Introduction

Identifying financial entities is quite a challenging task when designing applications that aggregate content from different vendors.

There are several ways of identifying a financial instrument that largely vary on the scope of the standard that describes them. There are international, local, trading venue specific and proprietary codes.

For instance, below codes are assigned to Apple Inc., an American technology giant, and its ordinary shares:

Scheme Code Scope
LEI HWUPKR0MPOU8FGXBT394 International
Business register RA000598 National, United States
ISIN US0378331005 International
SEDOL 2046251 National, United Kingdom and Ireland
CUSIP 037833100 National, United States
Valor 908440 National, Switzerland
Ticker AAPL Exchange-specific
RIC AAPL.O Quote-level, proprietary

If you are a Thomson Reuters Eikon user, type in AAPL.O CODES into the command bar, press Enter, and you will see something like this:

You have probably noticed that there is also another code scheme that I have not mentioned yet, but it is on the screenshot, Thomson Reuters Permanent Identifier or, simply, PermID.

So, what is PermID?

PermID is a open unique identifier developed by Thomson Reuters. Technically it is a 64-bit number that we use to identify an different kinds of objects, for example:

Type PermID Label
Organization 4295905573 Apple Inc.
Instrument 8590932301 Apple Ordinary Shares
Asset class 300281 Ordinary Shares
Quote 55838974642 Apple Ordinary Shares, NASDAQ
Person 34413400240 Craig Federighi
Currency 500110 USD
Industry code 4294952723 Economic Sector:Technology

PermIDs are distributed through permid.org where you can download the whole dataset as a series of flat files (n-triples, turtle formats) or use one of the APIs that is most suitable for your needs.

I am going to show how to use the matching functionality in Python.

PermID API

In order to run the code samples, you will need to obtain an API key either through the Developer Community portal or registering at permid.org.

Record matching API

In order to use the record matching API, we need to create a standard CSV file for input. Since in this example we will be matching organizations, here is what the template will look like:

LocalID Standard Identifier Name Country Street City PostalCode State Website
1 Ticker:AAPL&&MIC:XNGS              
2 LEI:549300561UZND4C7B569            

As you can see, we can use some standard identifiers such as: Ticker + MIC, LEI, etc.

This is how we will create our request, assuming that input.csv contains the above table:

url = 'https://api.thomsonreuters.com/permid/match/file'

headers = {
        'x-ag-access-token':'your_access_token',
        'x-openmatch-numberOfMatchesPerRecord':'1',
        'x-openmatch-dataType':'Organization',
        'Accept':'application/json'
}

with open('input.csv', 'rb') as input_file:
        content = input_file.read()

files = {'file': ('input.csv', content, 'application/vnd.ms-excel')}

response = requests.post(url=url, headers=headers, files=files)

Let's dump the results

data = json.loads(response.text)
if  result:
        header_list = list((result[0]).keys())
        writer.writerow(header_list)

        for item in result:
                writer.writerow(list(item.values()))

After running the tool, you should be able to see this:

Uploaded filename: input.csv
Request time:      168 ms
Status:            Success
Number of received records:           9
Number of processed records:          9
Number of error records:              0
Number of unmatched records:          0
Number of matched records, total:     9

Result

With these results you can easily build a console tool, as well as integrate matching into your data mapping pipeline.