Hosted by: Álvaro Cartea, the Oxford-Man Institute, Thierry Foucault, HEC Paris and José Penalva, University Carlos III, Madrid


Full schedule here


José S. Penalva, Universidad Carlos III, Madrid

Title: The Algorithmic Learning Equations: Evolving Strategies in Dynamic Games

Abstract: We introduce the algorithmic learning equations (ALEs), a set of ordinary differential equations which characterizes the finite-time and asymptotic behaviour of the stochastic interaction between state-dependent learning algorithms in dynamic games. We use the ALEs to prove that play converges to a Nash equilibrium or a correlated equilibrium depending on the type of signals the players receive and the algorithms employed by each player. Our framework allows for a variety of information and memory structures, including noisy, perfect, private, and public monitoring. We demonstrate our methodology in a repeated $2 \times 2$ prisoner’s dilemma game with perfect monitoring. Our analysis of the ALEs shows that algorithms can learn a reward-punishment mechanism to sustain tacit collusion and can also learn to coordinate in cycles of mutual collusion and mutual punishment.

Bio: José Penalva is Assiociate Professor of Finance at the Business Department of the Universidad Carlos III. He has an economics doctorate from the University of California, Los Angeles (UCLA) since 1997. He currently teaches Information in Markets and Market Microstructure, and Financial Mathematics.

His research interests are: the economics of information with a special emphasis on auctions and financial applications, market microstructure, and risk assignment and distribution in insurance markets. His publications include journals of international prestige such as Econometrica, Journal of Banking and Finance, Quaterly Journal of Finance, Review of Economic Dynamics and the Journal of Risk and Insurance. He gives presentations of his research at prestigious international conferences such as the European Group of Risk and Insurance Economists (EGRIE), the European Economic Association, the European Finance Association, and the European Econometric Society Meeting, among others. He has also participated in several edited books, and has co-authored (with Alvaro Cartea and Sebastian Jaimungal) the book “Algorithmic and High-frequency Trading” (Cambridge University Press, 2015)


Simon Mayer, Chicago Booth

Title: Antitrust, User Union, and Regulation in the Age of Digital Platforms and Big Data

Abstract: Firms’ production function in the digital era entails customers’ network adoption and data contribution. We model platform competition with endogenous pricing, user heterogeneity, network effects, infrastructure investment, and data collection and sharing, thereby providing a unifying framework to evaluate data-related regulation and antitrust policies. Similar to and interacting with network effects, data feedback, while improving service quality, may concentrate market power. Platforms strategically underprice initially but subsequently overcharge users, and can “collude” through data sharing.  Meanwhile, because users are dispersed, they do not internalize the impact of their actions (e.g., data contribution and sharing) on (i) future service or product quality which affects all users, (ii) concentration of market power, and (iii) platforms’ incentives to innovate and invest in data infrastructure. We show that data sharing proposals (e.g., open banking and data vendor) and user privacy protections (e.g., GDPR and CCPA) fail to address inefficiencies in data-driven competition and platforms’ ex-ante investment for data production. We propose and model user union as an effective solution for consumer protection: a representative governing body coordinates users’ contribution to the platforms and maximizes user surplus.

Bio: Simon Mayer is an Assistant Professor in Finance at HEC Paris. Prior to joining HEC Paris, he was a research fellow at the Fama-Miller center at Chicago Booth, and he completed his PhD in Financial Economics at Erasmus University and Tinbergen Institute in 2021. Simon’s research focuses on corporate finance, financial intermediation, and FinTech.


Alberto Rossi, Georgetown


Title: Selecting Mutual Funds from the Stocks They Hold: a Machine Learning Approach

Abstract: We combine individual mutual fund holdings and a large number of stock characteristics (factors) to compute fund-level exposures to factors on the basis of the stocks they hold. Fund performance is non-linearly related to fund factor exposures and their interactions. This feature proves important when we predict fund performance, as machine learning methods such as boosted regression trees (BRTs) significantly outperform standard linear frameworks and the BRT-generated forecasts encompass the ones generated by the predictors of mutual fund performance that have been proposed in the literature so far. Finally, factor exposures explain the vast majority of mutual fund performance.

Biography: Alberto Rossi is a Professor of Finance at the McDonough School of Business, Georgetown University. He is also the Director of the AI, Analytics, and Future of Work Initiative at Georgetown and an Academic Fellow of the Luohan Academy. His research interests include FinTech, Household Finance, Machine Learning, and Asset Pricing. His recent work studies how robo-advisors can help individuals make better financial decisions and how to predict stock market returns using machine learning algorithms. He has worked extensively in analyzing big data, and has collaborated with major brokerage houses, FinTech firms, and asset managers around the world.

