Why AI Is a Good Fit for Finance

Interacting with artificial intelligence was once the stuff of science fiction, but the reality is that AI and machine learning have become intertwined with everyday modern life.

Anyone who’s ever searched Google for an answer, or browsed Netflix for something to watch, has tapped into the advanced computational technology that powers this technology. AI and machine learning are in the toolkits for almost every company from fast food to retail. The same is true in the financial industry, which has numerous applications or emerging applications that are applying AI and machine learning tools and techniques.

Before diving into the details of how the finance industry is using AI, it’s good to understand the basics.

Artificial Intelligence 101

John McCarthy, one of the fathers of artificial intelligence, defined intelligence as the computational part of the ability to achieve goals in the world. Humans operate this way. Just going to get a cup of coffee requires a person to analyze the best route and consider that route against how much time there is for the trip. There are many paths available and a lot of data to consider in order to navigate that simple errand.

AI works the same way. It attempts to use data to determine the best way to achieve a goal.

Machine learning is slightly different from AI. Think about AI as the umbrella term. And under AI, we have several subtopics or subareas, machine learning being one.

In finance, the goal of AI and machine learning can be to make trades in the market, automate different risk management processes, or use data to make tax decisions, for example.

On the Uses of AI

Much of the power of these technologies is aimed at monitoring financial markets and detecting fraud. A bank will look for fraudulent activity by monitoring your purchasing behavior and seeing when there is a purchase that does not look like the typical purchases that you make. These are machine learning algorithms that have learned your purchasing behavior.

There are financial advisory services, modern subservices, that are often referred to as robo-advisory services. These machine learning algorithms make investments into securities like ETFs and automatically rebalance customer portfolios based on sound financial decision-making processes.

Customer service and marketing are other examples of how AI is leveraged in the financial industry. The technology can help understand a marketplace by analyzing the habits of customers and potential customers.

AI and Innovation

AI drives a lot of innovation in finance. It’s been adopted at breakneck speed over the past few years, and the environment today is radically different from just four years ago.

The rapid adoption is because AI and machine learning offer major benefits for companies that use the tools.

AI helps automate all kinds of processes and tasks, some of which could be very repetitive but require a certain form of decision making. Many of these algorithms are highly efficient. Automating repetitive tasks enhances existing workflows and provides better, faster service to people.

These tools are very good at extracting information from large amounts of data and presenting it or visualizing it in different ways, too.

Machine learning algorithms may also be more accurate, especially when compared with traditional statistical or computational tools. They’re often better at making the right recommendations, making more precise forecasts or predictions of outcomes, or helping make informed decisions.

The automation, decision making, and accuracy of predictions lead to cost savings and increased profitability for AI adopters.

AI Challenges

But AI and machine learning are not a panacea for businesses and financial institutions. The tools come with challenges.

For one, most of the algorithms are trained for very specific tasks. It’s not how you see AI in the movies, where we have robots moving autonomously, talking to someone and interacting, and then jumping in and driving a car. We’re far from that.

It’s also important to highlight that AI and machine learning are ineffective if they’re trained poorly. If they receive bad input, they’re going to make the wrong decisions, or biased decisions in one way or another.

Of course, with the right training, and the right algorithms working on their respective tasks, AI and machine learning offer more opportunities than challenges. But they need the right humans to operate and monitor. Finding those people is a challenge for companies across the spectrum.

A study by the European Commission in 2020 found that difficulties in hiring new staff with the right skills are a key internal barrier to AI adoption.

This is a rapidly evolving area, and we need more skilled people who understand the technology and understand how to interpret the results. A number of these AI algorithms are complex, and it’s unclear exactly what they’re doing. People often refer to these types of algorithms as black-box algorithms. That might be undesirable if you want to take a decision apart and understand what led to this decision.

Interpretability is important, and it’s an ongoing discussion in the research areas of machine learning about how we can best interpret complex decisions and models.

We’re constantly learning about this field. More and more examples of how people applied machine learning algorithms in different business areas are being shared. We can read the successes and the failures that have come from those cases and experiments to keep improving the algorithms.

What we know for sure is that AI and machine learning are here to stay in finance, and most other business categories. This new technology has clearly shown a competitive edge for its adopters.

---

About Petter Kolm

Professor Petter Kolm is the Director of Mathematics in the Finance Master’s Program and a Clinical Professor at the Courant Institute of Mathematical Sciences, New York University, and Partner at CorePoint-Partners.com. Previously, Petter worked in the Quantitative Strategies Group at Goldman Sachs Asset Management, developing proprietary investment strategies, portfolios, and risk analytics in equities, fixed income, and commodities. He was awarded “Quant of the Year” in 2021 by Portfolio Management Research and the Journal of Portfolio Management for his contributions to the field of quantitative portfolio theory. He is a frequent speaker, panelist, and moderator at academic and industry conferences and events.

---

This financial industry article is adapted from the GLG Webcast “Machine Learning in Finance: Hype or the New Frontier?” If you would like access to the transcript for this event or would like to speak with financial industry experts like Petter Kolm or any of our approximately 1 million industry experts, contact us.