Access to Credit: Part 2
This is part 2 of the two part series on access to credit. In part one, we went over the theory of credit rationing which is meant to provide further motivation for expanding credit markets in developing nations — specifically, finding solutions in the short run for expanding credit since property rights for collateral is a slowly changing process. If you haven’t read part one, you can find it here.
In part two we will be discussing some more theory on information asymmetry, summarizing and highlighting empirical results from field experiments on observing information asymmetries and how this ties into The Digital Reserve’s model which allows borrowers to select their own interest rate on their loan.
Credit rationing is predicated on the belief that there is information asymmetry in the market. There are three types of these asymmetries in credit markets, which are:
- Ex ante adverse selection
- Ex ante moral hazard
- Ex post moral hazard
Ex ante adverse selection is when a lender chooses the wrong borrower in a pool of borrowers where some are safe and some are risky. Ex ante moral hazard take place when a borrow deliberately uses the money in such a way the jeopardizes repayment. Finally ex post moral hazard take place when a borrower deliberately does not repay their loan although they have the ability to do so.
It has been incredibly difficult to observe information asymmetries in practice and different kinds of asymmetries have different implications for markets and policy. Ex ante adverse selection (or unobserved risk type) and ex ante moral hazard (also known as unobserved anticipated effort) fall under a category of hidden information, while ex post moral hazard falls under hidden action. Hidden information and hidden action call for different types of market and policy responses. According to Karlan and Zimmer, hidden information could require subsidy responses and loan guarantees, while hidden action could require legal reform on liability and garnishments.¹
In an attempt to disentangle and find a causal link between hidden information and/or hidden action Karlan and Zimmer designed an experiment. Using a South African lending company whose average APR is 200%, the experiment is designed to randomize interest rates along three distinct points in the lending process, which are:
- The mail offer
- The actual rate on the loan
- Interest rates offered on future loans if repaid on time
This experiment allows the researchers to compare groups of borrowers in multiple ways. The first comparison is groups that selected in at different mail offer interest rates but had equivalent interest rates moving forward in the lending process, and the second is those who were selected in at identical mail offer interest rates but faced different repayment incentives moving forward. The prior allows us to observe hidden information, while the latter allows us to observe hidden action.
How? Those who select in at different mail order interest rates are perceived to be different risk types. If they face the same repayment incentives moving forward, the only thing that is different is their initial risk type. If their default is different while facing the same repayment incentives, we will be able to attribute that difference to initial risk type. If however they are selected in at the same initial mail offer interest rate they will be perceived to be the same risk type, the only difference being their incentives of repayment. If a similar risk type lender has worse incentives to repay and they default more often as a result, we can attribute that difference to hidden action. Figure 1 below shows a graphical representation of the experiment.
The empirical results of this paper show that there is weak evidence of hidden information, while there is strong evidence of hidden action. This study showed that in this particular South African lending market between 13% and 21% of defaults were attributed to hidden action. Another interesting observation was that when the interest rates were lowered for those who selected in at a higher offer rate, only 3% of this group asked for more money.
So what does this mean in terms of this study? That changing the incentives of repayment induces a stronger behavioral effect than higher initial interest rates — meaning, at least in terms of this study, the theory that higher interest rates will lead to only riskier borrowers requesting loans at these rates does not tell the complete story. Also, the idea that lower interest rates will induce riskier borrowers to borrow more than they can afford is not completely true.
While we cannot speak in absolutes, we can try to make some assumptions. The first is that our starting interest rate (constrained within some interval e.g. 13% to 45%), while potentially constraining some borrowers from borrowing, does not necessarily attract riskier borrowers.
This assumption may seem like we are leading to the conclusion that we should charge higher interest rates. In actuality, we believe that the opposite is true. We believe that fair interest rates will reduce potential hidden information from unobserved effort, if there is any. Giving the borrower some power in determining their interest rate in The Digital Reserve Network may provide us with the positive effect mentioned above. In addition knowing that it is possible to select a lower interest rate with successful repayment may drastically reduce the hidden action that was observed in the study explained above.
Where should we take caution?
While the potential for the positive effects of The Digital Reserve and other micro-lending platforms in developing nations is huge, we cannot ignore that the negatives. Just as the paper has shown, hidden action requires strong policy responses. The problem is that many developing nations have weak legal institutions, poor property rights and not well established garnishment and litigation laws. This creates a window for borrowers to neglect their repayment obligations because of weak enforcement mechanisms. This reality requires a strong market mechanism where the legal institutions are currently incapable of doing so and the market mechanism should not be credit rationing.
Please post your contemplation, discussion and solutions below. It takes a community. Also, join our newsletter and telegram group to learn more about how we plan to provide solutions to the lending markets.
- Dean Karlan and Jonathan Zinman, Observing Unobservables: Identifying Information Asymmetries with a Consumer Credit Field Experiment, The Econometric Society, Econometrica, Vol. 77, №6 (Nov., 2009), pp. 1993–2008 Stable URL: http://www.jstor.org/stable/25621388
Author: Troy W — Entrepreneur , Public Policy and Statistics grad student, interested in research and applications of Blockchain.