Back in 2015 Bloomberg made a formerly in-house financial instrument identifier open for all to use.
ODI’s Head of Marketing and Membership Hannah Foulds chats to Richard Young – Industry Affairs at Bloomberg Global Data to find out why they made this information open, and what the benefits and challenges have been
Hi Richard, what’s your role at Bloomberg?
I sit within a department called Global Data, which sources and provides the data that goes into Bloomberg’s services, such as our terminals and enterprise data feeds. Within Global Data we have a team that is responsible for symbology and key financial reference data (counterparty and security identifiers that are used in trading). Financial reference data identifiers help trading to happen smoothly. To help achieve greater clarity in respect of identifying financial instruments, about ten years ago Bloomberg developed its own financial instrument identifier. In 2015 we turned this identifier into an open standard called FIGI (Financial Instrument Global Identifier). Part of my role is to create greater awareness of FIGI, and how it can best be used in financial markets.
In really simple terms, could you explain what FIGI is?
FIGI is a framework used to assign an identifier (a 12-character random alpha-numeric code) and associated metadata (descriptive information such as the name of the asset) to financial instruments of many different types e.g. equities, bonds, loans, indices, derivatives. FIGI is unique, never-changing, free to use, issue and redistribute. It provides an authoritative method for financial market participants such as; brokers and investment managers to accurately determine and record all of the financial assets they have traded and hold. The most basic thing that needs to happen operationally in financial markets is ensuring that you’ve correctly identified what and with whom you’ve traded – and FIGI helps with the ‘what’. It enables financial market participants to ensure that what they end up with after the trading and post-trade processing is complete is what was expected.
Our clients have always found the FIGI to be helpful, and we decided that it should not just be a standard for Bloomberg services, but also for people who may want to use it for their own operations separately from any Bloomberg offerings.
The reason this is important is that identifying the ‘what’ can be problematic because there are lots of identifier standards which all operate in various ways, and with different levels of associated metadata and granularity (levels of detail). Many of these other standards have usage restrictions and licence fees to pay as well, which can be costly. FIGI as an open data standard does not suffer from these problems, and helps users tie-together other identifier systems which they may be exposed to. Our clients have always found the FIGI to be helpful, and we decided that it should not just be a standard for Bloomberg services, but also for people who may want to use it for their own operations separately from any Bloomberg offerings.
A well respected technical standards body the Object Management Group (OMG) adopted FIGI as a technical financial standard in 2015, and maintains the specification for the FIGI, which can be found on the OMG website. Bloomberg’s role then became the entity that issues and maintains the FIGI, under a public domain guarantee, as Registration Authority and Certified Provider. It is important to note that OMG makes provision for more than one issuer (Certified Provider) of FIGIs, so in the future Bloomberg may not even be the only issuer. In order to provide access to the range of issued FIGIs, and associated metadata, we provided an open website so all codes and data could be easily accessed and consumed. This website is open to all, and allows for various methods by which FIGIs can be downloaded, including in bulk via an open API.
Increasingly we took the view that ring-fencing things like identifiers in a proprietary way or even, as some others do, charging for the use of codes, was not the best way forward for our clients in financial markets.
Why did you decide to make the standard open?
Increasingly we took the view that ring-fencing things like identifiers in a proprietary way or even, as some others do, charging for the use of codes, was not the best way forward for our clients in financial markets. We had a lot of interest from our clients and others wanting to use it beyond Bloomberg services, as it provided a useful method to uniquely identify financial instruments, in some cases right down to the trading venue level. By mapping FIGI to other identifiers clients were able to cross check across their various systems the instruments which they had traded. The FIGI can help market participants to reduce errors in their operations, and in the end that benefits everybody. Clients should be paying us for added value insight from the data that we make available, not for identifier codes.
In 2019 we continued to see an increased demand for FIGI with an average of 4 billion API mapping jobs per month.
Were there any unintended consequences from opening the standard?
