> In startups, why do the founders receive such a disproportionate percentage of equity?
There is a really straightforward answer to this - equity isn't distributed according to a theory of value and nobody ever claimed it was. I own some equity in the Australian company CSL. It is unfair to claim I've ever added any value whatsoever to the company at any point, so obviously I don't get that equity because of a value theory. Even on the basis that people anticipated my buying of that equity - I bought in because I believed the value had already been created.
Who gets equity isn't based on value, it is based on power and who has it right now. Value is for working out whether you, the valuer, should make a trade or not.
Adam Smith's commentary on wages and the considerations in their assignment is interesting.
In Book 1, Chapter X, "Of wages and profit in the different employments of labor and stock", "Inequalities Arising from the Nature of the Employments Themselves":
The five following are the principal circumstances which, so far as I have been able to observe, make up for a small pecuniary gain in some employments, and counter-balance a great one in others: first, the agreeableness or disagreeableness of the employments themselves; secondly, the easiness and cheapness, or the difficulty and expense of learning them; thirdly, the constancy or inconstancy of employment in them; fourthly, the small or great trust which must be reposed in those who exercise them; and fifthly, the probability or improbability of success in them.
It's an interesting set of factors, one that many economists and markets-advocates seem wholly unaware.
It does not reflect other arguments, notably marginalist analysis, or power dynamics as mentioned elsewhere in this thread. Still, it's worth some consideration, and explicitly references the roles of both risk and training.
All of those are very simple consequences of marginalist dynamics.
It has been a long time since I have read it, but it's nice to see Adam Smith was so close to the answer centuries before people started distorting the problem politically and got all those obviously wrong but mainstream answers.
That's been something of a recurring theme for me in reading Smith.
He's prolux and the language is somewhat alien to a modern reader. He does make errors and has some of his facts wrong; see his advice to the American colonies, and his origin story of money, especially. But he is a keen observer with a shrewd mind, and recognises that fundamentally the purpose of an economic system is the support of the people (and specifically the poorest) and the state itself, and that wealth confers power which itself is distortionary. These messages are suppressed, whilst others given incidentally are ascribed undue significance.
>>It is unfair to claim I've ever added any value whatsoever to the company at any point, so obviously I don't get that equity because of a value theory.
You don't receive any profit, realized or otherwise, unless the company appreciates in value, and if the company appreciates, that means you provided capital to what was an undervalued company, or a company that could produce a positive return on the investment of that capital.
Either way, your investment would likely have a net positive economic effect, and so the profit you received would be earned.
This thinking is almost exactly how most pricing works in derivative markets and is the real breakthrough in the original Black/Scholes/Merton model[1] that has been reproduced many times since. They showed how you could replicate the payoff from a European-style option by hedging using the underlying and a risk-free instrument (under certain unrealistic assumptions, notably constant vol and the ability to trade in continuous time). Their argument was that since the payoff of the two things (holding the option and performing this hedging strategy) was the same, the price must be the same, and that produced the famous Black-Scholes formula, which is a closed form price for the hedge portfolio under those assumptions.
Since then, pricing using a replication portfolio in this way has been a cornerstone of financial maths - the price of a thing and a perfect hedge/replacement for the thing must be the same.
> Since then, pricing using a replication portfolio in this way has been a cornerstone of financial maths - the price of a thing and a perfect hedge/replacement for the thing must be the same.
And the reason for this is arbitrage. If you can perfectly hedge your position in Asset A using Asset B (and vice versa), then you can make a profit by simultaneously buying the lower-priced asset while selling the higher-priced asset[1]. Thus causing the prices of the two assets to converge.
[1] Many of these assets are simply contracts, which means you don’t have to buy the asset before you can sell it: you simply create a new instance of by “writing something on a piece of paper” and sell that.
Is the implication here that if we had a Black/Scholes/Merton method for salaries we would see a readjustment of salaries across the board as people could identify their value as wins above replacement - probably a huge shock to CEOs.
