Oasis Loss Modelling Framework (Oasis LMF)
An open, free platform that runs catastrophe models from different vendors on one shared financial calculation engine, built around public data formats.
The Oasis Loss Modelling Framework (Oasis LMF) is an open calculation platform that runs catastrophe models from different vendors on the same engine. The initiative started in the early 2010s, backed by Lloyd's and reinsurance players who wanted out of a market dominated by two or three proprietary modelers. Its code is published as open source. The idea rests on a split of roles. The vendor supplies hazard and vulnerability, as event footprint tables and damage curves. Oasis handles the rest: reading exposures, sampling secondary uncertainty, applying policy and treaty terms, producing losses by event. Losses computed event by event are then rolled up into average annual loss and into occurrence and aggregate exceedance curves. The license, when there is one, covers the model; the engine stays free. Two public formats act as a common language, Open Exposure Data (OED) for the input portfolio and Open Results Data (ORD) for the output. A flood model from a specialist shop, a windstorm model from a university lab and an in-house model can therefore produce comparable exceedance curves, because the financial calculation is identical. Underwriters mostly use it to cover perils or countries that the large commercial models leave out. The caveat comes quickly. A shared engine says nothing about the quality of the hazard or vulnerability plugged into it, and every model needs the same validation as a proprietary one. Many cedents and reinsurers run it as a second opinion without making it their reference view of risk.
Textbook case, made-up figures: a reinsurer tests two river flood models on the same German portfolio, loaded once in OED format. The first gives an average annual loss of 12 million euros and a 1-in-200 loss of 410 million; the second, 9 and 520 million. With the financial engine identical, the gap can only come from hazard or vulnerability. The underwriter knows where to dig: the frequency of mid-sized floods on one side, the tail of the distribution on the other.
Oasis LMF, Oasis platform, open-source cat modeling framework, Oasis engine
Follow the research
One letter a week. First, the week's underwriting note, when one is published: a dated, argued decision, signed by Paul Bernardski. Then what the observatory published on cyber risk, political risk and insurance. No account, no tracking.
No spam, one-click unsubscribe, no data ever sold.