“[Naval Ravikant] is an entrepreneur and angel investor, a co-author of Venture Hacks, and a co-maintainer of AngelList. Previously he]was a co-founder at Genoa Corp (acquired by Finisar), Epinions.com (IPO via Shopping.com), and Vast.com (white-label classifieds marketplace).”
1. “The cost of starting a company has collapsed.” “As the cost of running a startup experiment is coming down, more experiments are being run.” “Three years ago, companies could for the first time get all the way through a prototype of a service before they even raised seed money. Two years ago, they could make it through launch before raising money. Now, they can start to get traction with a user base by the time they come looking for seed money.” A capitalist economy is an evolutionary system. Innovation and best practices are discovered by the experimentation of entrepreneurs who try to establish the fitness of their business. Products created by this experimentation which have greater fitness survive and other “less fit” products die. Entrepreneurs are essentially running experiments in this evolutionary system when they create or alter a business. Entrepreneurs are engaged in ‘Deductive tinkering” as they search for better products. Eric Ries describes the process in this way: “Learning how to build a sustainable business is the outcome of experiments [which follow] a three step process. Build, measure, learn.”
Why is experimentation so important? The answer is that experimentation is the best way to deal with one of nature’s solutions to dealing with risk, uncertainty and ignorance: a complex adaptive system. An economy is a complex system in that it is networked and therefore adaptive in ways that simple formalisms such as used in physics will fail to predict. Michael Mauboussin explains:
“A complex adaptive system has three characteristics. The first is that the system consists of a number of heterogeneous agents, and each of those agents makes decisions about how to behave. The most important dimension here is that those decisions will evolve over time. The second characteristic is that the agents interact with one another. That interaction leads to the third—something that scientists call emergence: In a very real way, the whole becomes greater than the sum of the parts. The key issue is that you can’t really understand the whole system by simply looking at its individual parts.”
In the case of complex adaptive systems like an economy or a business, the correct approach is to discover solutions via trial and error rather than try to predict. Nassim Taleb describes why the experimentation approach works well: “it is in complex systems, ones in which we have little visibility of the chains of cause-consequences, that tinkering, bricolage, or similar variations of trial and error have been shown to vastly outperform the teleological—it is nature’s modus operandi. But tinkering needs to be convex; it is imperative…. Critically [what is desired is to] have the option, not the obligation to keep the result, which allows us to retain the upper bound and be unaffected by adverse outcomes.” What Taleb is talking about is “convexity” which is something I have written a lot about recently. As an example of convex financial proposition, all a founder or venture capitalist can lose is 100% of what they invest in a startup and yet what they can potentially gain is potentially many multiples of that investment. People will inevitably say when the topic of complex adaptive systems comes up that nothing has emerged from research in this area that allows you to predict the future, which ignores the point that knowing what you can’t predict is one of the most valuable things you can know. Discovery which happens via experimentation via trial and error is a vastly superior way to deal with unpredictability than trying to predict what can’t be predicted.
It is important to note that Naval is talking about a specific type of business experiment. One way to look at the impact of business startups is to group them into two categories:
Some startups are an attempt to create entirely new categories of businesses at global scale (e.g., Uber, Salesforce.com or Airbnb).
Some startups are about incremental or local innovation, such as a new frozen yogurt shop or sushi restaurant.
The number of businesses trying to create value in category 1 through business experiments has substantially increased. Comparing a category 1 startup (e.g., venture-backed Uber or Airbnb) to a category 2 startup (e.g., new frozen yogurt store on main street) is apples and oranges. As an aside, I believe anyone doing something like opening up a new restaurant or main street retailer deserves a medal for bravery. The competition they face on a daily basis is brutal. These category 2 entrepreneurs are vitally important parts of the process that makes capitalism work creating jobs and paying taxes.
