CEO & CHAIRMAN DUALITY... ON THE DECLINE?

CEO duality, dual-hatting or the CEO also holding the role of Board Chairman, is once again becoming topical in the literature and more recently the press. What do we know about this corporate governance feature in a banking context? Narrowly in financial services and banking, Booth, Cornett and Tehranian (2002) investigated various board of director features for different regulated firm types, including banks, and found duality does not appear to serve as a substitute for outside director monitoring.  Others have examined this feature including Grove et al. (2011) to find that duality is negatively associated with financial performance, such as ROA and HPRs during the crisis period for 236 US BHCs. An alternative perspective is provided by Byrd et al. (2012), who probed the US thrift industry during the 1980s, present evidence that firms with duality were less likely to fail than those where duality did not exist. These authors believe such powerful CEOs were able to resist shareholder pressure to adopt riskier strategies.  

I have looked at this issue more recently with a panel of 140 BHCs for the 2012 to 2015 period. Using BoardEx governance and Bloomberg financial data, I found some pretty interesting and significant findings for duality and performance. In this sample, 52.75% of BHCs report duality. This is an interesting finding versus global averages across all industries reported at less than the 10% level (FT, MSCI, Bloomberg, Credit Suisse). Simple OLS regressions (robust to standard errors) to test the relation with ROA indicate today that a significant positive relationship exists in my BHC panel, with a set of other board and bank model control variables.

Last week, the FT (FT Reports, 1 Nov 2016, page 4) probes duality and finds that the practice is nonetheless, across many firm types, dying out, with the exception of the US. However executives are quoted in the article that even financial services firms should move away from duality given the complex business models at play.  So this becomes an interesting debate for US BHCs, in that while individual BHC performance may improve with duality, systemic concerns may override and support those in the non-duality camp.

Comments