The Odds of Everything Are 100%

Some of you have seen the news about unusual weather in New England. Tonight we have the heat back on in part of the house, and hot water for the first time since Sunday. David and I spent a couple of days racing up and down to the basement managing a variety of permanent and temporary sump pumps. We have had a small flood and a small fire and only a little damage and are really lucky compared to the guy in Massachusetts whose car fell into a sinkhole and the poor folks in Rhode Island whose houses and cars are underwater in conditions that are just unspeakable. It has been a hell of a March. I don't really mind eating cold food and taking cold showers... well, I do mind, I mind a lot, but if forced to choose, would choose the Internet. At least right now we we have that.

This storm, we are told, was either a one in 100 or a one in 500 year event. It follows two weeks on the heels of a storm that was also a one in 100 year event. It's been sixty or eighty years since we had weather like this in New England, depending on which storm you are talking about. What are the odds that we would have two of these one in 100 buggers within two weeks?

Actually, the odds of anything going wrong are 100% if you wait long enough. Try it for yourself. Think of a low probability event like a hurricane, or earthquake - one in a hundred odds - and mentally put millions of endlessly replenishable balls in a jar, 99% white balls and 1% of black balls. If you start drawing balls out one after another and continue long enough, the odds are 100% that sooner or later you will draw two black balls in a row.

Would Warren Buffett write flood insurance for two events like this in a single month? Yes, for a price that was at least double the estimated odds (and preferably triple or more). He would also limit how much capital he risked to an amount that wouldn't cripple Berkshire Hathaway.

This is how to think about risk -- don't bet the ranch if you can't afford to lose it. 

You, of course, aren't going to pay Warren Buffett 3x the risk price for flood insurance. So what's your alternative? Either buy national flood insurance, or live on a hill. (Take it from a former insurance analyst.)