What different types of replication are used in ETFs and which should you choose for your investment? This article provides answers to these and other questions.
Exchange traded funds (ETFs) are simple and very cost efficient products to invest in the capital market. Instead of a traditional active fund manager performing detailed analysis and picking single securities, ETFs try to replicate their predefined index and replicate its performance as closely as possible. The index sets the rules for how securities are selected and weighted.
The technique used by an ETF to replicate the performance of the underlying index is called the replication method. There are two basic types of replication methods and a combination of both: physical, synthetic and hybrid. The method chosen affects, among other things, how closely the ETF tracks the index (tracking error) and how much the ETF underperforms or outperforms its index over time (tracking difference). Understanding the differences in replication methods can help investors select the most efficient ETF for their needs, taking into account their investment objectives and risk tolerance.
With physical replication, the ETF replicates the performance of an index by directly buying the securities contained in the index. This approach is used by the majority of ETFs on the market. Physical replication can be realised either through full replication or through a sampling approach.
In full replication, an ETF holds every security in the index in the exact proportions as the index. This method is favoured for indices with fewer components or those with highly liquid and accessible securities.
Sampling is used when full replication is impractical, for example due to the lower size or liquidity of the underlying securities in the index. Instead of holding every security directly, the ETF holds a representative subset of securities that statistically mimics the index's performance.
The sampling approach is commonly used, especially for Fixed Income and Small Cap ETFs. The ETF issuer is responsible for defining the right securities which will be included in the ETF. Typically the top positions of the index are replicated directly with the same weights as in the index. The smaller index constituents are then filtered based on liquidity criteria and security characteristics.
Investors can benefit from lower trading cost within the ETF, while getting approximately the same performance as the index. Although there are differences between ETF holdings and index constituents, the impact on the total ETF performance is typically very low.
High transparency - Investors can clearly see what securities are held within the ETF. This transparency builds trust and allows for straightforward performance analysis.
Accurate tracking - Physical replication is known for its high degree of accuracy in tracking the index. By holding the same securities in the same proportions, the ETF’s performance closely matches the index’s performance.
Direct ownership - Physical replication avoids counterparty risk since it involves direct ownership of securities. More about this under 3.
Voting rights for equities - Because the ETF has direct ownership of the equity securities, it can exercise its voting rights at the Annual General Meeting.
Higher costs - The costs associated with buying and managing a large number of individual securities can be significant. This includes transaction costs, custody fees, and potential financial transaction taxes. ETFs using the sampling approach try to reduce these costs.
Liquidity constraints - Large transactions required to replicate the index can influence the market prices of the underlying securities, in particular for smaller companies or issuances. For indices with less liquid securities, full replication can become very difficult or even impossible.
Complex management - Managing a portfolio of numerous securities can be complex, especially for broad global indices with small cap tilts. This can lead to operational challenges and increased management costs.
Synthetic replication is an indirect method to replicate the index performance. Instead of buying the securities or assets directly, the ETF uses financial derivatives such as Swaps or Futures.
The ETF provider concludes a swap agreement with a major investment bank. The investment bank guarantees to deliver the index performance, including distributions such as dividends or interest payments, directly to the ETF. This means that the exchange-traded index fund does not necessarily contain the physical securities of the index, but can still realise its return. This saves transaction costs for investors.
Synthetic replication inherents counterparty risk. In the unlikely event of insolvency of the investment bank, it can no longer deliver the index performance to the ETF. To protect ETF investors, the EU has set very strict requirements for regulated Swap ETFs. Those Swap ETFs need to have a collateral basket of securities as protection against such adverse events. The collateral basket is delivered from the investment bank to the ETF. The size needs to be minimum 90% of the ETF’s Net Asset Value (the cumulative value of all securities) by law and has to be monitored daily. In many cases, individual Swap agreements between ETF issuer and investment banks are even stricter. These strict requirements and close monitoring of collateral from the ETF issuer makes sure that Swap ETFs are considered as safe as physical ETFs.
Tax-efficiency - In some cases, synthetic ETFs pay lower taxes than physical ETFs. This can lead to a structural outperformance for investors.
Access to hard-to-reach markets - This method allows exposure to markets or asset classes that might be difficult or costly to access physically, such as commodities.
Efficient tracking - Synthetic ETFs can provide highly accurate tracking of their indices due to the nature of derivatives, which can precisely replicate index returns (lower tracking error).
Counterparty risk - If the financial institution providing the swap defaults, the ETF may not receive the full promised index performance. Individual swap agreements and active management of the collateral basket reduces such risks.
Reduced transparency - Investors have less visibility into the ETF's underlying assets, as they only see the collateral basket as ETF holdings rather than the actual securities being invested.
Swap costs - Although often lower than the costs of physical replication, the fees associated with swap agreements can affect the ETF’s overall return, particularly if market conditions change unfavourably.
Hybrid replication is a method that combines elements of both physical and synthetic replication to track the performance of an index. This approach allows ETF providers to leverage the benefits of both methods while mitigating some of their respective drawbacks. Hybrid replicating ETFs directly hold a portion of the actual securities of the index for some exposures and use derivatives like swaps for others. This blend allows the ETF to choose the best replication method for each exposure individually and flexibly over time, which leads to optimised cost and maximum efficiency.
Example
A hybrid replicating ETF that aims to replicate the MSCI ACWI world index can utilise the advantages of physical and synthetic replication depending on the region. For US and China A-shares, for example, the ETF can utilise a total return swap, as this reduces the withholding tax on dividends from US companies to 0%. Chinese A-shares, on the other hand, could benefit from the ongoing structural dislocation in the market. Other parts of the world are physically replicated, as physical replication is more efficient for these regions and so-called securities lending can even lead to increased performance. Securities held by the ETF are lent out for a fee. Overall, hybrid ETFs thus have the opportunity to even outperform the replicated index.
Balanced approach - By combining physical and synthetic methods, hybrid replication can optimise the trade-offs between transparency and performance. This flexibility allows ETF managers to adjust their replication strategies as market conditions change.
Cost efficiency & tracking accuracy - Hybrid replication can potentially lower transaction costs compared to full physical replication, especially in markets where certain securities are expensive or illiquid. The use of derivatives can also enhance the tracking accuracy of the ETF, ensuring that it closely follows the index performance even if not all securities are held physically.
Risk mitigation - Holding some physical securities reduces counterparty risk compared to a fully synthetic ETF, while the use of swaps can reduce the overall cost and improve liquidity.
Higher complexity - Managing a hybrid portfolio can be more complex than using a singular replication method. Balancing the physical and synthetic components requires sophisticated risk management and operational strategies.
Lower transparency - While more transparent than fully synthetic ETFs, the hybrid approach can still obscure parts of the portfolio, particularly the synthetic portion, making it less transparent than full physical replication.
Exposure to counterparty risk - Although mitigated, there is still exposure to counterparty risk inherent in the synthetic portion of the ETF.
Markus Kirchler
Markus is a Senior Analyst in the Capital Markets team. His main areas of expertise are portfolio management and ETP trading. He holds a Bachelor's degree in Economics from the Free University of Bolzano.