This is an interesting discussion, but I find it a little odd to be complaining about the combined ratio when 2 things are true: 1) they have excess capacity and 2) they are clearly making a return that exceeds their cost of float.
Would you rather they stop writing policies that are clearly profitable for them? That may result in a lower CR (or not as fixed costs are spread across fewer policies) but would certainly result in less profit.
When under-utilizing capacity, they should write any policy with a positive expected value assuming very conservative investment returns. (And they certainly have the right to assume higher investment returns than Chubb!) Once they are closer to full capacity, they can start culling the least profitable policies. Until then, I'd prefer they write any policy they deem to be profitable not to focus solely on the CR. (In fact, I'd say when you're under-utilized, ignore sunk costs when considering any given policy. The marginal cost is all that really determines profitability of that policy when you know opportunity cost of not taking another policy is 0.)
I also see no discussion of how long the tail is on the policies. A weighted average 3-year tail policy at 105 CR sounds pretty darn good to me when the float is in the hands of Hamblin-Watsa (I'm guessing weighted average for Fairfax is a little over 3 years based on the table in the quarterly report). Chubb makes something like 4% on their float. Fairfax does a good deal better than that. The better the investment returns, the higher combined ratio can and SHOULD be tolerated (until full capacity), and this difference gets magnified the longer the tail.
Anyway, zero-cost or negative-cost float is a wonderful objective, but it's not a profit-maximizing rule. I'm not at all suggesting relaxed underwriting standards, but I would much rather judge them on how well they assess the risk than what the CR is. I.E. I'm more interested in how accurate their reserving is than what the bottom line CR is (and this has certainly been a problem in the past...). If their reserving is reasonably accurate one has to assume that the CR they're writing policies at has a sufficiently positive expected value. Or at least they've earned that right in my eyes. The problem comes when they have no idea what the expected value of the policies are...
And in fact, one question that comes to mind is why Chubb isn't writing more marginal policies. It doesn't look like they're anything close to full capacity. Doesn't it seem like they're leaving some profit on the table at a CR of 85? I don't know anything about Chubb, but in their case, it's likely that they're playing in specialty markets and the markets simply can't handle more policies. Obviously the less commoditized the insurance product, the more underwriting profit will be available... But anyway, the point is I wouldn't think an under-capacity CR of 85 is the ideal answer. If Fairfax started doing that, I'd be mad not happy.
A secondary point is that conservative reserving is good for the tax bill. I'd rather have as much potential profit as legally possible buried in the reserves. Over-reserving is a no-cost loan from the government, under-reserving is a no-cost loan to the government. I hope Fairfax is over-reserving!
And if the argument is that a good CR will make Fairfax look better to investors which will raise the stock price which will in turn let them buy companies with overvalued stock, who here is actually going to wait around for that outcome to play out? Once it's significantly overvalued in our eyes, aren't we all gonna sell? When you pay in stock, you pay in intrinsic value. If cash is worth more than the intrinsic value of the same nominal amount of stock, I'll invest my money elsewhere. (That isn't to say you cannot pay with stock, but it better be because the selling company didn't want the cash...)
Anyway, I like the discussion, and I'm no insurance expert, but I just think it's better to look a little deeper than the headline CR.