programming

Why I Don't Like to Build Trading Bots

Nathan Broadbent

15 Dec 2025 — 2 min read

Meaning vs extraction

Building products creates something new that didn’t exist before. It compounds usefulness. Trading mainly redistributes existing value. Even when it’s “clever”, it feels like shaving friction rather than adding substance. Efficient, but spiritually thin.

Agency and narrative

When I ship software, there’s a story I can tell myself: “I solved a real problem for real people.” Trading bots don’t give me that. The narrative is just “I was faster or luckier for a moment.” Humans care a lot about narrative identity, even engineers who pretend they don’t.

Asymmetric failure feels worse

Losing money building a product feels like tuition. I learned, I built skills, and I have artifacts to show for it. Losing money in trading feels like entropy. Nothing accumulates except regret and logs. Although you do learn more about AI and algorithms.

Ego plus honest self-assessment

Ego and fear are also involved, but maybe in a healthy way. I know algorithmic trading rewards a very specific profile: obsessive, adversarial, statistically ruthless, and comfortable with meaninglessness. I'm noticing “this might not be who I am.”

Misaligned feedback loops

Product work rewards patience, taste, empathy, and long arcs. Trading rewards short-term signal exploitation. My brain clearly prefers the former. When feedback loops don’t align with my values, they feel “gross”.

Quant envy without quant desire

I respect that real quants exist and have an edge, but I don’t actually want their life. This creates a mild cognitive dissonance that resolves as moral revulsion. My mind is protecting my identity.

Creation vs zero-sum

At a gut level, I'm allergic to zero-sum games. Even when markets are not strictly zero-sum, they feel that way experientially. SaaS feels positive-sum. That matters more than whether the economics textbook agrees.

The short version:

I'm a builder who wants durable impact, compounding meaning, and artifacts that survive me. Algorithmic trading optimizes for none of these. However, it still serves a vital role in price discovery and liquidity, and many people get very rich from it, even if I can't find much meaning in working on it.

Global Handicrafts: A 16-Year Rails Ghost Story

A couple of hours ago, I got a Google security alert: I created that Gmail address 16 years ago for my very first Ruby on Rails internship/job at Crossroads Foundation in Hong Kong. I had written a small integration that synced inventory from MYOB to a Spree store. I

ARC-AGI: The Efficiency Story the Leaderboards Don't Show

ARC-AGI is a benchmark designed to test genuine reasoning ability. Each task shows a few input-output examples, and you have to figure out the pattern and apply it to a new input. No memorization, no pattern matching against training data. Just pure abstraction and reasoning on challenging visual problems. It&

You're Going To Australia

It was ten past five on a Tuesday, and I received the booking confirmation for my stay at Rydges Hotel in Kalgoorlie, Australia. It sounded like a nice room. The only problem is that I did not make this booking. This booking confirmation was sent to my personal email address.

Error 404: Black Hole Not Found

I'm writing a short story and one of the plot points involves Sagittarius A* (Sgr A*)—the supermassive black hole at the center of the Milky Way galaxy. I wanted to read about Sagittarius A*, so I looked it up on Google. This is what I saw: At