 | |
Is it a secret though? I thought it's 101 of startups-VC interactions, and I learned it by reading pg's essays in about the same time I first learned what a "startup" is. I.e. the whole thing works because some people with money do high-risk investments in which nine companies out of ten fail, but the tenth one pays them back more than they invested in all of them together. | |
| |
But then you get the counter argument from people like Dave McClure who speak pretty actively against this style for most investors, implying they should be investing in way more companies at earlier stages (also I highly suggest avoid calling it a "shotgun" approach unless you want to piss him off).There's a reason 500 Startups [edited out "he" and replaced with firm name - although he is also a prolific individual angel] tops the league tables for most active early stage investments every year, because they believe a lot of VCs are wrong. Active investments by VCs since JAN2015 - http://imgur.com/QxVJVgZ | |
| |
It's the same core belief: returns are driven by outliers. The difference is how to find them. Peter Thiel suggests focusing only on finding those with the potential to do it. Dave McClure suggests looking at tons of companies because you can't tell.It's analogous to someone observing, "Over long time periods, companies that have low PE's outperform companies with high PEs" Warren Buffett picks individual companies with high PEs, while David Booth creates and index for all of them. Both have found ways to become billionaires. | |
|
| |
here's a rough summary of my track record (personal / professional)1) my angel investments ($300K portfolio, 2004-2008 vintage): 3 exits (Mint, Mashery, SlideShare) @ $100M+ out of 13 -- roughly 3.5X cash on cash in ~8 years 2) my investments at Founders Fund (~$3M portfolio, 2008-2010): 3 unicorns (Credit Karma, Lyft/ZimRide, Twilio) + 3 large wins (Wildfire, SendGrid, Life360) out of ~40 investments via FF Angel + fbFund -- roughly $50-60M appreciation in value over 7-8 years, >100% Gross IRR 3) 500 Startups main funds: $30M Fund I / 265 companies / 19% Net IRR / 2010-11 vintage, $45M Fund II / 325 companies / 23% Net IRR / 2012-13 vintage -- so far, 2 unicorns (Twilio, Credit Karma), 2 half-unicorns (Ipsy, Udemy), 30+ "centaurs" (>$100M+ value). Fund III is $85M / vintage 2014-15 / 650+ companies -- still pretty early but so far Net IRR trending ~20% our LPs have been happy with our results so far. my/our track record is likely upper quartile, and at least for my years at Founders Fund top decile. Peter and I may differ in approach & stage, but likely more in agreement than not about the #s, altho he would likely consider or strategy more brute force and inelegant than his. that said, I think made enough money for him at FF & found him 3 unicorns, so hopefully he doesn't think I'm an idiot ;) | |
| |
Wow, great stuff, and thanks for answering this personally ! | |
| |
Good question - we don't have much IRR data on 500 Startups' funds (most of them are still open), but they never have trouble raising more money for new funds, so I assume their LPs are happy. | |
| |
What does AUM mean?(Yes, can I have access to pitchbook?) | |
| |
Assets under management - basically how much cash these firms raise and then need to deploy through investments. | |
|
| |
> Is it a secret though?It's a secret to the majority of the people that are the market for a book like that. As opposed to those of us who are on top of startups who found out that information many many years ago. | |
| |
Well, he also uses secret throughout the book to mean "an idea that's not commonly accepted". | |
| |
Do most VC's think this way and is this actually a "rule", meaning it's the way you succeed?I'd think VC investing is like most other types of investing where you can take on different strategies depending on your goal. Some investors will take large risk in order for large reward, where other investors would rather take lower risk for a higher probability of some positive return. | |
| |
Yeah I feel that tends to play out based on what class of VC you are. E.g.seed/angels -> high risk / big portfolio (100s startups) / huge win-loss multiples (but invest small $) traditional VCs/Seed A-C -> medium risk / moderate portfolio size (10s) / medium win-loss multiples (but invest medium $$) growth private equity -> small risk / small portfolio (singles) / small win-loss multiples (but invest huge $$$) Of course the divisions are arbitrary and some investors put money in all categories, but I feel like the above holds true in general. | |
| |
I think the problem with taking the moderate-risk-for-moderate-reward strategy, as an early stage investor, is that a large number of your investments are going to fail anyway. And the returns of a company are not linearly correlated with its apparent risk at the early stage: a company might appear 2x as risky, but end up with 200x the returns if it succeeds. Early-stage investors that try to invest further down the risk/reward curve may find that the decreased risk doesn't end up making up for the decreased rewards, and so the fund as a whole becomes unprofitable.Larry Page was fond of saying that it's actually easier to work on big problems than on little problems, because a.) there's less competition and b.) you can get people to help you. I'm not sure that's actually true anymore - now that everyone wants to be Google - but it's illustrative of the non-linearity of returns. | |
