Research

Invited for a major Revision by MSOM

Rezaei, B., Dasu, S., Ahmadi, R.


Keywords: Joint Liability, Strategic Default, Monitoring, Social Networks, Graph Theory
Abstract:
(1) Problem definition: Microfinance institutions (MFIs), ubiquitous in the developing world, have had noticeable successes lending to borrowers at the bottom of the pyramid. One of the most practiced lending methods by MFIs is group lending with joint liability. We study the sustainability of group lending under joint liability contracts when borrowers are subject to strategic default. The benevolent lender (a non-profit) optimizes the borrower's welfare while covers her costs of lending, and social ties between borrowers serve as monitoring tools. (2) Academic/Practical relevance: The observation that social relationships promote cooperation is long-standing in the socio-economic literature. Group lending implicitly leverages borrowers' social ties by having self-selected groups of borrowers enter into a joint liability contract. We model borrowers' social network structures as patterns of peer monitoring and explains how different network structures may affect the lending outcome. To the best of our knowledge, our work is the first theoretical study of this problem. (3) Methodology: The methodology employed in this paper is the reverse game theory or mechanism design. (4) Results: The results show that a higher average centrality and connectivity in the borrower network, as well as a higher degree of borrower discount factor, allow for more favorable contract terms (i.e., a higher loan ceiling and a lower repayment amount). Further, we argue that while for smaller groups, the highly decentralized ring networks may have higher performance (i.e., a higher chance of game continuation and a higher repayment rate) than the highly centralized star networks, for larger groups, star networks outperform the rings. Intuitively, although decentralized networks are less susceptible to the failure of one member, they have a disadvantage in the level of monitoring they provide. This disadvantage manifests itself to a greater degree in larger groups. (5) Managerial implications: The setting studied in this paper can emerge in a variety of team cooperation or social dilemma contexts that deal with monitoring, and thus the results are valid wherever the interests of the individual and of the group are not aligned and agents exert externalities on one another.






Decision Analysis 14 (3), 204-225.

Rezaei, B., Dasu, S., Ahmadi, R. (2017)


Keywords: Microcredit, Joint Liability, Strategic Default
Abstract: We develop a model of repeated microcredit lending to study how group size affects optimal group-lending contracts with joint liability. In the setting being studied, a benevolent lender provides microcredit to a group of borrowers to invest in projects. The outcome of each risky project is not observable by the lender; therefore, if some of the borrowers default on their loan repayments, the lender cannot identify strategic default. The group will be entitled to a subsequent loan if total loan obligation is met. We characterize the optimal contract and determine the optimal size of the borrowers' group endogenously. We find that, although joint liability contracts are feasible under a smaller set of parameter values than individual liability contracts, joint liability has positive effects on the borrowers' repayment amount and welfare. Our analysis also suggests that group size should increase with project risk. Furthermore, we analyze the effect of partial joint liability, less severe punishment, and project correlation on the feasibility and characteristics of joint liability contracts. Our results show that, first, although partial joint liability has a negative effect on the borrowers' repayment amount and welfare, it can increase the loan ceiling of joint liability when collusion is not as likely, or when borrowers have high discount factors. Second, less severe punishment does not affect the borrowers' repayment amount or welfare, but decreases the loan ceiling of joint liability. However, these negative effects created by partial joint liability and less severe punishment on the borrowers' repayment amount, borrowers' welfare, and loan ceiling can be offset by forming larger groups. Third, we also found that project correlation allows a higher loan ceiling in larger groups.





Working Paper (to be submitted soon)

Rezaei, B., De Jaegher, K.


Keywords: Menu Pricing, Single-Crossing Property, Reference-Dependent Preferences
Abstract: We study optimal pricing strategy of a monopolist who faces consumers that have heterogeneous private tastes, have reference-dependent preferences, and are subject to loss aversion. There is asymmetric of information and monopolist does not observe the consumers’ valuations. Assuming that the monopolist can make consumers expect to buy the desired variety of the good, and that these expectations determine the consumers’ reference points, we obtain two main results. First, with expectation-based loss aversion, menu pricing is possible even if the single-crossing property is violated (high-valuation consumers do not have a larger marginal utility of quality than low-valuation consumers). Second, when firms face consumers with expectation-based loss aversion, menu pricing may become more desirable to the monopolist compared to selling only to high-valuation consumers.

Working Paper (to be submitted soon)

Rezaei, B., De Jaegher, K.


Keywords: Signaling Games, Single-Crossing Property, Reference-Dependent Preferences
Abstract: We develop a model of signaling with senders who have reference-dependent preferences and are subject to loss aversion. As is well known, the single-crossing property, saying that high types have lower marginal cost of signaling, is a necessary condition for the existence of separating equilibria in signaling games. However the single-crossing property is a very restrictive condition. In this paper, assuming that expectations form the sender's reference point, we show that with expectation-based loss aversion, the single-crossing property is no longer a necessary condition for separating equilibria to exist.