Physical vs. Synthetic Replication
Does the fund actually buy the underlying securities, or use a swap to deliver the index's return instead?
Physical replication means the fund genuinely owns the underlying securities — a physical S&P 500 ETF actually holds shares of Apple, Microsoft, and the rest, in roughly the index's weights (sometimes a representative sample rather than every single name, for very large or illiquid indexes).
Synthetic replication instead uses a total return swap with a counterparty bank: the fund holds a different collateral basket, and the bank contractually pays the fund the exact return of the target index in exchange for a fee. The investor still gets the index's return — just via a contract instead of direct ownership.
Physical
1
The Fund
2
Actual Holdings
Owns Apple, Microsoft, etc. directly
Synthetic
1
The Fund
Holds unrelated collateral
2
Swap Bank
Contractually pays the index's return
The real tradeoff
| Physical | Synthetic | |
|---|---|---|
| What you're exposed to | The actual underlying securities | A bank's contractual promise, plus its collateral |
| Extra risk introduced | Securities lending risk (if the fund lends out holdings) | Counterparty risk — the swap bank's own creditworthiness |
| Common use case | Most mainstream equity/bond index funds | Some hard-to-physically-access markets, and a few cost-efficient niches |
Key takeaways
- Synthetic isn't inherently unsafe — UCITS rules cap counterparty exposure and most swaps are collateralized daily — but it is a genuinely different risk profile worth knowing you're taking on.
- The fund's factsheet or KID (Key Information Document) will state its replication method plainly — it's not something you have to infer.