Essays on banking
Karanastasis, Dimitrios I.
Καραναστάσης, Δημήτριος Ι.
KeywordsEconomics ; Banking ; Shareholders ; Economic crisis ; Credit ratings ; Money market ; Earnings management ; Financial stability ; Monitoring ; Loan-loss provisions ; Dow Jones Sustainability Index (DJSI)
The work for this thesis was done during the period 2010 to 2016, after the worst financial crisis since the Great Depression, and in the immediate wake of the Eurozone debt crisis that has casted doubts on European integration. With the crisis largely blamed on bank excessive risk-taking and supervisors’ inability to contain it, and with the aim to prevent similar crises in the future, much of the policy discussions and regulatory actions have focused on ways to address both of these failures of market economies. One way concerns the strengthening of market discipline, i.e., strengthening the incentives and capacity of external stakeholders (depositors, shareholders and other holders of bank liabilities) to induce more prudent risktaking behavior by bank managers. Thus motivated, this thesis focuses on the channels and effectiveness of market discipline. Said differently, it explores whether, and how, market forces could be trusted to do a better job in containing risk-taking behavior by banks and thus effectively support and supplement supervisors’ efforts. From a policy-making point of view, the thesis explores whether market discipline can promote financial stability. The thesis comprises of three empirical papers. The first explores whether the expected government support of banks, implicit or explicit, weakens market discipline. The second explores whether the intrusive external monitoring by knowledgeable and influential external stakeholders fosters more prudent loan-loss provisions and, hence, more prudent accounting practices by banks. The third paper extends the second, taking into account the role of the institutional and social environment in an international setting. The results from all three papers are supportive of the notion that market discipline can promote financial stability. Details follow. The first paper explores whether expected government support weakens market discipline by bank shareholders, especially after the eruption of the global financial crisis in 2008. Two counter-veiling forces are at work. On the one hand, as it is suggested by the related literature, the expected support may give banks greater leeway to undertake risks, for it reduces the monitoring incentives of depositors and other bank creditors. Shareholders, recognizing the resultant higher probability of default as well as the possibility that they might be wiped out in case of bank default, have an incentive to intensify monitoring and exercise stronger market discipline. On the other hand, the expected support may also reduce the possibility of a bank run, weakening as a result the need for monitoring by bank shareholders and, hence, market discipline on their part. Expected support is measured as the difference of two bank credit ratings from Moody’s: an all-in rating, which encompasses expected support, and a stand-alone rating, which does not. We estimate a model in which a forward-looking measure of shareholder value, the market-to-book ratio, is the dependent variable and the measure of expected support is one of the explanatory variables. A negative coefficient of the expected support would be consistent with market discipline for it would indicate that shareholders are willing to pay less for banks with higher expected support The results, from a sample of about 250 banks worldwide, indicate that, far from weakening market discipline by shareholders, the expected support strengthens it, and more so for the riskier banks – i.e., those with lower stand-alone rating. The results also highlight the two counter-veiling effects of the expected support on market discipline. Specifically, as the size of the expected support increases, its negative effect on the market-to-book ratio decreases. The second paper sheds light on earnings management via loan-loss provisions (LLPs) and the associated trade-off between financial-statement transparency and financial stability, by exploiting time and cross-sectional variation. The time variation pertains to the welldocumented shift towards more forward-looking LLPs after the crisis of 2008. The working hypothesis is that the rules concerning LLPs effectively shifted in favor of forward-looking provisions. This shift is expected to be associated with stronger income smoothing and signaling, and more so for banks subject to weaker market discipline. The cross sectional variation pertains to the intrusive external monitoring by funds that are members of the US Sustainability Investment Forum (USSIF). The working hypothesis is that this process amounts to stronger monitoring and, hence, stronger market discipline. Thus, it is expected that income smoothing and signaling will be weaker for the banks in which USSIF funds invest relative to the remaining banks. Moreover, after the said shift in supervisory and regulatory preferences, income smoothing and signaling will increase by less for banks in which USSIF funds invest. The results, from a sample of more than 300 publicly-held US bank holding companies, over the period 1999 – 2014, confirm that the banks under the intrusive monitoring exhibit less earnings management. Since, however, this differential behavior got more pronounced after the regulatory shift, the results further suggest that USSIF funds induce provisioning behavior that goes well beyond the stricter application of the existing accounting and supervisory rules, thus ameliorating the aforementioned trade-off. The third paper explores whether the inclusion of a bank in the Dow Jones Sustainability Index (DJSI) reduces bank managers’ incentives for earnings management through loan-loss provisions. The inclusion in the DJSI depends on a rigorous bank assessment, conducted each year. The working hypothesis is that being included in the DJSI constitutes a credible signal to outside stakeholders of prudent behavior. An aspect of such behavior is less earnings management through loan-loss provisions. The results from a sample of 297 banks around the globe, over the period 2004 – 2010, indicate that banks included in the DJSI engage in less earnings management relative to the banks that were assessed but not included. This more prudent provisioning behavior persists when we control for the strength of private sector monitoring with country-level indices for private sector monitoring and the quality of external audits. Yet, it largely diminishes when we control for the strength of supervisory power and capital regulation. The last result poses an interesting question: Is stronger supervision a substitute of market discipline? Or, is the DJSI assessment process geared towards accepting more banks from countries with stronger supervision; All in all, the papers in this thesis highlight the important and decisive role market discipline plays in the financial system. They suggest that shareholders, probably driven by the objective difficulty of accurately assessing banks’ true condition and prospects, appreciate credible signals of prudent behavior. The results are also supportive of the power of external monitoring as a restraining factor in bank-managers’ risk-taking.