Professor Rossi’s work has been published in leading academic journals such as the Journal of Finance, the Review of Financial Studies, the Journal of Financial Economics and Management Science. 

Before McDonough, he was an Associate Professor with tenure at the R.H. Smith School of Business, University of Maryland. He also worked as an economist at the Board of Governors of the Federal Reserve System in Washington DC. He received his Ph.D. in Economics from the University of California, San Diego. 


Matthias Qian, University of Oxford


Title: Startups and Job Redesign: Evidence from the Third Age of Artificial Intelligence

Abstract: My job market paper examines the contexts in which VC-funded startups accelerate the AI-driven redesign of jobs. With the falling cost of prediction, a growing number of occupations are redesigned to include AI prediction tasks within their task bundle. Such job redesign is the main margin of adjustment within the labour market in response to the emergence of Artificial Intelligence, and within startups, over 20% of jobs are redesigned. By decomposing with NLP over 250 million online job postings into 1.2 billion task descriptions, I document that startups’ experimentation with job redesign has large local spill-over effects on incumbent firms. The effect is asymmetric, and incumbents have a much weaker effect on the prevalence of job redesign at startups. I find that the startup’s efforts to redesign jobs are associated with improved fundraising success. For incumbent firms, the redesign of jobs results in increased return on assets and sales growth. These findings establish an important role which VC-funded startups play in the dissemination of complementary work practices to AI adoption. We also highlight a new margin of experimentation for entrepreneurs, which increases fundraising success. 

Biography: Matthias is a Research Fellow applying innovations in machine learning and artificial intelligence to solve real-world problems. He believes that artificial intelligence will revolutionise the social sciences and help the discipline contribute more to society. He specialises in machine learning algorithms for natural language processing, clustering, time-series forecasting and outlier detection.

Matthias’ research lies at the intersection of theory and practice. He contributed to the asymptotic theory of a new class of algorithms which automatically model structural breaks in time-series data. His empirical work covers entrepreneurial ecosystems, financial markets forecasting, the rise of flexible work arrangements (the gig economy), and economic geography.


Jianqing Fan, Princeton

Title: How and When are High-Frequency Stock Returns Predictable

Abstract: This talk presents the studies on the predictability of ultra high-frequency stock returns and durations to relevant price, volume and transactions events, using machine learning methods. We find that, contrary to low frequency and long horizon returns, where predictability is rare and inconsistent, predictability in high frequency returns and durations is large, systematic and pervasive over short horizons. We identify the relevant predictors constructed from trades and quotes data and examine what determines the variation in predictability across different stock’s own characteristics and market environments. Next, we compute how the predictability improves with the timeliness of the data on a scale of milliseconds, providing a valuation of each millisecond gained. Finally, we simulate the impact of getting an (imperfect) peek at the incoming order flow, a look ahead ability that is often attributed to the fastest high frequency traders, in terms of improving the predictability of the following returns and durations. (Joint work with Yacine Air-Sahalia, Lirong Xue, and Yifeng Zhou).

Bio: Jianqing Fan is Frederick L. Moore ’18 Professor of Finance, Professor of Statistics, and former Chairman of Department of Operations Research and Financial Engineering at the Princeton University where he directs both financial econometrics lab and statistics lab. He previously held professorships at UNC-Chapel Hill, and UCLA. He has authored or co-authored over 250 articles on financial econometrics, statistical machine learning, analysis of Big Data, and various aspects of theoretical and methodological statistics and machine learning. His finance work focuses on the analysis of high-frequency data, option pricing, portfolio theory, risk assessment, high-dimensional data, and time series. He  was the joint-editor of Journal of Business and Economics Statistics, Journal of Econometrics, Annals of Statistics, and Econometrics Journal, and has served as associate editor of Econometrica, Management Science, Journal of American Statistical Association, and Journal of Financial Econometrics.

His published work has been recognized by the 2000 COPSS Presidents’ Award, the 2007 Morningside Gold Medal of Applied Mathematics, and a Guggenheim Fellowship in 2009, Academian of Academia Sinica 2012, P.L. Hsu prize in 2013,  Guy Medal in Silver, 2014, and Noether Senior Scholar Award in 2018. He is an Elected Fellow of the American Association for Advancement of Science,  the Society of Financial Econometrics, the Institute of Mathematical Statistics, and the American Statistical Association,  and a past President of the Institute of Mathematical Statistics. 


