9-10 June 2016 : DNB – CEPR Conference on Bank Equity over the Cycle
Location: De Nederlandsche Bank (DNB), Amsterdam
9-10 June 2016 : DNB – CEPR Conference on Bank Equity over the Cycle
Location: De Nederlandsche Bank, Amsterdam link
for Programme. link
for detailed information on the site of the CEPR.
On June 9-10 DNB has jointly organized with CEPR a conference on “Bank Equity over the Cycle”.
In the wake of the global financial crisis the regulatory reform has focused on an increase in capital cushions of financial institutions. The debate around the appropriate level of bank capital is multidimensional, involving agency problems in banks, asymmetric information, international coordination and arbitrage, bank governance, tax benefits of debt, government subsidies, systemic risks and externalities beyond the financial sector, and shadow banking. This conference contributed to this debate and took stock of the recent academic research and experts’ view on various questions about prudential capital regulation that are critical to regulatory policy. The main organizer of the conference was Enrico Perotti (from University of Amsterdam) and participants included academics (from United States, United Kingdom and Europe), policymakers (from European Central Bank, European Banking Authority, Single Resolution Board, Bank for International Settlements, and U.S. Securities and Exchange Commission), and practitioners (from ABN AMRO Bank and Credit Suisse).
The conference was opened by Jan Marc Berk (division director, EBO), who introduced the topic of the conference and highlighted the key arguments in favor of higher capital, as well as the challenges and potential effects of stronger capital buffers on real economy and financial sector.
Papers in the first session of the conference (chaired by Jakob de Haan) discussed the relationship between banks’ capital and their lending behavior. The session was opened by Frederic Malherbe (London Business School), who presented a paper (joint work with Saleem Bahaj, Jonathan Bridges, and Cian O’Neill, all from Bank of England) which provides a theory on how banks choose to issue capital and lend in the presence of capital requirements. They argue that changes in capital requirements are strongly correlated with changes in lending, but only during downturns, with the magnitude of this effect depending on the level of capital requirements, return on legacy assets and on new lending. Using UK banking regulatory data from 1989 to 2007, they find strong empirical support for this relationship. Andrea Polo (Pompeu Fabra) presented the second paper in this session (joint work with José-Luis Peydró, from Pompeu Fabra, and Enrico Sette, from Banca d'Italia) on the transmission of monetary policy. Using Italian securities and credit registers, they show that in crisis times, banks increase their holdings of securities when monetary policy is softer. This effect is stronger for less capitalized banks, which also reduce credit supply. However, the results reverse in normal times when less capitalized banks increase more the supply of credit and the riskiness of their loans portfolio. The third and last paper of this session was presented by Sebastian Pfeil (University of Bonn) (joint work with Nataliya Klimenko, Jean-Charles Rochet, both from University of Zurich, and Gianni De Nicolo from IMF). They argue that the capitalization of the entire banking sector is an important determinant for bank lending. The paper introduces a general equilibrium dynamic model with financial frictions to study the long-term impact of macro-prudential policies on output and financial stability. They show that in a framework in which aggregate bank capital determines the dynamic of credit, undercapitalized banks lend too much. This is due to the fact that these banks do not internalize the effect of their individual lending decisions on the future loss-absorbing capacity of the banking sector.
first day concluded with three panel discussions. The first panel (chaired by Jakob de Haan) focused on supervisory stress-testing. Mark Flannery (U.S. Securities and Exchange Commission and University of Florida) noted that while stress-testing has become an important tool for assessing how much capital a bank needs to support its risk-taking activities, there are some other important players in the financial markets that should be subject to stress-testing as well. He discussed the growing importance of mutual funds and the concerns over the massive retail money inflows into corporate bond funds. In his view, the potential risks to financial stability relate to prospective impact on market prices, particularly in a situation in which a large fund has to liquidate a large share of its holdings. Although desirable, the mutual funds’ liquidity stress-testing is a difficult exercise (much more difficult than in case of banks), as it depends on information about mutual fund’s holdings. Piers Haben (European Banking Authority) discussed the multiple use of stress-testing in the supervisory toolkit. He noted the role of stress-testing, first, as a valuable source of information for identifying business model vulnerabilities in risk management and problems with certain risk exposures, and second, as a mean for testing a bank’s ability to meet its capital requirements over the economic cycle. He highlighted the importance of stress-tests in providing a comprehensive view on an individual institution by considering not only the risks it is exposed to in isolation, thus allowing public authorities to assess potential problems in the financial system more broadly, and discussed the challenges in designing new stress-testing methodologies adequate for non-crisis period. Charles Goodhart (London School of Economics) noted that the conduct of annual stress tests gives the regulatory authorities their best available chance of dealing with fragile banks while there is still enough time to avert a, potentially contagious, failure. However, in his view there is an important question, not yet fully resolved, of how to provide back-stop funding to recapitalize the weaker banks once the results of stress-tests are revealed, in name of transparency, to the market (and the market turns against those banks that have failed the test). There are also some other remaining problems, such as: how to take account of second, amplifying effects, or how to reduce the implementation costs (in terms of time and resources consumed), or how to select the appropriate set of stresses for the purpose of the test, or what results to disclose and to whom.
