Home Slider How AI Builds and Sustains a Competitive Advantage in Banking

How AI Builds and Sustains a Competitive Advantage in Banking

by internationalbanker

By Vijay Doradla, Chief Business Officer, SparkCognition




We live in interesting times. Banking today is more complex, competitive and fast-moving than it’s ever been; it requires dealing successfully with extraordinary challenges, some of which have no historical precedent. However, those challenges also come bundled with extraordinary growth possibilities for organizations sufficiently smart and agile enough to cope.

Increasingly complex markets require a new approach

The fundamental challenge hasn’t changed: How to assess and hedge against risk to the necessary extent while simultaneously pursuing the best available investment opportunities.

Banking is very good business if you don’t do anything dumb. Warren Buffett, the chairman of Berkshire Hathaway and a legendary investor

Artificial intelligence (AI) solutions can, by developing and refining sophisticated, data-driven, ever-evolving models of complex financial ecosystems, accurately inform and guide investment professionals.

This is a question with a long history of analysis by the brightest and best minds, and these days, it’s increasingly finding answers not only from the academic and banking communities but also from the world of technology. Specifically, artificial intelligence (AI) solutions can, by developing and refining sophisticated, ever-evolving, data-driven models of complex financial ecosystems, accurately inform and guide investment professionals at a time when accurate insight and guidance are more essential than ever before.

Why is that necessary? What’s changed to spur what is now a nearly universal mandate in the banking community to leverage AI to best effect?

Among other factors to consider, there are:

  • New regulation constraints. When the push toward a risk-neutral business model (and client-driven model) has government muscle behind it, it becomes necessary to demonstrate that the complete investment inventory is properly hedged. Unfortunately, too much hedge against risk can itself be risky in the sense that it reduces potential revenues and margins and introduces the daunting possibility of losing clients to nimbler, more capable players.
  • A historical level of competition. When more players enter any given game, achieving and sustaining a competitive edge inevitably gets harder for all of them—the old and new alike. Today, the array of available investment services—both domestic and foreign—is larger than ever by a wide margin, and it’s getting larger by the day. This increasingly competitive situation, in turn, means investment professionals have less total time to respond to sudden market changes. A response that might have been fast enough in the old millennium may not suffice in the new—not even close.
  • New market complexity. Beyond even factors such as regulatory pressure or competition per se, there are now quite simply more variables—interacting in more ways and implying more risks and opportunities than ever before. Each variable, as it changes dynamically at an accelerating clip, translates into a shrinking timeframe window to act. The market, in short, is changing faster, and organizations that still rely on standard correlations across changing variables need to update their solutions and strategies because those correlations simply no longer always apply the way they once did.
  • New technologies (not always leveraged ideally). Some organizations have deployed and taken full advantage of the best available AI solutions to assess market risks and opportunities, predict and quantify emerging developments, and thus best serve their clients. Others, while they may have developed or deployed a limited flavor of AI, nevertheless lack key capabilities. Still others lack AI altogether. The gap between the technological haves and have-nots is swiftly growing to the point where it becomes a chasm for the have-nots. It is, if not a mortal threat, an obstacle that impedes their progress tremendously as they strive to build and sustain smarter and more effective business models in the modern investment arena.

When a request for quote (RFQ) comes in from a client or prospective client, how quickly can the organization respond? Delays may have a negative impact on the franchise’s reliability and could result in the client going elsewhere to a competitor that is, or at least is perceived to be, more agile, modern and on top of its game.

Just as important as speed is accuracy. A stunningly fast RFQ must also be informed by the full scope of the available data (data lake) and the discoverable but hidden trends and patterns that data implies to sharpen pricing and close the deal without introducing unacceptable risk into the process.

Moving past statistics to an adaptive, cognitive strategy

Many banks have, in the past, turned to statistical analysis to try to deal with such complexities. Tools of this type handle tasks that include mean-reversion analysis, mean-variance reduction to optimize asset allocation, statistical arbitrage and others.

Their very popularity, however, has diminished their power to create a lasting competitive advantage. As they have been widely deployed, their limitations have often been revealed under market pressure, and because they have become commonplace, they deliver less and less competitive advantage.

Statistical arbitrage, for instance, is not self-adaptive. The failure to learn from changing dynamics translates not just into diminished accuracy as a decision-making strategy but, in time, higher risks and eventually greater costs.

Statistical strategies also suffer from a relative lack of agility because of the typical lag time between arriving at new insight and acting on that insight. Meanwhile, agility is critical in a super-competitive market characterized by tiny windows of opportunity through which it gets harder by the day to leverage optimal hedging strategies and yet also maximize margin retention.

This is why organizations that are looking to create and benefit from a sustainable, competitive advantage—which is to say, all of them—are increasingly augmenting, or in some cases moving away from, purely statistical approaches. Instead, they are turning to more sophisticated AI-powered capabilities, such as cluster detection, reinforcement-learning-based automation, adaptive-model modification and dislocation detection. All of them can play a key role in the process of adjusting the investment equation as necessary, in or close to real-time, to deliver competitive advantage while also lowering costs and risks as much as possible.

This is, in part, because top-tier AI solutions are designed from the start to be highly adaptive. As a result, they enable just the kind of agility organizations need to outperform the competition. They can ingest and assess incoming data volumes of any size, correlate changes in that data to validate arbitrage opportunities and then act (or advise investment professionals to act) faster and more accurately than any alternative class of technology.

AI solutions can also help the organization orchestrate its policies and behaviors in carrying out various tasks at different logical levels—unifying them all in the pursuit of larger goals, as a conductor unifies an orchestra. To do this, they can aggregate and analyze data drawn from many different departments or business services, each with unique objectives and data types, ranging from voice trading to risk management to e-trading to leveraging Salesforce’s cloud platform. Then, based on the adaptive analysis, AI can provide different kinds of insight tailored to specific needs.

Perhaps, for instance, the organization wants to assess the performance of different classes of trades under varying conditions, determine asset toxicity, quantify real-time cross-asset allocation or establish whether company policies and infrastructure comply with applicable regulations. All of these possibilities and many more can be realized today using current technology.

The result is a holistic, end-to-end view of the business and everything it does in which AI has effectively become a collaborative partner in the service of prioritized goals.
The result is a holistic, end-to-end view of the business and everything it does, in which AI has effectively become a collaborative partner in the service of prioritized goals—not merely a limited tool applied in one dimension for one particular task.

Here’s another important point: All of the capabilities described above help advise individual investment professionals who, far from being replaced by AI, will find AI an essential source of insights they can leverage to make smarter decisions faster. Over time, as AI draws on more and more data to improve its modeling and forecasting, it will increasingly empower them to create more quantified value, reflecting favorably on both their organizations’ competitive postures and their client portfolios—as well as, of course, on their own careers.

All of this helps to explain why, going forward, it’s our confident expectation that AI-driven banking solutions will not only continue to be deployed on an increasing scale, they’ll become steadily both more powerful and versatile and woven more tightly into the very fabric of the business model.

And at that point, AI will create and deliver value to both banks and their clients that is far broader and deeper than the extraordinary value they receive from it today.

Vijay Doradla brings more than 20 years of diverse experience in innovation to his position as the Chief Business Officer of SparkCognition. He previously served as an engineer, industry analyst and investor in the field of technology, working with major corporations such as Bell Labs. He was also the Founder and Chief Executive Officer of SilverString Capital.


Related Articles

Leave a Comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.