This project develops data-driven trading signals by extracting and quantifying narratives from social media. Using advanced NLP and AI methods, we study how online sentiment, attention, and framing translate into market movements and investment behavior.

This project investigates how interest-rate differentials affect daily exchange-rate returns across G10 and emerging-market currencies. Using high-frequency FX and money-market data, we study whether currency sensitivities to short-term interest-rate differentials vary over time, change across monetary-policy and risk regimes, and differ between developed and emerging FX markets

Hate Radar is an AI-based platform for the real-time detection and monitoring of online hate speech. The project combines natural language processing, machine learning, and large-scale data collection to identify harmful content, track emerging narratives, and support evidence-based policy and intervention design.

This project develops open-source (R-based) tools for currency trading and hedging strategies tailored to small and medium enterprises and pension funds. We focus on univariate and multivariate approaches, drawdown management, and tail-risk protection to improve resilience against exchange-rate volatility.

This project studies how different outreach strategies affect tuberculosis screening uptake in rural Ghana. Using evidence from a randomized controlled trial, we examine whether personal appeals and financial incentives can increase participation in TB screening, with a focus on the relative effectiveness of social motivation, payments, and targeted public-health communication.

“Talking to Machines” explores how interactions with artificial intelligence affect human financial decisions. The project brings together an international team of academic and industry researchers to design AI-driven experiments, with findings published in leading social science journals and used to inform public policy and large-scale behavioral interventions.