• The fund hands over the returns of its stock basket to the bank.
• In return, the bank guarantees the exact performance of the index the ETF is tracking.

One key advantage of this approach is that dividends are not directly received by the fund but instead go through the bank, allowing tax-efficient treatment in certain cases. For example, in a swap-based S&P 500 ETF, the usual 15% U.S. dividend withholding tax can be avoided. Given an average dividend yield of 1.27%, this results in an annual outperformance of approximately 0.2% compared to a physically replicated ETF.

The main criticism of swap-based ETFs is counterparty risk, meaning that if the bank defaults, the ETF might not receive the promised returns. However, this risk is mitigated by EU regulations, which limit counterparty exposure to 10% of the fund's value and require collateral. Moreover, if the bank were to fail, investors who hold broad market ETFs already own shares of the same bank, making the additional risk less significant. Some funds combine this approach with direct stock holdings. See Hybrid replication. | | Physical replication | The fund manager buys the underlying stocks directly, instead of using a swap. For narrow indices, this usually means holding every constituent in the same proportion as the index ("full replication"). Most providers use sampling or optimization instead: holding a representative subset that closely tracks the index while avoiding the cost of trading small, illiquid positions. Physical replication is generally more transparent than swap-based replication, though trading costs and dividend withholding tax (see Domicile) can create a small drag versus the index. | | Hybrid replication | A combination of physical and swap-based replication within the same fund. Typically this means holding stocks directly where physical replication is efficient (e.g. developed markets), and using swaps where the tax benefit matters most (e.g. US equities, due to dividend withholding tax) or where direct access is harder (e.g. certain emerging markets). The goal is to combine the transparency of physical holdings with the tax efficiency of swaps, without taking on full swap counterparty exposure. See Swap-based replication for how the swap mechanism works and its risk limits. | | TER, Total Expense Ratio | The all-in fee charged annually by the fund manager.

If you invest 10 000€ in an ETF with a 0.07% TER, it means 7€ per year is deducted as a management fee. This includes payments to the index company. | | Domicile | The domicile of the fund can be seen from the ISIN code (IE = Ireland, LU = Luxembourg).

The domicile matters when the following criteria are met:

• The fund contains US companies.
• The fund has a physical replication method.

Ireland has a tax agreement with the US, which means it pays a 15% withholding tax, whereas other European countries pay a 30% tax.

If 60% of an All‑World fund is in the S&P 500 and the dividend yield is 1.15%, then reducing US withholding tax by 15 percentage points increases the fund’s total return by about 0.10% per year. | | AUM, Assets Under Manage-ment | Size of the fund - What is the value of the funds’ assets in money. AUM can grow through:

• people investing more money into the fund (ETF provider issues more shares)
• the value of assets is increasing.

Especially new and unpopular ETFs have low AUM, which means there is a risk that the fund provider will end / liquidate the ETF. For new ETFs, ExtraETF.com under “Chart” > “Fundsvolumen” or “Anzahl Fondsanteile” (number of shares), you can see how the popularity is developing. If it is going upwards - then the fund provider will most likely keep operating the ETF. | | Tracking Difference | The tracking difference is the key performance indicator of an ETF. Since an ETF aims to replicate an index, a good ETF may slightly outperform its benchmark through sampling, share lending, or optimization.

Tracking difference can be measured in two ways:

• Absolute tracking difference: The direct difference between the ETF and index performance.
• Relative tracking difference: The percentage deviation from the index.

Example: ETF performance: 24.7% Index performance: 24.5% Absolute tracking difference: +0.2% Relative tracking difference: (24.7 / 24.5) - 1 = +0.82%

A positive tracking difference means the ETF has outperformed its index, while a negative value indicates underperformance.

The comparison should be done of the Net Total Return (NTR) of the index, which assumes withholding taxes of the dividends. | | Spread /

Xetra Liquidity Measure (XLM) | The difference between the bid (buy) and ask (sell) price - the cost of trading immediately, paid each time you buy or sell. Smaller and newer ETFs generally have wider spreads, since there's less trading activity keeping pricing tight. Deutsche Börse publishes a more precise, order-size-adjusted version of this for Xetra-listed ETFs, the Xetra Liquidity Measure (XLM). For long-term monthly investors this is a minor cost - it applies only to the amount you're investing that month, not your whole balance (unlike TER, which applies every year to everything you hold) — and most automated savings plans execute away from the exchange this figure describes anyway. Spread is most useful as a signal of fund maturity: a wide spread on a new or small ETF flags thin liquidity, not necessarily a bad fund, just an unproven one. | | Transaction Costs (in KID) | An estimate of trading costs from portfolio changes, not a fixed fee. The methodology is a known weak point of EU rules: it can't cleanly separate real trading cost from normal market movement, and can produce negative values. EU rules floor these at zero, so a 0.00% figure may just mean flooring, not genuinely low cost. This mainly affects physically replicated funds; swap-based funds tend to score low here because the fund only trades a substitute basket (often lower-turnover, not required to match the index), while the actual index return comes through a derivative that doesn't count as a portfolio transaction under PRIIPs.

The KID's itemized table (ongoing, transaction, incidental costs as separate rows) is reliable read at that level. Be wary of any bundled total, in the KID's "Annual Cost Impact" or on a provider's website, since it mixes one stable figure with one volatile estimate. Don't add it to another fund's plain TER.

Once a fund has a few years of history, Tracking Difference is a better real-world measure of trading impact than any of these disclosures. |