Ronald Coase was a British economist and the 1991 Nobel Prize laureate in Economic Sciences. He is known for the question he posed in “The Nature of the Firm” (1937): if the market mechanism is so efficient, why do firms exist? Why don’t people conduct every transaction directly through the market? Raymond Hou (Raymond) treats this question and its answer as one of the theoretical foundations for his arguments about a “company of one” and “Super Individuals.” Raymond cites Coase not to conduct academic exegesis, but to use the idea of “transaction costs” as an economic vocabulary for explaining his choice to leave traditional employment and run a business on his own.
Coase’s question and Raymond’s translation
In courses and long-form writing, Raymond restates Coase’s central proposition: firms exist because “transacting directly through the market has costs.” Each transaction requires finding someone, negotiating, signing a contract, checking the work, following up on progress, and resolving disputes. When those transaction costs exceed the costs of internal management, organizations emerge. They draw scattered market transactions within a boundary and coordinate them through authority and systems, rather than renegotiating through the market each time. Raymond condenses Coase’s conclusion into the quotation: “The fundamental reason firms exist is to save on transaction costs.” He extends this logic to the nature of everyday work, arguing that work can be seen as a transaction: individuals exchange their time, expertise, and energy in the market for money and resources. A company’s function is to reduce the hassle of those direct market interactions so participants can focus on their work.
Reframing departments as an unavoidable compromise
Raymond’s most personal extension of transaction cost theory turns the question “Why do companies have departments?” into a proposition about human limits. He argues that companies divide work into marketing, design, engineering, accounting, human resources, and other departments not because specialization is ideal, but because one person can master only one or two fields in a lifetime. Limited learning capacity, time, and memory force work to be split among different people, then stitched back together through meetings, systems, and rules. In this framework, departmental silos are an “unavoidable compromise,” not a design worth celebrating. It also offers an explanation for why many people dislike going to work: the cause is often the communication costs, process gaps, and lack of information transparency involved in cross-department collaboration. This translation uses transaction cost theory not only to explain why firms exist, but also to support the argument for “why work that one person can finish should not be split among a group.”
Applications to a company of one and Super Individuals
Transaction costs provide a post hoc economic rationale for Raymond’s decision to leave a large company and run a business on his own. He describes his original motivation as reducing communication costs and seeking information transparency: if one person can complete the work, it should not be divided among a group. Coase’s theory supplies an economic basis for that intuition. Raymond also points out that the framework works in both directions. Companies are originally meant to reduce transaction costs, but once an organization grows beyond a certain size, the company itself can become “a higher transaction cost.” He therefore suggests that workers use this lens to break down their frustrations at work: distinguish how much pain comes from internal organizational friction, how much comes from the realities of the market, and how much from gaps in their own abilities. They can then decide whether to change environments or even start a business themselves. This line of thought connects with Raymond’s outsourcing principle: delegating tasks outside one’s areas of passion and expertise is also a way of reducing one’s personal transaction costs.
Extending the theory to the AI agent era
In 2026, Raymond connected transaction cost theory to discussions of AI agents, in response to the popular advice to “build a whole Agent Team.” His argument is that the fundamental reason humans invented departments—the limits of one person’s expertise, time, and memory—does not apply to AI in the same way. A large language model can call on cross-disciplinary knowledge and operate without time constraints. If transaction costs and capability limits are absent, imitating human organizations by building departmental walls around AI amounts to turning a “superhuman that does not suffer internal friction” into “a small company that does.”1 Raymond therefore argues that individual workers should provide one general-purpose agent with their knowledge, tools, and processes all at once, and let it assign the work itself, rather than defining an organizational chart for it. This makes Coase’s question take a reverse turn in Raymond’s hands: transaction costs once explained why firms exist; when technology pushes those costs close to zero, concentrating work in one individual (or one AI) can instead be a more efficient organizational form.
Relationship to other sources of thought
In Raymond’s map of ideas, Coase is placed alongside Brooks’s Law from The Mythical Man-Month (communication links grow with the square of a team’s size) and Conway’s Law (an organization’s structure shapes the form of its output). Together, these support the argument that division of labor has hidden costs and scale creates management debt. Coase answers “Why do firms exist?” while the latter two explain “Why does efficiency fall as firms grow?” Combined, they form a theoretical set Raymond uses to argue for the advantages of Super Individuals and companies of one. Related concepts are discussed in Super Individual and Business Thinking.
Sources cited
Footnotes
-
Hou Zhixun, “Why Do I Think Most People Don’t Need an Agent Team?” public Facebook post, 2026-05-06. View original ↩