Ledger Investing is proud to announce the release of BayesBlend, an innovative open-source model averaging library developed by our very own Nathaniel Haines and Conor Goold.

At Ledger, we use BayesBlend to combine predictions from multiple mathematical models, where each model reflects a unique hypothesis about how casualty insurance portfolios will perform into the future. This "data-driven hedging" gives us a better view than any given model in isolation. We continue to develop cutting edge analytics and commit to publishing our work to make it available to you.

Check out BayesBlend and see how it can enhance your data science projects: GitHub Link.

More on Ledger’s modeling approach here.