- Comment on "Bank Competition and Access to Finance:International Evidence" by Thorsten Beck, Asli Demirgüç-Kunt, and Vojislav Maksimovic
G21, L10
bank competition, bank concentration, financing obstacles
There are two commonly acknowledged, albeit conflicting predictions on the effect of banks' market power on firms' access to credit. The structure-performance hypothesis asserts that more market power leads to lower supply at higher prices. In contrast, the information-based hypothesis asserts that more market power increases bank lending to informationally opaque borrowers. In "Bank Competition and Access to Finance: International Evidence," Beck, Demirgüç-Kunt, and Maksimovic (2004, this issue of JMCB) (henceforth BDM) investigate which hypothesis survives empirical scrutiny. This is a research question that will never disappoint, as the results will be important no matter which hypothesis prevails in the end. It is not the first time the question has been addressed in the literature, but BDM have made the most comprehensive study on this issue. Policymakers in developing countries no longer have to rely mostly on results from the U.S. banking market when they formulate policies for their emerging banking markets. They can now learn a lesson from the experience of more than 6000 companies in 74 countries.
BDM's results suggest that in more concentrated banking markets (high market power), firms face higher financing obstacles. BDM interpret this finding as evidence in support of the structure-performance hypothesis and emphasize the negative effects of bank market power. In my comments, I would like to reinterpret these results, emphasizing the positive effects of market power, and express some concerns about the method of the paper. [End Page 649]
1. Market Concentration can be a Good Thing
According to the information-based hypothesis, banks with more market power can invest in costly relationships with informationally opaque borrowers because they expect to share in the surpluses of future business. Banking regulations that restrict the ways in which banks can interact with borrowers (restrictions on securities underwriting, equity ownership, etc.) would limit future business opportunities and consequently, the rents that banks could extract from borrowers. The end result would be less relationship lending. The specification that includes the Concentration x Restrict interaction in Table 5 provides strong support for this hypothesis. Market concentration alleviates firms' financing obstacles but high levels of banking restrictions eliminate the positive impact of concentration. In other words, the same level of market concentration implies less market power when there are activity restrictions and less market power leads to higher financing obstacles.
Other results in Tables 5 and 6 also support my interpretation of market concentration as a positive force. Market concentration reduces financing obstacles unless the country has poorly developed institutions, the government interferes in the banking sector, banks are unable to use a credit registry to exchange information about borrowers, or the banking system is closed to foreign banks or dominated by public banks. What is a policymaker in a transition economy to do? Reduce market concentration as BDM seem to suggest or deal with other problems that eliminate the positive impact of concentration? Is it more beneficial to develop the country's institutions or to pass laws that break up large banks? Unless the authors have a paradigm in mind in which the latter alternative dominates, their negative stance against bank market power seems unjustified.
2. Can There be More than Two Hypotheses?
BDM's two hypotheses are not necessarily the only plausible explanations for the results they report in their paper. In this part of my comments, I will provide two new hypotheses.
Sharpe (1990) has shown that a bank that is better informed about a borrower has a comparative advantage over uninformed lenders. Dell'Ariccia, Friedman, and Marques (1999) and Dell'Ariccia (2001) explain how the comparative advantage of incumbents affects the banking market structure. It is a winner's curse story in which a new entrant knows less about the creditworthiness of borrowers in a market than incumbents and therefore attracts lower-quality customers on average. Because new entrants can anticipate their disadvantage, the winner's curse becomes a barrier to entry. The incumbents' advantage comes from the fact that they have been around...