Essays on corporate governance of financial intermediaries

Uymaz, Yurtsev (2017) Essays on corporate governance of financial intermediaries. PhD thesis, Prifysgol Bangor University.

[img] Text
Signed Declaration Uymaz.pdf
Restricted to Repository staff only

Download (166kB)
UYMAZ PhD 2017.pdf

Download (3MB) | Preview


This thesis comprises four papers that examine the effect of information advantage of bank executives and CEOs on bank risk taking and performance and also investigate to reveal which CEO power variables, which denote information advantage to the CEO, influence the likelihood of bank fraud and the likelihood of detecting fraud. Paper 1 provides a theoretical, regulatory, structural, and historical analysis of US banks. The regulatory environment of banks has been changed dramatically as well as the structure of banks in the last three decades. Banks’ financial intermediation role and opaqueness that comes from greater risk-taking make them special in corporate governance applications. It is known that regulations have the direct effect on bank corporate governance with the hands of regulators. Paper 2 examines whether information advantage of the CEO can influence bank risk to add empirical evidence to hypothesised relationship from the perspective of the CEO power. CEO tenure and CEO network size that denote the sources of information advantage are used as the CEO power variables. The effect of CEO power on three measures of bank risk is assessed: Z-score, systematic risk, and systemic risk. Results from fixed effects and generalised method-of-moments (GMM) dynamic panel data estimations reveal that banks are more likely to take on more risks when CEO’s have a relatively long tenure and large network. The results of the robustness tests provide the same connection between CEO power and bank risk. Paper 3 explores whether institutional investors in publicly listed US banks can influence bank ownership structure and performance through a prior connection to newly appointed senior executives of the bank by employing a unique dataset. The impact of the connection on three measures of bank performance is assessed: non-interest income to total assets ratio, market beta, and Tobin’s Q. Institutional investors increase their shareholding in banks after the appointment of a connected executive. Results of regressions reveal that the presence of connected executives is positively and significantly associated with developments in market beta and non-interest income, and negatively and significantly related to developments in Tobin’s Q. The results as consistent with institutional investors with prior connections to bank executives having a significant information advantage relative to other shareholders in the bank on its likely future performance. Finally, paper 4 contributes the corporate governance literature that has little to say about the likelihood of banks engaging in financial fraud. The commission of financial fraud by banks as partly reflecting that bank’s culture, which is driven in large part by the bank’s senior executives, especially the CEO. A unique dataset on financial fraud in publiclylisted US banks is employed to test for a link between fraud and CEO power that creates information advantage. The results from probit and partially-observed bivariate probit estimations suggest that banks are more likely to commit fraud and more likely to be detected by regulators if they have powerful CEOs measured by length of CEO tenure, Chair/CEO duality, size of CEO’s network, and if the CEO is also a part-owner of the bank. Fraud also appears more likely to be committed by large banks with relatively poor balance sheets, raising the prospect that fraud (and powerful CEOs) can have adverse systemic consequences.

Item Type: Thesis (PhD)
Subjects: Degree Thesis
Departments: College of Business, Law, Education and Social Sciences > Bangor Business School
Degree Thesis
Date Deposited: 13 Dec 2017 10:28
Last Modified: 13 Dec 2017 10:28
URI: http://e.bangor.ac.uk/id/eprint/10848
Administer Item Administer Item

eBangor is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.