Wen Chen, Chinese University of Hong Kong-Shenzhen
Bo Hu, George Mason University
Liyan Yang, University of Toronto
Abstract: This paper rationalizes the use of LASSO for return predictions based on uncertain fat-tail priors and max-min robust optimization. Our theory excludes heuristic learning or restrictive prior assumptions in the statistical interpretation of LASSO by its inventor Tibshirani (1996). In our setting, agents (arbitrageurs) are uncertain about the scale of fat-tail shocks. In equilibrium, they ignore a range of ambiguous signals and respond linearly to almost unambiguous signals. Using this LASSO equivalent strategy, arbitrageurs amass extra market power which induces a "cartel" to protect their total profit from being competed a way. This result shows a new mechanism for limited arbitrage.
Yan Ji, Hong Kong University of Science & Technology
Abstract: The integration of algorithmic trading and reinforcement learning, known as AI-powered trading, has significantly impacted capital markets. This study utilizes a model of imperfect competition among informed speculators with asymmetric information to explore the implications of AI-powered trading strategies on speculators' market power, information rents, price informativeness, market liquidity, and mispricing. Our results demonstrate that informed AI speculators, even though they are ``unaware'' of collusion, can autonomously learn to employ collusive trading strategies. These collusive strategies allow them to achieve supra-competitive trading profits by strategically under-reacting to information, even without any form of agreement or communication, let alone interactions that might violate traditional antitrust regulations. Algorithmic collusion emerges from two distinct mechanisms. The first mechanism is through the adoption of price-trigger strategies (``artificial intelligence''), while the second stems from homogenized learning biases (``artificial stupidity''). The former mechanism is evident only in scenarios with limited price efficiency and noise trading risk. In contrast, the latter persists even under conditions of high price efficiency or large noise trading risk. As a result, in a market with prevalent AI-powered trading, both price informativeness and market liquidity can suffer, reflecting the influence of both artificial intelligence and stupidity.
Abstract: We explore the potential for automated market makers (AMMs) to enhance traditional financial markets, drawing on their success in the crypto-assets space. The increasing tokenization of assets and regulatory changes, including the SEC's initiatives to reshape retail order trading, underscore the relevance of considering AMMs in traditional markets. Our study establishes a practical framework to evaluate the viability of AMM liquidity provision in equities and assess if AMMs offer improvements over traditional markets. Analyzing U.S. equity trading data, we find that well-designed AMMs could save U.S. investors billions annually. These savings arise from the distinct characteristics of AMMs, especially the improved risk-sharing and the role of long-term asset holders as liquidity providers. Unlike traditional market makers, long-term asset holders in AMMs seek compensation only for incremental intraday risk relative to a buy-and-hold strategy. They utilize locked-up capital that would otherwise remain idle at brokerages. Small firms, in particular, can benefit by attracting more investors and capital through this approach.
Discussant: Chen Yao, Chinese University of Hong Kong
