«FRBSF Working Paper 2002-01 DRAFT Please do not quote without permission Operating Performance of Banks Among Asian Economies: An International and ...»
FRBSF Working Paper 2002-01
Please do not quote
Operating Performance of Banks Among Asian Economies:
An International and Time Series Comparison
Simon H. Kwan
Federal Reserve Bank of San Francisco
After controlling for loan quality, liquidity, capitalization, and output mix, per unit bank
operating costs are found to vary significantly across Asian countries and over time. Further
analysis reveals that the country rankings of per unit labor and physical capital costs are highly correlated, suggesting that there exist systematic differences in bank operating efficiency across Asian countries. However, this measure of operating efficiency is found to be unrelated to the degree of openness of the banking sector.
Asian bank operating costs were found to decline from 1992 to 1997, indicating that banks were improving their operating performance over time. Since 1997, the run-up in operating costs coincided with the Asian financial crisis, suggesting that banks were incurring additional costs in dealing with their problem loans while output was declining simultaneously.
Moreover, the labor cost share is found to decline significantly between 1997 and 1999, indicating that banks were able to cut their labor force after the financial crisis but were less flexible to reduce physical capital input. Furthermore, significant differences in labor cost share are detected across countries, suggesting that different countries have different bank production functions. The variations in labor cost share are significantly positively related to the country’s financial services wage rate, suggesting that banks using relatively more labor in a particular country is due to the labor force productivity, rather than labor being cheap.
Please send correspondence to: Simon Kwan, Research Advisor, Federal Reserve Bank of San Francisco, 101 Market Street, San Francisco, CA 94105, U.S.A.
Telephone: 415-974-3485; Fax: 415-974-2168; E-mail: email@example.com The findings, interpretations, and views expressed in this paper are those of the author only and do not necessarily represent the views of the Federal Reserve Bank of San Francisco or the Federal Reserve System. Thanks are due to the anonymous referee, participants at the Federal Reserve System Conference on International Economic Analysis and the 2001 Australasian Finance and Banking Conference for helpful comments. Irene Wang pro
1. Introduction In banking research, there is a large body of literature studying the efficiency of financial institutions, including both scale and scope economies and an increasing focus on X-efficiency.
These studies strongly suggest that X-efficiency in banking is large, typically accounting for twenty percent or more of costs, and dominate scale and scope efficiencies [Berger and Humphrey (1997)]. While banking efficiency research has been conducted quite extensively for U.S. commercial banks, and to a lesser extent, for European financial institutions, relatively little research has been done to investigate banking production among Asian financial institutions.
Furthermore, from an international perspective, there is very little research on comparing banking production across Asian countries.
A number of Asian economies, such as Hong Kong, Singapore, and South Korea, are considered highly competitive and demonstrate somewhat different productivity dynamics, including the substitution between labor and capital, than western economies. At the same time, there are considerable variations in the structure of the financial services industry, the scope of banking supervision and regulation, the development of local financial markets, and the openness of the banking industry to foreign competition across these Asian countries.1 Moreover, the coincidence of banking and currency problems associated with recent Asian 1 See for example, Beck, Demirguc-Kunt, and Levine (1999) and Demirguc-Kunt and Levine (1999) for cross-country comparisons of financial developments, Barth, Caprio, and Levine (2001) for the regulation and supervision of banks around the world and cross-country comparisons of the structure of the financial services industry, Claessens and Glaessner (1998) and Barth, Caprio, and Levine (2001) for the openness of the banking sector to foreign competition.
Asian economies [Glick and Hutchison (2000)].
To shed light on the micro production of banking outputs in Asia, this paper empirically examines the banking industry’s per unit operating costs in seven Asian economies, including Hong Kong, Indonesia, Malaysia, Philippines, Singapore, South Korea and Thailand, from 1992 to 1999. First, the cross-country comparison provides insights into the relative efficiency of the banking industry among these Asian economies. Second, the time series analysis reveals how banking production, and in particular the choice between labor and physical capital, evolve over time; it also captures the effect of the Asian financial crisis on banking production.
The rest of this paper is organized as follows. Section 2 briefly discusses existing banking literature that is related to this study. Section 3 describes the data used in the analysis.
The empirical analysis and findings are presented in Section 4. Section 5 summarizes and concludes this study.
2. Literature Review The efficiency of the financial services industry has long been a focus of banking research. The amount of attention that banking efficiency research has received is understandable. Their findings have obvious implications for bank management who seek to improve operating performance, and for policy makers who concern about banking competition and bank safety and soundness. Research that shows a positive relation between finance and growth [see for example, Levine (1999), Levine and Zervos (1998), and Beck, Levine and Loayza (2000)] prompts additional studies to focus more narrowly on the banking system
of this line of inquiry is to compare the efficiency of the banking system across countries.
Berger and Humphrey (1997) survey over 100 studies that apply frontier efficiency analysis to financial institutions in 21 countries. In studying cross-border banking performance, Berger, DeYoung, Genay, and Udell (1999) review the literature that provide international comparison of banking efficiency. In short, studies of the efficiency of Asian financial institutions are relatively few, compared to research on U.S. and European banking. Altunbas, Liu, Molyneux and Seth (2000) study the efficiency and risk in Japanese banking. Okuda (2000) estimates the cost function of Philippines domestic banks for the pre-Asian crisis period.
Leightner (1999) uses linear programming techniques to evaluate the performance of Thailand’s finance and securities companies over the 1990-1995 period. Huang, Fu, and Huang (1999) examine the efficiency of Taiwan’s farmers’ credit union.
