Betting on the Ponies: non-Unicorn Investing

Twenty fucking five to one
My gambling days are done
I bet on a horse called the Bottle of Smoke
And my horse won

– The Pogues

I have decided to angel invest. Any advice?

I spent some time decades ago in the horse-racing world, as a guest of someone who was actually in the horse-racing world. Two things: (1) it’s not as glamorous as it sounds, and (2) everyone has a system. Everyone. And, as you might expect if you thought about it for a minute, you can always tell the people who know what they’re doing: they’re the ones that don’t tell you about their system.

Everybody else likes to talk non-stop about their system and the sophisticated statistical modeling they’ve done on it on their laptop using Excel. Most of them have never ridden a horse. Most of them have not looked–really looked–at enough horses to know what a great horse looks like. In fact, they’ve probably only ever seen one or two great horses, because truly great horses are exceedingly rare.

You’re rambling, old man, I don’t care about horses.

Well, me neither. But I care about betting systems. And horse-racing is the ecosystem with the worst betting systems in the world. Ridiculous statistical models built by people who don’t understand statistics or models used as psuedo-rationalizations for rules of thumb and rabbit’s feet. In the end, the bets of 95% of bettors are a frantic attempt to avoid betting on any horse that resembles in any way any nag they’ve ever lost on before. But they need a win, they have to go home with at least one win if they ever want to be allowed to come back. So they spend all day making small, stupid bets, waiting to make the big bet on the sure thing, once it’s obvious it’s a sure thing. Yeah, it’s 1:10 odds because every other bettor at the track also knows it’s a sure thing, but that sure thing isn’t a horse, it’s a…


Yes, that’s right. A freakin unicorn.

As far as I can tell Aileen Lee popularized this term in a Techcrunch article last year. Some days I wonder why she hates us, and some days I thank the gods that she didn’t decide to call them Princesses, or worse. But there was in her analysis the rationale for The System as well as the reason for its absurdity.

Aileen looked at venture exits over the previous ten years, and found 39 companies that had been valued at $1 billion or more in the public or private markets (that number seems to have gone up a bit in the eight months since she wrote the

.) One in every 1,538 venture-backed companies passed this threshold. That’s not a lot. It took more than seven years on average from founding to exit. Also:

Facebook…accounts for almost half of the $260 billion aggregate value of the companies on our list…The 1990s gave birth to Google, currently worth nearly 3x Facebook; and Amazon, worth ~ $160 billion. The 1980’s: Cisco. The 1970s: Apple (currently the most valuable company in the world), Oracle, and Microsoft; and Intel was founded in the 1960s.

In other words, the returns on venture investments resemble a power law, where the most successful companies account for the bulk of dollars returned. If your venture fund didn’t invest in one of the most successful companies of your era you were not one of the top venture funds. You have two options as a venture fund, if you want to attract and retain investors in your fund: have consistently good returns year-in and year-out, or try to invest in unicorns.

What system would you choose if you’re a venture investor? Consistently good returns sounds like the better strategy, but it’s easier said than done. A friend who was a Wall Street trader made his bank hundreds of millions of dollars each year for seven years. In the eighth year he lost a couple of million dollars and was fired. Investors in investors believe in a hot hand and can be less than forgiving. It’s almost impossible to never have a bad year in venture.

So you try to invest in unicorns. The easy way to do this is to have the expertise, the network, and the reputation to be the one the founder chooses after it’s already obvious that they have a shot at the gold ring. It’s like betting on the sure thing but getting reasonable odds. In these cases the founders have their pick of VCs and they often pick the VCs who have backed unicorns before. Everyone believes in the hot hand. That’s why you see aggressive newer VC firms paying high prices to get into the late rounds of rocketship companies: once the unicorn goes platinum, you get to tell people you were an investor. No one asks if you made 2x or 200x. I’ve seen venture firms showing companies on their portfolio page when they bought the stock after the IPO1.

The less easy way is to index. That is, if you invest in enough good companies, one of them is bound to be a unicorn. Here, I’ll lay it out. One company in 1,538 venture-funded companies becomes a unicorn. If you invest in 1,538 companies you have a 60% chance that at least one of them is a unicorn. You’ll only get index returns, of course, which puts you square in the middle of the pack, but you can always point to your unicorn.

