Knowledge that Transforms

Bank liquidity shortages associated with the growth of wholesale-funded credit intermediation has motivated the implementation of liquidity regulations. We analyze a dynamic stochastic general equilibrium model in which liquidity and capital regulations interact with the supply of risk-free assets. In the model, the endogenously time varying tightness of liquidity and capital constraints generates intermediaries’ leverage cycle, influencing the pricing of risk and the level of risk in the economy. Our analysis focuses on liquidity policies’ implications for households’ welfare. Within the context of our model, liquidity requirements are preferable to capital requirements, as tightening liquidity requirements lowers the likelihood of systemic distress without impairing consumption growth. In addition, we find that intermediate ranges of risk-free asset supply achieve higher welfare.

This paper sheds new light on the role bank executives played in the financial crisis. It examines whether they foresaw the poor performance of their own bank by analyzing their insider trading patterns. Insider trading during 2006 predicts stock returns during the crisis: a portfolio strategy based on insider trading information earns a risk-adjusted return of over 40% during the crisis. Further, banks with a high exposure to the housing market and banks with a low exposure exhibit different insider trading patterns starting in mid-2006, when US housing prices first decline: insiders of high-exposure banks are 20% more likely to sell stock than insiders of low-exposure banks. This pattern is more pronounced for CEOs than other insiders. However, insider trading patterns of high- and low-exposure banks do not differ before 2006. Replacing high-exposure banks by too-big-too-fail banks yields similar results. This evidence indicates that insiders of high-exposure and too-big-too-fail banks revised their assessment of their banks’ investments following the reversal in the housing market.

We investigate why some banks are more exposed and contribute more to systemic risk in the global financial system than others. On average, European banks contribute more to global systemic risk than banks in the United States while banks on both continents have a comparable average exposure to systemic crises. In addition to being larger, European banks appear to contribute more to global systemic risk, because of the lower quality of their loan portfolios and their higher relative interconnectedness with the rest of the global financial system. Finally, more stringent capital regulations are found to decrease the average exposure of banks to systemic risk.

Retail banking is a complex industry in which depository institutions bundle various services and may have market power. We use a recent regulation as a natural experiment to provide broad evidence about competition and the importance of bundling in retail banking. That regulation, which resulted from the Durbin Amendment to the Dodd–Frank Act, capped debit card interchange fees for banks with over 10 billion in assets. Using a difference-in-differences identification strategy, we document and quantify the resulting decline in interchange income for treated banks. We further find that treated banks offset more than 90% of the lost interchange income through increases in deposit fees for account holders. We argue that the ability to adjust deposit fees indicates (i) that treated banks have market power with respect to their account holders and (ii) strong complementarity between debit card transactions and deposit accounts. These results are robust when limiting the sample to banks near the asset threshold or using control banks with low direct competition with treated banks. Treated banks neither reduced costs nor strategically avoided the 10 billion threshold.

We study how contingent capital affects banks’ risk choices. When triggered in highly levered states, going-concern conversion reduces risk-taking incentives, unlike conversion at default by traditional bail-inable debt. Interestingly, contingent capital (CoCo) may be less risky than bail-inable debt as its lower priority is compensated by a lower induced risk. The main beneficial effect on risk incentives comes from reduced leverage upon conversion, while any equity dilution has the opposite effect. This is in contrast to traditional convertible debt, since CoCo bondholders have a short option position. As a result, principal write-down CoCo debt is most desirable for risk preventive purposes, although the effect may be tempered by a higher yield. The risk reduction effect of CoCo debt depends critically on the informativeness of the trigger. As it should ensure deleveraging in all states with high risk incentives, it is always inferior to pure equity.

Using U.S. banking data between 1993 and 2014, this paper investigates the relationship between CEO optimism and bank liquidity creation. It finds that banks with optimistic CEOs create more liquidity over the entire sample period. In addition, the positive effect of CEO optimism on liquidity creation became stronger during the subprime crisis of 2007‒2009, and this stronger effect was mainly driven by banks with high capital ratios and large banks. These results imply that CEO optimism is likely to encourage banks to create liquidity, especially during banking crises. The results presented in this paper hold when subjected to various robustness checks.

We test whether adverse changes to banks’ market valuations during the financial and sovereign debt crises affected firms’ real decisions. Using new data linking over 5000 non-financial Italian firms to their bank(s), we find that increases in banks’ CDS spreads, and decreases in their equity valuations, resulted in lower investment, employment, and bank debt for younger and smaller firms. These effects dominate those of banks’ balance-sheet variables. Moreover, CDS spreads matter more than equity valuations. Finally, higher CDS spreads led to lower aggregate investment and employment, and to less efficient resource allocations, especially during the sovereign debt crisis.

The U.S. bank stress tests aim to improve financial system stability. However, they may also affect bank credit supply. We formulate and test opposing hypotheses about these effects. Our findings are consistent with the Risk Management Hypothesis, under which stress-tested banks reduce credit supply−particularly to relatively risky borrowers−to decrease their credit risk. The findings do not support the Moral Hazard Hypothesis, in which these banks expand credit supply−particularly to relatively risky borrowers that pay high spreads−increasing their risk. Results are generally stronger for safer banks, banks that passed the stress tests, and the earlier stress tests.

We study the effect of changes to bank-specific capital requirements on mortgage loans with a new loan-level dataset containing all new mortgages issued in the UK between 2005Q2 and 2007Q2. We find that a rise of a 100 basis points in capital requirements leads to a 4% decline in individual mortgage loan size. Borrowers with an impaired credit history (verified income) are not (most) affected. This is consistent with the origination of riskier loans to grow capital by raising retained earnings. We then examine the interaction of capital requirements and monetary policy. This suggests that a monetary policy tightening may mitigate the loan contraction associated with higher capital requirements, as maturity transformation allows the affected bank to profit from a steeper yield curve and raise capital through retained earnings. Overall, our findings in this paper indicate that retained earnings are an important channel of adjustment to capital requirements.

We show that banks’ risk exposure in one asset category affects how they report regulatory risk weights for another asset category. Specifically, banks report lower credit risk weights for their loan portfolio when they face higher risk exposure in their trading book. This relationship is especially strong for banks that have binding regulatory capital constraints. Our results suggest the existence of incentive spillovers across different risk categories. We relate this behavior to the discretion inherent in internal ratings-based models which these banks use to assess risk. These findings imply that supervision should include a comprehensive view of different bank risk dimensions.