3 ms·

> Sure, but humans care more about the median value than the expected value. See also the St. Petersburg paradox. Expected value is infinite, but general consen

> Sure, but humans care more about the median value than the expected value. See also the St. Petersburg paradox. Expected value is infinite, but general consensus is that it's irrational to pay more than a couple of dollars to play the game. Not sure I agree with that. Simple thought experiment: would you prefer "$100" or "51% chance at $0, 49% chance at $1000"? I strongly suspect you'd prefer the latter, even though the median for that scenario is $0. My argument isn't that every startup might be the next Uber and that as a result you should take whatever stock you can get; it's that on average, startups increase in value (otherwise both startups and venture capital would quickly die), so stock has a positive EV. Not in the "99.9% chance it's $0, .1% chance it's Uber" sense, but in the 50% chance 0, 40% chance 2x, 10% chance 10x sense (which is pretty reasonable, especially if you work for 5 or 10 startups during your career). > How should I judge the exit multiple? Especially for a late-stage company, isn't it generally reasonable to estimate it as 1x? VCs are generally looking for 3x-4x returns on average, and they'll never invest in a startup if the expected value isn't at least 3x-4x based on some estimated probability distribution. So the expected exit multiple is usually at least 3x or so. The later stage you go, the more predictable and narrow the exit distribution becomes. For example, pre-product, a VC might be investing on the 2% chance that something will go up 200x. Post-launch but pre-revenue, a VC might be looking for a 5% chance of a 60x. Series A or B, it might be more like 40% chance that the VC at least gets their money back and 30% chance they'll make 10x. Growth stage, it might be more like 60% chance the VC doesn't lose any money, 30% chance it's a 5x. In all of these cases, expected value might be 2x-4x, but variance drops off drastically as a company matures. (Seed VCs are usually aiming for more like 4x-5x for their funds, while a late stage VC is aiming for 3x, so the lower variance is compensated by lower expected returns.)

geofft 11y ago
> Simple thought experiment: would you prefer "$100" or "51% chance at $0, 49% chance at $1000"? I strongly suspect you'd prefer the latter, even though the median for that scenario is $0. $100 or $490 or $1000 isn't really enough to make a meaningful difference in my life decisions, so I'll take the higher-expected-value option, simply because the money is strictly bonus. If I want to plan with the money, though, I want the lower-risk option. In particular, compare a $90K + equity salary with a $120K + no equity one, where there's a small chance that the equity is worth well over $100K in three years. With the higher salary, if I want to spend it now, I can get an apartment closer to work, I can visit my family more or take more vacation, etc. Alternatively, if I save it, I can take a year off of work and fund my own work, I can put a down payment on a house, etc. It's certainly true that if I win the equity game, I can take well over a year off of work, I can put a larger down payment on a better house, etc. But if I don't, I have to keep working and living in an apartment. And since I have no guarantee of it, I certainly can't spend that expected value now. From the purpose of planning, if I actively want to do one of these things in three years, I'd rather have a guarantee of being able to do it somehow, than a chance of being able to do it very well and a chance of nothing. Once I have a guarantee of being able to buy a house in three years, I can do things like plan to start a family then. If I have such a goal in mind, it's clearly better to choose the lower-risk thing, if it provides enough money to enable this sort of long-term planning. $100 vs. $490 will hardly make any difference to how many people I can invite to the wedding.
- lpolovets 11y ago
  I hadn't looked at this from that perspective. That's a great point! =) The only thing I'd add is that 5-10 years ago, a startup salary was a big cut from a BigCo salary. Like maybe 30-50% lower. These days, the difference between a startup salary and a Google salary might be something like $125k vs $150k -- and $125k is already enough money for taking vacations or having a down payment after 5-10 years of saving. Perhaps if there's a downturn, startup salaries will become low again, but that's not the case right now.
  - seattle_spring 11y ago
    That's not quite the whole story though. Sure, base salary might be $125k vs $150k, but in reality it'd be: - Startup: $125k base salary, small potential for an OK payout but a large potential that you'll take away that salary and nothing more after 4 years. - BigCo: $150k salary, $200k RSU over 4 years, 15% bonus every year Not counting the inevitable refreshers, bonus multipliers, raises, etc, you're really looking at $500k / 4 yr at a startup vs $890k from bigco. That's not an insignificant difference, and I feel like I'm being pretty conservative with the BigCo numbers.

geofft 11y ago

> Simple thought experiment: would you prefer "$100" or "51% chance at $0, 49% chance at $1000"? I strongly suspect you'd prefer the latter, even though the median for that scenario is $0. $100 or $490 or $1000 isn't really enough to make a meaningful difference in my life decisions, so I'll take the higher-expected-value option, simply because the money is strictly bonus. If I want to plan with the money, though, I want the lower-risk option. In particular, compare a $90K + equity salary with a $120K + no equity one, where there's a small chance that the equity is worth well over $100K in three years. With the higher salary, if I want to spend it now, I can get an apartment closer to work, I can visit my family more or take more vacation, etc. Alternatively, if I save it, I can take a year off of work and fund my own work, I can put a down payment on a house, etc. It's certainly true that if I win the equity game, I can take well over a year off of work, I can put a larger down payment on a better house, etc. But if I don't, I have to keep working and living in an apartment. And since I have no guarantee of it, I certainly can't spend that expected value now. From the purpose of planning, if I actively want to do one of these things in three years, I'd rather have a guarantee of being able to do it somehow, than a chance of being able to do it very well and a chance of nothing. Once I have a guarantee of being able to buy a house in three years, I can do things like plan to start a family then. If I have such a goal in mind, it's clearly better to choose the lower-risk thing, if it provides enough money to enable this sort of long-term planning. $100 vs. $490 will hardly make any difference to how many people I can invite to the wedding.

lpolovets 11y ago
I hadn't looked at this from that perspective. That's a great point! =) The only thing I'd add is that 5-10 years ago, a startup salary was a big cut from a BigCo salary. Like maybe 30-50% lower. These days, the difference between a startup salary and a Google salary might be something like $125k vs $150k -- and $125k is already enough money for taking vacations or having a down payment after 5-10 years of saving. Perhaps if there's a downturn, startup salaries will become low again, but that's not the case right now.
- seattle_spring 11y ago
  That's not quite the whole story though. Sure, base salary might be $125k vs $150k, but in reality it'd be: - Startup: $125k base salary, small potential for an OK payout but a large potential that you'll take away that salary and nothing more after 4 years. - BigCo: $150k salary, $200k RSU over 4 years, 15% bonus every year Not counting the inevitable refreshers, bonus multipliers, raises, etc, you're really looking at $500k / 4 yr at a startup vs $890k from bigco. That's not an insignificant difference, and I feel like I'm being pretty conservative with the BigCo numbers.