It is important to understand how much bigger category 2 is than category 1. The number of startups that obtain venture finance for the first time in a given year is small. There are ~ 700-800 firms per quarter in the US which raise venture capital for the first time. This data is from PitchBook:
Part of Ravikant’s point is that there are a huge number of startups that are able to conduct experiments based on the personal capital of founders even before seed. I’m not aware of anyone who has compiled a reliable estimate of how many startup experiments fail to raise seed capital or bootstrap the business to success, but it is a significant number. The same problem exists for trying to estimate how many startups raise some “friends and family” money. I have been told that there is no reliable estimate of how much non professional seed investing goes on in any given year.This activity is just to diffuse and distributed. Nevertheless the lower cost of conducting an experiments has increased the number of experiments and the level of innovation is higher as a result.
Compare venture capital-backed business to a figure that includes category 2’s businesses like yogurt stands and convenience stores:
“Firms are individual businesses, while establishments include multiple outlets for existing firms. The Brookings report specifically discussed entrepreneurship, which is why it used the numbers for firms. “The distinction between a new Chase Bank branch opening in your neighborhood versus a brand new community bank is critical — particularly when studying entrepreneurship”…. The most recent data from Census, released in September 2014, that showed firm deaths in 2012 (424,864) still outnumbered births (410,001). That’s a difference of 14,863, and the figures showed that the gap had been narrowing each year since deaths first outnumbered births by 90,670 in 2009. But, it turns out, Census had released new numbers for 2013 in September of this year. And in doing so, it revised its figures for past years. The latest statistics now show that in 2012, firm births (411,252) were higher than firm deaths (375,192) by 36,060. And the same held true for 2013, when births outnumbered deaths by 5,666. We can expect that figure to be revised, too, when Census releases new figures in the future. …the latest numbers show that more firms are opening than closing.
My trip to Omaha for the Berkshire meeting this year was interesting in terms of how many business, especially in West Omaha, are franchises.
A paper prepared for the Federal Reserve Bank of Cleveland, in August 2014 said there had been a shift away from brand-new businesses toward new outlets of existing businesses, a trend that many Americans may have seen in their own communities. “The Shifting Source of New Business Establishments and New Jobs,” Aug. 21, 2014: We find that while new firms have been forming at a slower pace over the past 33 years and creating fewer jobs, there has been a simultaneous rise in the number of new establishments opened by existing businesses (which we will call new outlets). … Markets that used to be served by independent entrepreneurs creating businesses are now increasingly being served by the expansion of existing businesses.
There are a lot more business in 2 category in absolute numbers but it is category 1 that generates the most fundamental innovation. Do I wish there was more business entry in category 2 and that they were more successful more often? Definitely. Los of jobs are create in category 2. But that is not the same set of issues as are involved in category 1 which is what Naval is referring to.
The additional point that must be considered is that innovation in category 1 can cause both more business failure and more new business starts in category 2. The outcome varies since any given innovation can increase profit or not. Whether profit is generated by any given innovation is determined by the presence of a moat. Am I saying that some innovation at an aggregate level produces no profit or even less overall profit? Yes. Charlie Munger explains the phenomenon best:
“When technology moves as fast as it does in a civilization like ours, you get a phenomenon which I call competitive destruction. You know, you have the finest buggy whip factory and all of a sudden in comes this little horseless carriage. And before too many years go by, your buggy whip business is dead. You either get into a different business or you’re dead—you’re destroyed. It happens again and again and again. There are all kinds of wonderful new inventions that give you nothing as owners except the opportunity to spend a lot more money in a business that’s still going to be lousy. The money still won’t come to you. All of the advantages from great improvements are going to flow through to the customers.”
