Can a personalized price comparison tool improve consumers’ choices in the market for personal loans?

By Erik Berwart, Sean Higgins, Sheisha Kulkarni & Santiago Truffa
Posted on February 25, 2021

Shopping at more than one bank when seeking credit? Not a bad idea: banks tend to offer vastly different rates to the same consumers for similar loans. In the US auto loan market, for instance, the average borrower obtains a car loan that is 1.3 percentage points more expensive than the best rate she could ontain in the market (Argyle, Nadauld, & Palmer, 2020). In the US mortgage market, consumers lose at least $1,000 in interest payments by not shopping around (Woodward & Hall, 2012), not shopping could be due to low levels of financial literacy (Lusardi & Mitchell, 2014), the complexity of financial products and offers hiding the total price of a loan (Alan, Cemalcilar, Karlan, and Zinman, 2018), or high search costs (Agarwal et al., 2020; Argyle, Nadauld, & Palmer 2020).

So far, researchers have been testing a variety of interventions to stimulate consumers to shop around. However, these interventions – based on providing information – have shown limited success. For instance, in the UK, providing savings account holders with information about similar accounts with higher interest rates did not result in consumers switching their savings to these more generous accounts (Adams, Hunt, Palmer, & Zaliauskas, forthcoming) – despite the low switching costs and an average annual benefit of £123 (€163 in 2015) in annual interest income they could have had. Similarly, in Mexico, informing credit card holders about the substantially lower interest rates available by switching to a comparable credit card had no effect (Ponce, Seira, & Zamarripa, 2017).

Discovering the benefits of personal information

Why do information interventions fail to change households’ financial behaviour? For our TFI research project, we hypothesize that offering aggregated, general information does not work, especially not when consumers have made their financial decision, like already having opened a savings account or already having picked a credit card. We aim to test the effects of providing potential borrowers with personalized, just-in-time information instead, relevant to the specific loan they are looking for.

Our research focuses on Chile, a relatively high-income country, where total bank-issued consumer and mortgage debt stands at 103 billion euros as of October 2020, according to the Chilean financial regulator. In the last decade, Chilean regulators have approved reforms aimed at improving consumer decision-making in the market for personal loans. Both quotes and contracts from banks, for instance, must include the effective interest rate (APR), total and monthly payment costs, and a breakdown of contingent and non-contingent fees in a standardized format.

Despite these regulations, we find that Chilean consumers — much like consumers worldwide — rarely shop for personal loans. We take a closer look at search patterns from loan seekers by over 2.3 million potential borrowers between 2013 and 2016. After receiving an interest rate quote and loan offer from one bank, only 3% of consumers searched for an additional quote from a different bank in the next 30 days. After an initial rejection, only 10% of consumers searched for a loan at another bank in the next 30 days.

Our study will recruit participants online, targeting ads to users who search for keywords related to consumer loans in Chile. We will show them ads offering them tools to evaluate and compare the costs of their loan in the market. Our expectation is that this recruitment strategy will be particularly effective in times of lockdowns and stay-at-home orders, as more people are conducting their financial operations online rather than in person. After answering a short survey, participants will view a loan price comparison tool, providing personalized, just-in-time information. Specifically, using real-time data from the financial regulator and personal information entered by the consumer (including monthly income, place of residence, desired loan type, loan size, and maturity), it displays information about interest rates given to similar consumers for similar loans in the previous six months. By interacting with the tool, the consumer can test how different interest rates change their monthly and total loan costs (Figure 1).

Figure 1. The price comparison tool: consumers can change their personal and loan characteristics and receive personalized information about the interest rates that similar consumers have received for similar loans over the last six months.

Using the financial regulator’s continuously updated data on all loans in Chile, we will track the eventual loan outcomes for all participants in the study. Subsequently, a follow-up survey will help us measure how our tool impacted people’s search across banks when making their loan decision. At how many and which banks did they search? We will thus measure the impact of the loan price comparison tool on changes in consumer search behaviour, and ultimately, how that search behaviour influences financial health and well-being.