Christine Parlour, Berkeley

Title: Battle of the Bots

Abstract: Settlement on decentralized ledgers is transparent and batched.  The settlement also allows settlement agents to expropriate profitable arbitrage trades.  Arbitrage may be socially beneficial or wasteful. We model the effect of an alternate, private settlement on  arbitrage.  We document payments from arbitrageurs to private settlers that exceed 1 million USD per day

Bio: Christine A. Parlour is the Sylvan C. Coleman Chair of Finance and Accounting at Berkeley Haas. Most of her work is in institutionally complex areas, such as market microstructure , FinTech and Banking. Her current work focuses on changes in the payments system and the effects on bank balance sheets and FinTech especially Decentralized Finance.   She has written for major finance and economics journals. She has been on the Nasdaq Economic Advisory Board and is currently on the steering committee for the New Special Study of Securities Markets.  She is the current president of the Western Finance Association, past president of the Finance Theory Group and co-director of Berkeley Center for Responsible, Decentralized Intelligence (RDI).


Huan Tang, LSE

Title: The Supply and Demand for Data Privacy

Abstract: This paper investigates how consumers and investors react to the standardized disclosure of data privacy practices. Since December 2020, Apple has required all apps to disclose their data collection practices by filling out privacy “nutrition” labels that are standardized and easy-to-read. We web-scrape these privacy labels and first document several stylized facts regarding the supply of privacy. Second, augmenting privacy labels with weekly app downloads and revenues, we examine how this disclosure affects consumer behaviour. We exploit the staggered release of privacy labels and use the non-exposed Android version of each app to construct the counterfactual. After privacy label release, an average iOS app experiences a 14% (15%) drop in weekly downloads (revenues) when compared to its Android counterpart, with an even stronger effect for more privacy-invasive and substitutable apps. Consumers in the US, UK, and France respond more negatively, suggesting that they are most averse to data collection. Moreover, we observe adverse stock market reactions, especially among firms that harvest more data, corroborating the findings on product markets. Our findings highlight data as a key asset for firms in the digital era.

Bio: Huan Tang is an assistant professor of finance at the London School of Economics. Her areas of expertise are FinTech, banking, and digital economy. In her recent work, she combines novel data and theory-motivated empirical approach to understand the benefits and costs of FinTech innovations for households and firms. She is also interested in data privacy issues brought about by the digital economy. She is the recipient of the 2020 AQR Top Finance Graduate Award. Huan Tang received a PhD in Finance in 2020 from HEC Paris.


Roxana Mihet, Swiss Finance Institute and HEC Lausanne

Title: Consumer Privacy and the Value of Consumer Data

Abstract: We analyze how the adoption of the California Consumer Privacy Act (CCPA), which limits consumer personal data acquisition, processing, and trade, affects voice-AI firms. To derive theoretical predictions, we use a general equilibrium model where firms produce intermediate goods using labor and data in the form of intangible capital, which can be traded subject to a cost representing regulatory and technical challenges. Firms differ in their ability to collect data internally, driven by the size of their customer base and reliance on data. When the introduction of the CCPA increases the cost of trading data, sophisticated firms with small customer bases are hit the hardest. Such firms have a low ability to collect in-house data and high reliance on data and cannot adequately substitute the previously externally purchased data. We utilize novel and hand-collected data on voice-AI firms to provide empirical support for our theoretical predictions. We empirically show that sophisticated firms with voice-AI products experience lower returns on assets than their industry peers after the introduction of the CCPA, and firms with weak customer bases experience the strongest distortionary effects.

Bio: Roxana Mihet is Assistant Professor of Finance at HEC Lausanne and Faculty Member of the Swiss Finance Institute and Research Affiliate of the CEPR. Roxana examines the modern data economy and how new technologies, such as artificial intelligence and big data, are impacting the macroeconomy and financial markets. Her research has won various prizes, including the ECB Young Economists’ Competition, the Cubist Strategies Prize for Outstanding Research at the WFA, and the Sandoz Foundation – Monique de Meuron Program Fellowship. Prior to HEC Lausanne, Roxana completed a Ph.D. in Business Economics at NYU, Stern School of Business, an M.Phil. in Economics at the University of Oxford, and a B.A. in Economics and Mathematics at the University of Chicago. She also worked, over the years, in the Research Departments of the IMF, BIS, EBRD and Norges Bank on macro-finance issues.


Olivier Dessaint, INSEAD

Title: Does Alternative Data Improve Financial Forecasting? The Horizon Effect

Abstract: “Existing research suggests that alternative data is mainly informative about short-term future outcomes. We show theoretically that the availability of short-term oriented data can induce forecasters to optimally shift their attention from the long-term to the short-term because it reduces the cost of obtaining short-term information. Consequently, the informativeness of their long-term forecasts decreases, even though the informativeness of their short-term forecasts increases. We test and confirm this prediction by considering how the informativeness of equity analysts’ forecasts at various horizons varies over the long run and with their exposure to social media data.”