Is it possible that automation provides that arbitrage opportunity?
The gap between the perfect fungibility of two contracts for asset A or B and the hopelessly imperfect fungibility of two employees is too wide for this to hold. (I think.)
Well one interesting implication of your insight here is that the more fungible/replaceable your skills are as an employee, the easier it is for the market to accurately price your labour and the less pricing power you have.
Conversely niche skills would mean a large variance in compensation due to price uncertainty ("I can't just go out and hire an exact replacement for sokoloff!"), and if there is demand for those niche skills it should mean high comp overall. And that's what we see in general in the labour market.
Agree. I think an additional corollary is the wider gap between say $50K/yr and the value possible to create (as roughly capped by the gross margin fraction of sales/employee), the more variability you’ll see.
For a Subway sandwich artist, not only is the production process standardized (making “exact [practical] replacement” possible), but the value is also capped by how many sandwiches are sold on a given shift. No sandwich artist will make 5x as many as another. Rarely will excellence in an individual employee result in a sharp uptick in sales.
For a company that sells online ads, cloud computing, streaming services, or software that cap is so much higher as to be practically irrelevant for many roles.
But ... Amazon for example redesigns i are ouse picking to make employees fungible. I am pretty sure project management is becoming totally fungible - the reporting and chasing and so forth, can be replaced by automation and is becoming so despite best efforts of millions of "mqnagers"
Something intrigues me about this idea.
The film / book "Moneyball" showed how to re-create great players "in the aggregate".
I think that we will soon enter a phase of "MOOP" - measuring our daily activities on a deep basis, phone calls to colleagues and clients, tonality, agreements and actions etc.
And at this point redesigning an organisation to recreate CFOs or salespeople "in the aggregate" becomes ... possible?
And if you can do that you get close to the ability to create that fungible option - and close to valuing the job.
Your first line has some auto-correct line noise in it, but I think I get the gist of your post.
Moneyball to me was a lesson and framework on how to make great teams, not great [individual] players. Focusing on under-valued qualities allowed the A’s to build a pennant-winning team on a low budget, by selecting free-agent players the baseball operations groups at other teams were over-looking/under-valuing.
Coming back to software, I think many of us know the three people we’d go form a startup with and the 50% of people who might be okay but we’d never sign up to start a company with. I don’t know if that is amenable to a Moneyball-style spreadsheet analysis though. If it is, VC firms would pay billions in aggregate for a working system.
Otherwise known as the Shapley value [1], which also takes into account the interaction between participants.
> The Shapley value is the average marginal contribution of a feature value across all possible coalitions. [2]
EDIT: There are efforts to decompose that attribution of value into Synergy, Redundancy and Independence components [3], which I believe could also have meaning the context of business.
There might be some phenomenon he is pointing out here, but his logic in doing so makes no sense. Even his internal logic makes no sense.
He starts out by talking about the value of a product (commodity) determined by the labor theory of value. Then he jumps to how much someone is compensated in salary. Then he jumps to a company's value and the ease at which someone at the helm can raise capital.
In terms of comparing two systems of value, this is jumping all over the place. How value is added to a specific nascent commodity going through an assembly line is a different thing than the ability of one person versus another to raise capital and how that affects a company market cap and is different than what they should be compensated.
People like Eugen von Boehm-Bawerk wrote well-reasoned arguments against the labor theory of value, which might be right or wrong, but at least they have a clear line of reasoning. This piece talks about value but jumps all over the place. There might be an interesting phenomenon being pointed to, but there is no well reasoned argument.
Here's my argument against the labor theory of value: if you spend one hour making mud pies, and I spend one hour making sandwiches, then our products have equal value according to the labor theory of value.
Not really, since you are swapping an easily available commodity ‘mud’ for bread and fillings. You should specify that the value added to the mud would be unequal to the value added to the sandwich ingredients to make the point you want to. Since you begin with different initial conditions you ‘muddy’ the conclusions.