| |
As an investor, I can say you nailed it with this explanation. | |
| |
I think it's a "rule" as opposed to a strategy like "don't sacrifice your pieces in exchange for nothing" is a rule in chess. It's possible to succeed without following that guide line, but it's very difficult.VC firms produce modest returns for a very high risk. You've probably only heard of one out every hundred companies that have been venture funded. Low risk investments just don't produce enough return. | |
| |
> only invest in companies that have the potential to return the value of the entire fundI have a feeling this is much, much easier said than done. How do you even determine "potential" of a startup, when, according to Paul Graham, "the best ideas look initially like bad ideas". | |
| |
It's incredibly hard to identify successful companies early on. But following Thiel's rule isn't actually all that hard.The key is to look at the total addressible market for a company or product. It has to be large, or growing quickly, or both. For example, AirBnB might have looked like a bad idea, but the hospitality market is huge, so if it did work out then they could grow to become a giant company. On the other hand, you could create a transformative product for blind people, and build a business around it that makes you a multi-millionaire. But VCs will never invest in your company because there just isn't a big enough market. You might 3x or 5x an investment but you'll never deliver the kinds of giant returns VCs need in order to make their LPs happy. Of course, I'm talking about traditional VC firms - like the one the blog post's author runs. There are all kinds of investors out there with different motivations. | |
| |
> But following Thiel's rule isn't actually all that hard.> The key is to look at the total addressible market for a company or product. It has to be large, or growing quickly, or both. For example, AirBnB might have looked like a bad idea, but the hospitality market is huge, so if it did work out then they could grow to become a giant company. Again, this just feels like post-hoc rationalization. The guy who wrote this blog post declined to invest in AirBnB despite Paul Graham himself practically begging him to. So maybe it actually is hard? | |
| |
Just because he decided not to invest doesn't mean he didn't think AirBnB had the potential to become big. It doesn't mean he didn't think they were going after a big market.He might have just thought the team wasn't very good, or the product wasn't quite right, or any of the other reasons investors pass on companies. The potential was there, but it just wasn't very likely given the details of the company. I think what Thiel is getting at with his rule is to not bother with companies that don't have the potential to ever get huge, given their product and market. That rules out a huge number of businesses, so following it prevents you from wasting a lot of time. | |
| |
> ... build a business around it that makes you a multi-millionaire. But VCs will never invest in your company because there just isn't a big enough market. You might 3x or 5x an investment but you'll never deliver the kinds of giant returns VCs need in order to make their LPs happy.Ah, a lifestyle business. Just kidding -- I'm a fan. I found PG's take on them in the footnote of Black Swan Farming (http://paulgraham.com/swan.html) refreshing: > Nor do we push founders to try to become one of the big winners if they don't want to. We didn't "swing for the fences" in our own startup (Viaweb, which was acquired for $50 million), and it would feel pretty bogus to press founders to do something we didn't do. Our rule is that it's up to the founders. Some want to take over the world, and some just want that first few million. But we invest in so many companies that we don't have to sweat any one outcome. In fact, we don't have to sweat whether startups have exits at all. The biggest exits are the only ones that matter financially, and those are guaranteed in the sense that if a company becomes big enough, a market for its shares will inevitably arise. Since the remaining outcomes don't have a significant effect on returns, it's cool with us if the founders want to sell early for a small amount, or grow slowly and never sell (i.e. become a so-called lifestyle business), or even shut the company down. We're sometimes disappointed when a startup we had high hopes for doesn't do well, but this disappointment is mostly the ordinary variety that anyone feels when that happens. | |
| |
When Facebook came around, wouldnt the addressable market have looked tiny? Basically US based college campuses.Or was Zuckerberg already envisioning opening it up and spreading it far beyond college (and maybe high school) campuses? | |
| |
Peter Thiel said recently in an interview[0] that when he made his initial $500K investment in Facebook at a $5M valuation in September 2004, he thought it would be big on college campuses, but didn't anticipate how successful it would actually become.0: https://www.youtube.com/watch?v=ryFB6mvy4uE#t=3m18s | |
