This paper proposes a new design and pricing method for catastrophe (cat) bonds that aims to be both transparent and closer to actual losses. Cat bonds are securities that let insurers transfer extreme disaster risk to investors. The authors introduce a ‘‘blended’’ trigger that sits between simple parametric triggers (based on a few physical measurements) and full modeled-loss triggers (which use complex, closed-source vendor models). The trigger is based on a cost random field that cleanly separates three parts: the physical hazard, a vulnerability function (how exposed things are to damage), and the exposure (what is at risk).

To test the idea, the researchers built a case study for historical windstorms in Germany. They used wind speed data from past storms and fitted a max-stable random field, a statistical model commonly used for spatial extreme events, at the resolution commonly used in the reinsurance industry. Where industry insured-loss and exposure data were available, they used those observations to calibrate the vulnerability piece of the model. This makes the trigger flexible: it can act more like a traditional parametric index when exposure is not included, or move toward a modeled-loss style trigger when exposure data are incorporated.

On the pricing side, the paper gives a fully transparent formula inside a contingent-claims framework. In plain terms, that means both sponsor and investor can independently compute the bond price from the same loss model and data, rather than relying on black-box vendor models. For valuation the authors follow an established approach that treats financial markets and insurance technical variables as independent and they use the Wang transform to move to a risk-adjusted, or ‘‘risk-neutral,’’ pricing measure. They also treat the market spread as an input to the pricing exercise.