The Trade That
Built Everything

I need to tell you something that changes everything about this story. I am not building a quant fund in Bangladesh. I am building an entire discipline from scratch in a country where it does not exist.

There are no quant APIs. There are no broker APIs for algorithmic trading. There are no quant jobs, no quant roles, no quant community. No one around me who understands what I do. I have been grinding completely alone for over two years. Dr. Bhuyan was the first person who looked at my work and immediately understood its value.

When I started, there was no API. Not for the Dhaka Stock Exchange, not for any broker in the country. I built the first DSE trading bot by automating a brokerage web interface through Selenium — because that was the only way to place trades. Browser-level automation on a desktop computer in Dhaka. No cloud, no infrastructure, no tools. Just a laptop and stubbornness.

The DSE bot worked. The thesis was correct. But the market went flat — sideways for months with no movement, no volatility, no opportunity. That is when I pivoted to crypto. Not because it was trendy, but because it was the only market where someone in Bangladesh could actually trade algorithmically. Interactive Brokers was the gateway. First time in my life I had real tools.

Every tool I have used — Python, Postgres, Supabase, Oracle Cloud, Binance API, Foundry, Solidity, cadCAD, Stable Baselines — I learned alone. Every concept, every strategy, every optimization technique. No courses, no mentors, no team. Just documentation, trial, and error.

You are not just reading about a fund. You are reading about someone who built an entire discipline from nothing — in a country where the concept of quant trading did not exist before him. That is not a story about returns. That is a story about what humans do when they have no infrastructure and no permission.

The January 2026 Trade

In January 2026, I made the largest trade of my life. A leveraged position on silver perpetual futures — most of my net worth. I had spent weeks building the thesis: US executing coordinated resource nationalism across Venezuela, rare earths, semiconductors. China restricting silver exports. The dollar being used as a weapon. My models said 85–90% probability of success. Expected return: over 1,000%.

The thesis was right. The trade was not.

Silver crashed 30% without warning. My position was liquidated. I lost most of what I had built.

That night I made a vow: I would never place a manual trade again.

The honest version: I already knew how to build a system. I had taught myself Python, built backtesting engines from scratch, run thousands of simulations, constructed the full stack — data infrastructure, optimization engine, walk-forward validation, cloud deployment, monitoring dashboard. I had done all of it.

What I did not do was use it for the silver trade. The thesis felt so clear, so well-researched, that I got lazy. Setting up a proper bot means months of analysis, simulation, optimization. I skipped it. I traded manually. I told myself conviction was enough.

It was not. The loss taught me that cutting the process short — even once, for even the strongest thesis — is how you lose everything. The vow was not to learn something new. It was to never again pretend I could afford to be lazy about what I already knew.

October 2025

For most of 2025, crypto was in a slow bleed. XRP, the asset my bot trades, peaked at $3.40 in mid-2025 and spent the rest of the year falling. By year end it was near $1.88. Nearly fifty percent gone from its high.

Across that entire period — while the underlying asset lost half its value — the bot kept making money.

Then came October 10th.

Trump announced 100% tariffs on China. Within 24 hours, $19.37 billion in crypto positions were liquidated. 1.6 million traders wiped out in a single day. Bitcoin fell 18%. XRP and altcoins dropped 60–80%. Yahoo Finance called it "the most cursed month in crypto history."

At 2:39 AM on October 12th, my stop-loss triggered at −17.41%.

I was asleep.

The system exited. No hesitation, no "maybe it comes back," no one watching a screen at 2 AM hoping for a reversal. The rules executed. Two days later, the system re-entered on its own. I finished October with a profitable month — five trades, net positive.

That is the point of building a system rather than trading with conviction. I do not follow individual trades. Most days I am entirely unaware the bot has done anything until I check the balance. I know exactly how it behaves in every market condition — but I do not need to watch it.

The model had predicted 15–20% maximum drawdown. The actual was 17.41%. The underlying asset lost fifty percent of its value across the year. The bot finished profitable. The difference was not intelligence or courage. It was the presence of a system that did not care what I believed — and did not need me to be awake.

The Numbers

+117% return over 18 months of live trading.Sharpe 2.47. Profit Factor 2.21. Win Rate 54%. 30 closed trades on 3.5x isolated margin. 99.3% Monte Carlo probability of profit.

Before going live, I validated across 517+ assets. 2-year backtest. 8-fold walk-forward (87.5% folds profitable). 10K Monte Carlo simulations. Every parameter stress-tested. Every edge verified.

I tested LSTM, Random Forest, and Hidden Markov Models on DSE equity data before concluding that rule-based DC+VWAP-EMA generalizes better on thin markets. Walk-forward validation showed ML overfitting — so I chose the approach that works, not the approach that looks impressive.

The Halal Problem

The bots use margin. Leveraged trading. Every time someone sees how much I am making, their first question is: is this halal?

I always believed there was a way to reconstruct these instruments — the same mathematical elegance, the same economic function — without the interest component. That the problem was not finance. It was the assumption buried inside it.

Where Baraka Came From

I looked at DeFi protocols — on-chain finance, smart contracts that execute without counterparty trust — and I saw what they could become: financial infrastructure built from mathematics, not from institutional permission — and restructurable, at the protocol level, to remove interest entirely.

The mathematics already existed. Ackerer, Hugonnier and Jermann published a framework in 2024 that, at the parameter ι=0 — interest set to zero — produced a fully functional perpetual futures mechanism with no riba embedded anywhere. Six research papers. Deployed on-chain. Zero interest hardcoded at the protocol level. Not a fatwa layered on top. Built from first principles.

The Partnership

Dr. Rafiq Bhuyan — Fulbright Scholar, 80+ peer-reviewed papers, PhD Economics Concordia Montreal, former Purcell Chair at Le Moyne College — saw what I was building. Before I had finished explaining, he said he wanted to be my co-founder. He offered to put $50,000 of his own capital into my system to manage.

He proposed all of this without hearing my plans. The work was the pitch.

What We Are Building

Web3 is becoming the infrastructure layer of global finance. Every transaction, every instrument, every settlement — eventually on-chain.

Muslims will face a choice: participate in systems built on interest, or have nothing. We are building the alternative. Baraka: the world's first Shariah-compliant perpetual futures protocol.

Alongside it, Arcus Quant Fund manages capital systematically — the same rigor, the same accountability — for clients who want returns without the noise.

I am twenty-five. I lost most of my net worth on a trade I believed in. I built the system anyway. I found the alpha anyway. I found the partner, the protocol, the mission — all of it arrived, in sequence, as if it was always going to.

Now we build in public.