These are the two main VC strategies: (1) have a reputation of being the go-to investor in a certain type of company so you get first shot at investing in companies that are more likely to be unicorns; and (2) invest in enough companies that you have a decent probability of being an investor in the next unicorn.

There. Fixed your investing strategy for you.

Of course, being the go-to investor in any category of company is pretty hard. You’re competing with VCs who spend a lot of time and money on burnishing their brands. You’re competing with investors who built and sold the iconic last-generation company in that category. You’re competing with people who spend all day every day trying to make themselves more famous.

Investing in hundreds of companies is no cake-walk either. First off, you have to have the time and organizational skills to meet thousands of companies, the patience to cull them down, the fortitude to read all those legal documents, and, not least, the money to write the minimum-check-size check for all of them. Since the minimum-check-size is at least $35k, that means to just get 100 companies in the portfolio you need to lay out some $3.5 million. And with only 100 companies, your chances of getting a unicorn are between 6% and 7%. If you up that to 500 companies, your odds are 27%-28%. That would cost $17.5 million.

If either of these strategies is available to you, read no further, just keep doing what you’re doing.

Yes, well, I can do better than random, because I know what a good company looks like. So I don’t need to invest in 500 companies to get a unicorn.

Oh, you think you’re a picker. Those 1,538 companies that your unicorn is just one of? All of those were deals backed by professional VCs, whose job it is to think about how to make the right picks.

Well…VCs. They invested in Color.

And Twitter, and Facebook, and Uber. They all seemed like pretty stupid ideas. Until they weren’t.

Look, here’s what I know, after 17 years of investing and 13 years of lie-awake-at-night-thinking-neurotically-about-my-mistakes soul-searching: picking is bullshit. No one can pick.

What about Sequoia, or Union Square Ventures, or Andreesen Horowitz? Can’t those people pick? They probably can, a little bit. But that’s my point. Imagine going to a bar at 2am on a Saturday night and meeting a financial analyst who spends eight hours a week in that bar and 156 hours a week trying to figure out if AAPL is going to trade up or down over the next year. What do you think he thinks when you tell him you’re buying the stock because you really like Apple’s wearables strategy? He thinks you’re an idiot. He has to put up with armchair quarterbacking every time he glances away from his Bloomberg. He traded on that news 30ms after it was announced. He traded on the probability of that news four months beforehand. And even then, he probably only moved his odds from 50/50 to 51/49. That matters when you have half a billion dollars in the stock; it doesn’t matter to you. He spends his life doing this. He talks to Luca Maestri on the telephone, on a freakin landline probably. He has people stand on line at the Apple Store for him to get the new products first and then take them apart and test them for rare metals so he can corner the world supply of indium. And what do you have? You have a good feeling about wearables. Who do you think is selling the AAPL stock you’re buying? He is. Why is he selling it to you? Because you’re the greater fool.

Don’t compete with that guy.

What about the angels in Twitter or Uber? Can’t those people pick? How about that monkey that typed Shakespeare? Can’t that monkey write? Recognize survivorship bias. (That doesn’t mean those angels did it wrong, they probably did it pretty all right, but it wasn’t picking they did right.)

What about Paul Graham, who said you shouldn’t spend your time thinking about price, you should spend it trying to get into the best companies? Doesn’t that imply that he thinks you can pick?

You mean Paul Graham of the incubator that backed 630 companies, each at a $225,000 pre-money? Oh, ok. If you convince 630 companies started by Stanford engineering grads to take a $225,000 pre-money, then you can go telling other people they should be pickers too.

But don’t confuse this with reality. You think you’re an amazing picker. But that’s mainly because you think picking is ridiculously easy. But for that to be true it would mean that all the other VCs out there can’t pick because they’re idiots. And that, my friend, is just not true. VCs may not know lots of things you know, they may not always be valuable as business partners, but they do know one thing very well and that is how to invest in startups. And that’s what we’re talking about.

So you’re saying it’s all just luck.

