Artificial Intelligence: the future of financial intermediaries

For some years now, Machine Learning models have begun to be used by major financial institutions around the world, and could radically transform the entire financial sector in the future. In a recent MIT Sloan Management Review survey of more than 2,500 companies globally, 84% of respondents said that implementing AI in their processes, products or services could be a crucial advantage in their against its competitors. At the same time, the data from the World Retail Banking Report by Campgemini and Efma tell us that in 2020, only 39% of credit institutions have begun to extract value from the enormous amounts of data available to them.

In 75% of cases, the management of bank data, which often reach volumes such as to be defined Big Data, is hindered by manual processes which, in addition to slowing down the entire system, can give rise to errors that affect the validity of the models used and, consequently, the quality of the information available to decision-makers worsens. To be able to fully exploit such a massive data heritage conventional tools are not enough, but it is necessary to use up-to-date systems, which use Machine Learning algorithms to correctly process information.

The growing complexity of financial systems, together with the acceleration towards the Digital Transformation that the pandemic has inevitably brought with it, has led the new players in the sector to focus on the User Experience to expand their market share. In the world of Retail banking, the new participants face traditional credit institutions by proposing a model focused on the final consumer, through the use of highly personalized and customer-friendly services, often based on a business model free from the presence of physical branches, but which makes use of online and mobile banking platforms. In a similar context, the competition in the sector shifts to how to best use the data at its disposal to be able to develop solutions that can meet the new needs of consumers, who are increasingly experts in the use of digital technologies. In addition to the advantages offered for improving the User Experience, Machine Learning algorithms are able to automate many of the particularly labor-intensive activities, reducing costs and improving productivity and efficiency in all company divisions.

For a bank that wants to approach the world of Artificial Intelligence, an important first step is certainly to give the right importance to AI, trying to develop a long-term strategy, scalable for each different business area. It is no coincidence that the “Frontrunners”, i.e. the financial companies that have obtained the highest returns from the implementation of AI, assign a high strategic value to the development of these technologies, while among the “Starters” — that is, those who have just adopted this type of solution and/or has yet to profit from it — only 2% consider AI as a strategically crucial element for the future. The resistance to innovation on the part of some executives can also affect the budget allocated for this type of activity, undermining its ability to be truly effective. In this sense, starting from relatively less ambitious projects, being guided by those who already have proven experience in the sector, can be a great way to overcome the initial skepticism and experience the benefits offered by Artificial Intelligence.

Here are some of the possible applications of Artificial Intelligence in the banking sector:

• The construction of advanced credit rating models, which exploit the entire data assets available to the bank to produce a more accurate assessment and thus reduce the default rate;
• The use of clustering algorithms to segment customers, which make it possible to propose specific measures for each group, improving dedicated products and the User Experience;
• The use of Predictive Analytics algorithms for forecasting cash flows or other fundamental quantities;
• The identification of patterns in transactions that allow fraud and illegal transactions to be detected more quickly;
• The use of Natural Language Processing (NLP) techniques for the implementation of chatbots and virtual assistants, or for the automatic reading of documents.
• The development of automated investment, asset management and financial advisory strategies (Robo-advisor), able to optimize performance, meet customer needs and considerably reduce the costs associated with this type of activity.

According to Gartner, by 2030 80% of institutions offering traditional financial services will cease their activities or in any case will no longer be able to compete with new players, who will dominate the market through digital platforms and with new business models. Some might consider it a somewhat drastic vision, but it is certain that with the widespread diffusion of the data-driven economy paradigm, the adoption of Artificial Intelligence tools will be increasingly important.