4 ms·

Only in SV could Elizabeth Holmes lose $4.5B she never had to begin with

10 points by smalera 10y ago

algirau 10y ago
They don't discuss if the preferred shares are participating or not. Makes a huge difference. If non-participating, typically investors can elect to convert preferred into common.
paulddraper 10y ago
This makes some excellent points, but this is not one of them: > A related problem is the way startups are valued. Theranos wasn’t valued at $9 billion because it was bringing in massive amounts of revenue or held vast assets. Rather, venture capitalists valued the biotech company so highly because it had the potential to be worth that, or more. It held $9 billion of expectations. Obviously, this is not the same thing as cash. This implies that valuations are intrinsically an overestimate of wealth. $9 billion is the statistically expected value (https://en.m.wikipedia.org/wiki/Expected_value https://en.m.wikipedia.org/wiki/Expected_value), which weighs outcomes by their likelihood. The best case is far, far more then $9b and the worst case is far less (assuming rational investment...ha ha). Now, there can be trends of humans broadly overestimating value (i.e. a bubble), but there's nothing intrinsically wrong with counting illiquid or unstable assets. There's a chance that the future is better and there's a chance it is worse; the estimate takes the middle of the road. Omitting these assets would be an underestimate.
- dragonwriter 10y ago
  > $9 billion is the statistically expected value No, $9 billion is the subjective expectation of the party or group of parties willing to pay the most. (Now, no doubt, they apply some kind of statistical formulae to conclude that that is the statistically expected value, but its also clear that other potential investors are also doing that with different models, and in the case of disagreement, the ones assessing the valuation the highest are, ceteris paribus, the ones with whom a deal will be made. Valuations derived from public market caps have some problems, and may not be justifiable rational expectations, but they are at least some kind of broad consensus. But private-deal-based valuations are inherently inflated by the way deals get made.)

10 points by smalera 10y ago

They don't discuss if the preferred shares are participating or not. Makes a huge difference. If non-participating, typically investors can elect to convert preferred into common.

paulddraper 10y ago

This makes some excellent points, but this is not one of them: > A related problem is the way startups are valued. Theranos wasn’t valued at $9 billion because it was bringing in massive amounts of revenue or held vast assets. Rather, venture capitalists valued the biotech company so highly because it had the potential to be worth that, or more. It held $9 billion of expectations. Obviously, this is not the same thing as cash. This implies that valuations are intrinsically an overestimate of wealth. $9 billion is the statistically expected value (https://en.m.wikipedia.org/wiki/Expected_value https://en.m.wikipedia.org/wiki/Expected_value), which weighs outcomes by their likelihood. The best case is far, far more then $9b and the worst case is far less (assuming rational investment...ha ha). Now, there can be trends of humans broadly overestimating value (i.e. a bubble), but there's nothing intrinsically wrong with counting illiquid or unstable assets. There's a chance that the future is better and there's a chance it is worse; the estimate takes the middle of the road. Omitting these assets would be an underestimate.

dragonwriter 10y ago
> $9 billion is the statistically expected value No, $9 billion is the subjective expectation of the party or group of parties willing to pay the most. (Now, no doubt, they apply some kind of statistical formulae to conclude that that is the statistically expected value, but its also clear that other potential investors are also doing that with different models, and in the case of disagreement, the ones assessing the valuation the highest are, ceteris paribus, the ones with whom a deal will be made. Valuations derived from public market caps have some problems, and may not be justifiable rational expectations, but they are at least some kind of broad consensus. But private-deal-based valuations are inherently inflated by the way deals get made.)