beachlawyer2003, I'm not a fan of float based models in general, and in today's world of low interest rates (you can thank bubbles Bennie for that) float has even less value than it normally would.
WEB estimated float at 71.1 billion, as of 6/30/2012, . Float, a liability, has to be balanced by assets garnered exclusively from the insurance section of the balance sheet. I see: 36.8 billion in cash and cash equivalents earning maybe 0.17%, 30.5 billion in fixed maturity securities (intermediate term notes) earning maybe 1.65%, and that leaves 3.8 billion of float remaining that can be applied to the equity securities which may, at best, earn 3% in dividend interest. So we have 71.1 billion dollars of Other Peoples' Money with a weighed average yield of maybe 0.96 percent, earning maybe 0.68 billion in cash per year. To put a value on that future stream of income, I'll capitalize it by dividing by 4%, my growth adjusted discount rate, that values the float at about 17 billion, or a little less than seven dollars per B share. And that generously assumes that BRK's insurance operations will always run in the black, that they will always generate zero cost float. Let's assume that's only true 70 percent of the time, about seven years out of every ten. The value of float now drops to slightly less than five dollars per B share. Not insignificant, but not a big deal either on a presumed intrinsic value of say, 87. And the proceeding financial fantasy is just that, nothing more than a fantasy, if shareholders never receive even so much as a penny of that hypothetical value in the form of a cash distribution from the company at some point in the future.
Rational Walk has a very nice PDF of a 10 year spreadsheet detailing BRK's cost of float. You can find it by googling the sentence below, verbatim.
Berkshire Hathaway Historical Cost of Float Summary - 1999 to 2008
I recalculated the ten year average from their spreadsheet and I got a slightly better value than they reported. I get an average cost of float of -0.33%. A negative cost is good, it means BRK got paid. If you capitalized that ten year average by dividing by 4% (my growth adjusted discount rate) then a dollar of float is really only worth about eight cents, not the dollar that the KoolAid Klowns routinely assume in their page four - two column valuation. In my opinion, that's probably the single greatest reason why BRK's market price is almost always considerably lower than what they want to believe it is worth. If you want to build a margin of safety into your valuation guesstimate, simply ignoring float completely would be a good way to do it.
Giving float no value is just as silly as valuing it dollar for dollar as equity. Clearly, its value lies somewhere in between. In low interest rate environments, it is worth less. But given BRK's underwriting prowess (the negative cost of its float), float cleary has substantial value. If someone gave you $70 billion to use for an undetermined period of time which could last decades or longer, wouldn't you be better off than not? How much better off? Think you could earn more than than 8% before having to pay it back?
Jad, that is precisely my point. But you recognize that rates will not stay artificially (and that is exactly what this low rate environment is) low forever.
More to your point, I'm not a fan of ANY "model". Models are for people who want simple answers. With models, you get what you pay for. Float is one of many elements of the company's value. In a normal interest rate environment, $70 billion of OPM is a very significant element of value.
I failed to mention in my previous post that I did not correct the guesstimated interest and dividend income generated by the float for taxes due. Somebody has to pay those taxes, either the corporation, or the shareholder, or both. So the after tax value of float is even less than I implied.