OK, here's another refutation of the labor theory of value. A woodworking crafter with twenty years of experience takes five hours to make a bookshelf. A person who just started woodworking today takes twenty hours to make the same bookshelf, but it is out of square, out of plumb, is wobbly and has several nail holes in it. According to the labor theory of value, the noob bookshelf is worth more because the laborer took twenty hours to make said shelf, instead of the five hours of the efficient crafter.
OK here's my third refutation of the labor theory of value: you and I both have 1,000 hours of experience making bookshelves. You spend 50 hours making a beautiful bookshelf, and you carve scrollwork into the legs of the shelf and beautiful scrollwork into the front of each shelf, and it is level, square, and plumb. I spend 50 hours making a bookshelf, but it has one leg that is much thinner than the rest, and there is no scrollwork because I don't know how to do that, and I didn't really measure anything, I just kind of eyeballed it, so everything is out of plumb and out of square, and the bookshelf is rickety, at best.
According to the labor theory of value, our bookshelves are worth the same amount, because you put 50 hours into making yours, and I put 50 hours into making mine, and we have the same prior experience.
Another follow-up question, then: How did you manage to spend 50 hours making it? Did you sit there, painstakingly, making one leg thinner than the rest for some reason? Did you sleep? Did you spend 20 hours of it browsing HN? In either case, this would likely fall under the category of unproductive labor: https://en.wikipedia.org/wiki/Productive_and_unproductive_la...
There's also the distinction of "socially-necessary labor", which feels (to me) like Marx's parallel to subjective value - the more socially-necessary, the more valuable the labor (a nice table vs. a crappy one).
In any case, the value is not just a measure of "hours put in" as you're making it out to be here; clearly I put more labor into it than you if mine looks/performs better in the same amount of time.
The problem I have with this is that it presupposes that you can change 1 thing and view the difference. What if the only reason a company is successful is because of a unique idea that 3 people came up with together? Are each of them worth 100% of the company- take any of them away and it’d be worth nothing. Any company or endeavour is made up of complex interactions of many different people it’s impossible to differentiate with respect to just one variable.
> The problem I have with this is that it presupposes that you can change 1 thing and view the difference.
You should have other problems with it. Value that would be lost in the absence of a thing is unrelated to the value gained by its presence, because you have more choices than "pay for the thing" or "do without".
The theory of value espoused in this post would tell you that you should pay more for the oxygen you consume than you do for your computer.
(There is some self-contradiction in the post - it defines counterfactual value thus:
> what if pay is determined by asking: “How much would this company be worth without this individual?”
But then it provides an example of something very different:
> Baseball managers use the Counterfactual Theory of Value all the time, using a statistic called “Wins Above Replacement.” It predicts how many more wins a player gives their team, compared to whoever would replace them.)
I am no more willing to pay for the oxygen I breathe than you are. It's not just free, it's also worthless. Paying for it would only make sense in a context like scuba diving, where I'd be in an environment in which finding my own oxygen was difficult.
The cost is very low (it's everywhere), which drags down the price to zero. People do value oxygen a lot though, it's essential to survival. If someone was to take away all the oxygen,we would see how much people value it: they would go to great lengths to obtain some. This is not usually put to the test, of course.
Edit: this question was raised by Adam Smith, known as the diamond-water paradox, or paradox of value:
Ceteris paribus is actually a central assumption of much economic analysis. The Latin is "all things being equal", and the notion is that a single parameter or input is considered at a time.
Labour theory isn't wrong for the reason he states. Causing people to work harder isn't actually a problem with the theory. It's wrong for other reasons.
The moneyball theory of value is not terrible, is actually pretty sound. But this kind of analysis works well in a context like sports because there's only so many players in a team and the team is basically just struggling in a win or lose world against other similar teams. Your wins come from someone else's losses.