2. “Success rates are definitely coming down but that is because the cost of running a startup experiment is coming down…so more experiments are being run. In the old days, we would have one company spend $10 million to figure out if it has a market. Today, maybe that same company could do it under $1-2 million. The capital, as a whole, may make the same or better returns, but yeah, if the failures don’t cost a half of what they used to, you are actually saving money, it is a more efficient market.” More experiments inevitably means more failures on an absolute basis. In addition, as the rate of business experimentation rises there will inevitably be an increase in the number of poseurs trying to create new businesses and that will increase failure rates. A lower overall success rate caused by an increase in the number of experiments is a positive tradeoff overall since society benefits from the increased level of innovation. This net benefit for society is created even though most experiments fail and some experiments are being conducted on the margin by poseurs who have little or no idea what they are doing. Success is found by any given business via negativa. For this reason, some failure is essential to the capitalist process since it is what fuels success. What the collapse of the cost of running business experiments has done is radically increased the pace of the discovery process that creates innovation. Because the creative destruction process is now operating as if it has taken steroids, the rate at which profit is turned into consumer surplus has never been greater, especially in the technology sector.
A real economy is messy and there is lots of failure. Failure is in fact an essential part of the process of creating innovation and a healthy economy. A fantasy economy in which fully informed and perfectly rational agents interact with perfect efficiency is believed to reflect reality only by a few people suffering from extreme forms of psychological denial. A child of ten knows humans are not perfectly informed and rational. Failure is literally everywhere and is essential to making capitalism work given the economy is a nest of complex adaptive systems.
Despite the fact that experimentation is beneficial, it is interesting to think about whether there can be too much failure. Does the existence of too many startups in a given area like a city lower the overall benefit by diluting the talent available for startups of higher quality? I have argued previously that there is a Goldilocks “just right” level of startups that is unique for each city. That city A has a huge number of startups should not be the test of success but rather: what is the outcome of that activity? As many startups as are created in the Bay Area of California may not be optimal for a city like Seattle or New York. How big is the city? How much supporting infrastructure exists? What alternative employment opportunities exist? Does the city’s culture reward or punish failure? Is the culture in that city most supportive of missionaries or mercenaries? Does the city have a major research university?
At some level, the ability of an economy to grow is a brake on the overall level of success that can be created in a given time frame like a year or decade. To illustrate, Warren Buffett made a comment once that could be applied to unicorns waiting to go public as a group:
“Think about a [bunch of unicorns with a combined private valuation of] $500 billion. To justify paying this price, [they] would have to earn $50 billion every year until perpetuity, assuming a 10% discount rate. And if the [businesses don’t] begin this payout for a year, the figure rises to $55 billion annually, and if you wait three years, $66.5 billion. Think about how many businesses today earn $50 billion, or $40 billion, or $30 billion. It would require a rather extraordinary change in profitability to justify that [valuation].”
Regarding the amount that has been invested in breakthrough innovation, Mattermark calculates: “we found that 75% of the approximately $108 billion that investors deployed into these [software] companies is still locked up in private coffers.”
I am conformable in predicting that I do not know who the outcome will be. Lots of unicorns may go public in the next year or two with an IPO, but perhaps in many cases only with a down round. That company A can’t go public at a valuation greater than $1 billion does not mean that it can’t go public at some price. Or maybe large numbers of unicorns will go public at flat or even higher levels. Of course some unicorns may be bought by large already profitable firms for defensive reasons, which allows financial exits by startups to exceed aggregate GDP growth by some amount. We will see what happens soon enough. That’s part of the uncertainty and fun of this process.
3. “The funding market is so bifurcated because outcomes are so bifurcated.” “ Startup outcomes fall on a power law distribution. So startup financings look the same way. You’re un-fundable until you’re oversubscribed.” The financial returns from a tiny number of startups have always driven venture capital returns due the inherent convexity of technological innovation. This distribution of financial returns inevitably takes the form of a power law for the best venture capital firms. As just one example to illustrate, the ten biggest exits from 2014 look like this:
2005:
Ravikant explains: “The nature of the markets have in many cases become more consumerized. People have caught on to how network effects work.” Network effects are increasingly driving financial success which phenomenon causes investors to put more investment into a smaller number of startups and firms that are deemed to have momentum. As a result, some firms can easily raise billions of dollars if investors see potential network effects, while other firms can’t raise even small amounts of capital.
