limit they buy and the risk management and so forth. That may be what comes out of this.
McDONALD: Thank you, Laurie. We’re going to be coming back to this topic. Roger, welcome. You have a fascinating background. One of the reasons that you’re on this panel is you take a step back and take a larger look. Obviously there’s a threat of hurricanes. There’s been a lot of activity among hurricane prognosticators to determine: will there be more or fewer hurricanes and will they land? When you look at hurricanes and you look at the possibility of major catastrophic hurricane events in the United States what do you see?
PIELKE: One of the most important things to recognize when you’re looking at the big picture, long term, is there are a lot of moving parts. You have changes in the climate system. You have changes in society. It’s very easy to confuse those if you look at the damage record. One of the things we focused on in our research is trying to disentangle those different signals. We can talk about climate signals. We can talk about the effects of building and development.
I’ve prepared four quick slides that I’ll talk about. The first slide shows the increasing aggregate losses as kept by the National Weather Service. These are total losses, not insured losses and it’s adjusted for inflation. What you see is this steady stair step increase in losses. If you want to know why that’s occurring you just can’t look at the loss data. You have to disentangle it.
The second slide shows one measure of the intensity of storms making landfall. One thing that many people aren’t aware of is, over the long term, 80, 90, 100 years, there are no trends in either the frequency of hurricanes that make landfall or in their intensity that make landfall. What that says is that entire signal of increasing losses has to be due to something other than changes in storm behavior.
The third slide is a picture of Miami Beach, two pictures in fact. One is from 1926. The other is from 2006. The picture from 1926 shows a lonely solitary hotel. The 2006 picture shows thousands of buildings. One of the points I emphasize is that change in development in vulnerability and exposure on a year to year to basis makes a huge impact on the loss potential.
The final slide shows the results of our analysis of what losses would be if every hurricane season occurred with 2005 exposures. We’ve prepared a little calculator on the web that shows this updated to 2009. You get a very different picture. Just some summary statistics from that, we think that the 1926 Great Miami storm is potentially two or more times that of Katrina. Katrina is not the worst-case scenario.
Looking forward, it’s not unreasonable to expect by 2025 we might be talking about $500 billion hurricane losses.
McDONALD: You shy away from short-term forecasts, I gather. Why is that?
PIELKE: If you look at the track record of forecasts, of hurricane behavior inside of a season there is some forecasting skill potential such as related to El Nino. When the Pacific is cool there’s more landfall at higher intensity in the United States. This is well known. But this is after the hurricane season begins in June that you get some sense of what might be coming.
If you look at the longer time period, say one to five years, there’s no evidence that the scientific community has any skill predicting. One of the points I raise with folks is that you’re probably not going to beat Mother Nature. You can probably beat your competitors, have better portfolios than others. But if you’re trying to guess how many storms are going to make landfall next year or the year after you probably should recognize that you are guessing.
McDONALD: So if you were responsible for insuring, if you’re a risk manager, if you’re a broker, if you’re providing someone with insurance or watching out for the welfare of your own organization and you wanted to determine your likelihood of having a hurricane, what do you do?
PIELKE: I would start with the long-term historical record. There’s really no scientific evidence to suggest that we should be deviating from that in our judgments of risk. There are plenty of business reasons why you would want to deviate from that. You may want to hedge in one direction or another, take a gamble. But you want to know when you’re relying on science and when you’re not. My advice to people would be to use that long-term historical record as a starting point and then make your judgments on altering that with open eyes.
McDONALD: I guess a shortcoming of insurance and reinsurance is they tend to think in 12-month cycles. You tend to think in centuries of cycles, if they’re even cycles, who knows. There’s a mismatch there, right?
PIELKE: Yes.
McDONALD: So basically you want to err on the high side?
PIELKE: Well it depends on what your goals are. You could judge to go out on a limb, take on more risk or not. But you want to know when you’re doing that. One of my concerns is that a misunderstanding of the loss potentials could lead people to misjudge their risks. I would think over enough trials if you misjudge your risk you’re going to make some mistakes in your expectations.
McDONALD: When you talk about hurricanes in the United States, the conversation always starts with Florida. You’ve got the whole coast there. Looking back the past five, 10 years, you saw some extreme years, you saw the last couple of years where there were hardly any events. It’s kind of peppered there. Par for the course?
PIELKE: You have to be real careful using any short time