It’s pretty simple, those probabilities are based on historical storm instances, but as global sea temperatures have risen, it makes the likelihood of major storms higher.
According to NOAA July was the hottest July on record. The mid-Atlantic where the US south east storms gain force was also significantly above average.
So any pricing model based on historical datasets might underestimate the future level and frequency of reinsured losses from major US hurricanes.
For UCITS funds that have a 25-40% risk concentration exposure to US windstorms; this raises additional questions over risk concentration, risk/ reward disclosures to investors and liquidity/ pricing of impaired bonds in the Sept-October period.