No. of Recommendations: 4
Grove 5: Net premiums were $88.257 in 2024, average underwriting margin in the last 20 years has been 3.46%[3], so I'm saying $3.058b in normalized underwriting earnings.
FWIW
That's a bit higher than the figure I use, but doesn't seem unreasonable. I like to lean to the conservative side, since the occasional gigantic loss is not an exception, but something that happens in the normal course of business. I would rather be pleasantly surprised rather than unpleasantly.
I've found two quick estimates of pre-tax underwriting profit that seem to have held up well, in that aggregate P/L over the years has tracked aggregate estimated profit pretty well in the same period, as well as percentages. The two are:
-1.3% of float
+2.7% of premiums earned
The two give fairly similar results, and I use a simple average of those two methods. Four quarters to 2025-Q3 came out at $2.327bn before tax.
I added another five years to include a gigantic loss that happened in 2001. From 2000 to 2024, the average underwriting gain was 2.30% of net premiums earned, close to your figure. I think this is TOO conservative, because it includes 3 years with losses (not just 2001, but also 2000 and 2002), just after acquiring GenRe and its book. Since then, reinsurance has gone from 40% of Berkshire's revenues to 7% of its revenues. So even if we got another terrible year like 2001, it should have about five times less impact on Berkshire than it did in 2001.
BTW, the other thing I should have mentioned is that, when counting the value of the equity portfolio (grove 2), I made no adjustment for the unrealized capital gains, which were $200b as of September 30 (10-Q, p.9), so there is about $42b in future tax liability. It is hard to know when that will come due, so to split the apple in half, one might dock $21b from the company's value and consider that the other half is float. That would take the P/E on groves 1,3 and 5 from 15.5 to 16.2, not completely insignificant.
