Warren Buffett has made such an impression on value investors and insurance investors, that they think that float is magic. ?Write insurance, gain float, invest cleverly against the float, and make tons of money.
Now, the insurance industry in general?has been?a great place to invest, but we need to think about float differently. ?Float is composed of two things: claim reserves and premium reserves.
- Claim reserves are the assets set aside to satisfy all claims that likely will be made as of the current date.
- Premium reserves are the assets set aside representing prepaid premiums that have not been earned yet.
Claim reserves can be long, short or in-between. ?Last night’s article dealt with long claim reserves — asbestos, environmental, etc. ?Those reserves can be invested in stocks, real estate, long bonds, etc. ?But most claim reserves are pretty short, like a year or so for most personal insurance auto & home claims — those typically get settled in a year.
The there are classes of insurance business that are in-between — workers comp, D&O, E&O, commercial liability, business continuation, etc. ?Investing the claim reserves should reflect the length of time it will take until ultimate payoff.
The premium reserves are very short. ?If premiums are paid annually, the average period for the premium reserves?is half a year. ?If premiums are paid more frequently, the average period for the float falls, but the premiums rise disproportionately to reflect the insurance company’s desire to have the full year’s premium on hand. ?It usually makes sense for policyholders to pay at the longest period allowed — thus, thinking about premium reserves as having a ?duration of half a year on average makes sense. ?Except auto — make that a quarter of a year.
Earnings financed by?float should be divided into two pieces — non-speculative, and speculative. ?The non-speculative returns on float reflect what can be earned by investing in high quality bonds that match the time period over which the float will exist. ?Short for premium reserves, longer for claim reserves. ?So, the value of float is this:
Present value of (investment earnings of high quality duration-matched assets plus underwriting gains [or minus losses]).
This is a squishy calculation, because we do not know:
- the number of years to calculate it over
- future underwriting gains or losses
The speculative earnings from float come from assuming that float will stay at the same levels or grow over many years, and so the insurer invests more aggressively, assuming that float will be a permanent or growing thing. ?He speculates by financing stocks or whole businesses using the float that could reduce, or that could become more expensive.
How could that happen? P&C insurance often gets very competitive, and the cost of maintaining float in a soft underwriting environment is considerable. ?Also note the claim reserves mean that the company took a loss. ?That the company earns something while waiting to pay the loss does not help much. ?Far better that there were fewer losses and less float.
Smart P&C insurance companies reduce underwriting in soft markets, and in such a time, float will shrink. ?Let aggressive companies undercharge for bad business, and let them choke on it, while we make a little less money.
Well-run insurers let float shrink – they don’t depend on float being the same, much less growing. ?If it does grow, great! ?But don’t invest assuming it will always be there or grow forever. ?That way lies madness.
Berkshire Hathaway has benefited from intelligent underwriting and intelligent investment over a long period. ?That is not normal for insurance companies. ?That is why it has done so well. ?Float is a?handmaiden to good results, but not worth the attention paid to it. ?After all, all?insurance companies have float, but none have done as well as Berkshire Hathaway. ?Better you should focus on underwriting earnings rather than float.
Underwriting insurance produces premium float. ?Underwriting bad business produces claim reserve float. ?Float is not an unmitigated good. ?Good underwriting is an unmitigated good. ?So focus on underwriting, and not float.
=-=-=-=-=-=-=-=-=-=-=-===-
Berkshire Hathaway has been in the fortunate position of having had wise underwriters, and and ability to expand into new markets for many years. ?Guess what, that was AIG up until 2003 or so. ?After that, they could not find more profitable markets to underwrite, and results began to deteriorate. ?They ran up against the limits of their ecosystem.
Buffett is a brighter man than Greenberg; he can consider a greater realm of possibilities in how to run an insurance conglomerate, and the results have been better. ?All that said, there is only so much insurance to underwrite in the world, and big insurers will eventually run out of places to write insurance profitably.
All that said — float is a sideshow. ?Focus on profitable underwriting — that is what drives the best insurers.
Thank you, this was very helpful. Would you happen to know of any data sources (ideally free) that compile the underwriting profitability for individual insurers?
I usually just go to the company’s financial reports at SEC EDGAR. Away from that, some of the better companies will do tables in their presentations comparing themselves against the industry.
How do you calculate the float of life insurers and how does it differ from P&C companies/
Life insurers don’t have float, or at least, almost no one thinks of it that way. The policies they write have embedded interest rates that the life insurer would have to meet in order to make money on the policy, along with enough money to cover the use of capital. Those rates are high enough, that together with risk-based capital rules, few life insurers invest significant amounts in equities and other risk assets — they make their money off the yield on their assets vs the implied yield on their liabilities as their investment gain, and mortality charges less actual mortality costs for their underwriting gain.
Note: in eras where interest rates were high, P&C insurers built expected investment gains into their underwriting, and the cost of float was high across the industry as many companies wrote insurance policies to make investments. As such, underwriting suffered, and overall insurance profitability was bad — only the few companies that held to strict underwriting discipline did well in that era — BRK and AIG would have been good examples then.
I’m sure my understanding of float is too simple, because it seems like all you are doing is investing the money you will have to pay out on claims later. Isn’t this potentially risky? I’m sure there are rules and regulations I am missing, otherwise, it would seem a lot more insurance companys could invest in stocks and potentially not have enough money to cover their claims costs later, if those investments drop dramatically.
This is why traditionally P&C insurers invest their premium reserves in money market instruments, claim reserves in a match portfolio of bonds, and only invest in risky assets to the degree that they have surplus.
Re: As such, underwriting suffered, and overall insurance profitability was bad.
I was wondering if the opposite is occurring now, i.e., in your judgment are historically low interest rates leading to better than historical underwriting quality? I notice there are many insurers trading around 10x earnings multiples. This coupled with high underwriting quality would make insurers a particularly attractive investment I’d think.
The P&C underwriting cycle has been pretty benign for the last 10-12 years, partly because disasters have been big enough to suck away excess capital. That said, capital seems be building up at present. Will everyone be good and do special dividends or buy back stock? If not, pricing seems to be weakening.
The low interest rates do force the insurers to be more disciplined about underwriting, because it is the only way to earn decent money now.