Skip to content
All library documents

Matching Cross-Listed Companies with CUSIP and WKN Identifiers

Article Quant Q&A · Author: moezelot
Read the original

Summary

The question concerns linking records for German companies listed in both Germany and the United States across datasets that identify securities with WKN and CUSIP codes. Matching company names and other record features can help, but the author asks whether exact identifier matches or a reference database can reliably connect the securities across exchanges, including for historical records.

An example using Volkswagen illustrates a limitation in the attempted lookup: OpenFIGI returns security details for the German WKN, while the queried CUSIP for the non-US company returns no result. This suggests identifier coverage can vary by issuer and identifier format, so a failed lookup does not by itself show that two records refer to different companies. The post raises a data-linkage problem rather than establishing a complete matching method or identifying a no-cost database that solves it.

Key ideas

  • The task is to link records for the same company across German and US listings.
  • Company names and other record features can support probabilistic matching.
  • Exact WKN and CUSIP matching may be useful when the identifiers are available and correctly formatted.
  • The example shows that identifier lookup coverage can differ across markets and issuers.
  • A missing database result is not proof that two records represent different companies.

Tags

EquitiesUS marketsStatistics
Full text
# Identifying the same company behind stocks traded in different exchanges (e.g. CUSIP and WKN in 1990ies)


# Identifying the same company behind stocks traded in different exchanges (e.g. CUSIP and WKN in 1990ies)












I would like to merge two data sources with companies headquartered in Germany between 1980-2000, but listed in Germany and the US. One source identifies "German companies" by their CUSIP, the other one by their WKN. There are names in both data sets, so I can approximately match on name (+ record link on other features). But I thought I could improve on that by matching exactly on (6dig) WKNs and (6dig) CUSIPs?! I guess I would need a resource which has linked stocks of the same company traded in different exchanges? For instance, I am looking for this correspondence

```
Volkswagen AG | WKN = 766400 | CUSIP = 928662
```

Academia no-cost solution required. :/ Thanks.

Edit:

There is OPEN FIGI. But they seem to do poorly on CUSIPs of companies not based in the US (adding 10 or 30 + check digit to CUSIP).

```
curl 'https://api.openfigi.com/v2/mapping' \
--request POST \
--header 'Content-Type: application/json' \
--data '[{"idType":"ID_CUSIP","idValue":"928662303"}]'
```

Returns `[{"error":"No identifier found."}]`

But `--data '[{"idType":"ID_WERTPAPIER","idValue":"766400"}]'` returns an ID, company name and ticker useful for the matching:

```
[{"data":[{"figi":"BBG000BCH458","name":"VOLKSWAGEN AG","ticker":"VOW","exchCode":"GR","compositeFIGI":"BBG000BCH458","uniqueID":"EQ0011575700001000","securityType":"Common Stock","marketSector":"Equity","shareClassFIGI":"BBG001S67CB7","uniqueIDFutOpt":null,"securityType2":"Common Stock","securityDescription":"VOW"},{"figi":"BBG000BCH4V9","name":"VOLKSWAGEN AG",
```

Any other DB which can identify multinationals (not headquartered in the US) by their CUSIP?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.