Bank mergers are one of the most widely observed phenom- ena during a financial restructuring process. As seen in various merger cases (most recently in the East Asian countries after the crisis), government encourages and supports banking consolidation in various ways, including financial and legal assistance. one of the main grounds for government involvement is that safer banking systems can be established through bank mergers. This argument is based on the presumptions that bigger banks are better diversified than smaller banks, and that diversification can potentially reduce bank risk. This paper undertakes an empirical study to examine these presumptions. I use international banking data - individual bank’s financial data from different countries (mainly OECD and APEC countries) - extracted from OSIRIS database. The data set consists of 530 banks from 30 countries.

The analysis takes two steps. In the first step, I examine the relationship between bank size and asset diversification to see whether the size effect in bank diversification exists. A diversifi- cation index (the Herfindahl index) is computed to measures bank asset exposure to different sectors. The Herfindahl index is the sum of squared weights corresponding to a bank’s asset exposure to different sectors. A lower value of the index corresponds to higher diversification. To see the relationship between size and diversification of bank, I examine the relationship between this index and bank size variable measured in log of total assets.

In the second step of the analysis, I examine whether bank risk is reduced by asset diversification. As measures of bank risk, standard deviation of bank’s ROA and Z-score are employed. These measures represent the stability side and the safety side - probability of bank failure - of bank, respectively. Using these measures I test whether a bank with better asset diversification (lower Herfindahl index) actually shows lower bank risk.

I find strong evidence that bigger banks are better diversified in their asset portfolios than smaller banks. I also find that better diversified banks show lower standard deviation of ROA, which implies that bigger banks enjoy diversification advantage in the stability of their asset returns. I do not find, however, the evidence that supports better diversified banks are less likely to fail. In other words, the empirical work fails to show that the probability of bank failure is decreased by enhanced bank diversification.

In summary, this paper shows that bigger banks are stable in terms of the variability of asset returns using their diversification advantage, but this does not necessarily guarantee that bigger banks are safe. These findings suggest that regulatory efforts on monitoring bank safety should not be loosened as banks grow in size.