Where AI and Post-Trade Meet

Narasimha Kodihalli News

Artificial Intelligence (AI) is a technology that is rapidly becoming relevant in the post-trade financial services space. AI is a branch of computer science dealing with the simulation of intelligent behavior in computers. Operationally it’s taken as the capability of a machine to imitate intelligent human behaviors or human behavior patterns.

Intelligence is only as good as the underlying data, which is only fully usable when it is standardized in a market context. In the derivatives market, there is a large knowledge base, but in order to apply algorithms to put it into context, analyze, derive patterns, and make intelligent decisions, we need smart data across pre-trade and post-trade processes. This is where AI comes into picture; Convert your information into smart data and apply algorithms to drive intelligent business decisions.

Once you have smart data and algorithms to drive business operations decisions, the next question is how do you apply it to become tactically or strategically aligned, be regulatory compliant, and reduce your total cost of operations in both pre-trade analytics and post-trade processing? There are various tools and frameworks available in the market to address this but only a very few firms in capital markets are really taking advantage of AI techniques at this point. One prime example of such an initiative is FIBO (Financial Industry Business Ontology), an industry initiative developed by EDM Council members.

Common misconceptions about AI and FinTech

The biggest misconception is that people often think of AI as robots taking over everything that humans do. This is clearly not going to be the case. Robots will not replace humans completely, however machines will come to replace operations on Wall Street. This is only a matter of time.

A very common misconception about FinTech is that people think of it as being just about B2C and P2P payments. AI will drive the ability for exotic derivatives—currently only understood by investment bankers—to be made available to the common man as an investor. That gap can only be bridged when FinTech is viewed holistically as both B2B and B2C.

Market drivers for increased use of AI

The biggest driver is regulatory compliance. The capital markets overhaul is far from finished. Firms need to leverage sustainable, sophisticated technology, and AI is their best option from a proactive and strategic standpoint. In Collateral Management context, creating smart data models to capture bespoke nature of legal contracts and processes around it is by far the best solution for an ever-changing regulatory landscape.

The other major driver is the need to save costs. Recent paper undertaken by Pricewaterhouse Coopers (PwC) for DTCC-Euroclear Global Collateral Ltd shows the annual cost of operations and technology infrastructure for collateral management operations for a sell-side bank was about $7 million in 2015, and it is expected to grow to $27 million by 2020. This is an exponential increase in the amount of money market participants will spend. How do they control that? The only answer is advanced technology such as AI. Research undertaken by Morgan Stanley | Oliver Wyman indicates the wholesale banking industry can save $15-20billion if deep digital adoption with robotics and automation is embraced.

In the derivatives industry as a whole, the cost per transaction is very high, partly because counterparties don’t take a holistic approach in their communications. The only real way to reduce these costs is to have a sophisticated protocol, such as RDF and OWL, enabling market participants to communicate with each other directly. Imagine collateral management systems are speaking different languages, and they need the utility to serve as their interpreter.

Using AI helps participants reduce a lot of inefficiencies, enabling them, for example, to perform trade reconciliation before margin operations. This is seldom practiced by the market, but would make a lot of sense to do it in that order, and AI techniques can enable that.
converted by Web2PDFConvert.com

Where AI can be applied? Hint – collateral management

Payments is one of the more developed areas within FinTech and settlement management is a classic use case for both blockchain and AI.
Another application is tracking the changes to regulatory requirements, which is a constantly moving target. The rollout of variation margin requirements in March under BCBS/IOSCO was chaotic because both the market and the regulators were underprepared. The only way to advance these integral plans is to deploy sophisticated technology like AI and natural language processing, which means ISDA master agreements and credit support annex contracts (CSAs) need to be rationalized and digitized. Only approx. 5% of contracts in the market were amended for March 2017 deadline. Firms will be able to reduce a lot of operational overheads if they prioritize this shift.

Margin call processing and dispute resolution are another good use case for AI in the OTC derivatives space. Regulations are driving an explosion in the number of margin events, and disputes will be common. Firms could plausibly teach a machine the mundane tasks a manager performs to resolve them, saving firms large amounts of time, money and energy.

The application of innovative technology and including AI towards collateral management operations is particularly timely given the rapid evolution of this space as collateral moves from a back-office function more upstream to the middle and front offices. For instance, coupling up post-trade processing and pre-trade analytics is becoming a big focus for market participants. With massive data volumes, participants need to make very fast and accurate decisions. The front and back office need to come together, and there are already a lot of initiatives to address that. There is an argument that there will not even be a middle office a couple of years from now. As they begin to leverage AI technology, market participants need to break down the silos between front, middle or back office, and between different derivatives products.

Looking forward

The next couple of years will see significant growth in the number of concrete implementations of AI. Globally, 62% of the $544 trillion notional outstanding reported by dealer is centrally cleared. This provides a very ripe ground for applying AI techniques.
According to CB Insights, the top 22 FinTech companies were valued at $74 billion at the end of Q3 2016. FinTech innovations and investments are only bound to grow in coming years.

In the next few years, we will see a move towards more intelligent delineation and tracking of risk, so that we won’t again have a situation where a banking system collapses because various individuals are unable to pay their mortgages. Much has changed since 2009, but there hasn’t yet been enough effort to track that lineage and measure global financial risk. That is where the market is headed over the next few years.