Why I follow markets as an engineer, not as a fortune teller

I do not follow markets because I believe I can predict every turn. I follow them because they are one of the most demanding systems I know.

A market brings together businesses, technology, incentives, regulation, psychology and expectations. Every price is an output, but the system that produced it is never fully visible. That feels familiar to me as an engineer. I am used to looking at a complex result and asking what inputs, dependencies and feedback loops sit underneath it.

The difference is that software usually behaves according to rules someone wrote down. Markets contain people, and people change their behaviour when they see what everyone else is doing. That makes them frustrating, but also endlessly interesting.

I care more about the machine than the ticker

When I look at a company, the share price is not the first thing that holds my attention. I want to understand the machine behind it. What problem does the company solve? Why do customers keep paying? Where does the revenue come from? What constrains growth? Which part of the story depends on something outside the company's control?

This is close to how I approach an engineering problem. Before changing a system, I want to know how data moves through it, where decisions are made and what happens when one component fails. A business has similar questions. Its architecture includes products, people, capital, distribution and trust.

Financial statements help make parts of that architecture visible. Revenue can show demand. Margins can reveal pricing power or operational strain. Cash flow can challenge a story that looks convincing at first glance. A balance sheet can show whether a company has room to make mistakes. None of these numbers predicts the future on its own, but together they make the discussion more concrete.

I build my own indicators

That systems mindset is also why I do not limit myself to the indicators a platform happens to provide. When I want to examine a relationship in a particular way, I enjoy building the indicator myself. The process forces me to define what I am actually trying to measure, which data belongs in it and where the result can become misleading.

A custom indicator is not valuable simply because it is custom. It still needs a reason to exist. I might use one to combine signals that are usually viewed separately, normalise information so companies or periods can be compared, or make a change in momentum easier to inspect. Building it myself means I know which assumptions sit underneath the output. I can change those assumptions, test edge cases and discard the indicator if it adds noise instead of clarity.

This is one of the parts of following markets that I enjoy most. It turns passive observation into a small research and engineering problem. The goal is not to create a secret line that predicts the future. It is to build a better lens for one specific question.

Expectations matter as much as quality

One of the hardest lessons in markets is that a great company is not automatically a great investment at every price. The market is not only judging what a business is today. It is continuously pricing a version of what people expect it to become.

That is where engineering intuition needs an adjustment. A technically impressive product can still disappoint investors if expectations were even more impressive. A company can report good results and fall because the market expected exceptional ones. Another can report weak results and rise because reality was less bad than feared.

I find that gap between reality and expectation fascinating. It makes valuation feel less like finding a single correct number and more like testing a range of assumptions. What needs to be true for the current price to make sense? How much growth is already implied? Which assumption carries most of the risk? Those questions are more useful than pretending a spreadsheet can remove uncertainty.

Options made uncertainty impossible to ignore

Options interest me because they make time, uncertainty and asymmetry explicit. With a share, it is easy to focus mainly on direction. With an option, being broadly right is not enough. Timing matters. The size of the move matters. Volatility matters. The path matters.

That creates a useful kind of discipline. An idea can be correct in the long run and still be expressed badly. A trade can have limited downside and large potential upside, but a dramatic payoff diagram does not make the probability attractive. Leverage can amplify an insight, but it can amplify impatience and overconfidence just as easily.

I like that options refuse to let uncertainty remain abstract. They put a price on it. They also remind me that risk is not simply the chance of being wrong. Risk includes how wrong I can be, how long I can stay wrong and whether I can survive long enough to learn from it.

AI has made the market more interesting and more difficult

My work in AI naturally shapes what I notice in the market. I care about model capability, infrastructure, data, adoption and the gap between a compelling demonstration and a system that creates durable value.

I also use AI throughout many of my own evaluations. It helps me work through larger amounts of information, compare reports, structure an initial company analysis and identify claims that deserve a closer look. I can ask it to challenge a thesis, surface alternative explanations or show which assumptions are doing most of the work in a scenario.

The value is not that AI gives me a final answer. I do not want to outsource conviction to a model that can sound certain while missing context. I use it to widen the evaluation and make the research loop faster. Important facts still need to be traced back to primary material, calculations need to be checked and the final judgment remains mine.

That perspective can be useful, but it can also create bias. Knowing a technology well can make a person too enthusiastic about its potential or too critical of companies simplifying the story for a broader audience. Technical understanding does not automatically produce investment insight.

So I try to separate several questions. Is the technology real? Is the product useful? Will customers pay for it repeatedly? Can the company capture enough of the value it creates? Is that value already reflected in the price? A strong answer to the first question does not guarantee a strong answer to the others.

This is especially important during a period when almost every company can attach AI to its narrative. I am less interested in how often management says the word and more interested in what changed operationally. Did the product improve? Did customer behaviour change? Did costs move? Did a new capability become difficult for competitors to copy?

A thesis should be allowed to fail

Engineering taught me to test assumptions, inspect failures and update the system. Markets make the same principle emotionally harder. Once money, time and identity become attached to an idea, new evidence can feel like an attack rather than useful information.

I am not immune to that. Nobody is perfectly rational while watching an opinion get repriced in public. That is exactly why I value a written thesis. Writing down what I believe, why I believe it and what would change my mind creates something I can evaluate later. Without that record, it is too easy to move the goalposts and tell myself I believed something different all along.

A falling price does not automatically mean the thesis is broken. A rising price does not automatically mean it was correct. The useful work is comparing what actually happened with the mechanism I expected. Sometimes the market reveals a flaw in the analysis. Sometimes the business needs more time. Sometimes the original idea was sound but the price left no room for error.

Following is not the same as acting

Markets produce a constant stream of reasons to do something. There is always a headline, a chart, an upgrade, a warning or a new opinion delivered with complete confidence. I think one of the most valuable skills is learning that attention does not require action.

I can find a company interesting without needing to own it. I can study an industry without having a view on next week's price. I can change my mind without immediately replacing one strong conviction with another.

That patience connects back to engineering. Good systems are rarely built by reacting to every signal with a new architecture. They improve through observation, careful changes and feedback. The same mindset helps me treat the market as something to learn from instead of something I need to defeat every day.

Why I keep coming back

I follow markets because they bring many of my interests into one place. Technology matters, but so do business models. Numbers matter, but so do people. A logical argument matters, but so does the humility to accept that the world can produce an outcome I did not anticipate.

Most of all, markets keep me curious. They give me a reason to learn how companies work, why industries change and how expectations become prices. They reward depth but punish certainty. That tension is uncomfortable, and I think that is part of its value.

I am not trying to become a fortune teller. I am trying to become a better observer of complex systems, a clearer thinker under uncertainty and someone who can change his mind without losing the lesson.