Maverick Trinity Cascade: how three regime-specific champions share one autonomous engine

Maverick Trinity Cascade: how three regime-specific champions share one autonomous engine

A common pattern in autonomous trading systems is the single-strategy engine. One champion strategy gets crowned, runs everywhere, and either prints or bleeds depending on whether the current market regime suits its bias. When regimes change, the engine that was crowned during a trending phase loses money during chop and the operator either accepts the drawdown or pulls the engine offline.

Maverick rejects this pattern. The current architecture runs three regime-specific champions in a cascade, each owning a different market state. This note describes how the cascade works.

Three tiers, three lenses

The Trinity Cascade has three tiers. Each tier handles a distinct regime and is engineered for it.

Tier 1 — Deep Discount Reversal. This champion handles markets where price has fallen 50 to 75 percent from its all-time high and is showing reversal structure at a major support level. The signature setup is negative directional movement, an oversold oscillator, ADX divergence between the 4-hour and 15-minute timeframes, and a single morph entry around an eight-hour expected hold. The champion specifically does not chase. It waits for confirmation that the panic is exhausting.

Tier 2 — Compression Genesis. This champion handles volatility breakout setups where price has compressed into a tight Bollinger Band squeeze followed by a clean breakout. The thesis is that compression resolves into directional moves and that the early bars of the move offer the best risk-reward. This tier is currently dormant pending stop-cluster work but the architecture supports it as a slot for the breakout regime.

Tier 3 — Sovereign PDL Spring OTE Ratchet. This is the trend continuation champion. It enters at the optimal trade entry zone on a pullback after price has broken the prior day low, then ratchets the stop forward as the trend extends. This is the workhorse champion in trending markets and it is currently crowned.

Why the cascade matters

A single-champion architecture forces the engine to either trade everything through one lens or stay flat through anything outside that lens. The cascade lets each lens see only the regime it understands.

When a market is panicking after a long downtrend, Tier 1 takes the trade. When the same market is consolidating into a breakout setup, Tier 2 takes it. When the same market is in a clean trend, Tier 3 takes it. The handoff between tiers is governed by regime classifiers that read chop, ADX, and structure features, and the engine routes entries to the champion whose conditions are active.

Trend handoff and the chop governor

The trickiest part of any cascade is the handoff. A trend-birth thesis from Tier 3 must exit cleanly when chop returns. The Maverick engine handles this with what the architecture documents call the chop governor.

A trend-birth entry from Tier 3 must exit on a chop-14 reversal post-entry. A handbrake activates at plus three ATR versus entry. An absolute kill triggers if chop-14 climbs above 55. A gamma kill triggers if implied gamma exceeds 0.65. These are not soft suggestions. They are binding rails coded into the position manager.

This matters because the alternative is a champion that catches a beautiful trend birth, then sits through the inevitable chop that follows, then gives back all its gains when the chop resolves against the original direction. The chop governor forces the champion to hand off cleanly when conditions flip.

What this means for ARENA visitors

The ARENA public proof at arena.freedomcore.io shows the result of all three tiers running together. Each wallet page reflects the realised account outcome with the cascade active. The architecture is exposed deliberately because the result is only credible if the method is inspectable.

The architecture is also why the engine does not promise returns. Regimes change. A cascade designed for the current regime mix may underperform if the mix shifts. The honest answer is that the cascade improves the odds of catching the regime in which each champion was tuned to operate.

What is changing next

Two changes are in progress.

The first is making compression genesis live again. The champion is structurally sound but produced a negative replay in the last validation run because of stop-cluster behaviour near recent breakout pivots. The fix is a wider buffer plus a delayed entry trigger.

The second is integrating regime classifier feedback from SHADOW wallet behaviour. If many SHADOW wallets are simultaneously holding long positions in a specific symbol, that is information about market positioning that the cascade could use to weight its entries. The architecture for this exists but the integration is not wired yet.

The cascade is not a final answer. It is the current best answer to the problem of running an autonomous engine across changing regimes. Each champion is independently auditable. The handoffs are explicit. The result is published live on ARENA. That combination of inspectable architecture and live proof is what FreedomCore Maverick is built around.