#6 1967 · Berkshire Hathaway · Insurance
Buffett bought insurers for the money that sits between premium and claim
the problem
Premiums lay idle until claims came due
background
Berkshire Hathaway in the mid-1960s was a dying New England textile mill; Buffett had taken it over as a value bet, but manufacturing capital was structurally trapped — every dollar reinvested in looms competed against cheaper Southern and overseas mills for a shrinking margin. The obvious moves were to run the mills harder or liquidate; neither turned the company into a compounding machine.
In 1967 Buffett instead used Berkshire's cash to buy National Indemnity, a small Omaha insurer, for $8.6 million. Insurers already sat on float — premiums collected before claims are paid — but most of the industry invested it conservatively in bonds and treated underwriting, not the float itself, as the business. Buffett saw National Indemnity's investment portfolio as something he could redirect toward better returns, and ran the float aggressively rather than conservatively.
what everyone would do
Insurers already sit on float -- premiums collected before claims come due -- but the standard playbook treats it conservatively: park it in short-duration bonds to guarantee the cash is there when a claim lands, and make the real profit on the underwriting spread, not on what happens to the money while it waits.
what they saw
Buffett saw float as an interest-free loan, not the insurer's money at all: as long as the underwriting roughly breaks even, the cash sitting between premium and claim is capital an owner can invest anywhere, for as long as claims stay unpaid, without paying for the privilege.
the move
Buffett bought insurers explicitly for float — premiums held before claims — describing it as money held but not owned, investable in the meantime.
why it works
Because premiums arrive before the matching claims are owed, an insurer that underwrites near breakeven is effectively borrowing billions every year at zero or negative cost; redirecting that pool into equities and whole businesses instead of conservative bonds means each dollar of float compounds for as long as the policy stays open, and growing the underwriting book grows the float pool itself, so the investable base and the return earned on it expand together rather than trading off.
the payoff
Float grew from tens of millions to well over $100B and financed much of Berkshire's compounding.
where it breaks
It depends entirely on underwriting discipline -- write policies that lose money on claims and float stops being free capital and becomes an expensive, involuntary liability, which is what sank insurers that chased premium growth without matching pricing rigor. It also needs an owner with permanent, patient capital willing to hold volatile investments against money nominally earmarked for claims, a risk tolerance most regulated or short-horizon insurance managers neither have nor are allowed to take.
what came after
Berkshire's insurance float grew from $39 million in 1970 to $114.5 billion by the end of 2017, and in his 2004 shareholder letter Buffett wrote that "had we not made this acquisition, Berkshire would be lucky to be worth half of what it is today." Fairfax Financial, Markel and Exor later built similar insurer-as-investment-vehicle structures.
references
- [1]Warren Buffett and the Insurance Business: A 52-Year Love StoryYahoo Finance, 2019finance.yahoo.com
- [2]Berkshire Hathaway Chairman Warren Buffett on Insurance Economics and 2004 ResultsInsurance Journal, 2005insurancejournal.com