The concept of a web that links not merely sites but concepts dates back to the founding of the World Wide Web. David Saul, chief scientist at State Street, describes how the world of finance should aim to develop a common machine-readable language to describe its products In 1998 Tim Berners-Lee, the inventor of the World Wide Web, wrote a paper entitled 'Semantic web roadmap'. In the paper he explained that the web was designed as an information space, with the goal that it should be useful for not only human-to-human communication, but also that machines would be able to participate and help. He went on to say the following: "One of the major obstacles to this has been the fact that most information on the web is designed for human consumption, and even if it was derived from a database with well-defined meanings (in at least some terms) for its columns, that the structure of the data is not evident to a robot browsing the web. Leaving aside the artificial intelligence problem of training machines to behave like people, the semantic web approach instead develops languages for expressing information in a machine processable form." Sixteen years later, the use of semantic technology for the financial sector is being developed as one of many responses to the financial crisis of 2008-2009. The Enterprise Data Management (EDM) Council, an industry body, has been busy developing the financial industry business ontology (Fibo) – a project aimed at standardising the language used to describe financial instruments, business entities, market movements and corporate actions. David Saul, chief scientist at State Street, is a vocal proponent of the use of semantics in the financial industry. He argues that the trust that is needed between the different constituents – financial organisations, regulatory bodies around the world, vendors and consultants, and standards organisations – can and will be engendered by the use of semantic standards across the financial industry. He also says that putting data in a standard form through the use of semantics improves the efficiency and productivity of organisations such as State Street. Describing the products it provides to clients using a common semantic standard allows the clients to understand and use the products much more quickly, without having to translate and interpret the terms and details first. Saul also argues that regulators would benefit greatly as well from the use of semantic standards within the financial sector – demands for data would be simpler and more precise, and the results easier to use. At the moment, regulators receive a vast amount of information from a diverse set of organisations, and integrating and correlating it before reporting back to their governmental overseers is not a trivial task; common semantic standards, he says, would lift much of the burden from regulators and the industry. To understand why Saul wants the financial community to embrace semantic standards, Operational Risk & Regulation asked him to explain how semantics work and what the specific benefits would be. ORR: How do semantic standards work? DS: Today we don't think about it when we go to a website and there's a link in that site. We click on the link and we don't realise that there has more than likely never been any communication between those two sites; they're dependent upon a standard, which is the URL mapping and a lot of software that is invisible to them that creates those connections. What Tim Berners-Lee postulated with the semantic web is: what if we could extend that same level of transparent connectivity between pieces of data? In other words, if we have a piece of data that represented the name of a company, it wouldn't matter whether we spelt the name of the company out, whether we abbreviated it, whether it was a database, it was in a Word document, it was in a PowerPoint presentation and so on. The meaning that that represented – the name of a company, for example – would be constant. And if we could create a standard way of mapping those just as we map linkages today, that's the basis for the semantic web. Why is semantic mapping necessary? The web today operates almost flawlessly and it does it at scale, so those linkages are done billions and billions of times a day and they work with a high level of reliability. So we can now start to do the same thing with data. A lot of the work that people have to do when they move data from one place to another, whether it's internally within their company or whether they move it externally, is what I would characterise as wasted effort – translating the fact that these two things mean the same thing and that means the same as this and so on. We've created whole bodies of software that don't actually do anything except those kinds of translation. So the first advantage of semantic mapping is on the productivity side where we can create those linkages without having to go through extra translation software. How could this benefit the financial sector? The really big benefit is that today a lot of these translations are done by technical people, because the software with which to do this – whether it's schemas for databases or ETL software (extract, transform and load) – is not looked at by the business people who really are the ones who understand the meaning of the data in the first place. So you've not only got the software that's doing the translation, you've also got to communicate from the business person who understands the meaning to a technical person and then back again to another business person. What semantics brings to the table is the possibility that now the business people are the ones who are going to create those definitions and who better to do that than the people who really understand what it means? Do we have examples of defined standards that exist now in the financial sector? We have work that's going on in multiple places to define standards for meanings in the financial sector. An example is legal entity identifiers – the unique code designed for all market participants – which means we'll all be able to speak the same language when we talk about a company because when we define its legal identifier, we will now know we're talking about the same thing. Where is the work taking place to help develop semantic standards in the financial sector? It's happening at multiple levels. I've been directly involved with the EDM Council. Its membership is pretty much all of the major global financial institutions. And global is very important – we don't want this to be specific to one geography. The people who are defining the standards that we're using for the financial industry are business people – to my earlier point. So when we set out to define what the components are of a credit default swap (CDS) for example, perhaps looking at what the definitions are of the different kinds of collaterals that make up the different legs of a CDS, the people who are creating those definitions in the EDM Council are business people, people who actually understand far better than I do – as a technology person – what the different kinds of collateral are, for example. What is it about semantic standards that you believe will help to foster trust in the financial system? Coming from one of the major players in the financial services industry, I very strongly believe that if we try to do something like this on our own, given