I recently had a discussion on Twitter with Dave McClure (who I respect a lot) about the power laws in venture capital and my view is that a firm like 500 Startups that makes over 150-200 investments a year has a fundamentally different portfolio that a venture capital firm in the top decile of historical performance that makes < 10 investments a year. On recent survey survey reported that the average venture capital firm looks at 400 business a year and invests in only 5 businesses on average during that year. McClure says that his financial return data does not look like a power law, but my view is that this does not mean other venture capital firms that embrace more convexity and make fewer investments do not have a return distribution that looks like a power law. All the data I have seen over the years indicates that top venture capital firms have a return distribution that reflect a power law. McClure is making ten times as many investments a year. My thesis is that these views can be reconciled by looking at the composition of the two different portfolios. When financial exits reflect a power law and the head of the distribution is only a handful of companies, investing at a very high rate in any given year inevitably means you are in the tail for most investments. The 500 Startups portfolio is fundamentally different.
This is from a blog post by Jerry Neumann linked to in the notes below:
“At a given alpha, the more investments you make, the better, because your mean return multiple increases with the number of investments, as does the likeliest highest multiple. Dave McClure makes this case::
‘Most VC funds are far too concentrated in a small number (<20–40) of companies. The industry would be better served by doubling or tripling the average # of investments in a portfolio, particularly for early-stage investors where startup attrition is even greater. If unicorns happen only 1–2% of the time, it logically follows that portfolio size should include a minimum of 50–100+ companies in order to have a reasonable shot at capturing these elusive and mythical creatures.’
Peter Thiel flatly contradicts this:
‘Given a big power law distribution, you want to be fairly concentrated. If you invest in 100 companies to try and cover your bases through volume, there’s probably sloppy thinking somewhere. There just aren’t that many businesses that you can have the requisite high degree of conviction about.’
McClure believes he can find hundreds of companies with high enough growth to maintain his requisite alpha. Thiel thinks this is not possible. Venture capitalists have always faced this tension: the average growth rate of all small businesses in the US is closer to 7.5% than 30%. The pool of companies that can grow fast enough is limited.”
My view is that both McClure’s and Thiel’s approaches to venture capital can be successful just as Warren Buffett and Ray Dalio can be successful with different investing approaches. The financial out performance can come from different sources of mispricing for each of the two approaches. One interesting point: would VC firms in the lower two quartiles improve performance by investing more like 500 Startups? The old joke is that all VC firms are in the top quartile, or at least that is what they say to their limited partners.
The discussion in this blog post to this point begs a few questions: Why are financial outcomes in venture capital and business in general so bifurcated? What explains the power laws? This topic is worthy of its own blog post but in my view the simplest explanation is that there are many forms of feedback in a world that is a nest of adaptive complex systems and that feedback in all its forms creates the power law distributions. I have always loved this statement by Richard Feynman: “Imagine how much harder physics would be if electrons had feelings!” Duncan Watts describes how this manifests itself in this way: “individuals do not make decisions independently, but rather are influenced by behavior of others.” Humans are not like electrons- they have emotions, do things like herd and otherwise copy each other. That behavior can cause success and failure to feed back on itself, which produces outcomes that under the right conditions reflect a power law.
4.“The Internet is very efficiently arbitraged. Anything you can think of has been thought of and tried. The only way you’re going to find something is if you stick to it, at an irrational level and try a whole bunch of things.” The number of business experiments being conducted is increasing so quickly that the more obvious opportunity spaces for entrepreneurs are being exhausted with unprecedented efficiency and speed. There is no place to hide from the relentless pace of competition if the business person’s plan is to do something conventional. This places a premium on genuine product breakthroughs, often resulting from original research and development. In other words, convexity as a source of out performance is more important than ever.
5. “You get paid for being right first, and to be first, you can’t wait for consensus.” An investor can’t beat the market if they are the market. This is the point made so well by investors like Howard Marks. It is mathematically provable that being right will not lead to outperformance if the consensus forecast is also right. Howard Marks puts is this way: “To achieve superior investment results, your insight into value has to be superior. Thus you must learn things others don’t, see things differently or do a better job of analyzing them – ideally all three.” Ravikant is saying that you also need to do it first. If you are not first, the bet will probably become consensus before you can capitalize on that bet.
