Inhaltspezifische Aktionen

Alexandru Mandes

Assessing the Collective Behavior of Trading Algorithms and its Implications on Financial Market Systemic Risk


  • Bearbeiter: Dr. Alexandru Mandes
  • Titel: Assessing the Collective Behavior of Trading Algorithms and its Implications on Financial Market Systemic Risk
  • Kategorie: Promotion
  • Fachgebiet: Statistik und Ökonometrie
  • Status: Abgeschlossen WS 15/16
  • Bearbeitungs- Zeitraum: 10/2012 bis 03/2016
  • Gutachter: Prof. Dr. Peter Winker, Prof. Dr. Ludger Overbeck
  • Abstract: A purely methodological part of the thesis deals with model design of agent based modeling (ABM) for financial markets, by proposing a unitary design framework and by tackling the issue of size and complexity for ABM. Explicitly, we define a framework composed of three main building blocks, i.e. agent design, agent evolution and price discovery, and then discuss different facets of complexity, i.e. difficulty of description, difficulty of creation and degree of organization, as well as analyze different possible indicators along these three dimensions.

    The application part proposes an intraday financial market model, which implements a continuous double auction based on an event-based Java framework. Firstly, the order execution problem for order-book-centered ABMs is solved by means of a novel microstructure-based order placement strategy, built around an optimization problem that minimizes the risk adjusted execution cost while taking into consideration relevant market microstructure factors and intrinsic agent characteristics. Secondly, an explicit model of the time dimension is introduced, allowing for a correct implementation of trading strategies that are active at different temporal-frequencies. Using the proposed test-bed, the impact of high-frequency electronic liquidity provision strategies on the intraday market dynamics is addressed, with respect to market quality and systemic risk, under both regular and market stress conditions. New insights regarding the determinants of extreme volatility collective behaviors are provided – not only the inventory thresholds and the low latencies play an important role, but especially the number and the homogeneity of high-frequency strategies. Finally, two potential regulatory policies, i.e. minimum holding / quote resting time and financial-transaction taxes, are assessed and compared both with respect to their flash crash prevention power, as well as to their impact on market participants and market quality, shedding new light on the policy trade-offs.

  • Publikationen:

    Mandes, A., Gatu, C. and Winker, P. (2013). Convergence of Heuristic-based Estimators of the GARCH Model, In: Christian Borgelt et. al., eds., Towards Advanced Data Analysis by Combining Soft Computing and Statistics, vol. 285: 151-163, Springer, Heidelberg.

    Mandes, A. (2015). Microstructure-based order placement in a continuous double auction agent based model, Algorithmic Finance (4): 105–125.

    Mandes, A. (2015). Impact of inventory-based electronic liquidity providers within a high-frequency event- and agent-based modeling framework, Discussion Paper 15-2015, MAGKS Joint Discussion Paper Series in Economics.

    Mandes, A. and Winker, P. (2016). Complexity and model comparison in agent based modeling of financial markets, Journal of Economic Interaction and Coordination. DOI 10.1007/s11403-016-0173-0

  •  Vorträge: 

    • GENED Workshop and School on Networks in Finance and Macroeconomics, Kiel Institute for the World Economy, 28th - 29th April 2014
    • MAGKS Colloquium, Rauischholzhausen, September 9-10, 2014
    • 2nd Meeting of the German Network for New Economic Dynamics (GENED), Technical University Darmstadt, 29th - 30th September 2014
    • 7th International Conference on Computational and Methodological Statistics (ERCIM 2014) & 8th International Conference on Computational and Financial Econometrics (CFE 2014), University of Pisa (Italy), December 6-8, 2014
    • Doktorandenseminar der Sektion Finanzmaerkte (GGS), JLU Gießen, December 10, 2014
    • 20th Annual Workshop on the Economic Science with Heterogeneous Interacting Agents (WEHIA), University of Nice-Sophia Antipolis (France)
    • May 21-23, 2015
    • 3rd Meeting of the German Network for New Economic Dynamics (GENED), Ruhr University Bochum, 28th - 29th September 2015