The FCA conducted an online robo-advisory experiment, with nearly 3,500 participants representative of the UK population.
According to the analysis, many people said they were willing to pay more for the tool than for its monetary benefit, potentially suggesting a significant mental cost to consumers juggling debts and making repayment decisions.
The regulator pointed out that concerns about consumer debt and financial hardship were growing amid a severe cost-of-living crisis. “More people will find themselves borrowing to make ends meet and then potentially struggling to manage those debts.”
He added that most debt counseling services – free and paid – were designed to help consumers who were already in serious financial difficulty, but poor repayment decisions early in the debt life cycle could cause serious hardship to consumers later.
The watchdog investigated whether robo-advice is an inexpensive and affordable way to help consumers make better repayment decisions when time is on their side.
He found that subjects who received free robotic advice significantly improved their repayment decisions compared to the control group. Among those who accepted help from the robo-advisory tool, the average forgone savings fell by 19.5 percentage points, from 21.9% to 2.4%.
In terms of willingness to pay (WTP) for robo-advice, the regulator found that individuals’ WTP was, on average, greater than the monetary benefit they received from the advice. He said this could be because the subjects were too careful to avoid any possibility of errors rather than eliminating average errors they made unaided, or due to a desire to avoid mistakes. cognitive and psychological costs of resolving repayment issues on their own.
On average, participants were unwilling to pay more to receive an education that explained what the robo-advisor was doing alongside the robo-advisor itself, suggesting that participants did not place any value in this framework, at least to algorithmic explainability.
The regulator also asked if robotic counseling helped subjects learn strategies for optimal debt repayment. It found that better decisions were made after the intervention in all treatment groups, with the biggest difference being for the control group. This group had to overcome more debt management problems without help before reaching the post-intervention phase.
According to the FCA, no imitation learning was detected from the instructional guidance provided with the robo-advisor. To be effective, robo-advice interventions must be repeated each time consumers make choices, he said.
“The potentially very low cost of delivering robo-advice, which can in principle be done via personal devices without the scale constraints of traditional advising, means that repeat interventions may well be feasible.”
Reject robotic advice
Around 25% of consumers refuse the offer of free robotic advice, according to FCA research. Many of them continue to make costly mistakes in their decisions, according to the regulator.
“Some consumers may be reluctant to take advantage of algorithmic help even when told it will clearly improve them. This connects to an emerging literature on trust in algorithms, which explores if and when humans display ‘dislike’ for algorithms “.trust in algorithms can be an important demand-side enabler to help consumers harness technology to navigate complex environments.”
According to the regulator, a second reason consumers might reject free robo-advice is concerns about data privacy.
“People don’t seem to appreciate ‘explainable’ algorithms in this context. This idea connects to other emerging publications on explainable AI and may contribute to a better understanding of where and how explainability matters most to the consumer well-being.
The FCA concluded: “Overall, given the high stakes for many vulnerable consumers, the potential need for ongoing decision support to manage debt and the fact that optimal advice does not depend on preferences or individual risk beliefs, debt management may be a particularly promising area for robo-advisory.”