Connections or Performance: What Determines Turnover of Italian Bankers?
37 Pages Posted: 1 Mar 2007
We analyze top executives turnover in Italian Banks in the period 1993-2002. We relate executive's turnover to bank performance as measured by return on equity and non performing loans, and to local connections of the manager, controlling for other bank and manager characteristics. We classify banks in two groups according to their voting mechanisms. The first group, joint stock company banks, includes all banks (Commercial and Saving and Loans) with voting mechanism based on the number of shares owned. The second group includes all cooperative banks that have a per capita voting mechanism: Mutual, Rural and Cooperative banks.
First, we consider OLS regressions with and without bank fixed effects, and managers fixed effects. We find that top executive turnover is affected by bank organizational form. President and General Manager turnover are negatively related to return on equity in joint stock company banks, while in cooperative banks they are related to both performance variables. CEO turnover, which we observe only in the first group of banks, is ambiguously related to bank performance. Top executives in all types of banks are strongly locally connected. The relationship between local connections and turnover depends on the category of banks. In the banks chartered as Joint Stock Company only the turnover of the president is negatively related to connections, while for the Mutual, Rural and Cooperative Banks the turnover of both President and General Director is negatively related to connections. Consistently with previous literature we find that turnover is higher in banks affiliated to a group. Finally for both groups of banks and for all positions the presence of a contemporaneous episode of turnover in the same bank increases the likelihood of turnover, thus suggesting that a discipline mechanism is at work.
Then, we construct a measure of managers' tenure and we look at the probability of turnover conditional on tenure. We find that the probability of turnover increases with tenure for all positions and all types of banks, and that, for a given performance, the hazard function is lower the higher the degree of connections.
Keywords: Corporate Governance, Executive Turnover, Banking
JEL Classification: G21, G34
Suggested Citation: Suggested Citation