Learning from experience in financial services

28 August 2019

Experience is the most effective teacher, but how can consumers learn by experience when some of the most important financial products are bought just once or twice in a lifetime?

Imagine a world where you have no idea about how fast you’re driving. No dashboard speedometer. No road signs displaying the speed limit. Every day you return home and rifle through the mail looking for a speeding ticket. Your heart sinks as you see the familiar brown envelope and try to remember when you might have exceeded the speed limit.

Fortunately for us, we don’t typically learn about the speed limit through painful experience. We are taught the rules as part of learning to drive and receive helpful immediate feedback as we apply these in practice - you check your speedometer, compare it to the speed limit signs on the side of the road and adjust your behaviour accordingly. Your adjustment is reflected instantly in the drop on your speedometer. If you apply the rules and take in the information provided by the speed limit signs and your speedometer, it is painless and cost-free.

Now let’s think about how we learn in financial services. The equivalent to a speed limit sign in consumer financial decision making are risk warnings. These are meant to warn consumers about the risks of certain products, such as rising or falling interest rates. But evidence suggests they are often ineffective.

One reason they might be ineffective is that risk warnings are merely summary descriptions (or ‘representations’) of a situation. Psychologists call this learning from description. This path to knowledge is important: we access accumulated wisdom that we wouldn’t be able to learn if we only relied on our own individual experiences.

But this is only one of the paths to knowledge. The other path to knowledge is learning through experience, which involves an interaction with the environment. Compare learning to obey the speed limit with learning to ride a bike. No amount of ‘learning from description’ is going to be enough. In the end, you can only learn to ride a bike by experience – including a few scraped knees.

But while this type of learning by experience comes with some perhaps painful costs, it is also arguably a more effective path to knowledge.

Recent research suggests consumers give greater weight to information gained through experience than to information acquired through description. Moreover, consumers who learn from experience make better calibrated predictions about the future compared to consumers who learn from description.  And in one hypothetical investment setting, facing the same choice repeatedly, even without feedback, still led consumers to make better investment decisions. Practice makes, if not perfect, then at least better.

consumers give greater weight to information gained through experience than to information acquired through description.

As we’ve discussed, when it comes to the design of financial products and services, institutions typically take the speed limit sign approach. They provide some type of description upfront, via advertising, disclosure and/or warnings, about the product features and risks.  All are convenient tools because they’re relatively simple to implement and do not require changing the underlying structure of a financial product or service.

But research from Warwick and Nottingham University suggests consumers rarely learn how to effectively use financial products, and when they do, it is not the descriptive method of learning that really works, but practical experience. In other words, learning to use financial products is more like learning to ride a bike.

Let’s look at an example. Past research (done before the prevalence of direct debit), had shown the longer a consumer has a credit card, the less likely they are to incur a late fee. Paying a fee the previous month reduced the likelihood of paying a fee this month by 40 per cent.  The mechanism behind this learning was assumed to be that over time consumers were learning from experience and consistently taking corrective action, such as monitoring their spending more closely. However, researchers using more recent data found rather than taking sustained corrective action, consumers were simply switching over to direct debit, thereby avoiding the need to remember to repay altogether. With late fees, consumers are able to take a one-shot action (switching to direct debit) to avoid the penalty fee. In the case of other fees for credit cards or other products, there often is no such mechanism.

In theory, credit cards are an ideal financial product because they provide frequent opportunities to learn from experience.

In contrast, other products, such as pensions and mortgages, are typically ‘one-shot decisions’ which don’t give consumers many opportunities to learn from experience. Furthermore, they can be much more consequential. As well as being infrequent, one-shot decisions are often binding, diminishing or removing entirely the ability to respond to feedback. 

products, such as pensions and mortgages, are typically ‘one-shot decisions’ which don’t give consumers many opportunities to learn from experience.

Let’s look more closely at mortgages. Around 30% of consumers in the UK were found to have taken out a mortgage when a better one was available to them.

These households paid £550 more per year on average than they needed to.

The FCA found one of the main reasons consumers chose a more expensive mortgage was because they didn’t realize they might qualify for a better loan because the eligibility requirements for the better product were unclear.

A lack of transparency in the product eligibility requirements, combines with the relatively low frequency of the transaction to make it difficult for consumers to learn which mortgage is best for them. They do not easily understand their mistakes and have few opportunities to learn by ‘having another go’.

The knee-jerk response to this problem is often to improve financial education. However, formal financial education interventions (which typically use description based learning) have yielded discouraging results. In our analogy, this is like teaching people to learn to ride a bicycle by sitting in a classroom.

A meta-analysis using 201 prior studies found interventions that attempted to improve financial literacy had very little effect on financial behaviours. For example, in a mandated counselling program implemented in Chicago, high risk borrowers could choose to either attend mortgage counselling or take out a less risky mortgage. When faced with this choice many consumers opted to skip counselling and chose a less risky product. Having to invest more time and effort in the short term (for counselling) dissuaded consumers from making bad choices in the future, although not necessarily as a result of learning.

So how might we help consumers to better learn, without paying to learn by expensive mistakes? Presently, the design of financial products and services make it difficult for consumers to fully understand them and learning how to use them effectively can be costly.  

For firms, this means not just better descriptions of products - though that is important. It means incorporating the basics of what effective learning looks like from psychology – timely feedback and enabling consumers to learn from experience (practice!). This might take the form of ‘just-in-time’ financial education – where information is provided at the time individuals are actually dealing with the issue -  so people can incorporate it into their decision making. 

And where appropriate, it means using clever design and technology to automate behaviours, like repayments, rather than relying on increasingly overloaded memories to do the right thing. 

Gamification is another potentially promising avenue. For example, a repayment sandbox to see whether consumers could afford mortgage monthly payments on top of their existing expenditures. This would simulate the experience of paying a new/different mortgage, receiving notifications and real-time feedback – showing when your account would go overdrawn. One could also use this tool to 'fast forward' consumers through mortgages that are back loaded, to the point in time when monthly payments increase substantially. In this way, they could experience (in a safe sandbox environment) the full lifetime of the product they may be about to buy and compare the experience of other products.

For regulators, this might involve moving away from specifying the ‘means’ to achieve a particular outcome (eg risk warnings) and shift instead to performance based regulation that focusses more on the outcome.

Focussing on the outcomes in theory would incentivise firms (and their creative potential and resources) to find the best way to give consumers a fair chance to learn how to use a financial product - from both description and experience - and without making expensive and hard to reverse mistakes.

Get Insight in your inbox