Revolutionising retirement savings through the power of AI

Festina Finance has been an early adopter of artificial intelligence (AI) and machine learning to transform how its financial services customers engage with their clients and is now using the same technology to revolutionise complex areas like retirement planning.

Festina Finance has been using AI and machine learning to transform and automate services like mortgage advice and credit scoring for several years. By analysing a customer’s own data and responses, it can quickly understand how that customer will behave over a range of different scenarios and provide them with better financial products as a result.

Festina Finance has now taken the next step and applied the same science to the complex art of retirement planning. The company’s clients, such as PFA, PensionDanmark and users of Letpension or Naerpension, are already using the cutting edge technology via its 360-degree Festina Pension Advisor tool.

In the arena of financial services, AI and machine learning has been a game changer. By combining it with predictive analytics, it is enabling our clients to provide their customers with a personalised experience that tailors products and services to their individual needs.
Mikael Braagaard, CEO of Festina Finance

He points to the fact that while traditional financial planning is well known as being slow and manually intensive for both firms and clients, they are also inherently vulnerable to individual human biases towards products and a customer’s circumstance.

“By contrast, with machine learning we can build algorithms that predict a scenario based on the analysis of a vast number of actions and data about the customer,” Braagaard says. “This allows the technology to act like a human adviser would but at a much faster pace and because of the vast amount of data being processed, with a higher degree of precision.”

Machine learning vs brute force and human assumptions

Designing the right strategy for retirement savings is a complex task for human advisors since there are a vast number of potential withdrawal plans. Calculations will typically be made with the use of a so-called brute force algorithm, which solves the problem by trying every possible scenario and repeating the samples. As Festina Finance’s lead machine learning engineer Jonas Steen explains, there are a number of problems in taking this approach.

“The method is very time consuming as it deals with many variables,” Steen says. “Typically human advisors can offer a faster but less precise calculation based on assumptions about the potential outcome.”

“For example, the number of consumption patterns increases exponentially in relation to the complexity of a customer’s finances, especially with the quantity of different pension schemes.” Steen adds that assumptions about potential outcomes are equally problematic, as in practice you run the risk of no longer solving the real underlying problem in terms of the course of action for the customer.

Always learning, always improving

To address both of these downsides for traditional retirement planning methods, Festina Finance has designed machine learning-powered solutions that are always working to improve the accuracy of the financial forecasts produced. The algorithms at the heart of its 360-degree Festina Pension Advice tool empower human advisors to make quick and accurate aggregated calculations in a matter of seconds

“The algorithm uses statistics to find patterns in data and apply them to make highly accurate predictions,” says Steen. And every time Festina Finance’s algorithm crunches the numbers for another customer it gets better at assessing the optimum retirement plan for customers.

This means that for users of the Festina Pension Advice tool, they can deliver a first-class service to every customer, every time.