Ajit Jain Will Not Insure Elon Musk, his Mars Mission, Cyber Attacks, Nuclear War | Warren Buffett

Okay, let me start. Cyber insurance has become a very fashionable product these days over these last few years. It is at least a $10 billion market right now globally, and profitability is also being fairly high. I think profitability is at least 20% of the total premium has ended up as profit in the pockets of the insurance bearers. Now having said that, we at Berkshire tend to be very, very careful when it comes to taking on cyber insurance liabilities, for two reasons. One is it's very difficult to know what is the quantum of losses that can be subject to a single occurrence and the aggregation potential of cyber losses, especially if some cloud operation comes to a standstill. You know, that aggregation potential can be huge, and not being able to have a worst case cap on it is what scares us. Secondly, it's also very difficult to have some sense of what we call loss cost, the cost of goods sold could potentially be. It's not just for a single loss but for losses across over time. They have been fairly well contained. Out of 100 cents of the dollar premium, losses over the last four or five years I think have not been beyond 40 cents on the dollar, leaving a decent profit margin. But having said that, there's not enough data to be able to hang your hat on and say what your true loss cost is. So in our insurance operations, I have told the people running the operations, I've discouraged them from writing cyber insurance. To the extent they need to write it so as to satisfy certain client needs, I have told them, no matter how much you charge, you should tell yourself that each time you write a cyber insurance policy, you're losing money. We can argue about how much money you're losing, but the mindset should be you're not making money on it, you're losing money, and then we should go from there. So it is good, it's projected to be a huge business. My guess is at some point it might become a huge business, but it might be associated with huge losses, and our approach is to sort of stay away from it right now until we can have access to some meaningful data and hang a hat on data.

In addition to all what Warren has said in terms of the chance of something like this happening, the additional thing that concerns me about a nuclear situation is my lack of ability to really estimate what my real exposure is in the event of a nuclear event. When you're talking about other big exposures we have, earthquake and hurricane and cyber, I can with some reasonable degree of accuracy have a point of view in terms of how large exposures can be and how big our loss can be. When it comes to a nuclear thing, I sort of surrender. It's very difficult for us to estimate how bad bad can be. Very many different lines of exposures will be affected by it. And even though in almost all our contracts we try and exclude nuclear as a covered peril, nevertheless if something like that were to happen, I'm fairly positive that the regulators and the courts will hold it against the insurers and we will be, and they'll rewrite the contract and we'll be required to pay. For example, one thing which is already being talked about: we issue what are called fire policies, and these fire policies try and exclude nuclear as a covered peril, but there are several regulators who feel that if it's a fire policy and if the nuclear attack causes a fire, then how can you exclude fire? And you better include fire. So debates like that we will have to live with, and it'll be very difficult for the insurance industry to fight back both the regulators and the court systems in terms of what is covered and what is not covered. And there won't be any regulators or anybody else. So we'll leave it to a million years of reconstruction. Einstein said that he said, 'I know not what the weapons will be for World War III, but I know the weapons for World War IV will be sticks and stones.'