ETFs - Replication, Creation Units and Tracking Error

By EC Assets Research Team, Index & Vehicle Research · Published 28 May 2026

ETFs — An exchange-traded fund (ETF) is a pooled investment vehicle whose shares trade on a stock exchange like an individual stock. Most ETFs hold a basket of securities designed to track a published index - a mechanism called the creation/redemption process keeps the share price tightly anchored to net asset value.

Definition

An ETF is a fund structure that combines the diversification of mutual funds with the intraday tradability of stocks. Investors buy and sell ETF shares on an exchange at market-determined prices. Behind the scenes, the ETF holds a portfolio of underlying securities - usually mirroring a published index - and an authorised participant (AP) mechanism keeps the market price aligned to the underlying portfolio value.

Three elements define the structure:

In well-functioning ETFs, market price tracks NAV within a few basis points. The mechanism that enforces this is the creation/redemption process.

The Creation/Redemption Mechanism

Authorised participants (typically large banks and market makers) are licensed to create or redeem large blocks of ETF shares ("creation units", usually 25,000–100,000 shares) directly with the ETF issuer at NAV. This arbitrage is the structural pillar of ETF liquidity:

This arbitrage runs continuously during exchange hours. For deeply liquid ETFs (SPY, QQQ, IWM) the premium/discount is usually 1–3 basis points. For ETFs holding less-liquid underlyings (emerging-market debt, high-yield credit, some commodities) discounts can widen to 50–100 basis points during stress.

Replication Methods

ETFs achieve index exposure through three replication methods:

US and UK ETF markets are predominantly physical. European synthetic ETFs are more common (particularly for emerging-market and total-return swap structures), with strict UCITS rules on counterparty exposure (max 10% of NAV with any single counterparty).

Tracking Error

Tracking error is the standard deviation of the difference between ETF returns and benchmark returns over a chosen window. For a well-run developed-market index ETF, annualised tracking error is typically 5–15 basis points. Five drivers explain most of the variance:

A persistent positive tracking difference (ETF return > index return) suggests securities-lending revenue offsetting fees. A persistent negative tracking difference larger than the management fee suggests operational inefficiency.

ETF Liquidity Has Two Layers

A common mistake is to evaluate ETF liquidity by the on-screen volume of the ETF shares themselves. The actual liquidity has two layers:

1. On-exchange liquidity: Average daily volume of the ETF on the secondary market. Affects retail-sized trades.
2. Underlying liquidity: The aggregate liquidity of the basket the ETF holds. Affects institutional-sized trades, because APs can create new shares in unlimited quantity as long as they can source the basket.

For a 100 million USD institutional trade, the relevant question is not "how much does this ETF trade per day" but "can the AP build the creation basket". This is why an ETF with only 5 million USD daily on-exchange volume can absorb 100 million USD trades without significant market impact - provided the underlying basket is liquid.

When ETFs Stress

Three historical episodes have stressed the ETF arbitrage mechanism:

In each case, the ETF was reflecting genuine illiquidity in the underlying - it was not the ETF "breaking". The discount was real-time price discovery of underlying assets that no longer had two-sided markets.

Why ETFs Matter

ETFs are the dominant access vehicle for index exposure across global capital markets. Total global ETF AUM exceeded 14 trillion USD by 2025, with daily trading volume of roughly 200 billion USD across the major exchanges. Three institutional uses dominate:

1. Core beta exposure: Replacing mutual funds and individual stock baskets at lower fee, with intraday liquidity.
2. Cash management: Equity-index ETFs as a parking vehicle for unallocated cash, replacing T-bills for short-duration equity exposure.
3. Tactical positioning: Sector, factor and thematic ETFs for short-term tilts, easier to size and unwind than equivalent futures or individual stock baskets.

The category continues to evolve: actively managed ETFs, options-based income ETFs, and thematic ETFs now collectively account for over 1 trillion USD AUM. The structure has proved flexible enough to accommodate strategies that two decades ago would only have existed in mutual fund or hedge fund wrappers.

The In-Kind Creation/Redemption Mechanism

ETFs' tax efficiency comes from the in-kind creation/redemption mechanism. The mechanics:

When demand for an ETF rises, Authorised Participants (APs, typically banks) deliver baskets of the underlying securities to the ETF in exchange for new ETF shares. When demand falls, APs deliver ETF shares to the fund in exchange for baskets of underlying securities.

These transactions are in-kind exchanges, not cash sales. Critically, they do not trigger capital gains for the ETF. Compared to a mutual fund that must sell securities to meet redemptions (triggering gains for all remaining shareholders), the ETF structure produces materially better after-tax outcomes.

The ~50-100 bp after-tax advantage compounds substantially over decades.

The Tracking Error Question

[!example] Two S&P 500 ETFs from major issuers: SPY (oldest, $400B AUM) and VOO (newer, $400B AUM, lower fee). Over 5 years, the after-tax returns differ by ~5-15 bp annually despite tracking the same index. The differences come from: expense ratio gap (3-5 bp), securities lending revenue (variable), trading-cost differences, and minor in-kind portfolio composition timing. The cumulative impact over a decade can be 50-100 bp difference between otherwise "identical" products.

Institutional ETF selection considers these subtle differences alongside the obvious factors (size, liquidity, expense ratio). Sophisticated comparison tools (e.g., for trade-by-trade execution) can identify the marginally better choice.

References

1. Gastineau, G. L. (2010). The Exchange-Traded Funds Manual (2nd ed.). Wiley.
2. Hill, J. M., Nadig, D., & Hougan, M. (2015). A Comprehensive Guide to Exchange-Traded Funds. CFA Institute Research Foundation.