6. “The market has to be huge because everyone makes mistakes. You never quite get it right the first time. Companies that don’t do giant pivots are always doing micro pivots. You need a large enough market that you can pivot in and you still have a customer base.” The convexity of a business investment (potentially massive upside and small potential downside) has never been so important. Having a huge addressable market increases the convexity of the potential financial outcome since it increases optionality. Small addressable markets provide entrepreneurs with fewer options and are not as likely to be convex.
7. “[A $1 billion seed fund would] destroy the entire market and put prices up 20% overnight.” “We don’t think we can allocate that kind of capital without distorting the market.” “Even the $400 million [we raised] will be spread out over six to eight years. Maybe the first year, we’ll deploy $20 or 30 million as we figure out the model, and then scale it out.” The market for venture capital is top down constrained by the potential for financial exits. This scalability problem in the venture capital business that Fred Wilson and others have written about means that in a country like the US only about 800 startups per quarter raised venture capital for the first time. If seed stage valuations get too high, the market can become distorted since it can become hard to raise funds in later rounds. Angel and seed considered together is only slightly higher as this chart indicates:
As I noted above the economy is limited in its ability to generate exists at some level. Yes it can rise more over time with the right environment in place but number of financial exits doesn’t just suddenly jump 10x when overall GDP is increasing only at low single digit levels. One thing that may take the possible exit total higher is that some startup firms are purchased even though only consumers benefit from their existence and there is no net increase in GDP. This creates a weird asymmetry where startups can win without ever making a profit in some cases. The idea that sometimes only consumers benefit from innovation escapes many people. Some innovations have a moat and create profit and some innovations do not have a moat and create no profit.
8. “It’s just as hard to build a large company as it is a small company, so you might as well build a big company. It’s roughly the same effort.” This is yet another point related to convexity. The bigger the upside the greater the convexity.
9. “I use Warren Buffett’s criteria for assessing the Team: Intelligence, Integrity and Energy. You want someone who is really smart, very hard working and trustworthy. A lot of people forget the integrity part, because if you don’t have that, then you have a really hard working crook and they will find a way to cheat you.” “Intelligence and energy are easier to measure. Integrity is the most important factor.” An honest colleague or partners with integrity increases the convexity of an investment since trust gives you more options. Decisions can get made faster and with greater confidence. The work is fun. Life is better. People have actually does research to prove that people who life in high trust societies are happier. A lot of research has also been done on how trust is a key enabler of an economy. The studies show that “high-trust societies achieve higher economic growth due to lower transaction costs. Since trust protects property and contractual rights, it is not necessary to divert resources from production to protection.” These same ideas about the value of trust are fractal and exist at a company and personal level.
10. “Companies only fail for two reasons: The founder gives up or they run out money.” “Don’t be proud. Get the cash wherever you can. Cash is everything.” “Raise twice as much and make it last four times as long. Pretend that you don’t have the money in the bank, run lean. Assuming your unit economics are at least breakeven, keep your headcount low, raise money and stay in it for the long haul. It takes a decade to build a great company. There’s no shortcuts.” The only unforgivable sin in business is to run out of cash. What does cash give a business? Options. What do options create? Convexity! By now you have probably figured out that convexity is everywhere if you know how to look. “There is a whole set of companies that are not financeable by the venture community; service businesses, markets that are heavily played out, if you are fighting a war that has already been won…you better have some really core differentiation and traction. [Other disqualifiers include] not enough technical people on the team…if you are completely out of market… pre-launch companies tend to not do well…teams that have no credibility. The companies that fail to raise funding are the ones who use too many words and too few actions. Your biography is a record of your past actions. Your execution on your current business is a record of your current actions. Talking about what you are going to do in the future is almost pointless. Talking about what you can become is almost pointless. People want evidence. There is a lot of talk out there.” Most startups will not be able to raise venture capital. That is not the end of the world for those businesses. There are many ways to finance a business that do not involve venture capital.
