I followed the script. Studied hard. Got the degree and built a career. Then one meeting, one email, and my calendar was empty. The formula my parents trusted broke under structural change. What it left behind is a constant background hum of uncertainty.
My interest in trading on the XRP Ledger grew out of that uncertainty. Not to chase alpha or outsmart the market, but to own the risk I take.
From Curiosity to Conviction
What followed the layoff was the modern humiliation ritual: applications into silence, behavioral screens, rounds of technical interviews, positive feedback from the team and then nothing.
Between interviews, I needed something to work on. I started exploring how fear and greed move crypto markets in ways that fundamental analysis could not explain. The more I explored, the more I noticed who was doing the talking, and how little of it was backed by anything. Analysts hid behind charts. Influencers sold certainty. Neither offered anything reproducible, just speculation.
I tried to quantify sentiment. I scraped websites, built pipelines, trained models, and attempted to turn social media commentary into a tradable signal. The inherent noise in social media made extracting anything actionable impractical.
The failure was not wasted. I learned how to build the kind of machine learning pipeline that takes raw data and produces something a model can actually learn from. I had the tools. I was just pointing them at the wrong problem.
The conviction came when I stopped chasing sentiment and looked at what was already there: price itself. Patterns in price, volume, and volatility. Structure that machine learning could work with. Not a crystal ball, but a basis for knowing when to act and when to sit out.
I was not tinkering between interviews anymore. By year two, the runway was gone. The only way out was through.
Rules Over Instinct
Most systems do not fail in backtesting. They fail the first time the market does something the data never showed.
The market's adversarial nature is not a flaw. It is the mechanism. Capital flows toward those who manage risk and away from those who do not.
Every trade carries adverse selection. If someone is willing to take the other side of my position, I have to ask what they know that I don't. Markets shift constantly, so any pattern a model learns today may stop working tomorrow. The confidence behind any signal has an expiration date.
Automation does not eliminate my biases. It encodes them. That is why I am building something boring and interpretable. One rule as an example: if the model's prediction cannot overcome the spread, the cost of entering the trade, no trade is placed. No overrides. No gut calls. Most of the time, it will do nothing. That is the point.
The system does not optimize for frequency or certainty. It enforces when not to act. I am not claiming guaranteed profits, a finished system, or a permanent edge. Just the humility to keep iterating.
The XRP Ledger is what makes this viable. Market data is public, settlement takes seconds, and there are no exchange memberships or compliance costs gating entry. That infrastructure makes a one-person operation possible, but a thin order book means the risks are real. Wide spreads, sudden liquidity gaps, spoofed depth, artificial volume, self-fills, and price dislocations that no model fully anticipates.
This time, I own the risk and what defines it. Building an algorithmic trading system I run and answer for is the closest thing to solid ground I have found. It may not be enough. The losses are mine. So is everything else.