Sheri will be giving 4 lectures on Systemic Risk and Complexity Economics
Monday 6 July 2015
Morning (2 hours): Introduction to Systemic Risk and Macro-prudential Modelling: Market Prices Based
Methods v Network Analysis of Financial Interconnections
This lecture covers the desirable characteristics that one needs from systemic risk indexes. Early warning
signal (avoiding false negatives) is upheld to be high on this list. Market price based systemic risk indexes
lack early warning and may mislead about systemic risk as volatility indexes do. Bilateral asset-liability data
based network models for financial systems are needed to see how procyclical liabilities exceed financial
buffers, leading to system wide instability.
Afternoon (2 hours): Why Topology Matters in Financial Networks: Granularity, Stability and Tipping
Points
Many influential models have been misleading about the stability of financial networks by considering
only their connectivity and assuming homogeneity in the number and size of links of nodes. Real world
networks are highly heterogeneous in the number and size of links, increasing connectivity in such system can be dangerous. I show how the Robert May characterization of stability of a network system in terms of
its spectral properties as a dynamical system gives us a handle in determining financial stability and tipping
points. Both the index of stability of the financial network and the rank order of systemic importance and vulnerability can be obtained simultaneously as part of the eigen-pair solution.
Tuesday 7 July 2015
Morning (2 hours) : New Global Granular Macro-Net Models for Macruprudential Policy
The challenge is to marry, at a high level of data granularity, the financial networks with the real sectors of economies, along with a global/cross border dimension for both. Ignoring cross border imbalances in macro-prudential policy and systemic risk modelling is a fatal error. Hence, it is important to combine cross border imbalances with within country sectoral imbalances that have emerged with the growing size of the financial sector (including housing and mortgages) a la Carvalho-Gabaix-Acemeglu granular macro-economics. Specific analysis of the Euro zone core-periphery crisis will be done.
Afternoon (2 hours) : Complexity and Computability in Economics
There is a fundamental non-computability at the heart of economic interactions that economists have hitherto
ignored. It is the basis of organized complexity that produces highly contextual innovations and structure
changing dynamics as in arms races. It is produced by the interactions of highly sophisticated mentalizing
agents who can perceive opposition or contrarian/anti-coordinated structures. Fixed points involving
contrarian interactions cannot be computed. These can become launch pads for protean behaviours that can
produce novel objects and surprises. Evidence for such mutual mentalizing is given from latest discoveries in
neuroscience on mirror neurons by Vittorio Gallesse et. al. and also from the work of Scot Kelso, who
identified neuro markers for anti-coordination and opposition.