11. “When building a startup, microeconomics is fundamental, macroeconomics is entertainment.” “Getting real traction is hard. Raising millions of dollars is hard. Building a sustainable, long-term company is hard. Your pre-traction company has not achieved product/market fit and so it has a hard time hiring.” “There isn’t a shortage of developers and designers. There’s a surplus of founders.” Understanding microeconomics is essential if you want to be successful in business. The distinction between these two types of economics was explained at the 2016 Berkshire shareholder meeting by Charlie Munger who said: “Microeconomics is what we do, macro is what we have to put up with.” Munger and Buffett elaborated on that point during this interchange:
BUFFETT: “Charlie and I read a lot and we’re interested in economic matters and political matters for that matter and so we’re familiar with almost all of that macroeconomic factors. That doesn’t mean we know where they’re going to lead. For example, where zero interest rates are going to lead.”
MUNGER: “We pay a lot of attention to microeconomic factors. If you’re talking about macro we don’t know anything more than anybody else.”
BUFFETT: “Yes, he summed it up. In terms of the businesses we buy – and when we buy stocks we look at it as buying businesses, so they’re very similar decisions – we try to know all or as many as we can know of the microeconomic factors. I like looking at the details of a business whether we buy it or not. I just find it interesting to study the species.”
MUNGER: “Well there hardly could be anything more important that the microeconomics, that is business. Business and microeconomics are sort of the same term.”
12. “Desire is a contract you make with yourself to be unhappy until you get what you want.” “Seems like too many people, public and private sector, are making a living slicing the pie rather than baking it.” “Figure out what you’re good at and start helping other people with it; give it away. Pay it forward. Karma sort of works because people are very consistent. On a long enough timescale, you will attract what you project.” “If you are young, one of the best things you can do is build a brand around yourself.” Yeah, this last set of quotes is a grab bag of ideas but there is a lot to grab in what he says! This post is already at 4,000 words, so I will let his words speak for themselves.
Notes:
Anatomy of a Fundable Startup https://vimeo.com/25392719
Business Insider Interview: http://www.businessinsider.com/interview-naval-ravikant-co-founder-angellist-and-co-maintainer-venture-hacks-2011-8
Nextweb: http://thenextweb.com/entrepreneur/2011/02/22/naval-ravikant-talks-in-depth-on-twitter-bubbles-new-york-and-start-fund-interview-part-2/#gref
Pando: https://pando.com/2015/05/08/naval-ravikant-to-vcs-you-can-lie-to-your-lps-but-dont-lie-to-yourselves/
Profile: http://dartmouthalumnimagazine.com/articles/avenging-angel
SJBJ: http://www.bizjournals.com/sanjose/blog/techflash/2015/06/angellist-chief-ravikant-its-a-bubble-but-not-just.html
CNBC: http://www.cnbc.com/2015/10/13/
CB Insights: http://www.forbes.com/sites/niallmccarthy/2015/01/21/the-10-biggest-u-s-venture-capital-exits-of-2014-infographic/#660ae6322917
WSO http://www.wallstreetoasis.com/blog/the-current-state-of-startups-from-naval-ravikant
Power Laws in Venture http://reactionwheel.net/2015/06/power-laws-in-venture.html
Embracing Complexity https://hbr.org/2011/09/embracing-complexity
Firm deaths and births: http://www.factcheck.org/2015/11/a-gop-talking-point-turned-false/
Mattermark: https://mattermark.com/75-capital-invested-unicorns-still-locked-private-coffers/?utm_campaign=Mattermark%20Daily&utm_source=hs_email&utm_medium=email&utm_content=33087858&_hsenc=p2ANqtz-_oxddUidtX8i2-IRVA-CTZo5zp4YHD4joHvSJxGTvMmW_BIsyQ-oOi0eEEA3MTIfMS4Ur31j6mG9d1UDAHF9tJgcgkNw&_hsmi=33087858