This is the second in a series of articles highlighting applications of artificial intelligence.
Financial Services is all about data- it’s one of the most data-rich industries out there. Banks and Insurance companies are always processing and analyzing data in the hopes of providing better service and making better decisions. With all this data, there are an inordinate number of potential AI applications, including:
The term “banker’s hours” has been an anachronism for a long time. Clients are demanding 24/7 service, and firms are providing an omnichannel selection of applications which will allow their clients to get the information they need when they need it.
Chatbots are being added to the call centers, websites, and mobile apps already available to clients. Chatbots are a great addition for several reasons. First, many clients prefer to interact with their financial institution in a conversational manner rather than having to fill out a form or use a search engine. Second, chatbots are a much less expensive solution when compared to fielding a call or even responding to emails.
The natural language processing at the heart of chatbots can help financial firms provide the service their clients are demanding while effectively managing their costs.
The vast majority of trading by the major banks and brokerage firms is transactional in nature — they are just fulfilling the requests of their clients. However, all of the larger firms are also engaged in “proprietary trading”, where they are investing the bank’s own funds, or a select number of investor’s funds, with the intent of making money through shrewd trading. Of course there are also innumerable mutual funds and hedge funds chasing trading profits.
Traders are under incredible pressure to deliver results. They will look at any solution that will help them make profitable trades. Given this profile, they are often early adopters of new technology, and AI was no exception. Traders have been using AI technology for over 20 years. Unlike some other areas, “explainability” is not a requirement in trading — if an AI developed trading strategy delivers consistent results, most traders are not really that concerned about the “why”.
Today, trading firms and departments are hiring few MBAs. They are looking to Computer Science and Physics grad students to help them develop predictive models that will deliver profits on an ongoing basis. Foundational to these models are machine learning and deep learning.
Of course, this is all highly secretive. People don’t discuss, much less publish, their financial models. A firm may admit that they’re using AI, but that will be about the extent of it. The rule of the road in trading is that if something is working, don’t talk about it.
The firm that has been ahead of the curve in using technology to generate profits is Renaissance Technologies. They’ve had consistently spectacular results, and you can count on the fact that AI has contributed to their success.
Financial Services is all about managing risk. The better picture you can develop of your risk, the more money you can make, and the more confident you can be in your decision. Whether you’re deciding on whether to make a loan, how to price an insurance policy, or whether to issue someone a credit card, you need to base your decision on some sort of risk model.
Risk is highly complex and multivariate. This is the type of environment where AI shines. From practitioners of the ancient discipline of actuarial science to credit analysts, folks who are charged with creating risk models are leveraging AI technology to build more accurate models that are generating higher profits.
“Regtech” is a subset of fintech focused on applying technology to help financial institutions meet their ongoing and often expanding regulatory requirements. This can involve anything from automating the reporting of compliance documents, improving anti-money laundering surveillance, or enabling improved compliance with data security regulations.
AI has a growing role in regtech. Something like anti-money laundering, where you are monitoring customers and transactions to try to pinpoint suspicious activity can seem like an impossible task. You are attempting to find the proverbial needle in the haystack. AI can help you build a more effective predictive model that can “red flag” money laundering while reducing the historically very high level of false positives. In this case AI can both reduce costs, and improve regulatory compliance.
This is just a subset of the endless AI applications within financial services. From banking to brokerage to insurance, AI offers new solutions to improve customer satisfaction, reduce costs, increase revenue, and ultimately increase profitability in the financial services industry.
Ken Tucker is a business consultant specializing in AI and Analytics.