What an ETF Overlap Result Cannot Tell You from Partial Holdings

A holdings tool can calculate a precise percentage from incomplete information. Precision in the arithmetic does not make the underlying dataset complete. Before interpreting an ETF overlap result, find out how much of each fund the supplied list covers, whether the weights are original fund weights and whether both lists refer to comparable dates.

This case shows why a short list of major holdings should not silently become a complete fund. It uses fictional companies and assumes positive weights in unleveraged portfolios. The goal is to make uncertainty visible without inventing the missing holdings or replacing unknown weights with guesses.

Preserve the original weights

The supplied portion of Fund A contains 12% AAA, 8% BBB and 5% CCC. It covers 25% of that fund. The supplied portion of Fund B contains 10% AAA, 6% DDD and 4% EEE, covering 20%. These are the original percentages of each full fund, not percentages of the supplied lists.

The only observed shared company is AAA. Its contribution to a minimum weight overlap calculation is the smaller of 12% and 10%, which is 10 percentage points. The observed lists therefore establish at least that amount of shared holding weight under the stated calculation. They do not establish that the complete funds have only 10% overlap.

Other shared positions may be missing from both lists or appear in one list but not the other. There is no honest way to label the unobserved portion as zero simply because it was not supplied. An absence from the input is a data coverage issue, not evidence that the fund has no exposure.

See what normalization changes

Some tools rescale each supplied list to total 100%. In Fund A’s observed subset, AAA becomes 48%, BBB 32% and CCC 20%. In Fund B’s subset, AAA becomes 50%, DDD 30% and EEE 20%. The overlap between those normalized subsets is then 48%.

That 48% can be a valid description of the normalized supplied baskets. It is not a measured overlap of the complete funds. It answers a different question because the denominators changed from total fund assets to the listed portions. Presenting it without a coverage explanation would create a much stronger impression of knowledge than the inputs justify.

Keep a useful lower bound separate from an estimate

With original full fund weights, the observed 10 percentage points in AAA is a known contribution to overlap in this fictional example. Full overlap could be higher. The available information does not identify a unique complete result, so do not fill the gap by averaging the coverage percentages or assuming the missing holdings resemble the visible ones.

Even a formal upper bound would depend on assumptions about whether the lists are exclusive subsets, how identifiers match and whether the remaining positions could include names already seen in the other fund. A broad bound can be mathematically valid and still too uninformative for a portfolio decision. Sometimes the correct next step is better data, not a more elaborate score.

Check dates and definitions together

Two complete holdings lists from materially different dates may also give a misleading comparison. A fund can rebalance, receive cash, change an index or adjust positions between observations. The calculated result describes those two supplied snapshots; it does not automatically describe the current funds.

Read the provider’s treatment of cash, derivatives, short positions and share classes. A simple long only holding overlap formula is not a complete economic exposure model. An instrument label can refer to a derivative exposure rather than direct ownership, and a negative weight requires different interpretation from a positive stock holding.

Use the analyzer with an explicit coverage note

The overlap analyzer provides a fictional walkthrough that separates supplied baskets from real fund data. When entering your own information, retain the source date and the coverage percentage. If normalization is used, describe the output as a comparison of normalized supplied holdings. Do not relabel it as a verified current fund overlap.

After obtaining complete compatible lists, calculate combined portfolio company weights using your allocations. Pairwise fund overlap and investor level concentration remain different measurements even when both inputs are complete. The combined exposure case study works through that second calculation with explicit dollars and weights.

Know when to stop calculating

If the result could change substantially when missing holdings arrive, treat it as provisional. A larger font, extra decimal place or attractive chart cannot fix missing coverage. Record the unanswered question, identify the information needed to resolve it and avoid making an allocation decision solely from an incomplete score.

For broader context, read how to compare index funds. Costs, tracking, coverage and concentration are separate dimensions. A useful research workflow makes those dimensions visible and permits an unresolved answer when the evidence is incomplete. That is more valuable than turning every input into a definitive percentage.