The sports metaphor is also applicable because sports are reproducible. It's the same little world over and over again, with the same people. And there's just an enormous amount of stats to back up or reject some hypothesis. Goalkeeper is good or bad? Let's check how many shots he faced and what quality they were.
In the general market this is not the case. The firm only launches its first product once, in a world where you don't get to repeat the conditions. We also do have a history of each contributor's actions. Basically it's very hard to make comparisons.
One thing that is important here is whether the counterfactual is the marginal employee, or nobody.
In the sports roster case, a soccer/baseball team would be in trouble if they lost one of their average-talent players if they didn't have anyone on the pitch but there might be a plentiful supply of low-cost players willing to work for a similar amount.
This means that the counterfactual of 'lose average player and replace with noone' suggests that the player has a very high counterfactual theory of value, but 'lose average player and replace with similar average player' suggests that the counterfactual value is not high.
There might be a star player such that they are not replacable by anyone, or by anyone also able to demand a high salary from an alternative team, and they would have a very high value according to both counterfactuals.
It seems to me that many sports teams operate with the 'similar average player' counterfactual.
I think you could draw analogies to businesses and employee pay.
But all other stock allocations are an arms length transaction; so all you've done is reframe the question to "why do the founders retain so much equity?"
Why dont seed fund demand 16% instead of 8%? Why doesn't the 6th employee demand 5% instead of 3%?
>reframe the question to "why do the founders retain so much equity?"
The founders can only retain as much equity as the marketplace allows. Each side has a threshold to do a business transaction.
>Why dont seed fund demand 16% instead of 8%? Why doesn't the 6th employee demand 5% instead of 3%?
The parties can demand any percentage they want but the ultimate resolution is will the other side agree to it? In other words, the question is, "Why do angels _agree_ to 8% instead of 16%"? Because an offer requiring 16% would be rejected by the founder and lose to other angels only requesting 8%.
E.g., back in 1999 during the dot-com craze, VC Sequoia Capital offered MP3.com (founder Michael Robertson) $10 million for 45% of the company. He said no deal. They later negotiated it down to 20% ... which is in the more reasonable ~15% to ~20% range of other VC deals.
Lesson: Demanding a high 45% so that the founder only retains 55% instead of 80% doesn't automatically mean the founder will say "yes" to the reduced equity. People can still voluntarily choose not to do business with you at all. All "demands" are competing in the marketplace. Can an employee demand a higher 10% equity? Sure. Whether the founders _agree_ to it depends on the marketplace and the employee's particular leverage (e.g. a very rare skill).
I can believe it's useful as an unscaled relative measure but I can't see it having absolute value. It's totally contextually defined. And in the case of McAfee had net negative value over time. Or perhaps Bernie Madoff is a better example.
Essentially: what is the marginal contribution of some factor X. Including the case where X is some individual contributor.
The question Dave Perell asks needs to be augmented by one other: What is the cost, and value, of an alternative input factor?
If X provides 100 units of value at 20 units of cost, and Y provides 90 units of value at 9 units of cost, then the net advantage is to utilise Y (net benefit: 81) rather than X (net benefit: 80).
Of course in the real world, neither benefits nor costs are quite so readily assessed, but the principle remains: you want to account for both the productivity* AND the cost of the input or contributor.
Not really counterfactual theory, but opportunity cost. The cost of retaining the CEO to most of the shareholders (even a consensus of them) is high because they can’t realistically install a new one.
Opportunity cost gets conditioned on the investor. If you’re a maker, or a player with leverage on the board or something, your opportunity cost is different.
This seems like a roundabout way of saying something that seems fairly anodyne: A F500 CEO doesn’t provide 1,000x the value of a rank-and-file employee, they’re just (debatably, IMO) 1,000x more scarce.
