This is the second instalment of my ridiculously simplified perspectives. See my piece about startup strategy for the first post in the series. To complete the picture about why a startup strategy follows certain mechanics, we need to understand the other movers and shakers.
While this is a publication about startups, I am mostly reporting from the viewpoint of digital product startups, which are Venture Capital (VC) financed.
To clarify: digital product startups are mostly web- (or mobile-) based services, which are defined by quick iterations (vs. industrial products, which require longer research and development phases before they can be sold on the market). Venture Capital is an asset class (such as real estate or stock), which is defined by investing money directly into companies (ventures) for a share of ownership in those ventures.
This time I want to get a better understanding of the perspective of a Venture Capital Fund (VC). I am applying the same method as before: ridiculously simplifying concepts, so I am able to ask questions.
I will piece together generally available information, to get a rough understanding of the mechanics of a VC fund. We will see that - by applying common sense - there is a defined space for strategies. Understanding the basic rules of the game gives us a feeling of more or less useful moves.
Power Law
The goal of the startup game is to grow big. Real big. Whenever we observe a situation, in which a small number of things have a disproportionate influence (such as a few startups controlling the market; think Amazon controlling online retail and cloud computing, Google search and online advertisement), we can look through the lens of the power-law distribution.
The power-law distribution is a ridiculously versatile mental tool, and it’s unreasonably effective. I recently wrote a piece about the mental multitool, and I can’t believe I missed adding the power-law distribution to the toolkit. So, it looks like I need to write a part 2 of the mental multitool series.
The unreasonable effectiveness comes from the fact that we can find the power-law distribution everywhere. In nature and society. From population distributions, to finance, word frequencies, Meteorology, Psychology, Physics. Once you know what to look out for, it’s hard to not see it everywhere. It’s unreasonable, because it appears in so many unrelated areas. It is surprising that reality seems to follow basic rules (see also the golden ratio), but I believe it is rather the opposite. There are more or less hidden simple and general rules in nature, which converge in observed phenomena, such as the golden ration and the power-law distribution. The power law is even being used to detect forged data. If the manipulated data doesn’t follow the power-law distribution, something might be wrong with it.
The power-law distribution is also called the 80-20 rule, which generally states: 80% of the consequences come from 20% of causes. So, let’s just use the rough 80-20 distribution as our first assumption. So for our VC case this becomes:
A result of this is that the focus of your VC investor will be figuring out whether your venture falls into the 20% category. Strategically it wouldn’t make sense for them to put any more effort and resources into all the other portfolio companies.
To be more precise, they most likely will focus on only ~5% of their portfolio. The reason for this is the betting strategy, which we will focus on next.
Betting on the winning hand
The VC game has a lot of similarities to a poker game (primarily Texas Hold’em No-Limit, because this is the variant I know best). Every player starts with a stack of chips - the fund. The basic strategy is to get out of losing hands quickly and cheaply, and bet high or be all-in on strong hands. The fun part is that a player only has limited information at any point. The same goes for the VC game. And you have to pay (invest) to play.
The poker analogy
For those who haven’t played poker before, I will give a quick overview without going into too much detail. I recommend that you learn playing poker, if you don’t know the rules yet, and play a couple of hands with friends. It is a brilliant game which offers many life lessons. While there is some luck involved, it is ultimately deeply strategic, rewarding consistent strong decision making. And the game trains you to get a better feel for how short-term ups and downs (swings) do not matter in the long run (thousands of hands).
Poker is played in hands (the hands are divided into rounds). The cards the player competes with are also called hands. Depending on the variation, cards are dealt so some information about players’ hands is shared (sometimes there is no information). The players’ hands are ranked, some hands being stronger than others. Independent of the variant, information is always incomplete until hands must be revealed.
Players try to win a hand by either pushing everyone out of the hand, or by having the strongest hand at the reveal. They do this by betting in rounds. The specific rules of betting are dependent on the variant played, as well as how additional information is revealed in the rounds.
The strategic play is done through betting. Players prize their hand, putting a certain amount of chips/money into the pot. Once a bet is placed, it is lost to the player unless they win the whole pot (or a split pot if there is a tie). Other players must match the bets to stay in the game, or they can “fold” and forfeit their claim to the pot. Players can also choose to “check” and not bet anything, or “(re-)raise” betting even more than the previous bet. The rule is: the highest bet must be matched to stay in the game.
As you can see, it is not necessary to have the strongest hand. Risking more than others are willing to bet can also lead to winning the pot if no other player remains.
