4 ms·

I'm a VC. I think a lot of this boils down to the asymmetric relationship between the upside and the downside. If a company doesn't do well, the investor loses

I'm a VC. I think a lot of this boils down to the asymmetric relationship between the upside and the downside. If a company doesn't do well, the investor loses all/most of their investment. It doesn't matter how ambitiously or unambitiously the startup acted, the VC's losses are roughly fixed. On the other hand, the difference between a company having a very good (e.g. 8x) vs. fantastic (e.g. 50x) outcome is huge. If increasing burn helps a company grow faster, acquire more market share, and improves the chances that the company will 50x instead of 8x, then that's worth a higher chance of failure to the VC because they have some diversification. It may or may not be worth it to the founders, who only work on one startup at a time. So basically, if the two possible scenarios are the following.. 1) 60% chance to 0x, 30% chance to 3x, 10% chance to 8x 2) 80% chance to 0x, 15% chance to 3x, 5% chance to 50x .. then VCs strongly prefer the second scenario. There's a slightly higher chance of a loss, but the upside is much, much greater.

kordless 10y ago
A preference to viral models, backed by unbounded trust, sets the intent (rules of the game) to produce viral business models at the direct expense of less viral, but more holistic customer-centric models. Investing in high growth models seems like a great idea when sitting around drinking beer, but I'm willing to bet the long term effect of this practice is going to come back and bite us in the ass sooner than later, if it hasn't already. There's a reason why seaweed reduces cow gas. We have to use seaweed on cows now because we removed the cows from the grass. There's a similar reason behind why our privacy has been thrown out with the bathwater!
didgeoridoo 10y ago
Given the expected values in your example, I don't think anyone would fault a VC (or any actor with a non-pathological marginal utility curve) for taking option 2. Now, what about the two bets: 1) 80% 1x, 20% 5x (EV: 1.8x) 2) 99% 0x, 1% 100x (EV: 1.0x) How many VCs do you know who would take 1 over 2, even though it's the "obvious" money-winning bet? I'd love your thoughts on why this is. My semi-cynical thesis is that VCs need home-run logos on the Portfolio page of their website to keep their spot at the dealflow table. Let's break the VC world up into three types: A - Brand-name VCs that funded yesterday's massive unicorns, and are regarded as clairvoyant masters of the universe B - Unknown VCs with a half-decent return on their last fund, but didn't back anyone you've heard of C - VCs that swung for the fences a few too many times and imploded Founders want to work with "A" VCs (for both rational and irrational reasons), so the first names they call are the ones whose past big bets paid off. "B" VCs are much lower on the call list, and only get access to the deals that "A" VCs all passed on. "C" VCs are obviously not available to chat. It is arguably rational for a VC to seek to become an "A" and end up as a "C", rather than to muddle along as a "B" until she's getting pitched every week by hair salon and gym owners with A Great App Idea and her LPs wonder what she's doing with their money.
- lpolovets 10y ago
  Lots of great points in your comment. First, I don't think any rational VC would take #2 over #1 in your example, but the caveats are: a) people aren't good at estimating small probabilities, so they might think that #2 should be 97%/3% instead of 99%/1%, and then #2 is "better". And b) They overestimate how much of a "sure thing" something is, so maybe what seems like 80%/20% in scenario #1 is really 50% 0x, 30% 1x, and 20% 5x. I can't remember the origin of the quote, but I've repeatedly heard that "trying to build a $100m company is not that much riskier than trying to build a $10b company, so you may as well shoot for $10b." Also, generally VCs won't take bets where the EV is <3x or so. A good fund will have a net return of 3x+, and you can't get there by making a lot of 2x bets. Furthermore, 99% 0x vs. 1% 100x is very lopsided. Probably too lopsided for most investors with portfolios of 20-40 companies per fund. If someone made that bet for all 30 companies in their fund, the fund would be 0x almost 3/4 of the time. But if the bet was more like 95% 0x and 5% 100x, then that's kind of interesting. Making 30 investments like that would produce a terrible return 21% of the time (0 wins), a good return 34% of the time (1 win), and a terrific return the rest of the time (2+ wins). That said, prestige is (unfortunately) a factor. So I could imagine someone making an investment with an EV of 4x that is likely to be a unicorn over an investment with an EV of 5x that is unlikely to be a unicorn. The last thing I'll add is that A/B/C grades for investors are context-specific. An 'A' like Sequoia or Accel might be generally pretty great for everyone, and you won't go wrong taking money from those firms. But in certain areas, lesser-known funds might be very strong company-building partners. For example, there are specialty funds that are focused on SaaS companies or frontier tech, and if you operate in one of those spaces then those funds might be just as good as taking money from Sequoia. Finally, below is an interesting set of tweets from an LP who has been investing in VC funds for several decades. TLDR: it's basically impossible to build a great fund return on top of lots of small wins; you need a few huge wins instead. https://twitter.com/HorsleyBridge/status/657287940456886272 https://twitter.com/HorsleyBridge/status/657287940456886272 https://twitter.com/HorsleyBridge/status/657288063329013760 https://twitter.com/HorsleyBridge/status/657288063329013760 https://twitter.com/HorsleyBridge/status/657288245105917952 https://twitter.com/HorsleyBridge/status/657288245105917952 https://twitter.com/HorsleyBridge/status/657288489331855360 https://twitter.com/HorsleyBridge/status/657288489331855360