No. Go back and read it again. All I said was that picking is bullshit. Investors do a lot more than pick. But let’s drill down on picking first, because there’s more to picking than just picking a company. When you land in Vegas you can head to New York-New York to play the slots or you can head to Casino Royale to play craps. You can’t pick the outcomes in either game, but you’ll come out a lot better in the latter case. Pick the game with the best odds over the long run.

So you have a system.

Yes, I have a system. The system is not guaranteed to improve your outcomes. But I guarantee it will improve your odds.

I thought you said that you could tell the people with working systems because they didn’t talk about their systems.

Touché. But this isn’t parimutuel. I’m not betting against you. In fact, you working with a bad system hurts me.

I was working with a lawyer on a deal 15 years ago. The lawyer on the other side of the deal was newly minted. I said to my lawyer “we should be able to get pretty good terms, since he’s new.” My lawyer, who was one of the top ten deal lawyers in the world, according to American Lawyer magazine, sighed and said “we’ll get the exact same terms we would have gotten if he was an experienced lawyer, but it will take three times as long and he’ll convince his clients you’re trying to screw them.” That was exactly what happened. And that’s what happens in deals I’m in with newly minted angels who think it’s all so easy.

There are enough angels out there who will make my deals three times as painful while convincing founders that every venture investor is blood-sucking scum. I don’t want you to be one.

Well, that was mildly insulting.

Let’s move on.

Here’s my system for investing when you don’t plan on having 500 portfolio companies and you’re not the person that every entrepreneur thinks of when asked “who would be your dream investor?” This is not an avoid-the-unicorn strategy; I think several of my portfolio companies could have $1bn+ exits. But my strategy is not predicated on that. It’s a guide to getting great returns when targeting unicorns will not work.

1. Don’t look for unicorns

Michael Moritz once said “I rarely think about big themes. The business is like bird spotting. I don’t try to pick out the flock. Each one is different, and I try to find an interestingly complected bird in a flock rather than try to make an observation about an entire flock.”2

Sequoia has a reputation not for turning lead into gold, but for turning gold into platinum. They can get a great company an awesome outcome. This means that any entrepreneur who has a great company, who can raise money from any VC on earth, raises it from Sequoia. Pretty sweet for them, eh?

You are not Sequoia. Don’t take Michael Moritz’s advice. It will not work for you.

You should look for a flock. You should invest in markets you know. The two things you need to check off before you invest in any company, because they are the key difference between viable and non-viable, are (1) does the problem the company is solving cause their prospective customers great pain? and (2) can the team build the proposed solution with reasonable time and money?

People screw up on how painful certain problems really are all the time. They mistake problems that are slightly painful to a lot of people with problems that are excruciatingly painful to a few people. Solving the former is not a VC-backable company; the latter might be, depending on how many those few people are and how much they have to spend. Knowing which type of problem you’re hearing about is hard. The only way to know it is to know the customers really well. Eat with them, drink with them, work with them. Know what they love, know what they hate, know what will get them to get the checkbook out. Know what will get the to get the checkbook out today. Be the customer, Danny. If you are, you will not invest too early, you will not invest too late, and, most importantly, you will not invest in companies that launch products no one wants.

Knowing the customers is not easy. It is time consuming. And I’m not going to give advice on it here. But knowing the customers is an incredibly valuable asset, and if you have it you should husband it for all it’s worth. That means investing in many companies that address that same group of customers.

The second one, being technically viable, seems like it’s rarely screwed up by investors. How many companies are funded that can’t build their proposed technology? The straightforward fact of being able to build a technology or not is pretty easy to determine beforehand, so companies that can’t build their product are often the victims of not being able to hire enough developers, not of trying to build something unbuildable. This sort of technical viability, whether the product can be built or not, can be due diligenced.

But in a larger sense, building the product is often the least of it. Getting the product to work means effectively interconnecting it into the ecosystem. If you were trying to build a better bank, say, you had best know not just what you need to build in-house, but what systems you have to connect to and what they can and can’t do. Knowing this often requires a pretty in-depth knowledge of the ecosystem as a whole, who the players are, what their products do, and what it will take to get them to do what you need. Knowing this is, again, a valuable asset. Using it just once is a waste.

