I am a doctoral candidate at the London Business School. I am an empirical researcher who works on climate risk and sustainability. I apply econometric and deep learning methods on big and alternative data to address questions around this theme. I have also served as a senior consultant for data science and investment management organisations. I am on the 2025-26 academic job market.
solo authored
I compile seven million residential real estate transactions in the United Kingdom and recover discount rates used by homeowners to value dwelling sustainability. To do this, I calibrate the present value of energy savings from subsequent improvements in dwelling sustainability to the observed price premium. I show that homeowners accept lower returns for greener dwellings, evidenced by the declining structure of discount rates with increasing dwelling sustainability. Moreover, I exploit the spatial, temporal, tenurial, and vintage variation in premium to demonstrate that homeowners price sustainability following economic principles. My estimates provide direct measures for rates used to discount climate investments.
Presentations: Society of Finance Studies (SFS) Cavalcade, Northern Finance Association (NFA), ESG Conference at Cornell SC Johnson College of Business, Business Schools for Climate Leadership Conference at IESE Business School, Ageing and Sustainable Finance Conference by Leibniz-Centre for European Economic Research, 31st Finance Forum by Spanish Finance Association, INFORMS Annual Meeting, and Trans-Atlantic Doctoral Conference at London Business School
Grants: SFS PhD Travel Grant, NFA PhD Travel Grant
with Victor DeMiguel and Javier Gil-Bazo
To study the impact of green-transition regulation on firm value, we analyze stock returns around legislator tweets about climate change. Green stocks significantly outperform brown stocks in the one-to-ten-minute window around pro-transition tweets. The cumulative daily-average green-minus-brown portfolio return around pro-transition tweets is 6.9% higher than around anti-transition ones. For tweets that mention environmental regulations, the spread increases in the weeks before a congressional vote and is stronger before close votes and when Congress is split. Our findings suggest that the green transition impacts the relative performance of green and brown stocks, at least partly, via a regulatory channel.
Presentations: Asset Pricing and Machine Learning conference at Gothenburg University, INFORMS Annual Meeting, Workshop on Unstructured Data and Language Models at Michigan Ross*, ESG Workshop at Toulouse Business School*, Spring Workshop at ESADE*, Dauphine University*, WHU Otto Beisheim School of Management*, Conference in Sustainable Finance at the University of Luxembourg, and 2nd Exeter Sustinable Finance Conference*
(* co-authors)
Grants: INQUIRE Europe Research Grant
We examine the high-frequency returns and volatilities of more than 200 factor portfolios around 1.2 million tweets posted by members of the U.S. Congress. We note that while the changes in volatility are predictable in the one-to-thirty-minute window around tweets, there is no return predictability. For the five Fama-French and momentum factors, we predict changes in volatilities with an average accuracy of 60.35%. Our methodology uses tweet embeddings from large language models to estimate volatility changes across all factors simultaneously, avoiding separate queries per factor and thereby improving computational efficiency while reducing framing bias. Overall, legislator tweets reduce uncertainty, evidenced by declining factor volatility immediately after.