LML External Fellow Imre Kondor has recently presented at a number of events and published several papers which are now available to view. We have noted below Imre’s latest work and to access the full catalogue please visit Imre’s page on our website here.
Imre Kondor
Papers and Talks from September 2016 – December 2016
Papers
I.Varga-Haszonits, F. Caccioli and I. Kondor: Replica approach to mean-variance portfolio optimization, accepted for publication by JSTAT, available at: http://arxiv.org/abs/1606.08679
Fiaschi, I. Kondor, M.Marsili and V. Volpati: The missing assets and the size of shadow banking: an update, available at: https://arxiv.org/pdf/1611.02760
Talks
On Top and Beyond Networks
A talk given at the Workshop on “Approaches to evolution of complex systems” held at CSH, Vienna and KLI, Klosterneuburg, November 10-11, 2016 and at the IMT School for Advanced Studies, Lucca, Italy, December 7, 2016.
The Instability of Risk Measures is Due to Their Unboundedness
A seminar at the Sant’Anna School of Advanced Studies, Pisa, Italy, December 5, 2016
Summary
The paper by Varga-Haszonits et al. is part of a larger project which applies methods borrowed from statistical physics to various problems in finance, including portfolio optimization. Optimization of large, institutional portfolios suffers from the “curse of dimensionality”: the number of different assets in the portfolio (the dimension) is never sufficiently small compared with the available data. This leads to instability, with the estimation error becoming very large and the optimization hopeless. We have demonstrated in a series of recent works that the standard statistical methods of regularization cannot efficiently resolve this problem, because the risk measures applied in asset management, risk management and regulation are not bounded. This allows the investor to assume large compensating long and short positions. The resulting high leverage is the origin of financial risk. The talk given at the Sant’Anna School in Pisa covered these works, plus some new, unpublished results, which are being written up presently.
The paper by Fiaschi et al. is an update on a previous work where a parallel was pointed out between the “missing assets” (the deviation of the rank distribution of the largest banks from a Pareto law) and the size of shadow banking.
The two talks in Vienna and Lucca dealt with the possible directions in which the popular network paradigm can be extended.
Leave a Reply