Oxford-Man Institute of Quantitative Finance, University of Oxford

The OMI is delighted to share its April 2025 Newsletter. Link to the PDF: https://lnkd.in/girHMe79 #machinelearning #quantitativefinance 

The OMI is excited to be hosting our fourth Machine Learning and Quantitative Finance Conference on the 5th and 6th of June 2025. The conference provides an opportunity for junior and senior academics to discuss their latest work machine learning and quantitative finance. We have a great lineup of speakers: Xuedong He, Giulia Livieri, Andrew Lo, Lin Peng, yuantao shi, Mihail Velikov, Xiao Xiao, and Ruixun Zhang. Register on the conference website: https://lnkd.in/g3ExF7mY

We are happy to share a new working paper “The Limited Virtue of Complexity in a Noisy World” by Álvaro Cartea, Qi Jin, and yuantao shi. Highlights: •⁠ ⁠In a high-dimensional factor space, increasing model complexity under proper regularization can enhance the predictability of asset returns; however, the Sharpe ratio of a portfolio of assets and the R-squared of the prediction of the asset returns decrease monotonically and are convex as the noise level in factors increases. •⁠ ⁠⁠When only a subset of factors is observed, there is an optimal level of complexity beyond which incorporating additional factors can degrade portfolio performance due to the effect of noise in the factors. •⁠ ⁠⁠They underscore a limited virtue of complexity in financial forecasting, where the performance of portfolios depends on the noise level in factors, and where more complex models do not necessarily lead to better performance when factors are not perfectly observed. Link to paper: https://lnkd.in/gE8RFqCs

We are happy to share a new working paper “Equilibrium Reward for Liquidity Providers in Automated Market Makers” by Alif A., Philippe Bergault, and Leandro Sánchez-Betancourt. Highlights: •⁠ ⁠They find the equilibrium contract that an automated market maker offers to their strategic liquidity providers (LPs) to maximise the order flow that gets processed by the venue. •⁠ ⁠⁠They formulate the problem as a leader-follower stochastic game, where the venue is the leader and the LPs are the followers. •⁠ ⁠⁠They find that under the equilibrium contract, LPs have incentives to add liquidity to the pool only when higher liquidity on average attracts more noise trading. Link to paper: https://lnkd.in/giY4xMkV

We are happy to share a new working paper “A Simple Strategy to Deal with Toxic Flow” by Álvaro Cartea and Leandro Sánchez-Betancourt. Highlights: •⁠ ⁠They study the trading activity between a broker and her clients. •⁠ ⁠⁠They derive the optimal trading strategy that balances maximising the order flow from the broker’s clients and minimising adverse selection losses to the informed traders. •⁠ ⁠⁠They use the optimal strategy to introduce an algorithm that bypasses the need to calibrate individual parameters, which makes the strategy suitable in real-world trading environments. Link to paper: https://lnkd.in/ggSNcCNd

The OMI is delighted to share its March 2025 Newsletter. Link to the PDF: https://lnkd.in/gNaFAEvd #machinelearning #quantitativefinance 

Congratulations to Gerardo Duran-Martin, Leandro Sánchez-Betancourt, Alexander Shestopaloff, and Kevin Murphy for publishing their new paper “A unifying framework for generalised Bayesian online learning in non-stationary environments” in TMLR. Highlights: • ⁠They provide a unifying framework to study Bayesian online learning in dynamic and non-stationary environments.  • The key ingredients of the framework are: - a model to map inputs and outputs, - a "notion" of non-stationarity represented by an auxiliary random variable, and - a conditional prior that updates beliefs based on this auxiliary variable. • The framework defines a broad spectrum of methods in the literature, and offers a unified approach to develop new methods. Link to paper: https://lnkd.in/gQiSk9aU

We are happy to share a new working paper “Alpha in Analysts” by Álvaro Cartea and Qi Jin. Highlights: •⁠ ⁠They examine the investment value in sell-side analyst price targets by treating each analyst as a portfolio manager and use their price targets to construct 12-month implied return forecasts. •⁠ ⁠⁠Their analysis shows that the average analyst does not generate statistically significant alpha relative to the returns of a long-only portfolio benchmark, but a subset of analysts exhibits persistent alpha. •⁠ ⁠⁠They introduce a “fund-of-analysts” framework that first predicts analyst performance and then dynamically allocates weights across analysts based on predicted analyst performances. •⁠ ⁠⁠They show that this meta-portfolio strategy can yield significant alpha over long-only benchmarks Link to paper: https://lnkd.in/gs5qWbSf

Congratulations to Álvaro Cartea, Mihai Cucuringu, Qi Jin, and Mungo Wilson for their new working paper “Volume Shocks and Overnight Returns”. Highlights: •⁠ ⁠Intraday volume shocks are positively correlated to subsequent overnight stock returns, while no significant effect is observed during the following intraday session. •⁠ ⁠⁠The relationship between volume shocks and stock returns holds true for stocks regardless of their market capitalization and regardless of subsamples throughout the history. •⁠ ⁠⁠The findings question conventional explanations that attribute the relationship between abnormal trading volume and subsequent stock returns solely to investor attention and change in cost of capital. •⁠ ⁠⁠Linear regression and machine learning techniques (tree and deep learning models) are used to forecast volume shocks, which they use to construct tradable portfolios. They show that non-tradability is not the cause of the observed relationship between volume shocks and subsequent overnight stock returns. Link to paper: https://lnkd.in/g-JAFF_Z

The OMI is delighted to share its February 2025 Newsletter. Link to the PDF: https://lnkd.in/g2fWJRHR #machinelearning #quantitativefinance