2. An entire flock

When Edison invented a viable light bulb he knew that for the viable product to become the mainstay of a viable business, the country needed to be electrified. He knew his customers did not have electricity at home. Electrification was more than a product, it was a whole ecosystem of products. People like Edison did not build companies around single products and then forsake all they learned for some other technological field, they built many companies and many products based around a fundamental secular technological shift. As an investor, you should do the same.

We should be so lucky to see another tech movement like electricity or mass manufacturing or the internet again in our lifetimes. The really big ones come once every couple of generations. But the big-but-not-enormous ones come all the time. Mobile, programmatic advertising, big data, and 3d printing are all examples of technological movements that are big enough to sustain an ecosystem of companies; that is, many groups of companies where the groups are not competing with each other, even though the companies in each group compete with each other.

A few advantages to this. These companies may sell to the same customers, or to each other, and so if you know the customers of one, you know the customers of many. The people who run all these companies will know each other, at least by reputation. The tech advances of one company will affect the other companies, the industry news will be filtered through a few sources, the business models will tend to depend on similar KPIs, etc. In other words, the knowledge you need to effectively invest in these companies is all overlapping: if you invest in one company in the sector you will be much more knowledgeable about other companies in the sector (even when they don’t compete); you will be much more able to help your investments with advice about their industry, their business, their competition, future fundraising, and potential exit opportunities; you will have a higher profile in that industry; and you will be much more likely to see promising entrepreneurs in that sector because you will be known to be interested in it.

In addition, your support of any company in the sector will help the sector as a whole. The biggest challenge for any new sector is not competition from within the sector, it’s getting the sector itself taken seriously enough to compete with whatever it’s replacing. Your support of a company in the flock increases the value of the flock, including any other investments you have in the flock.

3. Let people know you’re investing

This is good advice for pretty much everyone doing anything. People don’t often show up with opportunities unless they think you’d be interested in those opportunities, because people don’t like to waste their own time. Let people know what opportunities you’d be interested in. Early on in my angel investing I’d ask founders, as a favor, if they’d include me in the press release about the financing. It wasn’t an ego thing, it was marketing. Other entrepreneurs and VCs read press releases or the news stories based on them and they are the best source of dealflow.

If you’re investing in a sector, find some way to talk about the sector in public. Get the news out about the technical innovation that’s going on. And always mention that you’re an investor in the sector. Write about the sector, on a blog or in the trade press. I wrote pretty frequently for AdExchanger, the programmatic adtech industry’s trade magazine. This got my opinions about where the industry was going in front of some pretty influential people, got my name out, and let me support AdExchanger, who I felt was helping the programmatic industry get taken seriously.

At some point the fact that you invest in a sector becomes well-known enough to be self-sustaining through word of mouth.

4. Things other people are confused by

It’s sort of pointless to read the New York Times restaurant reviews. Once a place gets a good review it’s impossible for you or me to get in. You need to find things before everyone else knows about them. If you’re in a room with more than 20 people, you’re in the wrong room.

With a few noticeable exceptions (Foundry Group, for instance) VCs like to invest in things that other VCs like. This makes some sense: to the degree that new industries can be analyzed, VCs who analyze them will generally come to the same conclusions. It also reduces future financing risk. This can seem like herd behavior, but it’s not really. Not any more than a line of cars is herd behavior. They’re not following each other, they’re following the road.

The problem is, if everyone wants to invest in the same types of companies it makes it hard for us angels to get into the best companies and put enough money to work in them. (Unless it’s a sector you’ve already been investing in and it suddenly gets hot and you’re the go-to person because you know so much. Then you’re in hog heaven.) Sometimes too many competing companies get funded, making it hard for even the best of the bunch because the customers are confused by the noise.

You need to find a sector that not every VC wants to invest in. But it needs to be for the right reasons. If VCs tell you they don’t like a sector because they don’t know how big the market can be or the sector is too crowded, that may be an interesting sector. If they tell you they don’t like the sector because there is no customer pain or the market is not big enough, you might want to do some more due diligence. The key difference is whether the VC doesn’t like the sector because the facts are against it or because the facts are unknown. VCs are good at analyzing situations, they like reading research and sizing markets and talking to customers. VCs hate going off the edge of the map.

