Business Is an Endless Fight Against Entropy

Bryne Hobart of The Diff puts it this way.

At a basic level, a company has to at least break even economically in order to survive. In other words, it needs to earn profits that exceed its cost of equity. But over time, various toxins accumulate and the business starts to suffer from shortages of things that are essential for survival.

Early-stage startups have a few hiring cheat codes they can use. Founders can tap into their personal networks, and if the early employees own equity in the company, they also have strong incentives to bring in the smartest people they know. There are downsides, of course. For example, the organization can become overly homogeneous, which tends to increase volatility rather than lower the expected outcome.

Since most startups are destined to fail anyway, higher volatility can actually help maximize outcomes. What matters most is that this approach makes early headcount expansion fast and efficient. In the very early stages, the cost of capital is highest and there is huge uncertainty about how many people the company will ultimately need, so being able to scale in a relatively predictable way at low cost is a major advantage.

But as the company grows, it becomes harder to sustain this model. The hundredth hire is far less likely than the tenth hire to fundamentally change the trajectory of the company, and because that person receives a much smaller equity grant, the impact on their net worth is limited. As the company scales, hiring becomes more complex, more expensive, and more energy-intensive.

One of the biggest bottlenecks that emerges as a company grows is the problem of asymmetric information and decision fatigue. In a small organization, information flows quickly. When people work closely together for long periods of time, it is almost impossible to hide things. And if you are the founder or report directly to the founder, there is not much room for political maneuvering either. But in any company with a formal org chart, there is also an informal org chart that describes how work actually gets done. You can call this "politics" or you can call it "efficiency"—the interpretation depends on where you sit in the organization. As the company grows, this process inevitably creates a kind of tax in the form of conflicts of interest. The larger the headcount, the more often choices that are good for individuals diverge from choices that are good for the company as a whole. When this shows up as blatant office politics, at least it is easy to spot. A more common and harder-to-measure problem is that uncomfortable but important information moves up the org chart more slowly. For example, messages like "the launch timeline for this new product is unrealistic" or "customers are deeply unhappy with the product and are likely to churn" can travel relatively easily from employee to manager. But if that information reflects poorly on the manager’s competence, it is much more likely to get stuck and never reach the people above. It is a kind of off-balance-sheet transaction: instead of reporting the problem, people quietly try to fix it. This happens constantly in every organization. The moment you type "Sorry for the late reply" in an email, something like this has already occurred.

An even more serious issue is decision fatigue at the executive level. Barack Obama once said this:

Whether an organization operates top-down or bottom-up, decisions about exceptions inevitably pile up on the desks of the top executives. Every time you add another layer to the org chart, the number of issues that ultimately need to be escalated to the CEO goes up. In other words, as the company grows, the number of decisions the very top of the organization has to process in a single day increases exponentially.

This problem is solvable—and it is one that absolutely must be solved. Marc Andreessen of a16z has said that the core skill that separates executives who can scale from those who cannot is whether they know how to manage managers. Effective delegation means that subordinates must sometimes be able to make decisions that are the exact opposite of what their manager would have chosen. For a startup to evolve from a founder-controlled operation into an increasingly independent organization, the founder has to give up a certain amount of control. There are only a few ways to deal with this. The company’s growth ceiling can be set by the founder’s capacity for micromanagement. Or the founder can try to hire people who are almost as capable as they are—a task that becomes harder the more exceptional the founder is. Another option is to run everything by checklists while stripping out principles and goals, but that rarely produces a resilient organization.

In the end, a company is like a living organism. It takes in data, processes it into information, and then reacts based on that information. The proportion of useful information that gets lost in this process is impossible to measure, but it is critically important. This connects to the idea of viewing a company as a kind of financial entropy-reduction machine. What prevents operating margins from falling all the way down to the cost of capital, and what extends the actual lifespan of a company beyond what would theoretically be possible, is ultimately the system that ensures information flows properly and that decisions can be made at lower levels of the organization.

Survival is, at its core, a battle against entropy. That is why it is so apt to compare companies to living things. We are not merely imposing order; we are building systems that can sustain order effectively. Ultimately, we all run up against the hard limits of the laws of thermodynamics, but that is still a long way off. Scalable ways of reducing entropy tend to rely on massive energy sources like the sun. Whether it is agriculture or solar power, whether we harness solar energy directly or burn fossil fuels that store it, a company’s growth ultimately depends on how efficiently it can use energy.

Investing can be simplified as the act of providing capital to companies that generate operating margins well above their cost of capital—that is, to companies that are pushing back against entropy. In theory, as time passes, the operating margins of such companies converge toward their cost of capital. This is because investors keep supplying capital either to those companies or to competitors that can produce substitutes, increasing supply until the margin on selling those products converges to the market rate of return.

You can think of an efficient market as one in which no company earns excess returns. In reality, many companies do generate excess returns, while others earn subpar returns that help balance the system overall. Stock prices and interest rates adjust to reflect whether companies are earning excess returns and how large those returns are. Capital is the catalyst that pushes companies toward entropy, and investing is the act of supplying that catalyst.

In that sense, Eugene Fama’s efficient market hypothesis is wrong. Capital and the behavior of the investors who provide it relentlessly push companies’ excess returns toward zero, but companies, in turn, fight entropy just as relentlessly in an effort to generate those excess returns.

The constant game of second-guessing and tug-of-war between companies and investors ultimately produces a market that, on average, is made up of distorted prices.