Aeolus launches research project to innovate hurricane prediction

Aeolus Capital Management Ltd., the Bermuda based reinsurance and insurance-linked securities (ILS) focused investment fund manager, has announced a new research project focused on hurricane prediction, looking at the medium-term time horizon.

Aeolus Capital Management has partnered with academics from the State University of New York (SUNY) at Albany Department of Atmospheric and Environmental Sciences, with the research collaboration set to explore the use of technologies such as machine learning’s application to hurricane forecasting and prediction.

This hurricane-focused research effort will be led by Dr. Pete Dailey, Head of Research at Aeolus Capital Management, a renowned expert in climate catastrophe model development and analytics.

A quantitative study, the research hopes to develop an enhanced understanding of the leading indicators related to hurricane development and tracking.

Specifically, the research will look at the medium-term time-horizon, which is the period between short-term weather forecasts (days) and climate projections (several months/years).

This is the time-horizon where those managing portfolios of insurance-linked securities (ILS) and reinsurance contracts can likely benefit most from a research-led exploration of how hurricanes develop and move across oceans, so could also yield the biggest gains for an asset management company like Aeolus.

Dailey explained, “We are excited to launch this research partnership with Dr. Robert Fovell and other faculty and graduate students at the University at Albany. By applying machine-learning techniques such as Self-Organizing Maps (“SOMs”) to hurricane data, the goal is to identify weather and climate patterns that signal risk before hurricanes strike.”

Importantly, Dailey also noted how research of this kind can be additive to Aeolus’ reinsurance investment management operations, furnishing it with information to aid in better decision-making.

“The insights we hope to take away from this study should advance our scientific understanding of hurricane risk and, ultimately, lead to better portfolio management decisions,” he said.

Dr. Robert Fovell, Professor of Atmospheric and Environmental Sciences added, “This is an important study into hurricane risk projections that looks to bridge the gap between short-term weather forecasts with lead times of a few days and medium-term climate projections with lead times of several months, with a specific focus on landfalling hurricanes. We look forward to our partnership with Aeolus to help improve our understanding of hurricane formation and movement.”