Hyperliquid: What Our Signals Do

Four trades. Seven months. A profit factor of 5.57.

That's either the start of something real or a lucky streak in thin data. We genuinely don't know yet — and we think you deserve to hear that upfront.

HYPE is a governance token for Hyperliquid, a decentralized perpetual futures exchange. We added it in April 2026 with the most caveats of any asset in our system. Here's exactly what the data shows, what it doesn't, and why we included it anyway.

What the Backtest Showed

We tested HYPE across both daily and 4-hour candles with ADX thresholds from 15 to 40. The daily timeframe failed completely — no configuration produced a profit factor above 1.0. Every combination lost money. HYPE went through a 60% drawdown from October 2025 through January 2026, and daily MACD was too slow to react. Every bullish crossover was a bear market bounce that reversed within days.

Switch to 4-hour candles with ADX>35, and the numbers change dramatically:

Profit factor: 5.57
Win rate: 75% (3 wins, 1 loss)
Alpha versus buy-and-hold: +41.4%
Mean quarterly alpha: +4.30%
Worst single-trade loss: -2.38%

The ADX>35 filter completely avoided Q4 2025, when HYPE crashed 47.6%. Zero long exposure during the worst quarter. ADX was below 35 the entire time — correctly flagging weak trend conditions.

The cost: Q1 2026, when HYPE rallied 54.3%. ADX was also below 35 during most of the recovery, so the strategy sat in cash. You can't have the crash protection without occasionally missing the bounce. Remember the weather forecast analogy from Why ADX Matters? ADX>35 is a very strict forecast — it only says "real storm" when conditions are extreme.

Why 4h and Not Daily

HYPE is the second asset (after ATOM) that breaks our daily-only rule. The reason is structural.

During HYPE's October-January drawdown, daily MACD generated six consecutive losing trades. Each one was a 1-3 day relief rally that MACD interpreted as a trend change — but on daily candles, the signal takes 2-3 weeks to generate and resolve. By then, the bounce was already over. The 4-hour timeframe reacts faster. Signals fire and close in 2-5 days, keeping losses small when they happen.

We covered why daily is usually better in How Our Strategy Works. HYPE is the exception that proves the rule — some assets move too fast for daily signals to keep up.

The 4h Early Warning

The most actionable finding isn't the trading strategy. It's the crash detection.

Unfiltered 4-hour bearish MACD signals captured 82.4% of HYPE's major crashes (>10% drawdowns) with an average lead time of 73 hours. That's roughly three days of warning before the worst of the move hit — and 74 hours ahead of the equivalent daily bearish signal.

The tradeoff: a 58% false positive rate. More than half the 4h bearish signals didn't lead to a real crash. That's why we show these as a passive indicator on the Brief screen, not as push notifications. If you see a 4h bearish signal on HYPE, it's a data point worth noting. Not a sell signal.

What We Don't Know

We need to be direct about the confidence level here.

Four closed trades over seven months. The profit factor of 5.57 has an enormous confidence interval — it could settle anywhere from 0.5 to 8.0 as more data comes in. ADA, by comparison, has 70+ signals across 20 quarters. HYPE has 4 signals across 3.5 quarters.

We don't know if the ADX>35 threshold is capturing a genuine pattern in HYPE's trend behavior, or if we're looking at a small sample that happened to work. We don't know if the 82.4% crash capture rate on 4h bearish signals will hold as the dataset grows. We don't know if HYPE's volatility profile will shift as the Hyperliquid ecosystem matures.

We plan to re-run this study at 14+ months of data, around Q4 2026. If the numbers hold, the provisional label comes off. If they don't, we'll adjust or remove HYPE from the signal system. That's how we handle every asset — the data decides, not our preferences.

What to Expect

HYPE generates roughly 7 signals per year on 4h ADX>35. That's less frequent than most of our assets — you might go weeks without a signal. When one does fire, it means ADX has climbed above 35, indicating a strong trend is in play.

Bullish signals only fire when ADX>35 — weak trends are filtered out. Bearish signals always fire, regardless of ADX. Same long-only approach as our other assets: buy on bullish crossover, exit to cash on bearish.

HYPE is the newest, least-tested asset in our system — promising early numbers on a tiny sample, with honest uncertainty about whether they'll hold. We'll show you exactly what happens as the data accumulates.

This is educational content, not financial advice. Past performance does not guarantee future results. Based on 7 months of 4h data from Polygon.io (Sep 2025 – Apr 2026). Results are provisional and will be re-evaluated with additional data.