An Investigation Of Market Structure In The Chinese Banking Industry

The paper investigates the degree of competition and market structure of the Chinese banking industry with structural and non-structural estimation comprehensively.Concentration n-bank concentration ratio (CRn) and HHI Index, which measure concentration, is adopted on the basis of structural estimation. According to the conclusions of empirical research, the Chinese banking industry is moderately concentrated from 1998 to 2007. The concentration level is still high, however, it shows the decreasing trend in this decade. If the two indexes are regarded as measures of monopoly power, then the Chinese banking industry is under monopolistic competition. And the degree of monopoly was decreasing in this decade, that is to say, the degree of competition was increasing in this decade.Nevertheless, structural method is flawed apparently as a measure the degree of competition and is criticized by some scholars in terms of both theoretical and empirical results. With emergence of contestable market theory, people hold different views towards competition. The theory assumes that there are potential new entrants and low barriers to entry and exit at the market and even a monopoly may be forced to be efficient. The paper analyses the Chinese banking industry with this theory and concludes that there are high barriers to entry and exit in the industry, so it's difficult to form the contestable market in China.In view of the flawed structural method, the paper then use a non-structural estimation technique, which is Panzar-Rosse (1977) method, this part is the center of the paper. For the sake of data availability and specific conditions, the model specification is so flexible that there are two problems on some previous Chinese papers. One is the proxy variable is not good enough, the other is that there is a distorting effect on H statistic as Bikker etc. (2006) point if the model specification is based on a price equation. The paper proves these problems indeed have a distorting effect on empirical results, which is doubtful because of misspecification. In order to resolve these problems, the paper references a lot of literatures and adopts the most frequently-used proxy variables. What'more, the model specification is revised based on the revenue equation. Then the paper re-estimates data of the Chinese banking industry with the revised model specification, it's a great contribution of the paper.The paper collects a mass of banking financial data and adds the sample of city commercial banks based on the previous sample of nationwide commercial banks. So the data sample includes 16 nationwide commercial banks and 50 city commercial banks and the period is from 1998 to 2007.The paper indicates that nationwide commercial banks are under conditions of perfect competition from 1998 to 2007 and they are in long-run equilibrium. From the original definition of H statistic, the paper assumes that the four biggest banks in China still posses too much market share by comparison with else nationwide commercial banks, however, they can't earn more with the same cost. So the state of "perfect competition" is just an illusion. As nationwide commercial banks perform more efficiently, competitive conditions of these banks will shift from "perfect competition" to monopolistic competition.And the paper suggests that city commercial banks are under conditions of monopolistic competition from 1998 to 2007 and they aren't in long-run equilibrium. By comparison with city commercial banks, the level of competition is higher at the market of nationwide commercial ones and the phenomenon can be explained by the contestable market theory.Finally, some advices are given below. At first, government shouldn't judge market structure of the banking industry just by market share, etc. Moreover further banking industry reform and going public should be encouraged to improve their own performance. At last, it's preferred to reduce barriers to entry and exit at the market.