| |
He has said himself that while working on Facebook in the early days, he thought the whole world needed something like it. I believe the college-first approach was more of a tactic, but also a somewhat obvious one after his earlier expeditions with "who's hotter?" or whatever. If he had launched "Facebook for the world" from the start, user reaction would see it just like MySpace and ask themselves why? Going with colleges first gave it a feel of exclusivity allowing it to grow the way it did. | |
| |
Another thing about startups I've realized is you want to start with something small, but you could also see being expanded further.For example, snapchat started with LA teenagers. Facebook started with harvard, then ivy league colleges, etc. Uber was licenced black car services in SF only at first, etc. | |
| |
Always.And it's funny - the startups that try to start with "we're revolutionizing the world" end up over-promising. The ones like you mentioned actually do. Not only just in starting in one market, but focusing on one customer segment, or one feature, or one vertical. | |
| |
pg discusses some of these things in his essay "do things that don't scale" http://paulgraham.com/ds.htmlThis is really hard advice to take ..and in my limited experience also hard to convince investors of. But I think he's right ...and it amounts to getting in the game / get out of the building etc. | |
| |
I don't know what the conversations around FB's early investment were like, or if expanding to a billion users was always part of the plan. My understanding is that very early on (i.e. the first six to nine months after it started) they wanted to be a kind of information hub for school campuses, with course listings and whatnot.Even if that's true and they pitched investors on a kind of online campus hub for students, they woukd still have been going after a pretty big market - they could sell software/functionality to schools and/or they could sell advertising (reaching young people is quite valuable for brands since young people tend to have less fixed opinions and loyalties as consumers). It may well be that investors thought it could grow to compete with MySpace. Others might have just thought being an essential part of every student's life would be a good enough outcome. (After all, there's plenty of VC in "ed tech" these days). | |
| |
The hub for all college students is actually a pretty big addressable market. It's certainly not as big as Facebook ended up being, but it's still big enough to interest VCs. | |
| |
Isn't Facebook's market their advertisers? Do regular users, nin-advertisers ever pay Facebook for anything? | |
| |
Of course. After all, Fred Wilson refused to invest in airbnb even when Paul Graham begged him to. | |
| |
YC (for example) works around this problem by investing in founders, judging for perseverance and hunger. Most other firms probably have similar strategies. | |
| |
> "the best ideas look initially like bad ideas"And they look even worse to those of us in the peanut gallery (and the pundits) who don't have access to all of the facts that someone who is actually investing has. They at least have answer to questions. A bit like investing at a higher level in the stock market (and taking major positions which often allows you to glean info from people that work at the company). | |
| |
Which brings to another unspoken rule of investments in general - "Looking at the history, every good investment looks good and every bad investment looks bad". Everything is explainable that occurred in the past but not with the information that was available at that time. | |
| |
> only invest in companies that have the potential to return the value of the entire fundNo it doesn't. If the investment is likely to break even it makes no difference, or can at least reduce losses. So long as there is a certain, critical number of potential high-earners, compared to total investment, you're ok. | |
| |
Yup, VCs tend to favor go-big-or-go-home ventures. There are also lots of seemingly tiny things that can be made bigger too... and those are better pitches/founders to astute investers like Thiel whom realize small people are all talk and big people understate themselves.Reminds me of running into some tipsy Sequoia guys the night before the WhatsApp announcement... no lie,I knew who they were and what happened (figure it was a gigadeal) the second they walked in to a certain donut shop. It's definitely party-worthy when the fund is above water and all other exits are IRR gravy. | |
| |
isn't this a self fulfilling prophecy? if i only invest in all-or-nothing startups, doesn't that mean that my returns are going to look like i only invested in all-or-nothing startups? | |
| |
It's almost impossible to have a different VC model though.If you want to remove the variability in your outcomes, you have to do substantially more due diligence on each investment and decline a lot more. When your model depends on each investment doing okay instead of a few home runs, there's a lot more pressure on each investment to not fail. Eventually this forces you to only accept the least aggressive plans and you become a bank. | |