First preliminary analysis

The first of our two preliminary studies was a randomized pilot from November 2019 to March 2020, in which we tested our online ads campaign, survey, and a prototype of the loan price comparison tool. Roughly 4% of the people who saw our ad clicked on it and consented to answer questions in our survey, and 53% of them completed enough survey questions for us to continue the experiment.

In this pilot, we also looked at the characteristics of our expected sample, comparing them to characteristics of the general population of loan recipients in Chile. Figure 2 shows the various characteristics of pilot participants who took out loans up to three months after participating in the pilot, and compares them to the characteristics of all borrowers in Chile who received loans over the same time period. The percentage of women, the percentage of residents of the Metropolitan Region (the capital of Chile), and the distribution of annual income look quite similar for both groups.

Meanwhile, pilot participants do seem to be slightly younger than the general population of borrowers, which is to be expected when taking into account the online recruitment process. The overall similarity between our pilot participants and the general borrower population in Chile suggests that the impact of our tool will be similar to the expected effects for others who do not search for loans online, but who could be shown the tool in other contexts.

Figure 2. Comparison of borrower characteristics, all borrowers vs. pilot participants who took loans

Second preliminary analysis

In our second study, we qualitatively tested the price comparison tool in focus groups in October 2020. 65 Chilean participants provided us with comments on the user experience, design, and perceived usefulness of our tool. The results from these focus groups prompted us to add more guidance, tips, and directions to our tool, as some participants had trouble understanding the graphs and data.

While most participants thought the price comparison tool conveyed useful information, many required assistance in reading and interpreting the graphs. Some participants from low-income backgrounds also reported that they usually just accepted the first offer they received without considering the possibility of shopping around.

Next steps

After making additional improvements to the price comparison tool and survey questionnaire based on insights from the pilot and focus groups, we will launch the experiment later this year and plan to report preliminary results by the end of 2021. Stay tuned!

References

  • Adams, P., Hunt, S., Palmer, C., & Zaliauskas, R. (forthcoming). Testing the Effectiveness of Consumer Financial Disclosure: Experimental Evidence from Savings Accounts. Journal of Financial Economics. http://web.mit.edu/cjpalmer/www/AHPZ-Savings.pdf
  • Agarwal, S., Grigsby, J., Hortaçsu, A., Matvos, G., Seru, A., & Yao, V. (2020). Searching for Approval (NBER Working Papers 27341). https://doi.org/10.3386/w27341
  • Alan, S., Cemalcilar, M., Karlan, D., & Zinman, J. (2018). Unshrouding: Evidence from Bank Overdrafts in Turkey: Unshrouding: Evidence from Bank Overdrafts in Turkey. The Journal of Finance, 73(2), 481–522. https://doi.org/10.1111/jofi.12593
  • Argyle, B., Nadauld, T., & Palmer, C. (2020). Real Effects of Search Frictions in Consumer Credit Markets (NBER Working Papers 26645). https://doi.org/10.3386/w26645
  • Kulkarni, S., Truffa, S., & Iberti, G. (2020). Removing the Fine Print: Standardization, Disclosure, and Consumer Loan Outcomes [Unpublished manuscript]. https://www.sheishakulkarni.com/s/DisclosureandDefault-August.pdf
  • Lusardi, A., & Mitchell, O. S. (2014). The Economic Importance of Financial Literacy: Theory and Evidence. Journal of Economic Literature, 52(1), 5–44. https://doi.org/10.1257/jel.52.1.5
  • Ponce, A., Seira, E., & Zamarripa, G. (2017). Borrowing on the Wrong Credit Card? Evidence from Mexico. American Economic Review, 107(4), 1335–1361. https://doi.org/10.1257/aer.20120273
  • Woodward, S. E., & Hall, R. E. (2012). Diagnosing Consumer Confusion and Sub-Optimal Shopping Effort: Theory and Mortgage-Market Evidence. American Economic Review, 102(7), 3249–3276. https://doi.org/10.1257/aer.102.7.3249