The 4-Signal Confluence Method™

There is no course.

There was never going to be one. Every element on the previous screen was invented — the countdown, the seat count, the students, the testimonials, the publications, the one-in-twenty dollar lottery, the account balances, the car, the beach. You have seen all of it before on pages that were not joking.

You were not the 20th person. Nobody is the 20th person. That's the point of the 20th person.

The chart went up and to the right forever and never once went down. No strategy does that. If you scrolled past it without flinching, that's worth sitting with — it's the single most common thing a fake track record gets wrong, and it's the easiest one to spot.

And even if the course existed, you wouldn't follow it. Not because you're not capable — the rules you'd actually follow fit on an index card. You wouldn't follow it because it requires doing the same unremarkable thing for twenty years while louder people appear to do better than you.

That's the entire barrier to entry. Not intelligence. Not access. Patience.

So here's what it's actually built from. For free. No email required.

The record

$100,000 → $16,900,156 across 20.5 years. Positive in 16 of 20 complete years. Worst year −15.9% (2015). In-sample 2018–2026; 2006–2017 was out-of-sample. Hypothetical, price-only, frictionless — no fees, slippage, financing or taxes.

Equity curve vs. buy & hold

Log scale — on a linear axis SPY flattens into the baseline and the comparison stops meaning anything. Both lines start at $100,000 in 2006. Note 2015 and 2011: the system loses money in years SPY was roughly flat. That's the cost of a trend filter, and it's supposed to be there.

Year by year

* partial year. Look at 2008 and 2022 — the system was flat while SPY lost 38% and 20%. That's the entire proposition. It gives up ground in choppy flat years to not be there for the crashes.

The configuration as traded, 2019–2026

$100k → $1,351k over the window actually traded, covering COVID, the 2022 bear and the 2025–26 volatility. Note the volatility line: this is a more volatile ride than SPY, not less. It simply gets paid more for it.

Monthly texture

246 months, 147 of them positive. Median +1.5%, mean +2.4%, monthly σ 7.9. Best month +22.7% (Apr 2026), worst −18.5% (Feb 2018 — Volmageddon, caught in-trade; the gate was calm at entry and correctly blind). Longest losing streak: four months, ending Sep 2015 — the same episode as the −31.4% record.

December is the weakest month here, not the strongest. There is no Santa rally in this book. September has the worst hit rate on the page at 40%. And 99 red months still compounded to 169×, which is the only seasonality lesson worth keeping: read this as texture, not as a timing edge.

What the volatility gate cost

Entries are sized down when volatility is elevated. Over twenty years that triggered thirteen times. Here is the honest accounting, including the part that makes it look like a mistake:

The gate cushioned six losing entries and trimmed seven winners. It saved 28 points on the losers and gave back 59 on the winners — a net cost of roughly two CAGR points, and $2.4M left on the table.

Turning it off makes more money. It also produces a −44.66% drawdown, which is a path almost nobody survives holding. The gate is insurance, and insurance costs. If that trade-off looks obviously wrong to you, you have just identified why you'd struggle to run this.

The four sleeves

Equity beta, growth beta, monetary hedge, managed futures. Nothing exotic. Nothing you can't buy in any brokerage account.

The ingredients

Entry — every condition true, confirmed across consecutive closes

1. Price above a simple moving average
2. RSI, Wilder smoothing, above a threshold
3. Price above the midline of a rolling range
4. Stochastic %K above a threshold

Exit

A fixed count of consecutive closes back below that midline. Not a feeling. Not a gut check. A count.

Volatility gate

A graded VIX filter applied at entry only. It decides how large you're allowed in. It has no opinion once you're in.

Four textbook indicators. None proprietary, none exotic, all of them already in whatever charting package you own.

The numbers are missing on purpose

Every lookback, every threshold, every close count, every leverage step — gone. Not because they're a secret. Because handing them over would be worse than useless to you.

The indicators above are public. Anyone can name them in ten seconds. What took years was roughly 1,500 build-era configurations tested across two decades of data, checking what survived out of sample, and throwing away almost all of it. That work is the system. The ingredient list is not.

Copy my settings and you'd be holding numbers you can't defend, fitted to a market you didn't study. The first losing streak would end you, because you'd have no way to tell whether the model was broken or behaving exactly as designed. You'd have nothing underneath it but my word.

So do the work. Take these four indicators, test the settings yourself, write down every version that failed, and find out what you can actually sit through when it's your money. You'll land somewhere different from me — and it'll hold, because you'll know why every number is the number.

That's the part nobody sells. It's also the only part that matters.

The graveyard

Around 1,500 configurations were built and tested. Every candidate got a body and a verdict. Ten of them are below — a selection, not the full ledger. A strategy page with no failures on it is telling you what got left out.

Why this won't help you

In twenty years of backtesting the system never had more than four consecutive losing months. That sounds reassuring until you sit through the fourth one — which happened in 2015, alongside the worst drawdown in the record — at which point every instinct you have will tell you the model is broken and this time is different.

It also spends long stretches doing nothing at all. No signal. No position. Just waiting, while the timeline fills with people apparently making money every day.

Most people quit somewhere in there. Not from a blown account — from boredom, or from the quiet conviction that they could do better. That's the failure mode. It's not dramatic and nobody posts about it.

The actual product

Nothing is for sale here. There's no email capture, no Discord, no upsell, no waitlist.

And to be accurate about what this is: everything above is a snapshot. Five exhibits, pulled out of a binder that runs to about forty-five pages. The rest is more of the same — more tables, more rejected candidates, more angles on the same twenty years, plus the operating documents that make it runnable day to day.

None of that is for sale either. It's just not interesting enough to publish, and it wouldn't change your conclusion. What's here is enough to judge whether the record is honest and whether the trade-offs are ones you could actually live with. That's the only thing a page like this should be asking you to decide.