This is inspiring, especially if you want to create something good, clearly seeing how bad that part of the world is now. It seems that it gets only worse each year, so opportunities should open up. But then the latter doesn’t feel like it should. Then you realize that your picture is missing a perception of value, a shift in which costs a fortune of scale and belief, so it takes an Elon Zuckerberg to do that. This is uninspiring.
This doesn’t account for frequentist past looking vs Bayesian future looking value. At some point, the value created by an individual is no longer what it used to be, but ownership compensation persists, while cash compensation tapers off.
The labor theory of value is an old marxist economic theory which is not used by serious social scientists in their field.
The modern approach is called the opportunity cost, and by definition is a counterfactual scenario. It is triggered by a choice you actually have. Otherwise it's called science fiction.
This article does a very average job at explaining the whole thing... Even the Elon Musk example is not accurate. The value of Tesla in the ''without Elon'' scenario is wrong, because in that case Elon would be *replaced* by another overconfident nerd. Just think of Apple without Steve Jobs ; he got replaced by something different, and the firm is still standing strong today...
Good counterfactual thinking is super important, but I do not feel that the author is a master of the art.
Startup valuations have nothing to do with theory of value. Theory of value, like the marxist labor theory of value are theories about the metaphysical status, origin an mechanisms of value. It is an explanation of value in general not a heuristic people would use to evaluate things. If labor theory is right (and it isn't) the answer of why founders get so much more is because they are exploiting the other guys working there.
Tesla would exist without Elon Musk. Elon Musk literally simply just bought the company, along with the right to call himself the "founder". Sure without Elon Musk Tesla might be more accurately valued, IE valued significantly less, but it's unclear how much social value is actually created by his showmanship.
There's a difference between exchange value and use value.
Exchange value is use value for the company, because it can make stock deals to obtain things for the same nominal(dollar) amount while giving up much less equity.
Similarly, if the stock is undervalued, it becomes more costly to raise capital in public markets. So hiring a CEO with poor market perception limits a company's options, even if they're internally very effective.
I dont mean to take a position on Musk per se, but I do think there's a reasonable argument that Tesla might actually do better without him. He does bring a certain amount of risk.
I raise this in part because one of the problems with counterfactuals is evaluating a state of affairs that you dont have access to. In fact, evaluation of that alternate state is necessarily biased by the currrent state.
>Elon Musk provides another example. Today, Tesla is worth more than $1 trillion. Without his will and personality, Tesla probably wouldn’t exist.
Yes it would. Mark Tarpenning and Martin Eberhard founded Tesla. They were later forced out by Elon.
They were arguably quite poorly compensated for their part in starting this company.
This speaks to the fundamental problem with this theory - it's all very well saying that Elon is worth his money because Tesla would be nothing without him, but what if it actually would have been wildly successful without him and he mostly rode the wave with his $6 million series A investment?
Of course, you cant prove what "would" have happened, you only know what actually did, so your valur is measured by whomever tells the most compelling story.
You can't prove it, but Musk is not a one-hit wonder. He's transformed the space industry by sheer force of will (as a part time founder & CEO), as Tesla has done for EVs, so it's quite reasonable to assign the success of Tesla to him.
Getting funding, finding the most talented engineers, deciding what to build and motivating people are the high-level skills needed from a founder/CEO. I don't see any argument here against his importance to all companies that he's been involved in. Nobody's claiming he's building rockets all by himself.
In this case it makes sense but in general you have to also ask about the environment. Was money cheap in that era? Was there a favourable media environment for that type of business? Was politics favourable? Etc.
There is a really straightforward answer to this - equity isn't distributed according to a theory of value and nobody ever claimed it was. I own some equity in the Australian company CSL. It is unfair to claim I've ever added any value whatsoever to the company at any point, so obviously I don't get that equity because of a value theory. Even on the basis that people anticipated my buying of that equity - I bought in because I believed the value had already been created.
Who gets equity isn't based on value, it is based on power and who has it right now. Value is for working out whether you, the valuer, should make a trade or not.