HypeDailySingleSmaBreakoutLong

HYPE Daily Single-SMA Breakout Long (Hyperliquid)

A minimal long-only single-instrument trend-following strategy on Hyperliquid's native HYPE-USD perpetual using daily (1D) bars. The strategy uses ONLY OHLCV data and a SINGLE simple moving average — no funding rates, no L/S ratios, no liquidations, no pairs, no multi-timeframe — to keep the verification surface as small as possible after a string of `verifying_stuck_3_timeouts` failures. The thesis is that HYPE has structurally bullish demand drivers (Hyperliquid's USDC fee buyback program, exchange volume growth, AI-driven flow) which produce persistent multi-month trends well-suited to a slow daily SMA crossover. Daily timeframe + only 1-2 trades per quarter means fees are negligible (HL ~0.09% RT << expected per-trade move of 20-40%). This is a deliberately diversifying hypothesis: zero overlap with the 5 strategies already in pipeline (all Binance) and the 10 recent failures (all Binance/Binance Spot, mostly 4H/1H/8H, all using more than one signal source).

Deliberately minimal to keep the verification surface tiny after a string of verifying_stuck timeouts: one indicator (a single SMA), OHLCV only, single instrument, single daily timeframe, no supplementary data, no extra legs. calculate_signal returns the continuous close-to-SMA distance so the verifier sees a real, bar-varying signal that crosses zero on the breakout, while the discrete above-SMA boolean gates should_enter and position_size (avoiding the gated-to-0.0 anti-pattern). Entry above and exit below the slow SMA captures HYPE's thesis of persistent multi-month uptrends with only ~1-2 trades per quarter, so HL's ~0.09% round-trip fee is negligible against expected 20-40% trend moves. Venue is HYPERLIQUID with the correct native symbol HYPEUSD.HYPERLIQUID (not HYPE-USD/HYPEUSDT) and $10 min notional; the strategy is long-only with no leverage, so leverage stays 1.0 and the leverage_set_but_unused gate does not trigger. Zero overlap with the Binance pipeline strategies and the recent multi-signal failures provides genuine diversification.

Not testable yet: HYPE is a new token (launched ~late 2024) with only ~1.5 years / ~480 daily bars, and the strategy produced just 11 trades — far too sparse and too short to support optimization (walk-forward's 3 windows would get ~3-4 trades each and the holdout near-zero). The +71.7% headline is illusory: it is dominated by a single +80.0% day (2025-07-31, largest_win $79,973) with win_rate 27% and the remaining trades mostly small losses, so the result is a single-outlier point estimate (Sharpe CI -3.72 to 2.75 straddles 0, kurtosis 105, skew 9.95, tail_ratio 0.03), not a validated edge. The code is correct and trades as specified (11 entries, 0 dropped), so this is an insufficient-data/sparsity blocker, not a code defect — iterate cannot help and optimization would fit noise. Not worth 2 hours. META-NOTE: revisit a HYPE daily trend strategy once the instrument has multi-year history; the simple SMA-trend mechanism is reasonable, but it cannot be validated on ~1.5 years of single-outlier-driven data. FAILURE PATTERN: trend/breakout strategies on newly-launched tokens (HYPE, ~1.5yr history) produce too few trades (~11) over too short a sample to validate, with the headline return dominated by one outlier move — best-of-N optimization and walk-forward/holdout are meaningless until the instrument accumulates multi-year, multi-regime history.

Minimal long-only daily single-SMA trend follower on Hyperliquid's HYPEUSD perpetual, pure OHLCV single feed. Enters LONG when daily close is above the 50-bar SMA and exits when close falls back below it. The signal is the continuous distance from the SMA (close/SMA - 1) recomputed every bar; the above-SMA state gates entry. Sized at 50% of equity notional, long-only, leverage 1.0.

Implemented correctly and trades as specified: 11 SMA-breakout entries, 0 dropped; minimal single-signal design; negligible fee drag (0.72% of gross), huge capacity

Insufficient history: HYPE launched ~late 2024, only ~1.5 years / ~480 daily bars — cannot support walk-forward (3 windows) or a meaningful holdout

Far too sparse: only 11 trades total; optimization would fit noise on ~3-4 trades per window

Single-outlier-driven: the +71.7% total is dominated by one +80.0% day (2025-07-31, largest_win $79,973); win_rate 27%, the other trades are mostly small losses — strip the outlier and it's a net loser

Statistically meaningless: Sharpe CI -3.72 to 2.75 (straddles 0), return_kurtosis 105, skew 9.95, tail_ratio 0.03 — the headline Sharpe 1.33/PF 2.70 are artifacts of one trade

This was a deliberately minimal long-only daily single-SMA trend follower on Hyperliquid's HYPE, designed to shrink the verification surface and diversify off Binance. It was implemented correctly (11 entries, none dropped, negligible fees) and posted a flashy +71.7% with Sharpe 1.33 and PF 2.70 — but on only ~480 daily bars since HYPE's late-2024 launch, and the entire headline rested on one +80% day against a 27% win rate. The analyst abandoned it before optimization as untestable: too short and too sparse for walk-forward or holdout, with metrics that are single-outlier artifacts rather than a validated edge, recommending the idea be revisited once the token has multi-year, multi-regime history.

Trend/breakout strategies on newly-launched tokens cannot be validated until the instrument accumulates multi-year, multi-regime history — ~1.5 years yields too few trades and a headline return dominated by a single outlier move, making walk-forward, holdout, and best-of-N optimization meaningless; the simple SMA mechanism is reasonable and worth revisiting once HYPE has longer history.

The analyst abandoned it at backtest review as an insufficient-data/sparsity blocker rather than a bug — HYPE launched ~late 2024, so ~480 bars and 11 trades cannot support walk-forward (3 windows would get ~3-4 trades each) or a meaningful holdout, and stripping the lone outlier leaves a net loser, so optimization would only fit noise.

A deliberately minimal long-only single-SMA daily trend follower on Hyperliquid's native HYPE-USD perpetual (pure OHLCV, single feed) — going long when the daily close is above a 50-day SMA and flat when it falls back below — built to keep the verification surface tiny after a string of verification timeouts and to diversify away from the all-Binance pipeline.

Over only ~1.5 years / ~480 daily bars it made just 11 trades with a +71.7% headline return, Sharpe 1.33, profit factor 2.70, Sortino 9.62, and 33.1% max drawdown — but the result was a single-outlier point estimate: win rate 27%, the +71.7% dominated by one +80.0% day (2025-07-31, largest win $79,973), with the Sharpe CI straddling zero (-3.72 to 2.75), kurtosis 105, skew 9.95, and tail_ratio 0.03.