Aha, glad somebody is seeing the same thing I'm seeing.
The problem with Google is that it is too large, and too successful. Their idea of having experimental projects is a good one, the problem is that Google makes so much money that they don't commit to them. They become too dilettante about them.
The fix is to do spin-offs, say keeping a minority stake at less than 20%. This generates capital for Google, gives them a share of the upside, but they don't have too much influence on the spin-off. So it's Death or Glory for the spin-off. They are forced to make whatever it is they invented work, or face extinction. There is a smaller management team, focussed on success.
This strikes me as a much better proposition than what we see at the moment, with Google just dabbling around pouring money into something they'll eventually get bored with.
You'd think that what with all the big brains at Google, someone would have thought of this. Maybe someone at Goldmans should make a pitch.
1) An enormous pile of cash would accumulate on Google's balance sheet. 2) The spinoffs would be subject to normal commercial rules about risk, rate of return and such. 3) The spinoffs couldn't be "brand Google." Which is a not-inconsiderable thing.
Clearly though, it's a good call.
Spin "the other bets" out as a separate company.
Capitalize it separately.
Literally just don't have it as part of the parent company.
The only advantage of having the company inside the "parent" increasingly becomes accounting tricks, such that excess profits can be shuffled to the "other bet" as a trick to conceal profits from shareholders.
Who might want the capital as a dividend.