The Architecture of a Mispricing

This is not a system built around owning great businesses for decades. It is built around identifying moments where the market has the wrong expectations about a business's future, positioning before those expectations are corrected, and capturing the move that follows. That distinction matters more than almost anything else in this chapter.

Investing Versus Exploiting a Mispricing

When most people talk about investing in equities, they mean buying a share of a business they believe is fundamentally sound, holding it for years, and collecting the return the business generates as it compounds its earnings over time. That approach is legitimate. It has made people very wealthy. It is also not what this system does.

What this system does is closer to exploitation than investment in the traditional sense. The starting point is not "is this a great business?" The starting point is "has the market priced this business incorrectly relative to where its expectations are heading?" Those are different questions with different answers, and they lead to completely different decisions about what to buy, when to buy it, how much to pay, and when to get out.

A great business at a full price offers no edge. The market has already done the work. The expectation for future performance is already embedded in the price, which means even if the business continues to perform exactly as expected, there is little return left to capture. The opportunity in this system lives specifically in the gap between what the market currently expects and what is actually developing. When that gap is wide and the evidence is strong that the market is about to be forced to revise its view, that is a setup worth sizing into.

This is the difference between an investor and an opportunistic speculator. The investor asks whether the business is good. The opportunistic speculator asks whether the market is wrong, and by how much, and for how long that wrongness can persist before price corrects it.

This Is What Works for Me

It needs to be said directly: this system reflects a specific way of participating in markets that has developed over time through experience, failure, observation, and refinement. It is not a universal prescription. There are traders and investors who generate exceptional returns through approaches that look nothing like this one. The objective here is not to argue that this is the only way or even the best way for everyone. It is to describe clearly what has worked, why it works, and how it is applied.

The core of what works for me is finding situations where a company is materially undervalued or misunderstood relative to what it is in the process of becoming, getting positioned before that becomes obvious, and holding through the development of the thesis. The edge is in the early identification, not in the execution of what is already known.

Where the Asymmetry Tends to Live

I see the greatest asymmetry in companies that are either unproven or in the process of proving themselves. The reason is straightforward: these are the companies the market is most likely to have priced incorrectly. When a company is early in its story, institutional coverage is thin or nonexistent, earnings history is short or negative, and analyst models carry wide ranges because there is not enough data to build a confident one. That uncertainty creates a discount. And that discount, when the company begins to prove itself, is what produces the violent repricing.

Most of the names I work with are not household names. They are not companies you would read about on the front page of a financial news site. They are businesses that most participants have not looked at yet, which is precisely the structural condition that allows them to be mispriced in the first place. By the time a company is widely covered and frequently discussed, the discovery has already occurred. The asymmetry that made it interesting earlier is gone, and what remains is a debate about fair value between people with roughly the same information.

That said, this preference for smaller and earlier-stage companies is a tendency, not a rule. The market's ability to misprice companies does not stop at a certain market capitalization. Amazon in the early 2000s was already a known business with significant public awareness, but the market's expectations for what it would become over the following decade were nowhere near what actually developed. Nvidia entering 2023 was a well-established semiconductor company that the market had not yet priced for the scale of demand that AI infrastructure would create. Both represented situations where the gap between current expectations and future reality was enormous, and both produced extraordinary returns for participants who were positioned early enough and held through the development of the thesis.

The point is not to avoid large-caps. The point is to go where the mispricing is. That tends to be in earlier-stage companies more often than not, because uncertainty creates discount, and discount combined with a developing thesis creates asymmetry. But when the conditions align at any size, the system applies the same way.

Why Concentration Follows From This

If the edge is specifically in early identification of a mispricing, and if those situations are genuinely rare rather than frequent, then spreading capital across twenty or thirty positions at any given time is not diversification. It is a confession that you do not actually know which ideas are the best ones.

The concentrated approach, three to seven positions sized to matter, is a direct consequence of how rare high-conviction setups actually are when every filter in this system is applied honestly. Most ideas that come through the process do not make it all the way. The ones that do are the ones where every layer of the analysis points in the same direction. When that happens, the position needs to be large enough that if the thesis is correct, the portfolio reflects it. If the sizing is too small to matter, the quality of the analysis is irrelevant regardless of how right it turns out to be.

The Process Over Outcome Problem

One more thing before the chapters that follow: this system will produce losses. Positions entered with high conviction, correctly analyzed, correctly sized, and correctly managed will still result in losses when the thesis does not develop. That is the nature of operating in probabilistic environments. A position can be completely correct in every analytical respect and still lose money if the timing is wrong or if an external event changes the conditions the thesis depended on.

The only thing within control is the quality of the process: how rigorously the filters were applied, whether the invalidation was defined honestly, whether the sizing reflected the actual quality of the evidence. Individual outcomes are noise over any reasonable sample. Process quality across many iterations is what determines the result over time.