One obstacle in researching the efficiency of Asian banks is the lack of publicly available data for non-publicly traded Asian financial institutions. Together with the fact that in certain Asian countries, a few large banks dominate the banking industry, researchers may find too few observations to estimate the efficient frontier in a given country. Furthermore, X-efficiency is a relative concept which involves comparing banks to the best practice institutions. The comparison is meaningful only when all the sample banks have equal access to the same production technology. This assumption is unlikely to hold in a cross country setting. Thus, in this paper, rather than estimating a global efficient frontier, we rely on accounting measures to compare operating efficiency across countries after controlling for bank characteristics and output mix. Although using accounting variables is arguably less preferable to using the frontier
fairly strong and robust.
Claessens and Glaessner (1998) also compare bank performance across several Asian countries in their study of the internationalization of financial services. However, their focus is more on the profit side rather than the operating side of financial institutions. One similarity between their study and this paper is that we also try to link operating efficiency to the openness of the financial sector. This is motivated by the work of Levine (1996), Walter and Gray (1983), Gelb and Sagari (1990), and Classens, Demirguc-Kunt, and Huizinga (1998) which found that foreign bank entry seems to have a positive effect on domestic banking markets. Finally, Barth, Caprio, and Levine (2000) study banking systems around the world, focusing on the question of whether regulation and ownership affect performance and stability.
3. Data Description The data for this study comprise the population of commercial banks in Hong Kong, Indonesia, Malaysia, Philippines, Singapore, South Korea and Thailand that are listed in the IBCA bank credit rating agencies Bankscope database, which reports published financial statements from financial institutions worldwide. 2 To focus on commercial banks and to maintain consistency across countries, only commercial banks that make commercial loans and accept deposits from public are included in the analysis. Therefore, deposit taking companies, 2 Bankscope does not cover foreign branches and agencies that are wholly owned by foreign banks. Hence, most of the sample banks are locally owned. Nevertheless, it is still possible for foreigners to own stocks of a local bank, making the ownership status somewhat ambiguous. Unfortunately, data on ownership status are not readily available.
shows the ratio of total assets of the final bank sample to the total assets of all financial institutions covered by Bankscope as of 1999 by country. In the cases of Hong Kong, Malaysia, Philippines, Singapore, and Thailand, the final sample represents over 95% of all publicly held financial institutions in these countries. The final sample represents 77% of Indonesian financial institution assets and 60% of Korean financial institution assets.
Since Bankscope reports individual banks’ historical financial statements usually back to 1992, and a large number of banks have not yet released their 2000 financial statements, the study period is from 1992 to 1999.3 During the study period, banks that were acquired or failed are dropped from the Bankscope database so that the final sample contains only surviving banks as of 2000. It should be noted that all countries being analyzed are subject to the same survival bias, so that the comparisons across countries would still be valid.
The annual balance sheet and income statement are used to construct the variables for the empirical analysis. To the extent that the cost variables used in the analysis are expressed as a ratio to total assets, it is not necessary to convert the financial statement variables into a common currency for international comparison. (For descriptive statistics, we convert total assets into U.S. dollars.) Table 1 provides sample statistics by country for 1999.
Panel A of Table 1 shows that on average, banks in Korea are the largest among the seven countries, although the largest bank is located in Hong Kong. The average bank in Hong Kong, Singapore, and Thailand are of similar size, while Indonesia, Malaysia, and Philippines 3 Data for banks in Hong Kong are unavailable for 1992, and for banks in Singapore are unavailable from 1992 to 1995.
Singapore has the fewest. Panel B shows that both Indonesia and Philippines have respectively the highest mean and median total operating costs per dollar asset, while Singapore has the lowest observed operating costs. Panel C and Panel D show broadly similar pattern for the distributions of the two components of operating costs: labor cost and capital cost for the Asian countries. Panel E shows the ratio of loans (net of provisions for losses) to total assets across countries. The average bank in Thailand has the highest loan to asset ratio, while the average bank in Indonesia has the lowest ratio. The average loan ratio for the other five countries clusters around 50 to 60 percent. Despite the potential difference in banking powers across countries [Barth, Caprio, and Levine (2001)], the average loan-to-asset ratio does not seem to vary that much across countries.
Two pieces of country specific data are also collected. First, the average wage rate in the financial services industry by country is collected from the International Labor Organization’s Laborsta database. The annual wage data is available for Hong Kong, Singapore, South Korea, and Thailand for the entire sampling period. It is available only through 1995 for Philippines.
For Malaysia, the wage data is available for only a few industries. No wage data is available for Indonesia. Thus, Indonesia will be dropped from the analysis that uses wage data. We extrapolate the Philippines wage data to 1999 using the historical growth rate, and estimate Malaysia’s wage rate in the financial services industry using the sectoral wage distribution relationship in Singapore and Thailand, the two economies that seem to be closest to Malaysia by proximity and overall economic structure. Table 2 provides the average wage rate converted into U.S. dollar using prevailing exchange rate for the financial services industry in the six Asian
financial services industry are in the order of two to three times higher than those in Malaysia and Thailand, whose average wage rates in turn are almost twice as high as that in Philippines.
Between 1992 and 1999, Singapore experienced the fastest growth in finance wages, at a compound annual rate of 9.87%, while the other Asian countries’s finance wages were growing at around 4% with the exception of Philippines whose wage rate were growing at only 1.5%.
Second, to measure the degree of openness of the banking sector to foreign competition, we use the banking openness index by Claessens and Glaessner (1998), as shown in Table 3.