Herding and systemic risk in banking sector

Student thesis: Doctoral Thesis


Systemic banking crises can cause significant disruptions in the economy (Dell'Ariccia, Detragiache and Rajan, 2008; Reinhart and Rogoff, 2014) and result in non-trivial fiscal costs (Calomiris, 1999; Dewatripont, 2014). Consequently, a great number of studies in the banking literature have been dedicated to examining systemic risk in the banking system, in particular that of too-big-to-fail banks (Shleifer and Vishny, 2010; Laeven, Ratnovski and Tong, 2016). Nonetheless, the problem of the systemic risk posed by bank herding has received less attention.This thesis provides a rigorous empirical examination of the theory that banks herd to increase the likelihood of a collective bailout position should default occur (Acharya and Yorulmazer, 2007; Farhi and Tirole, 2012). The empirical examination is assessed by addressing three key research questions: 1) Do banks herd and can country-level factors explain herding consistent with the theory? 2) If yes, does herding pose a systemic risk? 3) How does herding affect the competition and profit of banks? These questions are addressed in three individual empirical chapters.The first empirical chapter (Chapter 2) investigates whether banks do herd and if country-specific factors affect herding consistent with the theory. The findings support the proposition that banks do herd and that the degree of herding varies across countries. The results also show that herding is dependent on several country-specific features such as exposure to fiscal costs, banking sector characteristics, and regulatory and supervisory quality. However, contrary to the theory, the effect of shareholder protection laws on herding is not significant, possibly because banks with dispersed ownership also receive bailout subsidies when banks collectively fail.The second empirical chapter (Chapter 3) investigates the systemic risk implications of bank herding. The findings show that the effect on systemic risk of the interaction between herding and deposits and that between herding and loans are statistically significant. The results suggest that negative externalities from excessive funding risk and liquidity risk taking may not have been fully internalised through existing prudential regulations.The third empirical chapter (Chapter 4) examines the effect of herding on the competition and profit of banks. Empirical evidence in this chapter indicates higher competition among banks that herd compared to the rest of the banking industry. Nonetheless, herding may still be desirable when competition in the banking industry is weak. The possibility of a higher profit from low banking industry competition allows banks to compensate for the erosion in profit caused by herding. Furthermore, in the face of herding, the adverse effect of increased competition on profit is larger for banks that are followed by others compared to those that follow the leaders.
Date of Award10 Dec 2020
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SupervisorKrishna Paudyal (Supervisor) & Devraj Basu (Supervisor)

Cite this