Financial liberalization has been adopted by most countries as their financial policy. However, various risks arise from financial liberalization, which present a challenge to financial institutions and supervising sectors. The Basle Committee on Banking Supervision announced its Amendment to the Capital Accord to Incorporate Market Risks in 1996 and 2001, allowing financial institutions to develop internal models to value the market risk of securities. This will be enforced in Taiwan by implementing the capital adequacy requirement by 2006. Facing competitive pressure and international regulations from the Basle Committee on Banking Supervision within the banking industry, it is important to evaluate the bank efficiency index and efficiency effect incorporating account risk. It is also very meaningful to investigate factors influencing the efficiency of Taiwan banks.
Bank efficiency research is abundant. There are two major streams in this research, parametric programming and nonparametric programming. In parametric programming, Bell and Murphy (1976), Ferrier and Lovell (1990), Berger and Humphrey (1991), McAllister and McManus (1993), and Beard et al. (1991, 1997) focus on estimating functional characteristics and economies of scale and scope while evaluating bank performance. However, Elyasiani et al. (1994), Favero and Papi (1995), and DeYoung and Hasan (1998) change their focus from a one-stage to a two-stage approach in analyzing efficiency effects. In the nonparametric programming approach, most researchers use the data envelopment analysis (DEA) to discuss total productivity and efficiency of banks, such as Aly et al. (1990), Grabowski et al. (1993), Favero and Papi (1995), and Fukuyama et al. (1999). Ferrier and Lovell (1990), Bhattacharyya et al. (1997), and Huang and Wang (2000) examine bank technical efficiency by comparing the parametric and nonparametric programming approaches.
Researchers have recently focused on the relationship between bank efficiency and risk when studying a bank's efficiency. There are two issues of bank efficiency and risks. One treats risk as exogenous so as to analyze efficiency effects, such as Cebenoyan et al. (1993), Elyasiani et al. (1994), Berger and DeYoung (1997), and Chang (1999), who analyze a bank's efficiency effects in terms of the bank incorporating risk effects, including nonper-forming loans (NPLs), allowance for loan losses, and risky assets. Results from these research studies show that the efficiency level is significantly correlated with risk indicators.
Another way is to implement risk indicators into the production process. Mester (1996) uses the stochastic econometric cost frontier approach to account for quality and default of bank output on bank efficiency in the Third Federal Reserve District. Mester identifies that banks in the United States operate at cost-efficient output levels. Hughes (1999) shows the importance of incorporating endogenous risk into the production analysis and measures bank profit efficiency and scale economies in U.S. commercial banking by incorporating risk and financial structure in the model. Altunbas et al. (2000) use data on Japanese banking from 1993 to 1996 to test the impact of risk and quality factors on bank cost using the stochastic cost frontier methodology. They find that optimal bank size is considerably smaller when risk and quality factors are taken into account. Hughes et al. (2001) explore how to incorporate bank capital structure and risk taking into production models by using 1994 U.S. data to search for bank efficiency scales. They find that the level of financial capital and the risk-taking have significant influences on scale efficiency estimates.
Based on these findings, most research studies on bank efficiency have primarily focused on economies of scale and scope, total productivity, and bank efficiency by comparing the parametric and nonparametric programming approaches and efficiency effect, but the interrelationship between risk and efficiency has received little attention in banking literature. …