One, of several, regulatory responses to the financial crisis has been to consider the extent to which bank failure can be explained by flaws in banks’ corporate governance arrangements.
In many jurisdictions this diagnosis has generated calls upon shareholders to act as effective owners and hold boards of banks to account, as well as calls to empower shareholders to enable them to do so. But what do we know about the relationship between shareholder power and bank failure? To date scholarly attention has been paid to the relationship between board independence and bank failure, but limited attention has been given to the relationship between bank failure and the core corporate governance rules that determine the ease with which shareholders can remove and replace management. In our paper, Shareholder Empowerment and Bank Bailouts, we examine the relationship between shareholder power — and, thus, managerial accountability — and the probability of bank bailouts.
This paper makes two main contributions. The first one is the proposal of an index to measure the extent to which corporate managers are insulated from shareholder pressure. Our index, which we call the management insulation index (MII) is what could be called a contingent index. After filtering out the governance arrangements most relevant to our research question, we identify six combinations of governance arrangements that can theoretically be considered distinct. Most of our index values are the result of different (but effectively equivalent) corporate governance arrangements. US firms provide an optimal setting in which to apply our index, due to the considerable design flexibility provided by corporate law in the United States.
Our second contribution is to apply this index to show that banks with less insulated managers were more likely to receive capital injections under the Capital Purchase Program (CPP), the main bank-recapitalization program under the US Troubled Assets Relief Program (TARP). Our evidence is based on hand-collected data of the governance arrangements of 276 banks from the applicable corporation laws and the banks’ charters and by-laws.
The paper considers three main interpretations of the identified relationship between managerial insulation and bank bail outs. The first one is the possibility that bank participation in the CPP is correlated with a bank’s need to recapitalize after finding itself in a fragile position during the crisis. Thus, one interpretation of our evidence is that shareholder empowerment leads to decisions that make banks weaker and less able to weather the crisis.
A second possibility is the exact opposite of the first story: perhaps those banks with high management insulation scores were so weak that they did not qualify for government support. Participation in CPP was subject to several conditions, making the weakest institutions ineligible for government investment. If banks with high management insulation scores were badly run, they may have been among those banks that had their CPP applications rejected.
A third possible explanation is that as CPP equity injections can be seen as a source of cheap capital, a decision not to participate in the CPP may be a symptom of poor governance. CPP participation came with strings attached, such as restrictions on executive compensation. It is possible that powerful executives would prefer not to participate. Thus, perhaps when banks are offered the opportunity to recapitalize cheaply, only the well-governed ones do so.
The first two stories are mutually exclusive. To address them, we modify our bailout variable in the following way. We identify those banks that plausibly did not participate in CPP because they were too weak, and treat them as if they were bailed out. We find that the link between the MII variable and the adjusted bailout indicator becomes stronger. This finding strengthens the first interpretation: shareholder empowerment and bank strength at the beginning of the crisis seem to be negatively associated. We also show that banks with high MII were more likely to refuse CPP funds after the investment was approved. This evidence is inconsistent with the hypothesis that high MII banks did not receive funds because they were weak. The balance of the evidence thus rejects the second story.
This leaves us with the first and third stories, which are not mutually exclusive. Both stories are plausible. Although we cannot perfectly discriminate between them, we offer some additional evidence that they are not equally supported by the data. To address the third story, we first identify those banks that did not apply for funds, or that received CPP funds but repaid them early (before the end of 2009). The latter banks chose to replace cheap government capital with more expensive private capital. In particular, banks that exited CPP redeemed preferred shares at par, while the fair value of those shares was below par. Thus, similarly to the decision not to participate, exiting CPP early could be a symptom of bad governance. However, we find that the link between the MII and CPP application (either including or excluding early-repayment banks) is both economically and statistically weak. This evidence is difficult to reconcile with the hypothesis that the negative relation between the MII and bailouts is mainly due to management’s desire not to be bound by the CPP restrictions.
To investigate the mechanism further, we estimate the effects of management insulation on two additional variables. The first variable is the proportion of non-interest income in total income. We show that banks with high levels of management insulation in 2003 were less likely to increase their non-interest income ratios in the years prior to the crisis (2003-2006). The second variable is the proportion of Level 3 assets in total assets, which is a proxy for asset quality. Level 3 assets are illiquid complex securities. We find that banks whose governance arrangements insulated them from managerial pressure were likely to have a lower proportion of Level 3 assets in 2008 than banks that were not as insulated. Overall, our evidence suggests that banks with empowered shareholders were more likely to be bailed out partly because they engaged in non-traditional commercial banking activities, such as investment banking and trading of complex securities.
The full paper is available for download here.