Usage has increased substantially on the OpenFIGI website over time, something that was hard to predict. So we have had to make enhancements, especially to the API facilities which have become more sophisticated. We’ve also expanded the FIGI coverage into other areas like derivatives, and we’re looking to cover the crypto-asset space, and some of the new investment opportunities out there. The way the market has moved means that we’ve had to adapt FIGI, and expand the open website to provide increasingly sophisticated automated look up facilities. As mentioned we have expanded the API, which includes facilities so people can map thousands of FIGIs to other identifiers. In 2019 we continued to see an increased demand for FIGI with an average of 4 billion API mapping jobs per month.
There is always that slight tension around what is a viable data set that makes sense to include in the open scheme, and what added insight is chargeable – as Bloomberg is essentially a data business
What challenges have you faced with opening the standard?
When making something open there is always a challenge because you need to decide where the cut off is for what’s open and what’s chargeable. You could always expand the dataset quite significantly, and start putting more insight data in there, what is usually considered ‘paid for’ added value. Calibrating the boundaries has taken a while to evolve. It has continued to evolve over time. There is always that slight tension around what is a viable data set that makes sense to include in the open scheme, and what added insight is chargeable – as Bloomberg is essentially a data business. It is the case however that the base specification for FIGI, which includes the core meta-data for the open symbology, is part of the standard, so that base level cannot change. It is just what else we choose to add in, and the tools on the web site that we make available, which continues to evolve.
It has been worthwhile, as the benefits have outweighed the resources to maintain and develop FIGI.
Has it been worth the effort?
It has been worthwhile, as the benefits have outweighed the resources to maintain and develop FIGI. We are helping the market, and providing access to a useful standards tool, and that has huge reputational benefit for us.
We’ve also been able to get involved in other open data initiatives. We became an issuer of the LEI (Legal Entity Identifier) in 2017, which provides an open identifier for entities. LEI solves the ‘who’ part of the core trading information required for successful processing that I mentioned earlier. So it really made complete sense to be part of that wider open data initiative as well. LEI is an International Organisation for Standardisation (ISO) standard which is overseen by a not-for-profit foundation called the GLEIF. In the case of LEI Bloomberg is one of a number of issuers of that code. In addition to the LEI we’ve been able to get involved in other data initiatives, such as ISO working groups to help shape other standards used in financial markets.
What advice would you give to others considering opening their standards?
You need a thorough understanding of how the standard is used and what the likely needs of users are going forward. Getting a clear picture of that isn’t always easy right at the start. You may have a clear view of how you think it is is, or will be, used, but you don’t always know until you see how things develop. Also there’ll be clients for it you haven’t even thought of – so you need to be agile.
There’ll be clients for it you haven’t even thought of – so you need to be agile
There will be things you need to implement to make it more usable, and ideally the sooner you can add these the better. We’ve had to add extra functionality to make open FIGI more usable over time.
You learn from experience about how people are using things, and what they might need to make it better. Be flexible and agile, the scope may change, and it’s likely to change. Within the scope, the way it’s presented, and the way people can consume it needs to be refined and improved over time
What are the next steps for FIGI?
We have plans to continue to increase the scope of FIGI. The plan that is most developed is to cover the crypto/digital asset space. Also parts of the derivatives market, to expand coverage there. We’re also refining the website automation so it’s easier to use. We need to keep it relevant, with new datasets and new functionality. We can’t stay static.
What’s it like working with the crypto-asset space?
We’re at the starting point of this. There are some major things like Bitcoin that already have identifiers, but lots of digital assets don’t, so FIGI is going to be moving into that space as well. There’s obviously technical ramifications in trying to identify these instruments to ensure accuracy, so it’s a bit of a learning curve. We’ve been consulting and working with others on how we can enter that space. It is a very decentralised world, so ensuring accuracy is a real challenge. Things can also change during the lifetime of a crypto-asset (so-called forks), so you have to cope with that and link those things together in a meaningful way. Agility is certainly something required here as well.