How do I calculate the holdings overlap between two ETFs?

The formula

For two funds A and B with holdings weights wA[name] and wB[name] in percent:

text
overlap(A, B) = Σ_{name ∈ A ∩ B} min(wA[name], wB[name])

An overlap of 40 means 40% of each portfolio's weight sits in names the other fund also holds (at the smaller of the two weights). Two share classes of the same index fund overlap near 100; a world ETF and an emerging-markets ETF near 0.

From the payload

python
import os, requests
H = {"Authorization": f"Bearer {os.environ['FUNDFACTS_API_KEY']}"}
def holdings(isin):
    d = requests.get(f"https://fundfactsapi.com/api/v1/funds/{isin}", headers=H, timeout=300).json()["data"]
    return {h["name"].upper(): h["weight"] for h in d["topHoldings"]}
​
a, b = holdings("IE00B4L5Y983"), holdings("IE00B3RBWM25")
shared = {n: min(a[n], b[n]) for n in a.keys() & b.keys()}
print(round(sum(shared.values()), 2), "% overlap over published holdings;", len(shared), "shared names")

This is exact over the positions stored in topHoldings (the ten largest for most funds, up to 50 for some issuers, even where the issuer publishes a full holdings file) and is what the endpoint does, with better name matching. Over top lists it is a lower bound on the true overlap.

The endpoint

bash
curl -s "https://fundfactsapi.com/api/v1/overlap?isins=IE00B4L5Y983,IE00B3RBWM25" -H "Authorization: Bearer $FUNDFACTS_API_KEY"

json
{ "pairs": [ { "a": "IE00B4L5Y983", "b": "IE00B3RBWM25", "overlap": 18.4, "sharedCount": 9, "disclosed": { "a": 24.9, "b": 21.7 }, "shared": [ { "label": "NVIDIA", "a": 5.48, "b": 4.61 } ] } ] }

disclosed states what share of each portfolio the stored holdings cover, so you can tell a near-complete comparison (both close to 100) from a top-ten lower bound like this one. Several ISINs return every pair. One request per ISIN; Starter and above. The MCP server's fund_overlap tool wraps it for assistants.

Reading the number

Overlap is about names, not about risk: two funds with 30% overlap can still be 95% correlated because they hold the same sectors and countries. Pair it with the sector and geography tables, or with POST /portfolio for the aggregate view. How to calculate ETF holdings overlap goes deeper into matching and edge cases.