Do ATMs influence bankprofitability in Japan?

Abstract

Holden and El-Bannany (2004) showed that Automated Teller Machines (ATMs) play an important role in increasing Return On Asset (ROA) in their analysis of banks in the United Kingdom. The present article investigates whether this conclusion also applies to Japanese banks. We found that ATMs do not have any influence on the ROA of Japanese banks; however, it was revealed that ATMs had positive effects on fees and commissions (income) from 2000 to 2003, and positive effects of ATMs on interest income were also seen recently. We conclude that in Japan, ATMs do not influence ROA, which includes the overall profits of bank transactions, but do contribute to particular businesses in that they can make the most of their abilities.

Acknowledgements

The author would like to express his gratitude for theuseful comments of Nobuyoshi Yamori of the Graduate School of Nagoya University, and of the editor and anonymous referees of Applied Economics Letters. Of course, all errors remain the sole responsibility of the author.

Notes

¹ A recent example is the consolidation of the Mitsubishi-Tokyo Financial Group and UFJ Holdings to form the Mitsubishi-UFJ Financial Group in October 2005.

² The regulations on bank branching were abolished in July 1997. The license system for branching and closing branches was revised to the current reporting system in April 2002.

³ The ‘trend examination of the use of communication media’ that was carried out by the Ministry of Public Management reported that Internet coverage in Japan was 37.1% at the end of 2000, but had risen to 62.3% by the end of 2004.

⁴ For example, Smirlock (1985) calculated the concentration ratio of the market using bank balance.

⁵ Rhoades and Rutz (1982) analysed the relationship between the concentration ratio of the market and the overall risk of bank management.

⁶ Porter and Millar (1985) mentioned that investing in IT can decrease the total cost of a firm and increase net profit.

⁷ For example, the credit rating of Shizuoka Bank given by Moodys in March 2003 was A1.

⁸ For example, the credit rating of Tokyo-Mitsubishi Bank given by Moodys in March 2003 was A2, which was the highest credit rating given to a megabank. The credit ratings of Daiwa Bank and Asahi Bank, however, were only Baa3.

⁹ In 2004 the number of ATMs of Tokyo-Mitsubishi Bank, which was the largest in 2003, was not disclosed by the ‘Nikkin annual report on data’ because of the reorganization of the Mitsubishi-Tokyo Financial Group and UFJ Holdings. The influences of these kinds of situations must also be considered.

¹⁰ We added lagged ROA to Equation 1(1) and estimated it. The adjusted-R ² improved a little, but the results mentioned later did not change significantly.

¹¹ The increasing ratio of bank balances and that of lending balances were added to Equation 1(1) as proxy variables of the growth rate of the banking market, and we estimated the equation using the pooled data from 2000 to 2004. As a result, the coefficient of ATMs _it is negative and significant at the 1% level, the coefficient of RISK _it and the increasing ratio of bank balances are negative and significant at the 5% level and the coefficient of BSIZE _it is positive and significant at the 1% level. That is to say, no positive effect of ATMs is found in this estimation either.

¹² The balance of investment trusts in the retail income of Yokohama Bank was 38.4 billion yen at the end of March 2000, but the balance at the end of March 2004 was 298.8 billion yen.

¹³ By using the pooled data from 2000 to 2004, the same estimation as in Footnote 11 was performed on the fees and commissions (income), giving the result that the coefficient of ATMs_it is positive and significant at the 1% level. This estimated result is assumed to be strongly influenced by the results from 2000 to 2003.

¹⁴ By using the pooled data from 2000 to 2004, the same estimation as in Footnotes 11 and 13 was carried out on the interest income. The coefficient of ATMs_it is positive but insignificant. We believe that this result was strongly influenced by the results from 2000 to 2002. Because the coefficient of increasing ratio of lending balances is positive and significant at the 1% level, it can be said that banks can raise high interest income in years in which lending is actively pursued.

¹⁵ Banks have been permitted to sell investment trusts as of December 1998. Mediating business to buy and sell stocks in bank branches has also been permitted since April 2004.