Embedded AI has a distinct business case for hedge funds: Bijesh Amin, Co-founder, Indus Valley Partners

In the current digital revolution era, there has been an uptake of futuristic technologies such as machine learning and AI. Organizations across all industries have started realizing the importance of such innovative technologies and are bringing up the use of these in their day-to-day operations. In an interaction with Express Computer, Bijesh Amin, Co-founder, Indus Valley Partners, shares his outlook on the adoption of these technologies and the way it would help the enterprises in their digital transformation journey

How big an impact will AI and machine learning have on the industry in 2017? What will be the key changes?
To begin with, it is important to understand that we have an interesting and unique background that provides us to see a lot of the market, wherein we have 150 clients with over $1 trillion in AuM (Assets under Management) collectively. Our franchise is comprised of bigger funds, more than $1 billion, and we have a splattering of smaller funds that rely on our regulatory technology and data warehouse platforms.

In terms of the impact AI and machine learning will have this year, given the variety of clients we have, we can say that it is not going to have a tremendous impact from an investing standpoint. Many in the industry use exploratory investments and portfolios run by machine learning or AI algorithms. But they haven’t committed a huge amount of capital to these projects yet.

Although, in regards to post-trade, there are more vendors and more technology consultant firms looking at AI and other leading-edge technologies such as Blockchain and cementing it in to their software and platforms. Innovative financial technology is clearly on the market’s radar and we are proud to say that we are pioneers in this area. We can only see this kind of technology increasing in usage across the environment as the industry continues to trade more algorithmically, increasing the velocity and volatility of trading. However, we do not expect to see a seismic shift this year, or even in 2018, but the evolution is taking place and only a few years down the line.

What will happen to those who fail to stay up to date with these tech evolution? How quickly is the industry moving?
We are not hugely worried about embedded competitors. What is more of a concern is the scope from disruption by innovative start-ups that can build models that perform traditional functions quicker, cheaper while being more intuitive to the end user. Our aim is to be build these disruptive models ourselves and not be disrupted in turn. We are building platforms from the ground-up using technology that is state-of-the-art, cheaper, utilizing open source, big data or new ways of visualizing data.

Initially, when Indus Valley Partners was born, one of the advantages that we had over banks building technology themselves was that we were more agile and less constrained than they were. In banks, things tend to be very siloed and systems reflect that. While building our first set of products and applications, we could look at everything from scratch and craft something that could natively support both derivatives and cash products, or incorporate foreign exchange or swap hedges.

Nowadays, a similar shift is happening where new business models and operating paradigms are emerging that treat financial data – in all its varieties – as a commodity that can be gathered, analyzed and displayed really quickly, really efficiently and really intuitively. It does not matter if the device is static or mobile, or if the data needs to be displayed on a watch or on a projector.

How is cloud computing and embedded AI “baked into” the platform? What has this meant for the hedge funds?
Historically, hedge funds have largely outsourced post-trade functions to prime brokers and administrators. However, they do often retain a middle office function of a noticeable size. This ensures that internal risk management can reconcile and check and value everything correctly before handing off to the prime broker for clearing/settlement.

This space is perfect for an embedded AI. A hedge fund could utilize machine learning model to support their trading platform with less (but not zero) operational headcount and codify a lot of the institutional knowledge in the platform itself. Operationally, it would mean that they could value, hedge, and finance more on a just-in-time basis rather than waiting for the end of the trading day to see a holistic view of their exposures. Embedded AI has a distinct business case for hedge funds.

Is it a good thing that the trade life cycle is becoming increasingly machine-driven?

There are two sides to every coin. The positive is that it does reduce risk because you are able to react and modify on more of a just-in-time basis. However, the downside is that some of the systemic risk that might be pushed downstream. For example, with regards to the market infrastructure, not all exchanges and post-trade broker systems are ready for the velocity of trading that an even more widespread adoption of machine-driven, algorithmic trading would elicit.

Are there any misconceptions about this reliance on machines? How are these disproven?

The biggest danger is overlooking the significance of operational personnel. Some might believe that they no longer require certain roles or staff because of this machine-driven evolution. However, many machines and humans work in conjunction with one another and ignoring one element could prove disastrous. For example, we have built a platform, in Salt Lake City, Utah, that relies just as much on the experience of the person operating as it does on the underlying AI in the software. The aim is to have the two working hand-in-hand with knowledge being codified into the platform over time.