Capital is lost in the wrong regime.
Not execution.
Not when to trade — but how much of you should be in the market right now.
84% max drawdown reduction coverage across 30+ instruments and 181,000+ bars of market history.
What you see in TradingView
The live product uses the exact same terminology shown below. What you see on the page is what you see in the overlay.
EXPANSION
RISK ENABLED
State: Ranging · Extended duration
High-quality structure. Direction is not implied.
Core outputs
Regime Atlas classifies market conditions and determines whether deploying risk is structurally justified.
What Regime Atlas is not
Self-qualification matters. If you're looking for any of the below, this is not the tool.
Proof, not promises
Most traders try to optimize entries. This model solves a different problem: whether deploying risk was justified in the first place.
On 12 of 31 assets, regime sizing improved both Sharpe and max drawdown simultaneously.
Example: ADA improved from Sharpe 0.63 to 0.73 while max drawdown improved from −92% to −71%.
On SOL, Sharpe improved from 0.57 to 0.78 while max drawdown improved from −96% to −78%.
On strongly trending instruments like SPY, Sharpe declines marginally (−0.017) because the model steps aside during transitional windows. In exchange, maximum drawdown improves from −56.5% to −41.4% — a 15 percentage-point improvement.
You trade some trending-market Sharpe for materially lower tail risk. That is the design intent, not a side-effect.
Without a regime filter, capital is exposed during structurally weak conditions. That is where most drawdowns originate.
Regime Atlas reduces that exposure. Not by predicting direction, but by restricting participation when conditions degrade.
You are not optimizing trades. You are filtering when trading should happen at all.
Who built this
I built this after repeatedly seeing capital erode during structurally weak periods — even when entries were technically correct.
Most losses I experienced — and later observed across other traders — did not come from bad entries. They came from deploying capital in the wrong regime.
Regime Atlas was built to solve that specific problem. Not to predict direction. Not to generate signals. But to answer one question consistently: is risk deployment justified right now?
The model was developed and tested across 30+ instruments over multi-year datasets covering equities, crypto, forex, and commodities. All parameters were frozen prior to validation. No post-hoc tuning. No repaint behavior. No optimization loop. What you see in the output is deterministic and reproducible.
This is not a fund, not a signal service, and not a black-box system. It is a structural classification layer built for operators who already have a process and want to control when capital is exposed.
Parameters frozen February 18, 2026. All market data after that date constitutes out-of-sample validation.
Pine / Python implementation parity confirmed. No forward bias. No post-freeze tuning. No repaint behavior observed.
Validation window and methodology are included in the full validation pack shared after qualification.
Founding access
Limited to the first 10 operators. Lifetime-locked rate through every future tier.
- Daily Regime Atlas™ core model overlay on TradingView
- Expansion / Risk Enabled / State / Score outputs on D1
- Access to the full validation pack (30+ assets, methodology, per-asset results)
- Direct line to the builder for methodology questions
- Lifetime price lock — €99 stays €99, even as the public tier moves up
- First-month refund if the model doesn’t fit your workflow
Honest answers
Does the model repaint?
No. Once a daily bar closes, the regime output for that bar is fixed and does not revise. Historical calls remain reproducible bar-for-bar on the TradingView overlay.
Does this replace my existing strategy?
No. You keep your entries, exits, instrument selection, and directional bias. Regime Atlas only determines whether planned risk deployment is justified.
Do you short?
No. This is not a shorting engine. It is a regime-weighting layer that adjusts participation based on structural conditions.
Why D1 only?
Regime shifts are structural, not intraday. Running the classification on D1 with a 252-bar warmup window produces the most stable and reproducible output.
What about the SPY trade-off?
We publish it openly. On strongly trending instruments like SPY, Sharpe drops marginally (−0.017) while max drawdown improves materially from −56.5% to −41.4%.
Is this AI or machine learning?
No. The model is rule-based and deterministic. It is not a black-box ML system and does not rely on post-hoc curve fitting.
Do I need TradingView?
Currently, yes. The core model is delivered as an invite-only TradingView overlay. A broader workflow layer is on the roadmap.