But some of the best investments of all time were in companies where the size of the ultimate market was completely unknown: Apple, Intel, Google. The whole Christensen-ian idea of disruption rests on the creation of a completely new market, after all. If you can size the market beforehand, then the company isn’t disruptive, by definition3.

VCs know this, of course, but most of them have as investors people who are certain that the VC is in fact a charlatan and are just looking for some sort of proof so they can stop funding them. Risk is measurable, so rational expected value decisions can be made. Uncertainty is impossible to price. Overcoming uncertainty requires believing in an idea, an arational act and, as such, one that is impossible to defend if it proves to be wrong. Unless the VC has really outstanding returns they need to defend their decision-making every time they raise a fund. Angels don’t. We can be wrong with impunity.

This is not to say that it is the shying-away-from by other investors that makes an investment good. You can’t invest in companies that everyone shies away from and will for an extended period of time. You will need to convince other investors to invest in the syndicate, and you will need to convince follow-on investors (with the help of some traction) when it’s time to raise the next round. If you’re way too early and not able to show progress towards de-risking with the seed money, you’ll be on the hook for the Seed-2.

Knowing whether the skittishness of other investors is a good thing or a bad thing is something I worry about whenever I look at a new sector. Knowing the customer and how quickly they will adopt something new helps, but you’ll also need real conviction to take the plunge.

That’s your advice, put all your eggs in one basket? Really?

As an angel, you are constraint driven. You don’t have enough money, time, or help. The one thing that sets great venture investors apart from competent venture investors is their position as a hub in the informal information trade. Great venture investors know everything that is happening in the venture world. They know how private companies are faring, they know deal terms of financings they weren’t involved in, the know undisclosed exit values before the deals are done, they know revenues and margins of private competitors of their portfolio companies, they know which executives are secretly willing to entertain job offers, they know about a startup’s fundraising before the startup is even started, they know which companies have secretly filed to go public.  They need to know this stuff, and they spend a very large proportion of their time having the right relationships to know it. They spend their money, time, and help on it. They couldn’t be great if they didn’t.

How do you, as an angel, get this information? You need to have a network of people who trust you, and you need to give them a reason to talk to you. You can see how that works for a Michael Moritz, but how could it work for you? What reason do people have to tell you things?

This is where targeting a sector helps. The people with the most information about a specific sector are the founders and venture investors in that sector. The venture investors are easy: the primary thing you can trade with them for information they have is information you have. And while you may have generally poorer sources of information than VCs who are paid to have information, you will have better information about the sector you specialize in because you spend more time in it. A professional VC will know a good amount about ten different sectors; you will know a ton about one sector. This makes you valuable to them. They will talk to you.

Founders have a different sort of information. They have the information that comes from working in, living in the sector 24/7 for years. They know about achievable revenues and margins, they know competitors’ products and strategies, they know which people at other companies are most valuable, they know about companies just starting that have valuable and useful products, etc. They hear from their salespeople, who hear from their clients what other companies’ salespeople are saying about their competitors. They know more about their industry than anyone. They are happy to share this information with you, if you are helping them. This help usually takes the form of advice or introductions. They want advice when the company is just starting and they need to figure out price and product features and how they pitch themselves versus their competition. All of this advice is sector-specific. They want introductions a bit later when they are looking for customers, partnerships, employees, and further funding. All of these things benefit from being sector-specific.

This all makes sense, right? You fit into the network of people because you add value to your part of the network, and you add value because you’re thickly connected to your part. This is best achieved by picking a very specific part of the network to specialize in.

Why does sector investing deliver better returns? Because if you really, deeply know a sector, you not only see the best deals, you might actually see all the deals. You will know all the people in the sector, and be able to judge who knows what they’re doing, who is truly visionary, and who is just a good salesman. You will know all the VCs who become interested in the sector and which ones are willing to make the bets that require some belief. Because you’re an expert, your investment in a company will be considered social proof. You will give good advice to founders. You will be contacted by the media to find out what’s going on in the sector and they will tell you what they have learned from others about the sector. You will know who is looking at buying companies, who is not looking at buying, and which buyers are willing to pay what. You will know what is working and what is not so that if a company has to pivot, you will know where to pivot to. You will be able to do what every investor promises: add value.

