A discretion-free DCA strategy that enters at three levels anchored to aggregated liquidation Max Pain. L1 marketable at price, L2 at the Max Pain itself, L4 at the deepest band. Cascading take-profits, a self-tightening trail, and a catastrophic stop placed outside the L4 deploy zone.
Most retail DCA strategies build their ladders on fixed percentages — buy every 2% down, every $1 of price. This treats price as homogeneous, which it is not. In leveraged markets, price moves through clusters of stop orders and liquidation triggers that are geographically dense around specific levels, and when price arrives at such a cluster, forced selling accelerates the move and reliably overshoots before reversing. Liquidation Max Pain — computable in real time from aggregated open interest across Binance, Bybit, and OKX — is the closest publicly observable proxy for where leveraged positioning will be forcefully unwound. Empirically, price revisits the long Max Pain during pullbacks more often than round numbers, moving averages, or named support, and the bounces from those visits are more violent than average. Zones places its entries at the four-level band around that anchor.
When a candidate clears conviction and proximity, L1 enters as an aggressive marketable limit at current price plus 50 bps. L2 and L4 sit as post-only limits at the Max Pain level and the deepest band — all three orders submitted as a single atomic batch so they hit the book in the same block. Weights: ~28 / 34 / 38%. The deeper levels carry larger size because they're objectively better prices.
Partial closes fire at +2.5%, +4.0%, and +5.5% from weighted-average entry, closing 30/30/20% of original size and leaving a 20% runner. A trailing stop arms at +5% and tightens after each TP fires, so each successive leg of the move is captured at a tighter giveback than the last.
If the trail and TPs don't fire, the position can still hit the catastrophic stop — placed 5% below the L3 zone reference, intentionally outside the L4 deploy band. L4 gets room to fill and average down before any forced exit. When price crosses, an IOC market-close exits all remaining size at once.
The original Zones design deployed at four levels: L1 / L2 / L3 / L4. After roughly two months of live operation, a pattern became visible in the closed-trade record. Positions that filled L1, L2, and L3 but never reached L4 had a mean realized PnL of −$10.12 across 22 historical trades. Positions that did fill L4 — meaning price moved deep enough that the strategy's full DCA conviction was warranted — had a mean realized PnL of +$2.22.
A fill at L3 indicated that price had pushed deep enough to threaten the position but not deep enough to trigger the deep-DCA L4 fill that materially improves average entry. Positions that stalled at L3 averaged down enough to commit meaningful capital but not enough to be priced for the reversion the strategy is built to capture. L3 was retired from the deployment in May 2026.
The freed allocation moved to L1 / L2 / L4 in renormalized weights (~28 / 34 / 38%). L3 retains one role — as the reference price for the catastrophic stop — but no longer carries any deployable size. The change improved expected per-trade economics without changing the conviction stack or the entry conditions.
A 48-hour time-based force close, originally a backstop against stale positions, was analysed against the same closed-trade record. Time-exit closes were nearly always P&L-positive but consistently truncated otherwise-running winners. The feature was disabled pending further validation; positions now run until TP, trail, or catastrophic SL fires.
Live data through the strategy's first 67 days. The numbers reveal a clear directional asymmetry that has shaped the current configuration.
The short side underperformed in the current market regime — liquidation cascades on the downside more often produce follow-through than mean reversion, while upside cascades tend to revert cleanly. Shorts have been disabled at the configuration level pending further regime analysis. The vault currently runs in long-only mode. Max single-position drawdown across the live window was 11.8% against position notional.
The strategy's edge is conditional on liquidation cascades overshooting their Max Pain anchor — the structural pattern that has held in the current market regime. If aggregated leverage profile changes materially, or if exchanges modify margin mechanics in ways that dampen cascade behaviour, the edge would degrade. The long-only restriction is itself an acknowledgement that the short-side variant of the same thesis is currently impaired.
The proximity gate rejects most candidates most of the time. A typical scanner cycle produces ten ranked candidates; in practice, only two to three clear conviction and zero to one is close enough to L1 (within 2.0% of the band) to actually deploy. Capital is often idle waiting for setups that meet both filters simultaneously.
If price breaks below L4 without bouncing — a regime where cascades follow through rather than reverse — the catastrophic stop fires for a meaningful loss. The 5% buffer below L3 provides room for L4 to fill, but it does not protect the position from trends that simply continue past the strategy's lowest deploy zone.
Zones uses modest perpetual leverage. The catastrophic stop and per-position safeguards reduce tail risk, but Hyperliquid's standard liquidation mechanics still apply at the position level. The full deposit is at risk of loss.
Deposit USDC into the Hyperliquid vault. The strategy trades for you. Withdraw any time — fees only on profit.
Same liquidation-cascade thesis, different timing. Max Pain enters at single-shot exhaustion across 18 coins with per-coin parameter tuning. Hybrid TradingView + Python execution.
Same Max Pain anchor, different execution. Bidirectional concentric grids placed around the zone — monetises the oscillation rather than the entry-and-exit through it. ~50 perps scanned, 5 concurrent grids.
$7/mo signals service for the buy-the-dip retail investor. No leverage, execute on your own broker.