Financial Networks
Project Leader: Ana Babus
Financial Connections and Systemic Risk
with: Franklin Allen, University of Pennsylvania, and Elena Carletti, European University Institute
The ripple effects of Lehman's demise throughout global financial markets revealed, once again, the intertwined nature of financial systems. While the events unfolded, it became clear that the consequences of such an interconnected system are hard to predict. The market freeze that followed added a new dimension to how we should think about systemic risk. This project proposes a theory of strategic connections between financial institutions and their role in creating systemic risk. In particular, Babus, Allen and Carletti describe a mechanism through which financial systems become fragile when institutions are connected via overlapping portfolio exposures.
Financial Crises: Theory and Evidence
with: Franklin Allen, University of Pennsylvania, and Elena Carletti, European University Institute
Financial crises have occurred for many centuries. They are often preceded by a credit boom and a rise in real estate and other asset prices as in the current crisis. They are also often associated with severe disruption in the real economy. This project surveys the theoretical and empirical literature on crises. The first explanation of banking crises is that they are a panic. The second is that they are part of the business cycle. Modelling crises as a global game allows the two to be unified. With all the liquidity problems in interbank markets that have occurred during the current crisis, there is a growing literature on this topic. Perhaps the most serious market failure associated with crises is contagion and there are many papers on this important topic. The relationship between asset price bubbles, particularly in real estate, and crises is discussed at length. The results of this project are to be published in the Annual Review of Financial Economics.
Strategic Relationships in Over-the-Counter Markets
This project provides a theory of dynamic formation of relationships in over-the-counter markets. Over-the-counter (OTC) markets capture a large share of the trade volume for various assets. Distinctively, in these markets trade is conducted through bilateral negotiations, rather than a Walrasian auction. Counterparties meet and set prices through a bargaining process that reflects the strategic, repeated nature of the interactions. Moreover, in some cases, such as in the market for bank loans, a good resolution of transaction often requires careful monitoring of the terms of trade. Trading assets in the presence of moral hazard and asymmetric information often relies on trust and reputation. When transactions in over-the-counter markets involve counterparty risk, agents may find beneficial to develop a network of relationships.
The Formation of Financial Networks
Modern banking systems are highly interconnected. Despite their various benefits, the linkages that exist between banks carry the risk of contagion. In this project Babus investigated how banks decide on direct balance sheet linkages and the implications for contagion risk, and modelled a network formation process in the banking system. Banks form links with each other in order to reduce the risk of contagion. The network is formed endogenously and serves as an insurance mechanism. The project showed that banks manage to form networks that are resilient to contagion. Thus, in an equilibrium network, the probability of contagion is virtually zero.
Global Stochastic Properties of Dynamic Models and Their Approximations
with: Casper de Vries, Erasmus School of Economics
The dynamic properties of micro based stochastic macro models are often analysed through a linearisation around the associated deterministic steady state. Recent literature has investigated the error made by such a deterministic approximation. Complementary to this literature Babus and de Vries investigated how the linearisation affects the stochastic properties of the original model, considering a simple real business cycle model with a noisy learning-by-doing effect. The solution has a stationary distribution which exhibits moment failure and unbounded support. The approximation, however, yields a stationary distribution with bounded support and all moments finite.