kordless 10y ago

A preference to viral models, backed by unbounded trust, sets the intent (rules of the game) to produce viral business models at the direct expense of less viral, but more holistic customer-centric models. Investing in high growth models seems like a great idea when sitting around drinking beer, but I'm willing to bet the long term effect of this practice is going to come back and bite us in the ass sooner than later, if it hasn't already. There's a reason why seaweed reduces cow gas. We have to use seaweed on cows now because we removed the cows from the grass. There's a similar reason behind why our privacy has been thrown out with the bathwater!

didgeoridoo 10y ago

Given the expected values in your example, I don't think anyone would fault a VC (or any actor with a non-pathological marginal utility curve) for taking option 2. Now, what about the two bets: 1) 80% 1x, 20% 5x (EV: 1.8x) 2) 99% 0x, 1% 100x (EV: 1.0x) How many VCs do you know who would take 1 over 2, even though it's the "obvious" money-winning bet? I'd love your thoughts on why this is. My semi-cynical thesis is that VCs need home-run logos on the Portfolio page of their website to keep their spot at the dealflow table. Let's break the VC world up into three types: A - Brand-name VCs that funded yesterday's massive unicorns, and are regarded as clairvoyant masters of the universe B - Unknown VCs with a half-decent return on their last fund, but didn't back anyone you've heard of C - VCs that swung for the fences a few too many times and imploded Founders want to work with "A" VCs (for both rational and irrational reasons), so the first names they call are the ones whose past big bets paid off. "B" VCs are much lower on the call list, and only get access to the deals that "A" VCs all passed on. "C" VCs are obviously not available to chat. It is arguably rational for a VC to seek to become an "A" and end up as a "C", rather than to muddle along as a "B" until she's getting pitched every week by hair salon and gym owners with A Great App Idea and her LPs wonder what she's doing with their money.

lpolovets 10y ago
Lots of great points in your comment. First, I don't think any rational VC would take #2 over #1 in your example, but the caveats are: a) people aren't good at estimating small probabilities, so they might think that #2 should be 97%/3% instead of 99%/1%, and then #2 is "better". And b) They overestimate how much of a "sure thing" something is, so maybe what seems like 80%/20% in scenario #1 is really 50% 0x, 30% 1x, and 20% 5x. I can't remember the origin of the quote, but I've repeatedly heard that "trying to build a $100m company is not that much riskier than trying to build a $10b company, so you may as well shoot for $10b." Also, generally VCs won't take bets where the EV is <3x or so. A good fund will have a net return of 3x+, and you can't get there by making a lot of 2x bets. Furthermore, 99% 0x vs. 1% 100x is very lopsided. Probably too lopsided for most investors with portfolios of 20-40 companies per fund. If someone made that bet for all 30 companies in their fund, the fund would be 0x almost 3/4 of the time. But if the bet was more like 95% 0x and 5% 100x, then that's kind of interesting. Making 30 investments like that would produce a terrible return 21% of the time (0 wins), a good return 34% of the time (1 win), and a terrific return the rest of the time (2+ wins). That said, prestige is (unfortunately) a factor. So I could imagine someone making an investment with an EV of 4x that is likely to be a unicorn over an investment with an EV of 5x that is unlikely to be a unicorn. The last thing I'll add is that A/B/C grades for investors are context-specific. An 'A' like Sequoia or Accel might be generally pretty great for everyone, and you won't go wrong taking money from those firms. But in certain areas, lesser-known funds might be very strong company-building partners. For example, there are specialty funds that are focused on SaaS companies or frontier tech, and if you operate in one of those spaces then those funds might be just as good as taking money from Sequoia. Finally, below is an interesting set of tweets from an LP who has been investing in VC funds for several decades. TLDR: it's basically impossible to build a great fund return on top of lots of small wins; you need a few huge wins instead. https://twitter.com/HorsleyBridge/status/657287940456886272 https://twitter.com/HorsleyBridge/status/657287940456886272 https://twitter.com/HorsleyBridge/status/657288063329013760 https://twitter.com/HorsleyBridge/status/657288063329013760 https://twitter.com/HorsleyBridge/status/657288245105917952 https://twitter.com/HorsleyBridge/status/657288245105917952 https://twitter.com/HorsleyBridge/status/657288489331855360 https://twitter.com/HorsleyBridge/status/657288489331855360