The second panel (chaired by Iman van Lelyveld) has focused on bank’s models over the cycle. The panelists (Thomas Broeng Jørgensen, from European Central Bank, Piers Haben, from European Banking Authority, and Willy Westerborg-de Haan, from ABN-AMRO) discussed various aspects of credit risk modelling. Their views converged on the idea that a complete credit risk management system should require both through-the-cycle and point-in-time estimates (as they correspond to different objectives for credit measurement), while they acknowledged the difficulties in having long time series that can be fed into the models. Moving forward, the panelists highlighted the role of risk models in complementing the macro prudential instruments and noted that they should distinguish between intra- vs inter-bank diversification.
third policy panel (chaired by Paul Cavelaars) discussed new forms of bank capital, namely contingent convertible capital (CoCos) and bail-in debt. Thomas Broeng Jørgensen (European Central Bank) noted that a lot of attention has been given recently to the conversion mechanism of CoCos (i.e., principal write-down vs debt-to-equity swap). While this is an important aspect of CoCos’design, for the going-concern CoCos is equally critical the role played by the trigger event that leads to conversion (in particular, how low a trigger may be set), as well as the degree of transparency over the seniority structure of the bank’s capital. Mauro Grande (Single Resolution Board) noted that from a resolution perspective it is critical for the bail-in procedure to be credible. He discussed the legal risk associated with bail-in (i.e., when one category of creditors is subject to bail-in, but others not), as well as financial stability considerations (in particular the difficulties in measuring the impact that bail-in may have on financial stability, as well as potential spillover effects). Paul Glasserman (Columbia University) discussed three related questions relevant to CoCos’ market. First, he argued that changes in the CDS market may provide useful information not only about risk in banks, but also about the credibility of recovery and resolution frameworks. Second, he noted that there are a few obstacles to implementing market-based triggers, obstacles that may be surmounted if the triggers are set sufficiently high and if conversion is credible. Finally, he discussed the opposing incentive effects faced by a bank as it nears a CoCo’s conversion trigger: On one hand, shareholders have a strong incentive to raise capital to stave off conversion. On the other hand, as conversion becomes very likely, shareholders would prefer to take on more tail risk. Sandeep Agarwal (Credit Suisse) noted that the market is absorbing well the supply of contingent convertible capital. The main issue for the potential investors is related with the correct valuation of these instruments. Currently, the market is not pricing various instruments conditional on their loss-absorbing mechanism. Looking into the future, the panelists noted that demand for CoCos will mainly come from institutional investors.
Papers in the second session of the conference (chaired by Natalya Martynova) discussed the impact of bail-in requirements and capital buffers on lending. John Vourdas (European University Institute) presented a paper (joint work with Misa Tanaka from Bank of England) on the optimal structure of banks' loss absorbing capacity. The paper introduces a theoretical model which studies the interaction between minimum capital requirements, TLAC requirements, and additional capital buffer requirements, and the impact of these requirements on moral hazard. The authors explain how that ex-ante moral hazard (i.e., inefficient monitoring) can be mitigated by sufficiently high bank’s loss absorbing capacity (i.e., TLAC plus equity capital buffer), while ex-post moral hazard (i.e., asset substitution incentives) can be prevented by an efficient bail-in procedure. Catherine Koch (Bank of International Settlements) presented the second paper in this session (joint work with Christoph Basten from KOF-ETH Zurich and Swiss Financial Market Supervisory Authority FINMA) and provided empirical evidence on the impact of the introduction of countercyclical capital buffers (CCB) on mortgage market in Switzerland. Using data provided by the online mortgage platform Comparis, the authors show that the introduction of CCB changes the composition of credit supply, with mortgage specialized and less capitalized banks being more likely to raise prices, and thus shift mortgages from less to more resilient banks. The final paper of the conference was presented by Filippo De Marco (Bocconi University) (joint work with Tomasz Wieladek from Barclays Bank). It provided UK’s evidence for the impact of capital requirements (and monetary policy) on bank’s lending to SMEs. The authors find that as a result of higher capital requirements banks cut lending, in particular to those firms with whom they have a short business relationship. Changes in monetary policy on the other hand impact only smaller banks which reduce lending to riskier SMEs.