I.e. Develop a business that generates large cash flows, then relentlessly plow that money into more profitable side ventures. Except in this case they didn't buy an insurance company and instead developed a product that generated those flows.
He takes low risk, profitable companies with proven track records and strong management teams. Geico, Fruit of the Loom, building products and materials etc. Its a very different mindset then what Google is doing when it invests in self driving cars.
Every company thats profitable re-invests those profits in some way. Trying to lump the way Google or Amazon uses their excess cash is different then the way Facebook does (Facebook builds moats mostly) and definitely different then the way Berkshire does.
(1) Core technology & monetization -> (2) Cash Flow -> (3) M&A to Fuel (1) -> (4) Goto 2
Looks different at that size than it does for smaller entities. And if you have a revenue stream generic enough to feed on a variety of inputs, then you have huge flexibility in what you can purchase. And you can likely afford to pay more because you've got efficiencies of scale on your side (including access to hyper-scale amounts of data).
The core of BH has ~20 employees. Warren Buffet always calls out his subsidiary CEOs for doing a great job, because BH couldn't exist in its current format if he had to manage day-to-day operations in any subsidiary.
The steps in Buffet's algorithm are the same, but the emphasis is different. Amazon builds moats and cashflow, while BH buys stable companies that already produce cashflow. BH doesn't need moats, only profit. Amazon, Google, FB need both moats and profit.
Buffet's algorithm looks more like:
- (1) Identify stable, profitable businesses that WB understands
- (2) if they're a good bargain, and it moves the needle on BH stock, buy
- (3) goto (1)Amazon, Apple and Microsoft have managed to build some new profitable businesses over time with Amazon having done the greatest leap. The advertisement giants seem to struggle to gain traction - lots of projects but revenue streams less impressive.
As such you can invest that money in assets able to throw off 96M in cash, like mattress sales or stock dividends, even if the return on purchase cost of the stocks is 2 or 3%. If anyone was to borrow 100M to invest they would not see a nice 6-7% return, but for most of past 40 years would see a negative return.
The cost of capital for Google will be different - most of their ad revenue is upfront but is short timescale (say a week) so it's not annual like BH. At best their cost of capital is probably zero. So their range of investments is smaller compared to BH.
And let's be fair - the brand is different. Google fiber is one thing. Google Coke? Google Mattresses?
Google is almost DNA bound to swing for the fences, and one suspects their will be no investor festivals for Serge and Larry in their 80's because the returns on that strategy don't often work.
But betting against Google is also a bad strategy :-)
On the other hand, Waze, Orkut, Nest. They spent billions on those and all they got was a lousy t-shirt.
This is the difference between mobile/tablet performance looking unprofitable vs. profitable because the conversions are attributed to them.
The Nest acquisition seems to be part of a long term robotics strategy, so probably too early to call on that one.
Orkut I'll give you--Google seems to have basically thrown in the towel on social networking.
The GP said that they are like Berkshire Hathaway - basically a conglomerate of profitable companies. This simply isn't true. Acquiring a company to incorporate their assets into yours is just M&A, nothing like BH.
Nest is a miserable failure.
I don't know a whole lot about the Nest story--seems the consensus now is that it didn't go well, but if Google retains engineers and IP that help them down the line with their robotics push, it's still probably worth something to them, though perhaps that will wind up being a lot less than they paid in 2014.
X who knows but they're supposed to be moonshots, but Fiber and Verily can hardly be called failures.
Berkshire is a conglomerate that looks for proven, established, cash-flow-positive businesses. See's Candy, for instance, requires no debt and is a sure bet every holiday season.
On the other hand, self-driving cars and WiFi balloons are extremely speculative and definitely aren't cash generating.