Here’s a letter from a reader on insurance topics:
Hi David. I’ve been following your blog. Just want to say thank you for willing to share your knowledge in the public domain.
I have a question for you – as you know, “climate change” is happening… whether human caused or not, it certainly feels like we are seeing more extreme weathers of late.
How do you see this affecting P&C insurers? Does this give them the chance to start rising prices? ?
Lastly, just wondering if you have an opinion about Markel and Lancashire and Allied world. I owned allied for a long time. Made some gains. But the recent blow up at tower and short attack at Am Trust prompted me to really stick with firms that have a much longer record. Which lead me to Markel and Lancashire. Not that this verifies these guys are clean. I’m not an accountant and nor do I think accountants can catch anything. Nonetheless, their long term record offers me a better sense of security in my mind.?
First, I *don’t* know that climate change is happening, except that it always happens.? Evidence for climate science is weak, like that for economics.? We don’t have a good model yet.? If we had a good model, we would have better predictions on hurricanes, which have been uniformly lousy for the last ten years.? And as for warm climates, the Earth has been warmer than now in the past, and far colder, if the history books are correct.
As to how it affects P&C insurers and reinsurers, for that we do have a simple and reliable model.? Look at industry surplus relative to the past — when it is high, as it is now, premium rates will be lower than the risk demands.? Most P&C pricing is weak now — I have been decreasing exposure to P&C insurers.
Markel and Allied World I know and respect.? Good companies both, though I own neither of them.? I’ve heard of Lancashire, but I do not know them in any detail.? To analyze, look in my On Insurance Investing series.
Thanks for writing.
“If we had a good model, we would have better predictions on hurricanes, which have been uniformly lousy for the last ten years.”
This is really the same as saying, that we can’t predict that a series of 1000 coin tosses will be pretty close to 50:50 because predictions of a single coin toss are uniformly lousy. Or that an investor who’s no good at predicting one day price moves can’t outperform in the long term.
Modelling a single event is considerably harder than doing it for the dispersion of events and long term trends, where we have some success. The concept should be familiar to an actuary!
It’s unfortunate that many people are blinded by their politics when it comes to this matter: studies have shown that people’s view on small vs. big government correlate very strongly with their view on climate change, which is a bit like blaming the oncologist for someone’s cancer.
A good read on the science (without policy prescriptions!):
http://royalsociety.org/policy/projects/climate-evidence-causes/