4 ms·
Nice article, but drastically oversimplified. Paul ignores two critical issues: Risk, and non-linear utility-of-money functions. These two factors become crit
by cperciva 19y ago
Nice article, but drastically oversimplified. Paul ignores two critical issues: Risk, and non-linear utility-of-money functions. These two factors become critical when there is a tradeoff between probability of success and the payoff of success.
Suppose, as a simple example, that I have a startup which I think has a 50% chance of succeeding and being sold for $1M, and a 50% chance of failing and being worthless. Now suppose that Paul selects me to participate in YC, but wants 10% of the company, and I think his help will leave the potential valuation unchanged but increase the chance of success from 50% to 55%. If I accept his offer, my EXPECTED return drops from $500k (50% of $1M) to $495k (55% of $900k) -- but I'd still accept the offer, because increasing my chance of getting that first $900k is worth far more than getting an additional $100k on top of that.
On the other hand, suppose a venture capital company comes along and offers to help me expand into a much larger market, where I'd have a 10% chance of the company being worth $100M (and a 90% chance of the company being worthless), in exchange for taking 50% of the company stock. If I accept the offer, my EXPECTED return jumps from $500k to $5M (10% of $50M) -- but there's no way that I'd accept the offer, because I really don't want to spend years of my life on something which has a 90% chance of being worthless.
It's important to understand the numbers, but in the end the numbers, at best, have to guide you rather than making decisions for you.
- ivankirigin 19y ago"I really don't want to spend years of my life on something which has a 90% chance of being worthless." How does that mesh with the fact that a failed startup is probably worthless (in the literal sense that you can't make money from it), and most startups probably have >90% failure rate? I know there is a learning experience in startups and that working hard on something fun is valuable, so worthless is really just talking about immediate money here.
- cperciva 19y agoI don't think my probability of failure is 90%. :-) This isn't as naive as it sounds: If you take VC with standard liquidation preference terms, the company needs to do really well before you get anything back -- so the amount of money you need to avoid "failing" is dramatically increased. In my case, since I don't intend to take any VC, there's a wide range between "failure" (making less money than I would have earned risk-free by working at the university for the same duration) and "success" (making enough money that I never need to work again). Also, on a more self-serving note: I'm a heck of a lot more competent than 90% of startup founders. Or even 90% of YC-funded-startup founders for that matter -- and YC-funded startups have distinctly less than a 90% failure rate.
- ivankirigin 19y ago"Or even 90% of YC-funded-startup founders for that matter" Not to deny this, as I don't know anything about you, but... that is a very bold statement.
- cperciva 19y agoOf course it's a bold statement. But if I wasn't bold, I wouldn't have started university at age 13, set three world records for calculating pi (a stunt, I admit), ranked in the top six mathematics undergraduates in North America, received a $100k+ scholarship to Oxford University (not the Rhodes, unfortunately -- their mistake), received a doctorate in computer science from said university, and become the security officer for the FreeBSD operating system.
- sanj 19y agoDid you win the Putnam? If not, please don't be "bolder" than this guy: http://en.wikipedia.org/wiki/Ravi_Vakil http://en.wikipedia.org/wiki/Ravi_Vakil
- cperciva 19y ago"Did you win the Putnam?" Yes, I did.
- pg 19y agoThat has to be the comeback of all time.
- ivankirigin 19y agoHAHAHA! Awesome.
- Alex3917 19y agoThis whole thread is hilarious.
- admoin 19y agoyep, immediately thought of that too. PG writes VERY good stuff most of the time and is very smart about a lot of things, but when he strays into areas in which he is not well versed (the unions essay comes to mind), he ends up writing pieces with obvious holes.
- pg 19y agoDo you have any specific holes you could point to as examples?
- admoin 19y agoI read the union essay a while ago, but your failure to sufficiently consider risk-aversion in this one was a major hole. I don't mean to really criticize that much, you write consistently great stuff.
- pg 19y agohttp://news.ycombinator.com/edit?id=35171 http://news.ycombinator.com/edit?id=35171
- admoin 19y agoThat works, except then I guess the hole was not making it clear enough that you were talking about utility rather than expected value. The problem with your response, and the reason that I don't really see it as a good explanation, is that your basic principle is far less useful if it requires the user to calculate their expected utility.
- pg 19y agoIn practice it's not hard, because nearly all founders' utility functions are practically step functions.
- rzwitserloot 19y agoIn this case, oversimplifying is warranted, mostly because the things glossed over are either sufficiently complicated that it's hard to make simple, or are things are already generally known. Take, for example, the 50% point. Once you hand over so much stock that the amount you and the people you implicitly trust hold dips below 51%, you've lost control. Clearly an issue outside of the 1/(1-n) equation, and yet not really relevant. Everyone knows this already. Then there's the general notion of not handing out too much stock to too many factions, but this too is more or less established knowledge amongst the target audience. Having said that, the nuance of factoring in odds of success is a worthwhile consideration. Hat off for explaining it!
- mulcher 19y agoI agree that simplifying is warranted! I can't understand complex things anyway. Plus if you can't explain it, you don't understand it.. here I feel like I got some modicum of insight.
- pg 19y agoActually not. That's why I was careful to speak of the effect of trading equity on the "average outcome" rather than e.g. "average valuation at liquidity." What I'm literally saying is, does the trade improve your odds of getting what you want? That subsumes both your risk aversion and your utility function for money.
- cperciva 19y agoYour math is still wrong, because of the non-linear utility of money. If I give up 6% of my company, it costs me 6% of any MONEY I might end up getting, but it doesn't cost me 6% of the UTILITY. If I have a utility-of-money function of sqrt($), you don't have to increase my chance of success by 6.4%; it's enough if you increase my chance of success by 3.2%.
- pg 19y ago"Outcome" means the output of the utility function, not the input.
- cperciva 19y agoThen you're still wrong. I quote from the article: "For example, suppose Y Combinator offers to fund you in return for 6% of your company. In this case, n is .06 and 1/(1 - n) is 1.064. So you should take the deal if you believe we can improve your average outcome by more than 6.4%." If by "average outcome" you mean "expected value of the utility function", and assuming that my utility-of-money function is sqrt($), I don't need to improve my "average outcome" by more than 6.4% for the deal to be worth accepting; it's enough if I can increase my "average outcome" by 3.2%, since that's how much UTILITY giving up 6% of the MONEY costs me.
- palish 19y agoIf a man is wrong, you don't need to tell him so. Just give it a rest and let others form their own opinions. Seriously, these issues, and a lot of other issues, are covered in "How to Win Friends and Influence People". Read it. If I could figure out a way to get you to feel like you came up with the idea to read it, I would, but I can't, so just read it.
- DonDodge 19y agoEntrepreneurs face some pretty tough questions at a very early stage. Should I take Angel or VC money? How much money should I raise? How much equity should I give up? How much equity should I grant to early employees? The math equation is correct, but the likely outcomes are nearly impossible to estimate. I have been on the management team of 5 startups and advised many others. There are some "norms" and guidelines for how much to raise at each stage, how much equity to give up, and even how much stock to grant employees as you grow the company. I wrote an in depth blog on these questions, too long to detail here, but Paul is on the right track. For more details see How much Equity for Investors and Employees? http://dondodge.typepad.com/the_next_big_thing/2007/08/how-much-equity.html http://dondodge.typepad.com/the_next_big_thing/2007/08/how-m...