Comments on: Robert Muir-Wood in RMS Cat Models: From the Comments

By: Markk

Markk — Thu, 11 Jan 2007 15:38:00 +0000

This whole issue seems odd. (or maybe I should say – what is the issue here?) From an insurance or re-insurance point of view all I would be looking for is the estimate of hurricane activity that I am most willing to bet on to plug into my damage models. This would be for the next year or two because that is what the contracts are for. I could look out a few more years if I think premiums are going to jump fast to load them, so people are only normally complaining not so they go through the roof.

If there is reasonable competition in the market, then there will be constraint on pricing so that it should – vaguely – reflect best estimates. Is RMS that powerful a player that there is little option? Otherwise other Re’s should be able to undercut and get market share.

Climate change may or may not be a major part of the risk, but I don’t think these estimates were done off climate models. I never heard of one predicting #’s of storms. So If there is contoversy about this estimate it isn’t about climate change models correct?

By: Hervé Grenier

Hervé Grenier — Wed, 10 Jan 2007 14:28:10 +0000

There is no link between total number of hurricanes in the North-Atlantic and number of US landfalls ? Humm, let’s see.
Over the period 1970-2006, the linear correlation between the 2 time series is 0.55, p-value ~1.e-4, it is 0.50 over the period 1950-2006, p-value ~1.e-5. So about 25% of the variance in the number of US landfalls is explained by total number of hurricanes in the basin and, relating both seems relevant. Using other measures (rank correlation) also shows a significant relationship.
Details on how to relate them is up to anyone who tries to build a hurricane risk model for the US, but I don’t see on which grounds such a statistical link should be neglected in assessing today’s US hurricane risk.

By: Rich Horton

Rich Horton — Tue, 09 Jan 2007 23:00:42 +0000

I’ll just add that I wasn’t basing my numbers on any articles, just NOAA’s collection of historical data.

Additionally, since the topic at hand deals with insuring against US landfalling hurricanes, I don’t see why we shouldn’t look at…well, US landfalling hurricanes.

I’ll also note that picking any smaller subset of 12 years out of the data and making larger claims seems of dubious value. I noted the large number of storms in the period 1944-1955, was followed by a relative dearth of activity. I will also point out if you look at the 12 year period from 1992-2003 there were only 16 storms hitting the US (1.33/year). There is nothing in the data you could use to infer there would be 12 storms hitting the US over the next 2 years.

By: Roger Pielke, Jr.

Roger Pielke, Jr. — Tue, 09 Jan 2007 21:36:34 +0000

Dan- Thanks for commenting. On the relationship of North Atlantic basin activity and landfalls in the NOAA historical data, see this short essay:

http://sciencepolicy.colorado.edu/prometheus/archives/author_pielke_jr_r/001035draft_paper_for_comm.html

Thanks!

By: Dan

Dan — Tue, 09 Jan 2007 20:52:02 +0000

Rich:

Significantly higher hurricane activity and US landfalling hurricanes are two different animals. The proportion of total storms:US landfalling storms is not 1:1. If you are reading articles that use this type of argument to convince their readership, you may want to check their sources and read more critically.

By: Roger Pielke, Jr.

Roger Pielke, Jr. — Tue, 09 Jan 2007 20:24:27 +0000

Coby-

Thanks . . .

Clearly hurricane activity is up when we look at 1995-2006 as compared to 1970-1994. Some scientists have argued that this is a direct consequence of greenhouse gas emissions.

However, you are indeed correct. I cannot imagine why the cause of this increase would be relevant to politicians or journalists, especially since the reality is that the jury is out among the scientific community.

The is particularly the case for the reinsurance industry, in which contracts are signed for 1-year terms.

Thanks!

By: Rich Horton

Rich Horton — Tue, 09 Jan 2007 19:16:20 +0000

Robert Muir-Woods states: “Cat modelers are simply the messengers relaying news concerning the significance of a period of significantly higher hurricane activity that has persisted in 9 out of the last 12 years and that climatologists, as polled at the most recent expert elicitation, expect to continue for a decade or more longer.”

LEt’s look on the NOAA data on the storms hitting the US since 1994.

2005: 6 (4 Cat3, 2 Cat1)
2004: 6 (1 Cat4, 2 Cat3, 1 Cat2, 1 Cat1)
2003: 2 (1 Cat2, 1 Cat1)
2002: 1 (1 Cat1)
2001: 0
2000: 0
1999: 3 (1 Cat3, 1 Cat2, 1Cat1)
1998: 3 (2 Cat2, 1 Cat1)
1997: 1 (1 Cat1)
1996: 2 (1 Cat3, 1 Cat2)
1995: 2 (1 Cat3, 1 Cat2)
1994: 0

26 storms in a 12 year period (2.1667 per year). 8 of those 12 years had 2 or fewer storms hit. Since historically we have averaged around 1.75 storms a year, I hardly think 2 storms a year is “significantly higher.”

Other 12 year periods that had above average numbers of storms includes:

1932-1944, 24 storms
1942-1953, 26 storms
1944-1955, 29 storms

It is interesting to note that in the five years following the 1944-1955 12 year period (29 storms, 2.41/year), only 7 storms hit (1.4/year).

By: coby

coby — Tue, 09 Jan 2007 15:55:30 +0000

Roger, I wouldn’t presume 100% agreement from you even though you highlighted this comment, so do you agree with the concluding sentence?

“There is a need to get journalists and politicians in Florida to focus more attention on the reasons for the increase in hurricane activity and, in particular, the role of climate change.”

Presumably you would feel it is completely irrelevant why hurricane activity is up, and I think your writing idicates even the fact it is up is irrelevant from a policy perspective.