This is where the analogy falls short. In poker you can only win what others are willing to risk (apart from forced bets in some variants). The winning play is therefore to find a counterpart (or more) with a similarly strong hand, each willing to bet as much as possible.
In the VC game the ventures should create real value. They serve customers and generate revenue. So the pot analogy doesn’t hold anymore. There are - believe it or not - real fundamentals behind (some) startups. So it’s not only about the players betting on the pot, but there is substantial value being added to the pot over time.
Bet sizes
Let’s get to the bet sizes.
Let’s assume - for the sake of argument - the VC will invest up to 500,000 (Dollars or Euros or Chips, it doesn’t matter as long as we use the same unit and keep the dimension) in the seed round. We can now derive ballpark numbers for how the fund will most likely look, and what their expectations for the game will be.
What is a seed round? What is a round? Similar to how the poker game is played in rounds, the startup game from the perspective of a VC is played in financing rounds. In simple terms, these rounds are the only times an investor can make a “play”, i.e. deciding if and how much they want to invest in a venture.
A seed round (for the purpose of this piece) is the first round. In reality, there are much more variations. It depends on the venture, the market, the individual deals, more or fewer rounds, ventures which raise much later, etc. Different VCs will also have different criteria - pre-product, product, pre-revenue, revenue - showing how much the founders have already accomplished. For the VC (as well as the poker player) more information means less risk.
We will just name the rounds seed, and then series-A through series-Z (in principle, but realistically there won’t be that many rounds).
Once we know the bet sizes, we can make educated guesses at how big the stack is (the fund size), what they expect, and how they will make plays.
Our model VC
In this example I am describing a simplified model of a VC. The purpose is to show the underlying mechanics and forces.
The model VC we are looking at will invest in Seed, Series-A, and Series-B. The next rounds are for bigger investors with a different risk profile. Those take promising startups (at this stage rather grown-ups) with a proven business model to the next level.
In the real word things are more complex and dynamic. Every variable can and will be negotiated.
How to win at VC
If we assume the failure rates are consistent, and no VC has magical insights to affect the success rate of their investments (I assume that practically all VCs will fall into the average category). Then the only difference is in how big their winnings will be. An early investor in the next Facebook or Google will make big bank.
So for the sake of argument, let’s assume the above mentioned investor assumes they have average chances of picking the right investments. Their game requires a MinMax strategy where they want to solve for the following:
Invest in a maximum amount of promising ventures early
Ignore the losers
Double-down on few winners
Deploy your bets as quickly as possible (finding investments and placing bets is a time-consuming activity; better to spend the time on 5)
Move every mountain, call every number in your contact list, to make the winners really big
Wait
So the VC needs to maximize 1, while having enough for 3, but making sure the total winnings are significantly larger than the fund.
Let’s do some basic maths.
MinMax
Our next assumption is that 90% of startups fail. It is a commonly cited figure, but most refer to this article on Failory. A good source is hard to come by, since definitions vary. It becomes even murkier if we only want to look at VC funded startups, instead of all small businesses. Not many startups get an investment from a VC.
Further with some wild guesses. So, our model VC will invest in 100 startups, it expects 10 to survive, 7 will pay back the investment, 2 will make some money, 1 will take off. We will start with these assumptions to get some baseline estimates. In the conclusion we will check to see if these assumptions makes sense.
A maximum investment of 500,000 doesn’t mean the VC will always invest 500,000. Similar to the poker game, you will size your bets based on the information you have. As an average seed investment sum, I will assume 300,000.
300,000 x 100 = 30,000,000
This is the minimum stack size to be able to participate in 100 seed rounds. But this is not everything. Our VC also needs to be able to double-down on the winners.
Let’s look at the next rounds. But first we need to understand how the bet sizes change from round to round.
Funding rounds
How are the rounds played in the VC game? It is all done around the “cap table”’ (short for capitalization table), and it shows who owns how much (and what) in the venture.
We will follow the story of a hypothetical venture, which represents the winner in our model VCs stable.
Our hypothetical venture has two founders, each owning 50% of the venture. This is represented by shares, of which both founders own 50,000 each. The number of shares don’t matter. I just picked these numbers for convenience.
The founders decide to raise money. They do this by issuing new shares in a seed round. They seek to raise 600,000 (Dollars, Euros, Chips), and they are willing to sell 20% of their company. This works out to issuing 25,000 new shares at a price of 24.00 per share (100k shares before, now 125k, which means 40% for each founder and 20% new shares; share price is how much they want to raise divided by the number of issued shares 600,000 / 25,000 = 24).
