Soomario Alpha is a market-relative strategy on Hyperliquid: one leveraged long in HYPE paired against a broadly diversified short basket across the altcoin complex. The thesis is relative value - long the outperformer, short broad altcoin beta. It is live now as Alpha 2.0, a small real-money forward test.
Alpha pairs a single leveraged long in the structurally strongest asset against a broad short basket spanning the altcoin complex. In a regime where the strong asset rises while the broad alt complex falls, both legs profit. It is one continuous book, two legs, one shared risk framework, inside a single cross-margin account on Hyperliquid.
A leveraged long in HYPE sized to the minority share of gross exposure, opened once and held, carrying a hard exchange-resident stop below average entry and a documented core floor. In the forward test it is deliberately a simple hold - the active recycle engine is reserved for a later phase so the short sleeve can be validated in isolation. Opened exactly once, no averaging down.
A broadly diversified basket of liquid altcoin shorts sized to the majority share of gross exposure. Each name builds in a small number of front-loaded tranches, one per completed overbought cycle, to a fixed, uniform per-name cap on filled cost basis. Breadth is mandatory: many small names, not a few large ones. The diversification itself neutralizes single-name squeeze risk.
Each name shorts on confirmed weakness. A multi-hour relative-strength measure, on closed candles, must push up through its overbought line, ride, then close back down through it - and only when the cycle peak cleared a hysteresis gate. It scales in at progressively better prices across separate overbought events, up to a small per-name maximum, then goes quiet.
In a broad altcoin melt-up the strength measure pins at overbought and never closes back down, so no new tranches fire. The entry logic goes silent in exactly the regime that would otherwise feed a squeeze. After any cover, a name waits a cooldown of several bars before it can re-short.
Every name builds only to a fixed, uniform cap on filled cost basis. Max risk per name is fixed before its first tranche - tranches pace the entry, they do not raise the ceiling. The cap is enforced by both the engine's tranche counter and an independent cost-basis guard.
Each short carries a reduce-only stop resting on the exchange, a fixed distance above average entry. These covers are never widened, never defended, never averaged. They rest on the exchange, so they execute even if the bot is offline, and a fill-slippage pad past the trigger ensures a violent gap actually fills.
The correlated, whole-complex squeeze is handled by a basket-wide breaker that measures the short sleeve's drawdown from peak as a percentage of total account equity. On breach it cancels covers, mass-covers the entire short sleeve, pauses new entries, and requires manual human re-enable. Two states only: running and paused. No auto-recovery.
A hard low leverage cap refuses new entries before it is breached, while existing positions and stops are never touched. Every cycle the bot reads positions, orders, equity, and prices from the exchange and reconciles its own state, so its view can never durably diverge from the exchange.
The validated history sits within one regime - a strong asset outperforming a falling alt complex. The regime that ends the trade, a coordinated altcoin melt-up where short correlations converge to one, has not occurred in the sample. Survival through it is modeled, not observed. The canary is deliberately small precisely so this untested tail is met at a size that can absorb it.
Alpha 2.0 verifies that the entry mechanic and the risk controls behave as designed under real fills, real funding, and a possible squeeze - things no backtest can confirm. Success is defined as the caps and stops holding, not as a profit target.
A clean, breaker-caught cascade that loses money is a successful test. A resident cover that fills within its modeled band is a pass. A breaker trip is a pass - the safety working. The bot instruments every cap and stop with pass/fail events.
A position averaged past its cap, a breached cover that never fired, a cover that filled late, or a drawdown that crossed the threshold without tripping the breaker. The verdict is fail if any failure event has ever been logged, else pass.
Alpha 2.0 is a small capped-size canary running real money to validate execution. There is no deposit today. A Hyperliquid vault is intended only after the forward test validates - anticipated terms are a 10% profit share on realized profit, no management fee, with depositors retaining custody of vault shares.
The edge is demonstrated within a single favorable regime. Its adverse regime - a broad, correlated altcoin rally - has not occurred in the validation sample. Survival through it is modeled, not observed. Past behavior does not predict future results.
The book is concentrated in a single long (HYPE) and runs leverage. Leveraged perpetual futures and cryptocurrency involve substantial risk of loss, including total loss of deposited capital. A short basket carries the risk of a coordinated squeeze in which diversification does not protect the book.
Individual shorts are expected to lose - the edge rests on breadth across many names. Funding is a structural cost the book pays to hold both legs, so a small book builds buffer slowly. The strategy depends on a live process and exchange connectivity.
No deposit, no button - just a small real-money canary being validated in the open. Follow progress on Discord or X, and watch for the vault launch once the forward test completes.
HYPE + AVAX directional. Concentrated long-bias bot, 10% profit share, no monthly fee.
Multi-signal accumulator with layered defense across 15 stocks and crypto. Live beta on Hyperliquid + OKX copy trading.
Discretion-free DCA at liquidation zones. Two-level entry stack, midpoint take-profit, retrace trail across a curated universe.