Two essays in banks’ risk-taking behavior

Abstract

This thesis consists of two essays. The commonality of the essays is that I study the banks' earnings volatility in terms assessing bank' risk-taking behaviour. In first essay, "Banks' Earnings Volatility and Earnings Predictability," uses the setting of banks to investigate Dichev and Tang's (2009) hypothesis that volatility of earnings has a negative effect on the predictability of earnings. Consistent with their hypothesis, I find that banks' earnings predictability is lower when they have higher earnings volatility. The findings suggest that there is a trade-off between banks' earnings stability and their risk-taking, and provide important implications for banks' risk-taking behavior. First, assuming that the purpose of corporate risk-taking is to assure a high level of future earnings, then the worst position for a bank is to have low earnings and high earnings volatility, which indicate that the bank took too much risk with a poor result. To improve its position and to avoid insolvency, the bank should reduce risk-taking. While the literature on corporate risk-taking suggests that firms with low earnings have higher incentives to take more risk, the first implication notes that banks with low earnings should decrease their risk-taking if they face high earnings volatility. Second, for banks with high earnings and high earnings volatility, their earnings are likely not persistent. To assure a high level of future earnings, they should also reduce risk-taking to lower their earnings volatility. Third, for banks with low earnings and low volatility, their future earnings are likely to be low. They should increase risk-taking to improve their future earnings. Fourth, the best position for banks is to have high earnings and low earnings volatility because they can assure a high level of future earnings. Since all banks would try to maintain or get into this best position, competition is likely to increase. How to protect their position becomes the top priority. The second essay of my thesis, "Is there an Optimal Risk-Taking in Banks?," examines how banks manage their risk-taking, in terms of high earnings volatility. Specifically, I address the following questions: First, is there an optimal risk in banks? Second, if there is an optimal risk then how quickly bank adjust towards the optimal? How the risk adjustment mechanism differs in short term and long term? Finally, I test the asymmetric effects in the risk adjustment mechanism, which one is costly for banks: over risk-taking or under risk-taking, and how banks manage their risk-taking during high earnings volatility. To address these questions, I propose an empirical model with two partial-adjustment mechanisms for bank risk-taking behavior, where the risk-adjustment occurs in the cross-section and the time-series variations of bank setting. The model is sufficiently rich for examining whether there is an optimal risk in banks and if so, what is the speed of adjustment with which banks move toward the optimum level of risk. Using the US banks data, I find that banks tend to follow an optimal risk target, and typical banks converge toward their target level at a rate of 23.78% per year. Furthermore, due to banking stability and regulatory concerns, there is an asymmetry effect in the speed of risk adjustment. That is, when there is excess volatility (risk), banks adjust back to the optimum level much faster. In sum, my second essay provides new evidence in the literature of bank risk-taking and contributes to the literature by developing a partial risk adjustment model to estimate banks' optimal risk-taking.