Point72, the hedge fund founded by Steven Cohen, has hired an artificial intelligence expert to build weather prediction models for its traders, according to a person familiar with the matter. The move aims to give the firm an edge in commodity trading, where weather patterns can significantly impact prices of agricultural products, energy, and other assets.
AI Expert Joins Point72
The new hire is a specialist in machine learning and atmospheric science, who will work with Point72's quantitative research team. The expert's role is to develop algorithms that can forecast weather conditions more accurately than traditional models, using vast datasets including satellite imagery, historical weather data, and real-time sensor readings.
Point72 has been expanding its use of artificial intelligence and data science across its investment strategies. The firm now employs over 100 data scientists and engineers, according to its website. The weather prediction project is part of a broader push to incorporate alternative data sources into trading decisions.
Impact on Commodity Trading
Weather forecasting is critical for commodity traders because it affects supply and demand dynamics. For example, a predicted drought in a major grain-producing region could lead to higher crop prices, while an expected mild winter might reduce demand for heating oil. By having more accurate forecasts, traders can position themselves ahead of market moves.
Traditional weather models are often limited by computational constraints and data gaps. AI models can process larger volumes of data and identify complex patterns that humans might miss. Point72's AI expert is expected to develop models that provide probabilistic forecasts, giving traders a range of possible outcomes and their likelihoods.
Competitive Landscape
Other hedge funds and trading firms have also been investing in weather prediction. For instance, Citadel has a team of meteorologists and data scientists working on weather models, and Two Sigma uses machine learning to analyze weather data. However, Point72's approach is said to be more focused on integrating AI directly into trading algorithms.
The hiring underscores the growing importance of alternative data in finance. According to a report by Greenwich Associates, 56% of asset managers now use alternative data, up from 35% in 2016. Weather data is one of the most commonly used alternative data sets, particularly for commodity and energy funds.
Challenges and Opportunities
Developing accurate weather prediction models is challenging due to the chaotic nature of weather systems. Even with advanced AI, long-term forecasts remain unreliable. However, short-term predictions (up to 15 days) have improved significantly in recent years thanks to machine learning. Point72's models are expected to focus on these shorter time horizons.
The firm has not disclosed the financial terms of the hire or the specific trading desks that will use the models. A spokesperson for Point72 declined to comment on the hiring.
