Hey Quant X Tribe,
What if the reason your strategy keeps failing isn't the strategy?
It's the sample size.
Last week, we had Emmanuel Raja Singam on our weekly QXC session. He's the founder of Tradomatics and RoboGenesis — 9 years building institutional-grade algo infrastructure. He's been on Money FM 89.3, featured in Business Times, Reuters, Yahoo Finance.
But the thing that stood out wasn't his credentials.
It was one number he dropped casually, like it was obvious.
4,000 to 5,000 backtests. Per trading board. Per cycle.
Here's what most traders do:
They find a strategy on YouTube. Or a Discord group. Or after watching price action for a few weeks. They test it on a few trades — maybe a week of demo data, maybe a month. It looks okay. They go live. Then the market does something slightly different and the whole thing falls apart.
That's not bad luck. That's an insufficient sample.
That's what happens when you test one angle of one hypothesis.
Here's what Emmanuel's team actually does:
They built what they call The Terminator — a machine that backtests 24 hours a day, 7 days a week, across 162+ asset classes. His internal KPI for the team: if Terminator stops, someone gets fired.
For a single trading board, they generate 4,000–5,000 hypothesis permutations — different indicator combinations, timeframes (1m, 5m, 15m, 30m), risk settings, entry/exit parameters — and run each one across at least 2 years of data. Then they dump all 5,000 results into Claude and ask: "Give me a pattern."
The machine runs. Claude analyses. The team decides.
Only then does a strategy reach a user.
And even after all that — they forward test for 6 months on live or demo before shipping.
Every 2–4 weeks, settings get updated. Because what worked in January might not work in March. Not because the strategy is broken — because the market regime shifted.
The insight isn't "use more indicators."
It's this: most traders are making permanent decisions based on temporary data.
One week of backtesting isn't a sample. It's a guess dressed up as evidence.
The edge doesn't come from finding a better entry signal. It comes from knowing — with statistical confidence — how your system performs across thousands of conditions, and which conditions it breaks under.
If you can't answer that question, you're not testing. You're hoping.
There's also something Emmanuel said about exits that stopped me cold.
His trading boards have no take profit.
Not because he doesn't care about profit — but because a static TP creates two failure modes: you set it too high and it never hits, or you set it too low and leave 150 pips on the table when the market runs 200. Both are losses. Just different kinds.
Instead, they use a dynamic EMA-based exit — the trade closes when the trend reverses, not when a number is hit. The exit breathes with the market.
Even a 20% improvement in exit efficiency compounds massively over a thousand trades. The entry gets all the attention. The exit is where the money actually lives.
So what does this mean for you?
You don't need The Terminator. You don't need 162 asset classes or a team of data scientists.
But you do need a process. A repeatable, structured, systematic process for testing your ideas before you risk capital on them.
That's exactly what we teach inside the Quant X Accelerator.
The I.B.O.T framework — Ideate → Backtest → Optimise → Trade — is the exact same pipeline Emmanuel described, scaled for individual traders. You ideate a hypothesis. You stress-test it. You find the optimal parameters. Then and only then, you trade it.
No guessing. No hoping. Just structured edge validation.
Ready to trade with structure, not guesswork?
Join our FREE Quant X Accelerator Masterclass. We'll show you what quant trading actually is — and how to build a strategy you can actually trust.
Inside: The IBOT Framework · How to backtest properly · Finding real edge before risking capital
To your growth,
Quant X Team
Where Data Becomes Alpha
Editor: Dareen Tan
