Bank-Deposit Contracts Versus Financial-Market Participation in Emerging Economies

Abstract

The financial sector of emerging economies in Africa is characterized by a noncompetitive banking sector that dominates any direct participation of agents in asset markets. We formally identify “market inexperience” as an explanation for agents’ willingness to pay high banking fees rather than to participate in asset markets. Whereas experienced agents choose ex ante investments that result, through trading on the future asset market, in the optimal (second-best) allocation, inexperienced agents are ignorant about the possibility that future market equilibria can improve welfare upon an autarkic investment. As a consequence, a monopolistic banking sector can exploit these agents because their only outside option is an autarkic investment project.

KEY WORDS:

• asset market
• asymmetric information
• bounded rationality
• demand deposit contract
• emerging economies

Acknowledgements

The author is grateful to Alex Ludwig and two anonymous referees for valuable comments and suggestions.

Notes

1. In addition to the articles cited, see, for example, Bolton (2002), Bryant (1980); Jacklin (1987, 1993), Levine (2002), Rajan (1992), Stiglitz (1985), Tadesse (2002), Wallace (1988, 1990), Weinstein and Yafeh (1998), and references therein.

2. Diamond and Dybvig’s (1983) original model of bank runs only describes the “possibility” of bank runs in the sense that there exist multiple Nash equilibria such that one equilibrium implies a bank run. Subsequent models of bank runs admit for a unique Bayesian Nash equilibrium such that bank runs occur with a strictly positive probability in this equilbrium (Goldstein and Pauzner 2005; Postlewaite and Vives 1987; Rochet and Vives 2004; Zimper 2006).

3. We impose the assumption of risk neutrality for three reasons. First, it is technically convenient because all formal results are easily obtained from simple linear optimization arguments rather than from first-order conditions as in Hellwig (1994). Second, it avoids the problem of Diamond and Dybvig’s (1983) original strongly risk-averse setup where information constraints are not binding (cf. Zimper 2013a); by contrast, in our risk-neutral setup, the first- and the second-best allocations are different. Third, it emphasizes that asymmetric information but not risk sharing drives the optimal allocation in our nonautarkic large economy. Also compare the Remark at the end of the section headed “The Optimal Allocation.”

4. That the individual probability of a depositor to turn out as a high type coincides (almost surely) with the fraction of high types in the population is, for a countably infinite population, justified by the law of large numbers together with the assumption that depositors’ types are independently and identically distributed. While such justification is not at hand for the continuous population of our model (Duffie and Sun 2007; Judd 1985), I simply follow here the literature and misquote the law of large numbers in the “usual way.”

5. We ignore here the possibility of a bad “ bank run” Nash equilibrium in which high-patience types would also withdraw in period 1. The possibility of such strategic bank runs may be, for example, excluded by “suspension of convertibility” or a central bank serving as “lender of last resort” (see, e.g., Diamond and Dybvig 1983).

6. Note that the bank’s demand-deposit contract is perfectly price discriminating with respect to the assymmetric information about agents’ patience types.

Additional information

Funding

Financial support from ERSA (Economic Research Southern Africa) is gratefully acknowledged