Yeah. The simplist model for BRK would be leverage from float and DTA and a modest amount of debt used to buy high quality businesses with predictable cash flows.
It used to puzzle me how WEB is attracted to the purchase of a great business that had been around 100 years with modest growth prospects at a not so wonderful price that might have an earnings yield of 7% or so. However, with BRK's internal leverage, that modest earnings yield then increases to 10% or so. Then, BRK earns a compound rate of maybe 4% per annum above what the great businesses in its stable earn. Then, the large cash holdings and regular generation of float enable BRK to occasionally pick up unusual bargains like BAC pref + warrants. Result: now the compounding machine's normalized earnings are perhaps compounding at 11% to 12% per annum, much more than the earnings of the other great businesses it owns.
Think about it. Here is this mega cap company, selling for less than 1.2 times Q3 BV that is still a machine that compounds earnings and BV at a much higher rate than some of the very best S&P 500 companies that sell at perhaps more than 4 times BV. We're talking about a company that has compounded BV/SH at the highest rate of all for the last 47 years!
Plus, BRK does another neat thing. A few years ago I puzzled over BRK's SEC filings that Warren himself had apparently personally signed. These listed at the time $8B or so "best of best" holdings like AmEx, Wells Fargo, Coke etc. I think these were the majority of BRK's pension plan holdings, things that appropriately would have a growing earnings yield of 7 % or so with enormous compounding potential by the time the pensions had to be paid out. Contrast that to the bond heavy assets of the typical pension plan that will probably earn about half that rate. 
twacowfca,
great analysis of BRK! AmEx, Wells Fargo, Coke… a 7% earnings yield, add the benefit of float and you get a 10% earnings yield, pick up occasionally some unusual bargain and you compound at 11% to 12%… It looks easy!
So why isn’t it copied more often? You know that I am thinking hard about LRE: why an outstanding (unique, I daresay!) underwriter like LRE keeps the large part of its investments in short-term bonds, instead of just copying what Mr. Buffett has shown works so well? I don’t understand: Mr. Brindle could very well go on concentrating exclusively on the underwriting business - as he should do, because that is clearly what he does best -… but why don’t hire YOU??!! 
If only with the goal to replicate what you have written about BRK’s way of investing! It doesn’t take Mr. Buffett to do that! If the reason is: it works for BRK, because of Mr. Buffett’s skills; well, then I don’t agree. I just don’t see how Mr. Buffett’s unique investing skills should be required, to achieve good results the way you have so clearly described.
giofranchi
Giofranchi, you ask a great question: Why doesn't Brindle, who is arguably the best property underwriter in the Bermuda / Lloyds markets invest his float in great businesses the way Buffett does? The short answer is that he has had a top tier record investing in equities in the past, but what he does now is much better for LRE's business.
BRK had a AAA rating until Warren's put derivatives went against him in 2008 and led to a downgrade as the market value of BRK's stocks also declined. However, this had little downside because BRK was still one of the most solid companies in the world. LRE liquidated their equity portfolio in mid 2008 before the stock market tanked to avoid such a possible consequence, because the prospect of a downgrade from A - to below investment grade would have been disagreeable to say the least. LRE was one of the few insurers to have positive investment returns in 2008, and in every quarter of their existence except one.
Lancashire is a short tail property insurer. This necessitates carrying mostly high quality, short duration liquid assets to be able to pay claims quickly if there should be a large claim. This limits investment returns, but it also gives them the huge optionality that goes with carrying a lot of cash and near cash. For example, they were able to buy back about 25% of their stock a few years ago at a price that averaged less than book value. That's the gift that keeps on giving. Recently, their extra cash has enabled them to leverage their sterling reputation to goose returns through sidecars.
Here's how the economics of a sidecar works for them: their long term return on equity at LRE has averaged 19%+. Brindle's returns for the syndicates that he and Charmin managed at Lloyds also returned 19%+ on average. All of these returns have been without a single down year, quite low volatility considering that their peers had a very rough time at Lloyd's in the 80's and 90's. Therefore, it's reasonable to assume that the central value of LRE's returns will continue to be about 19%+ per annum or perhaps a little better with the way they get more upside with less downside in the sidecars.
LRE puts a relatively small amount of capital into a sidecar and other investors fund the rest. LRE gets a commission and a management fee for managing the sidecar. This attenuates the downside to their investment in the sidecar if there is a loss. They also cede some of their most catastrophe exposed business to the sidecar. This also attenuated the downside from a large catastrophe to their regular business. Then, if the loss experience of the sidecar is low, LRE will be entitled to a percentage of the profit of the entire sidecar above the profit of their investment in it.
Lets assume that the long term expectation of LRE's investment in the risk assumed by the sidecar sidecar would be greater than their long term average return of 19%+ because sidecars are are only set up when rates in a particular sector are exceptionally high. However, Lancashire limits how much catastrophe exposure it is willing to take on even though their expectation is that the return would be great if they took on more risk.
Let's assume that the long term average expectation of the annual profit of the sidecar is about 20% for the outside investors. LRE's expected return on the sidecar may be perhaps 30%+ because of their overrides. Plus the risk is far less than if they had retained a higher level of catastrophe exposed business.
That return on extra cash not needed in their core business is a lot more than what Lancashire would expect to get from an investment in common stocks. However BRK's business is much more long tail. There is a long time available to invest the float until claims have to be paid. That's why having substantial investments in common stocks makes sense for BRK much more than for Lancashire.
