Damien Challet

Full professor (HDR). Associate editor for Quantitative Finance, Journal of Economic Interaction and Coordination, Journal of Statistical Mechanics: theory and experiments. Co-chief editor of Market Microstructure and Liquidity. Previously associate editor for Applied Mathematical Finance (2006-2018).  

Research interests

• Agent behaviour (trader, investment fund, programmer). Recently, my interest has focused on agent-resolved data and aims at understanding what triggers activity and characterizing collective optimization in very heterogeneous populations.
• Statistics: I have introduced a new way to build nonparametric statistics for location and distribution equality tests from the number of records of sample cumulative sums, called r-statistics. One paper is about their power and another about their efficiency as an estimator of Sharpe ratios.
• Network inference, especially lead-lag networks. See e.g. this preprint.
• Machine learning applied to financial data and fundamental economic data. E.g. Google Trends: see a comment and a paper. More recently, this paper mixes order flow prediction with random forests and lead-lag trader activity network inference.

Coordinates

Room: SC.113, Bouyges building damien.challet@centralesupelec.fr Tel.:  +33-(0) 1 75 31 67 55

Publications

• e-prints
• a(R)ticles a blog about my latest articles and R contributions.
• Publons: recent activity as a referee and editor.

Bio

See my CV.

Source code

There is no reason to open-source mathematical computations and not computer code when submitting a paper. My repositories may be found on github.com. Some recent contributions:

• sharpeRratio (), an R package available on CRAN that implements my moment-free estimator of Sharpe ratios (see this preprint and the published paper) . Note that versions prior to 1.2 were prone to a pessimistic bias for heavy-tailed data.
• SVN (), an R package available on CRAN which computes Statistically Validated Networks (Tumminello et al. 2011)
• bahc (), an R package available on CRAN which implements the correlation/covariance filtering method introduced in Bongiorno and Challet (2020), also available on PyPI.
• ReplacementBootstrap, an R package available on github.com repository that implements a max-entropy based bootstrap procedure, written in collaboration with Amir Sani. See this paper.

Teaching resources

Current courses

• Deep Learning in Finance (2021-), how to leverage Deep Learning to replicate or improve financial models (estimation, volatility control, asset selection, timeseries generation), Applied Mathematics option for third-year students of CentraleSupélec and students following the Master in Statistics and Finance of Paris-Saclay University.
• Agent-based modelling of financial markets (2019-), a course for first year CentraleSupélec students.
• Physique des marchés (2012-), a mechanistic approach to financial market predictability and algorithmic trading, Applied Mathematics option for third-year students of CentraleSupélec and students following the Master in Statistics and Finance of Paris-Saclay University.
• Financial Big Data at EPFL (Master in Financial Engineering) (2016-). The whole aim is to learn how to retrieve and wrangle annoyingly big data, use estimators adapted to financial data, find strategies with machine learning and apply these techniques to a meaningful financial problems.

Previous courses

• Big Data and Algorithmic Trading (2018-2020), Advanced Master in Quantitative Finance, Solvay Business School, Brussels, Belgium. In short, big data and machine learning applied to trading.
• “Big data”, Advanced Master in Quantitative Finance, Solvay Business School, Brussels, Belgium (2017)
• “Inefficient Markets” at HEC Lausanne between 2010 and 2015. It was about why financial markets cannot be efficient. The course plan was as follows: 1) financial markets deconstruction (behavioural finance), 2) reconstruction (agent-based modelling), 3) the art of speculative strategy backtesting. Lecture slides are available there.

 

Internships

I will gladly supervise Master theses on a topic that is related to my interests (and, incidentally, abilities). Please check that yours are compatible with mine before contacting me. Note that the whole administrative process takes more than 3 months before the start of the internship.  

Press Coverage (non-exhaustive list)

1. Minority Game: What stock investors can learn from the failure of Black Friday, Paul Ormerod, CITY A.M., 2 December 2015
2. Tweaking taps for a constantly warm shower, New Scientist, 16 February 2008, p. 18
3. The maths behind group showers Philip Ball, Nature (18/01/2008)
4. Showered in surprises Michael Banks, Physics World, Volume 21 No 2, February 2008
5. Openness makes software better sooner Nature Science update (2003)
6. Open wide… Maths.org (2003)
7. Le bazar et la cathédrale, ou comment l’Open Source élimine les bugs plus rapidement (2003)
8. Forscher: Freierhältliche Quelltexte verbessern die Qualitaet von Computerprogrammen Wissenshaft.de (2003)
9. Open Source liefert mehr Qualitaet silicon.de (2003)
10. Debugging in OSS Always Faster slashdot.org (2003)
11. Collective Effort Makes the Good Times Roll, Adrian Cho Science 2002 July 5; 297: 33
12. Whole better than parts Nature Science update (2002)
13. N/2 Wrongs Make a Right Physical Review Focus (2002)
14. Defects combine to make perfect devices Physics Web (2002