1. Banks’ Non-Interest Income and Systemic Risk Brunnermeier, Dong, Palia Discussant Loriana Pelizzon Loriana Pelizzon Ca’ Foscari University Venice and MIT Sloan

2. Systemic Risk Measures  Banks are “special” for many different reasons…  Recently they have increasingly earned profits from non- interest income  In these activities they have started to compete with other capital market intermediaries: Hedge funds, mutual funds, investment banks and private equities (whom do not have deposit insurance).  This has build up a sort of “Shadow banking system” or a “Shadow hedge funds system” (with deposit Insurance...)  => Resulting in higher systemic risk Objectives 30/09/2011 Slide 2

3. Systemic Risk Measures  Capture Systemic risk with CoVar or SES  Financial statement data from Compustat and from FR Y-9C  OLS panel regression with Quarterly FE and a series of control variables and: –Non-interest income ratio –OR –Trading income ratio –Investment banking ratio –Other non interest income Methodology 30/09/2011 Slide 3

4. Systemic Risk Measures 30/09/2011 Slide 4 Systemic risk is positively related with: 1.non-interest income 2.Bank size 3.Trading income(-), investment banking, venture income 4.1990 and 2007 Results Slide 4 04/05/2011

5. Systemic Risk Measures  Very interesting topic!  Very interesting paper! Comments 30/09/2011 Slide 5

6. Systemic Risk Measures 1.It is surprising that CoVaR (and therefore one of your systemic risk measure) is not related to leverage 2.Do you control for heteroschedasticity when you perform the quantile regression to determine CoVaR? 3.Why N. of obervations change so much among the different regressions? Comments 30/09/2011 Slide 6

7. Systemic Risk Measures  All the different systemic risk measures have some limits. Why not using out of sample analysis, i.e.: –losses during the different crisis periods as dependent variable or –the results of stress test analysis Comments 30/09/2011 Slide 7