Bio: Olivier DESSAINT is professor of finance at INSEAD. His research interests include corporate finance and behavioral finance. His work has been published in leading academic journals such as the Journal of Financial Economics, the Review of Financial Studies, and the Review of Finance. He currently teaches in the MBA and PhD programs. Olivier holds a PhD in finance from HEC Paris. Prior to joining INSEAD, he was a faculty member at the Rotman School of Management from the University of Toronto in Canada. Before joining academia, he was also an investment banker at BNP Paribas. He was a part of the M&A advisory teams in Paris and Madrid.


Jean Edouard Colliard, HEC Paris

Title: Algorithmic Pricing and Liquidity in Securities Markets”, and is joint work with Thierry Foucault and Stefano Lovo (both at HEC Paris). 

Abstract: We let machine learning algorithms play a market-making game à la Glosten and Milgrom (1985), and compare their behavior to the predictions of standard theory. We find that “algorithmic market-makers” learn how to deal with adverse selection, and to update their prices after observing the order flow. However, while theory would predict competitive prices in our environment, we find that “algorithmic market-makers” charge a mark-up over the competitive price. Competition reduces mark-ups, but they are still significant with 10 market-makers. Importantly, we find that mark-ups decrease with the extent of adverse selection market-makers face. This result suggests that regulators concerned with the possibility of algorithmic collusion in financial markets should counter-intuitively focus on assets and periods with low adverse selection.

Bio: Jean-Edouard Colliard is Associate Professor of Finance at HEC Paris. His main research areas are the regulation of financial institutions and the microstructure of financial markets, including topics such as financial transactions taxes, over-the counter markets, bank capital requirements, or the European Banking Union. Jean-Edouard’s research has been published in leading finance and management journals such as the Journal of Finance, the Review of Financial Studies, and Management Science.


Emilio Calvano, University of Rome – Tor Vergata

Title: Artificial Intelligence, Algorithmic Pricing, and Collusion

Abstract: Increasingly, algorithms are supplanting human decision-makers in pricing products and auction bidding. I will present a study of the behavior of algorithms powered by Artificial Intelligence (Q-learning) in a workhorse oligopoly model of repeated price competition, and discuss policy implications as well as avenues for future research. The study finds that the algorithms consistently learn to charge supra-competitive prices, without communicating with one another. The high prices are sustained by collusive strategies with a finite phase of punishment followed by a gradual return to cooperation.

Bio: Emilio Calvano (PhD, Toulouse School of Economics, 2008) is a full professor of Economics at the University of Rome – Tor Vergata, associate faculty at the Toulouse School of Economics, and research fellow of the Centre for Economic Policy Research in London (CEPR). He is an applied theoretical economist whose research mainly focuses on the theory of Industrial Organization. He has published in top international peer-reviewed outlets such as Science Magazine, American Economic Review, Management Science, American Economic Journal: Microeconomics, Economic Journal, and International Journal of Industrial Organization and is an associate editor at the Journal of Industrial Economics. His research interests include the economics of artificial intelligence, the economics of platforms, information economics, and competition policy. His recent work studies the impact of AI-powered algorithms (such as pricing software and recommender systems) on digital markets


Maxime Bonelli, HEC Paris

Title: The Adoption of Artificial Intelligence by Venture Capitalists

Abstract: I study how the adoption of artificial intelligence (AI) by venture capitalists (VCs) to screen investment opportunities affects the funding of innovation. First, I document a series of new facts using global data on VCs’ investments. After AI adoption, VCs tilt their portfolio toward startups similar to previously tried businesses. Among these, VCs become better at picking startups that will receive follow-on funding. However, VCs’ investments become less likely to result in major breakthroughs, e.g., IPO or patents. These results are consistent with AI using historical data informative about startups similar to previous businesses but not about breakthrough companies. Second, to estimate the causal effect of AI adoption on VCs’ investments I exploit a shock increasing VCs’ number of potential investment opportunities. This shock raises VCs’ incentives to adopt AI to automate screening. I find that VCs more exposed to the shock are more likely to adopt AI and I confirm its causal effect on VCs’ investments. My findings suggest that AI adoption by VCs might not enable more funding of breakthrough innovations.

Bio: Maxime is a PhD candidate in Finance at HEC Paris. He will be on the 2022/2023 academic job market. His main research interests include investment management, entrepreneurship and innovation, and labor and finance. Before beginning his doctoral studies at HEC Paris, Maxime worked as a quantitative researcher in the asset management industry for more than four years. He holds a Ph.D. in Mathematics (Inria), a Master in Economics (Ecole polytechnique – ENSAE – HEC Paris) and an Engineering degree in Applied Mathematics (Ecole Polytechnique Université Nice Sophia Antipolis).