So why don’t all venture capitalists do this?

Some do. Many venture firms have a “thesis” and some even stick to it. The thesis is sometimes the sector they specialize in. Look at Union Square Ventures (“large networks of engaged users…”) or IA Ventures (“big data.”)4 Other firms are conglomerations of partners, each of whom specializes in a sector or two. These firms mitigate the risk to their partners by diversifying.

And this is the rub with this model. You are putting your eggs in one basket. When my portfolio was primarily adtech companies, someone asked me what would happen if the government banned third-party cookies. The answer was, of course, that I’d be screwed. But even though this was a single point of failure, it was a knowable risk: I could put a weight on it. And this risk was far outweighed by the benefits of knowing the sector the way I did. I put the probability of the government passing legislation at less than 10%, and I felt that knowledge of my sector gave me a much larger advantage than that.

And then there’s the gambler’s ruin. Even with a positive expected value for each bet, you can still go broke if your bet size is too big. You have no less than zero, but the market has an effectively infinite initial stake. The classic techniques to ward this off apply. Some portfolio management tips:

  • Keep a standard bet size. No matter how much I like the company my initial investment is always the same5 Don’t increase your bet size because you’re feeling flush after a win;
  • Re-up in your winners. Almost no other game lets you do this, take advantage of it. Reserve money for follow-on investments: they may not have the same cash-on-cash returns as seed investments, but because you’ve been helping the entrepreneur, you have a much better idea of whether the follow-on is a good investment or not. My best cash multiple investments have been seed, my best risk-return investments have been Series A, my best IRR has been Series B;
  • Bet size should be small enough to allow you to get to thirty companies or more in your portfolio, assuming you follow-on at an average of 1.5-2.5x. Remember that if you invest seed-stage, you should assume you don’t get the bigger exits for 5-8 years. At least in the sectors I have invested in, the early exits are not that big and the big exits take time;
  • Don’t keep betting a strategy that has stopped working. This is a true of all investment strategies, they only work for a while. The world changes. In venture a thesis usually only works for 3-5 years. Plan to change the sector you specialize in every 3-5 years. My investing in programmatic adtech exposed me to the big data technologies my portfolio companies were using so when seed-stage adtech stopped being as exciting, I started investing in big data companies. Much of my knowledge and network were transferable.

So, does it work?

Oh, right, I’m supposed to credential myself.

I’m probably one of the few people in the world who started with not very much cash (I had a barely-six-figures net worth thirteen years ago and have made negligible earned income over that time) and have managed to make a good living from my gains as an angel investor. I guess that means it’s working, but humility is in order. As a former boss used to say, “money’s not money until it’s cash.” In that sense, I don’t know if it will work. There’s a chance I’ll be wiped out tomorrow because of some macro-event that I can’t control. Then I’ll be sending you my resume, at which you will laugh.

But that’s not what working means, in a betting system. Even with the worst-case scenarios weighted in, my expected value IRR is much, much higher than would have been available to me in other investments, and top-quintile of venture over the same period. And not because I’m some super-genius6 but because I haven’t tried to emulate what venture firms have to do to win, I’ve played my own game. In that sense, it’s definitely working.

All these articles about “How Reid Hoffman Invests” or “The Secret Venture Wisdom of Sequoia Capital” or “Seven Amazing Things Fred Wilson Does Before Breakfast” are worthless to you. They are entertainment. If you go toe-to-toe with Reid Hoffman or Fred Wilson or that Sequoia partner whose name I can never remember, you will lose. Play your game, the game you can win.

  1. Just looked and it seems they removed it. You see some funny things when you run a VC portfolio page tracker

  2. I can’t find the original source for the quote. It’s reprinted all over the place. 

  3. Because someone is certainly going to argue this, I’ll point out that Christensen flat-out says this in The Innovator’s Dilemma: “Markets that don’t exist can’t be analyzed.” Page xxi. 

  4. I don’t think it’s a coincidence that these are both East Coast firms. There’s a certain density of network that makes you viable; if the network itself is less dense, you create density by having a more specific thesis. 

  5. OK, I’ve had a couple of exceptions, but not by much. 

  6. Or are I