the recent crisis – and even though the vast majority of everything that takes place in financial services is open and above board – if we tried to do this on our own, I don't think it would get anywhere. Even if we tried to do it with an industry association, it would still be met with some mistrust. My feeling on this came about through working with the EDM Council and the Object Management Group – which is the standards process that we're following in order to bring Fibo out. I got to know people there – people on the regulatory side and then also some vendors who were building traditional database products and semantic technology – products that can already ingest Fibo and web ontology language format and produce executable rules. There were four of us at a meeting at the EDM and we starting talking about this problem and realised that it was to all of our benefits. How can it benefit each of the constituents in the financial sector? For us as a bank it has helped us in producing data analytics products as well as risk management. As for the regulators, they experience difficulties in integrating data from multiple sources. For example, one of the regulators that was with us in the meeting at the EDM had experience at the US Securities and Exchange Commission (SEC) with XBRL, which is a global standard for exchanging business information, and the submission of 10-K reports – the annual report required by the SEC detailing a summary of a company's financial performance. He talked about how they couldn't have done it without creating a standard for reporting. Then the vendor that was with us said he builds products that could be used in the work of semantic standards but doesn't know exactly what the requirements are. He was clear he wanted to build better products, but was also clear that he needed to know what he should be providing to the industry. We also had somebody from the standards group with us and a metaphorical light bulb went on at that point. We realised we all have common interests and if we can create a standard set of financial semantics and encourage its use by everyone in the industry, it's going to help all of us. What are some of the benefits for financial institutions specifically of semantic standards? The key here is synergy. What happens today is that we have a set of data, whether it's coming from transactional systems or operational data stores, and we need to create revenues. So we come up with a list of products and services that can generate revenue for us by taking this data, integrating it, correlating it and delivering it to our clients because there's value in it. We also need to do risk management to ensure that these transactions are properly capitalised, and we have a whole set of regulatory requirements that we need to meet for this. And before we ever get started doing any kind of IT development, we split those two apart and we set up one group that's going to develop the revenue-producing products and we set up another group that's going to do the risk and compliance. And they follow very parallel development paths. They say ‘okay in the first case I'm going to set up a data warehouse that has the information I need to deliver these value reports to my clients' and in the second case they say ‘I'm going to set up a risk repository, which has the data that I need in order to do these calculations'. The synergy that can come from semantic standards is that they are both sourcing their data from exactly the same places and they go through ETL programmes and transform the data. They then create database schemas and part of the translation is again going from business people to the technical people, and we have a data warehouse on one side and a risk repository on the other side. In a lot of cases they are using the same technology but they no longer map it to one another, and so what we find is that we go down the two paths and at the end when we try and bring them back together, we find we're going to have to spend even more money doing reconciliation because they may have selected the same data, but they interpreted it slightly differently or they took it at a different time. That reconciliation is really a waste. Semantic technology will enable us to create synergy at the beginning and say ‘look, both of these groups are looking at the same data'. And that technology has a lot more flexibility in it because you want, in case one, to respond to market changes, and, in case two, to respond to regulatory and compliance changes. If we can somehow create the synergy process – organisationally and technically – early on, the savings are tremendous. Not just the cost savings, which would be huge, but it also means at the end, you don't have to reconcile because you've been reconciling every step of the way and you won't have to do that any more. This would actually be transformational. Is this happening in any capacity today? When do you think it could be fully in place? It's starting to happen. I don't know that anyone has fully committed down that synergy path, but I'm hearing more and more from colleagues that they recognise that there is synergy in this data and this is a way to deal with the increasing cost of complying with regulations while at the same time helping on the revenue side of the equation, so it's starting. I think once people start to see some successes, I don't really see any other way. At some point without this in place some company is going to get to the point and say the cost of regulation has now exceeded the profitability of my company and it's no longer viable. No one ever wants to get to that point. What are the regulators currently doing to get semantic standards in the rule book? The Commodity Futures Trading Commission, the Office of Financial Research, and the Treasury are really encouraging the industry to do this work on semantics. It's a bit of a chicken-and-egg situation because we first have to do the work and standardise things before this can be codified in the regulations. At the same time, we need it codified and the standards set in the regulations to have everybody using it. And while the regulators have not yet said that they're going to require this to come out in Fibo, the fact that we now have this has improved the odds that that is what they are going to do. So if they now codify that in the regulations, then that's certainly going to speed up adoption. When do you expect to see movement on this from the regulators? I would think before the end of 2014 we would have a very good idea. I can't speak for the regulators, but I've got very good feeling coming back from them that they recognise that this effort helps them as much as it helps us and so I would think that by the end of the year we'd have a pretty good indication of whether this is going to become part of the regulation. If it does, that's certainly going to speed things along and we're doing our work and contributing to the EDM Council in anticipation of that happening....
Start a FREE trial or subscribe to continue reading:
Start a 4 week free trial
Try Risk.net's premium content for a limited period. Register now for your FREE trial to one of our leading brands.
*not available to previous trialists or subscribers.
Log In or Subscribe Now
Subscribe to Risk.net Business now to access all our premium news & features content for 1 year.
Pay by Credit Card for immediate access.