How We Built a Quantitative Options Signal Engine

It started with a losing streak

Like most retail options traders, I spent my first year chasing setups off Twitter, buying calls because a stock “looked ready,” and wondering why my account kept bleeding. I had alerts from three different services, a Discord full of “trust me bro” plays, and a portfolio that looked like a graveyard.

The turning point wasn't a winning trade. It was a question: What if I stopped trying to predict where stocks are going and started measuring when they've gone too far?

The spreadsheet era

The first version of what would become SigmaSnap was a Google Sheet. Every morning I'd pull price data for a handful of tickers, calculate rolling averages and standard deviations, and look for stocks trading at statistical extremes — two or more standard deviations from their recent mean.

The concept isn't new. Mean reversion is one of the most well-studied phenomena in quantitative finance. Prices that deviate significantly from their average tend to snap back. The question was never if it works — it was whether I could build a system disciplined enough to trade it consistently.

That spreadsheet had maybe 10 tickers. I'd manually check it a few times a day, scribble trade ideas in a notebook, and sometimes actually place them. It was slow, inconsistent, and riddled with the same emotional decision-making I was trying to escape.

But the signals were working. Not every time — but enough.

From gut feel to multi-factor confirmation

The raw statistical signal was a start, but it wasn't enough on its own. A stock can trade at 2-sigma and keep going. I learned that the hard way more than once. So the next step was layering in confirmation factors — other indicators that had to agree before a signal was worth taking.

I won't get into the specific factors we use (that's the engine's edge), but the philosophy is simple: no single indicator drives a trade. Multiple independent measures have to converge. When they do, the probability of a snapback increases significantly. When they don't, we wait.

This was the hardest part to get right. Every factor I added needed to actually improve outcomes, not just make me feel more confident. That meant backtesting everything against real data, throwing away the things that looked good in theory but didn't hold up, and resisting the urge to over-fit to past results.

Automating the whole thing

The spreadsheet became a Python script. The Python script became a scanning engine. The scanning engine became SigmaSnap.

Today the system scans 60 high-volume tickers every 15 minutes during market hours. When a stock hits a statistical extreme and all confirmation factors align, the engine constructs a complete options trade — structure, strikes, expiration, position size, profit targets, and stop loss. The whole thing runs without human intervention.

That last part matters more than people think. The biggest advantage of automation isn't speed — it's discipline. The engine doesn't get scared after a losing trade. It doesn't chase a ticker because it's trending on social media. It doesn't double down because it “feels right.” It runs the same process, with the same rules, every single scan.

Why debit spreads

Early on, I traded naked calls and puts. The returns looked great — until they didn't. One bad move could wipe out weeks of gains. Defined-risk structures were the answer.

SigmaSnap primarily signals debit spreads — bull call spreads and bear put spreads. The maximum you can lose on any trade is the debit you pay to enter. No surprises, no margin calls, no waking up to a blown account. The trade-off is capped upside, but when your win rate and risk management are dialed in, consistent defined-risk trades compound into real returns.

The backtest

Before we launched, we ran the production engine against a full year of market data — 252 trading days, 60 tickers, using 15-minute intraday data. Same logic, same parameters, same everything. No tweaking after the fact.

Ten out of twelve months were profitable. The two losing months were expected — no system wins every month, and if someone tells you theirs does, run. What mattered was that the system recovered and the overall trajectory was positive.

Transparency as a feature

Most signal services show you their best trades and hide the rest. We show everything — every entry, every exit, every win, every loss. If a signal gets stopped out, you see it. If we have a bad week, it's right there on the dashboard.

This isn't just a principle — it's a competitive advantage. When your system actually works, transparency builds trust faster than any marketing ever could. And when it doesn't work (because no system works 100% of the time), people respect honesty more than excuses.

What's next

SigmaSnap is live. The engine is scanning. Signals are firing. But we're not done. We're continuously evaluating the system against live market data, monitoring performance metrics, and looking for ways to improve — without overfitting or chasing short-term results.

If you're an options trader who's tired of guessing, tired of following someone else's gut feeling, and want a system that runs on math instead of hype — try the 14-day free trial. See every signal, every trade, every result. Then decide for yourself.

We built this because we needed it. Maybe you do too.