Our model VC invests 300,000 and receives 12,500 shares, the other 12,500 go to other interested parties. Rounds are quite often closed with multiple VCs. Seeing other players being interested in the venture raises confidence and reduces (perceived) risk for the VC.
So now our cap table looks like this.
Since there are now a total of 125,000 shares for the company, and there have been investors paying 24.00 for each share, the company is now valued at 125,000 x 24.00 = 3,000,000.
The venture spends the money on research and development, they are lucky, and start making some money. It all looks promising. Now the venture needs to show that they can provide their services to a larger customer base. They also need to show their business model works and is profitable in the long run.
So they decide to raise 5,000,000 by selling 40,000 shares (which is ~24% of the company). They find a new VC and all agree on a share price of 125.00. Our model VC buys 20,000 shares, and the new investor buys the other 20,000.
The total number of shares is now 165,000. At a price of 125.00 per share, we get a valuation of a nice 20,625,000.
At this stage, the value of the company is still fictitious. But since the investors were willing to pay this price, at least according to some players, this value is justified.
How does it look for our model VC? They paid a total of 2,800,000 (300k in the Seed round, and 2.5M in the Series A) for a 19.7% ownership in the company. Their shares are now worth a little over 4,000,000 (19.7% x 20,625,000 valuation), giving a fantasy profit of 1,200,000 (~43%). This is not too bad. Our model VC believed in the company early, so they participate in the growth of the valuation.
The company does well and is breaking even. The founders have proven that they can grow quickly and be profitable. Since it’s a digital product, they don’t have to build large supply chains. They want to invest in a new software development centre and some overseas offices, beginning with their international expansion. It’s time to bring in the bigger investors. Our Model VC now places their last bet, The following investors will take larger stakes at way higher prices.
For their Series-B, the founders are raising 15,000,000 by selling 60,000 shares (~27%) at a share price of 250.00 (225,000 total shares x 250.00 = 56,250,000 valuation). Our model VC buys 10,000 shares (for 2,500,000), while a new investor buys the remaining 50,000.
Our model VC now invested a total of 5.3M for 18.89% of the company. The current value of their shares is now 10,625,625 (~100%!).
Now the play for our model VC is over. They can now wait to cash out. Our venture will surely raise even more capital before the company goes public. Once it goes public (the Initial Public Offering, IPO), the shares are converted, and the investors (and the founders) can sell them to anyone on the stock market. There could be also a direct sale to an interested party. The buying price will be then split between the investors and founders based on their shares.
Since our model VC doesn’t buy more shares, their ownership percentage will decrease with each successive funding round (this is called “dilution”). We can assume that our model VC will end up with something around 9%. This assumption is shaky. It is hard to say what a specific company will do, how the world changes around them. But let’s use 10% for our final analysis in the conclusion.
Did the VC win?
To be able to see if our VC has a real winning strategy, we first need to estimate how big their stack has to be to play this game.
We started with 30,000,000 in seed (300,000 for 100 ventures).
We assumed that out of the 100 ventures, 10 will survive on average. Our model VC will speak with the founders and monitor their progress, so our VC gains more information over time. A reasonable assumption for our VC would be to rank their investments and place the bets on the best 10. In reality, they don’t invest in 100 ventures at the same time. So there is a lot of uncertainty involved. The VC needs to allocate some of its funds to the Series-A rounds in the beginning, though.
It needs another 3,000,000 (Series-A) for each of the 10 surviving ventures. Our hypothetical venture raised 2,500,000 from our model VC, but this amount depends on the performance of the venture and the individual deals. For our estimation, we will add some buffer. So we have another 30,000,000 added to the fund for our survivors.
Now for the big bets. Our hypothetical venture had a smooth ride. But the real world is brutal. There is no guarantee any of the ventures will make it big (at least 10,000,000,000 valuation big). In our conclusion we will see what our model VC expects. It could also be that our VC would have to finance the majority of the Series-B round. So our VC allocates 10,000,000 for three of the survivors. Which means another 30,000,000 added to the fund.
Let’s round it up to a nice 100,000,000. This is the size of the fund.
Continued…
This essay has become much longer than I anticipated. I really enjoyed organizing my thoughts around this topic.
I will publish a second part soon with the conclusion. Stay tuned.
As always, I am happy to receive feedback - especially since I took the liberty to make wild assumptions and allow inaccuracies.