Physical replication of ETFs

• Level: For beginners
• Reading duration: 5 minutes

The goal of an ETF is to replicate the performance of an index as efficient and accurate as possible. An ETF with physical replication, also referred to as direct replication or full replication, tracks an index by directly buying the underlying securities of the index. An EURO STOXX 50 ETF, for example, invests in the 50 largest companies in the Eurozone and weights them with the same percentage as the EURO STOXX index does.

Full replication

For large or illiquid indices, this replication method reaches its limits. Such indices are often replicated by the so-called sampling method.

Sampling

With the sampling method, only the most important or liquid securities are purchased because these usually have the greatest influence on the index performance.

The difference: Full replication vs. sampling

Optimized sampling

If the selection is made up using quantitative models, one speaks of optimized sampling. This method of approximation also tries to select only a few securities that have the greatest influence on the index performance. A sampling ETF also invests directly into the selected securities. The advantage of both sampling methods compared to full replication is, that the trading and management costs can be significantly reduced, especially for indices with many securities. One possible drawback is that a deviation between index and ETF performance may be more volatile, especially the more aggressively the underlying portfolio is optimized. Two key figures to measure the tracking quality are tracking error and tracking difference.