Thread by @RogerPielkeJr

🧵
@MunichRe today published their estimates for natural disaster losses in 2021, allowing me to update the time series of global disaster losses as a proportion of global GDP

This thread reports and discusses this update

TL;DR --> Figure below

Munich Re reports ~$270B in total weather and climate disasters losses for 2021, which is 0.28% global GDP

Previous 4 years:

2020= $194B, 0.23%
2019= $151B, 0.17%
2018= $150B, 0.18%
2017= $329B, 0.41%

1990-2021
Median= 0.21%
Average= 0.23%

Munich Re reports US saw ~$145B of the total global losses (~54% of weather/climate losses)

According to FEMA, in 2021 the US should have expected $141B in weather/climate disaster losses

In principle, US & global losses are in line with expectations

https://twitter.com/RogerPielkeJr/status/1472247466397356034?s=20

And here is an update of North American losses as a pct of GDP based on an estimate of 2021 from Munich Re

Note: 2020 and 2021 losses were about the same in dollars, but decreased as a pct GDP because GDP increased 2020 to 2021

Methodological note:
Reporting of insured losses are largely consistent across companies, for regulatory reasons

Reporting of total losses are not consistent
eg.
Swiss Re reports $233B in total wx/cx losses for 2021 & Munich Re reports $270B
These numbers are much squishier

Elephant in the room:
Reinsurance companies have incentives to over-report total losses (i.e., not insured) because that supports their business aims -- the different between insured and total is called the "insurance gap" & explains why the world needs to buy more insurance

I discussed some of the methodological shenanigans in disaster loss accounting here: forbes.com/sites/rogerpie…

In general @MunichRe @SwissRe have done a good job in reporting disaster losses over decades, but it is probably time for independent evaluation of their proprietary data

Here is a time series of insured vs total losses from @SwissRe suggesting that the "insurance gap" has been slowly closing