the fund, reducing its effective total cost of ownership below the stated TER. Vanguard distributes 100% of securities lending income to funds; iShares and SPDR return 62.5–70% to the fund and retain the rest. This is a meaningful difference in how efficiently lending income benefits investors. It matters most when comparing two funds that quote near-identical TERs on the same index: the fund that retains a larger share of lending income for the provider, rather than passing it through, is quietly a more expensive product than its headline fee suggests. This detail sits in the fund's securities lending policy or annual report rather than its factsheet, so it is easy to miss unless it is specifically checked.

Securities lending risk: if the borrower defaults and the collateral is insufficient to buy back the lent securities at current market prices, the fund (and therefore its investors) bears the shortfall. Major fund providers mitigate this with overcollateralisation (collateral exceeds the value of lent securities by 2–10%), high-quality collateral requirements, and counterparty diversification. The overcollateralisation exists precisely because collateral value and the value of the lent securities can move apart between the daily marking cycles that most lending programmes use, so the buffer is there to absorb that gap even in a fast-moving market, not merely as a fixed margin of comfort. No major ETF has suffered material loss from securities lending in practice, but the risk is non-zero, and it is one more reason to check whether a specific fund participates in lending at all before assuming two trackers of the same index carry identical risk profiles.

Tracking Error vs Tracking Difference

These two metrics are frequently confused but measure different things:

Tracking error (also called active risk) measures the annualised standard deviation of the daily return difference between the fund and its index. A high tracking error means daily returns vary significantly from the index — even if the fund is ahead of the index on average. Tracking error is a measure of consistency, not direction.

Tracking difference measures the cumulative gap between the fund's actual return and the index's return over a period (typically one year). This is the all-in cost measure that matters most to investors. A fund with a TER of 0.20% might have a tracking difference of 0.10% if securities lending income more than offsets the management fee. A fund with a TER of 0.07% might have a tracking difference of 0.15% due to transaction costs and dividend withholding tax.

JustETF and ETF.com publish tracking difference data for major ETFs, allowing investors to compare actual historical performance versus stated TERs. For long-term investors, tracking difference is a better cost comparison metric than TER alone.

Market Impact of Passive Investing: The Price Discovery Debate

The growth of passive investing — which by some estimates accounts for over 50% of assets under management in US equities — has prompted serious academic debate about its effects on market efficiency.

The concern: if passive funds simply buy what indices dictate and never perform fundamental analysis, who is left to assess whether individual stocks are priced correctly? If prices are set increasingly by flows into and out of passive vehicles (driven by index inclusion decisions) rather than fundamental value assessment, market efficiency may deteriorate.

The counter-argument: the remaining active management community, hedge funds, arbitrageurs, and informed traders provide sufficient price discovery even as passive share grows. And passive investing's growth largely comes at the expense of expensive, underperforming active management — a social improvement, not a harm.

The empirical evidence suggests that index inclusion effects (the jump in a stock's price when it is added to the S&P 500, for example) have grown as passive AUM has grown, consistent with reduced price efficiency at the margins. Whether this creates investable opportunities for active investors is debated.

For practical purposes, most investors should not change their approach based on market-level efficiency arguments. Index funds remain the most cost-effective, reliable mechanism for accessing broad market returns for the vast majority of investors.

All investments can fall as well as rise. Index fund returns track the relevant index before costs; actual net returns depend on tracking efficiency, costs, and tax treatment. Synthetic fund structures carry counterparty risk. This guide does not constitute personal financial advice.

How Global Investments Can Help

Our team can review your existing index fund allocations to assess tracking difference, securities lending policies, and replication method, ensuring you are accessing market returns as efficiently as possible. We can also advise on the appropriate mix of physical and synthetic replication given your tax position and risk tolerance. Contact us to discuss your portfolio.