Backtest a Funding-Rate Rule in One Prompt

Funding rates are now live in alerts and backtests

Stingray takes a trading rule you write in plain English, builds the alert, and backtests it against history before you turn it on. Funding rates are now available as a trigger inside that flow, the same way price and volume already were. The live values come from Hyperliquid, with 5-minute resolution in backtests.

Text me when BTC funding on Hyperliquid goes negative and BTCUSDT rises 0.5% in 1 hour

Negative funding is shorts paying longs to keep a position open. When the price also rips upward, those shorts are trapped, paying carry to a position that’s moving against them. That’s the cleanest setup funding gives you. Express it in one sentence:

Create a draft alert that fires when (1) BTC funding rate on Hyperliquid is below 0 bps/hr AND (2) Binance BTCUSDT moves UP at least 0.5% over 60 minutes. Keep it as a draft, do not activate it.

Stingray turns the sentence into a rule that waits at least 1 hour before firing again (so a setup that stays true for hours triggers once, not dozens of times), and runs it against history before you turn it on.

37 fires, +8.3% compound, beats BTC by 3.4pp over 28 days

Over the last 28 days, 2026-04-20 to 2026-05-18, the rule had 150 raw 5-minute windows where both conditions held (HL funding from the Hyperliquid onchain feed, BTCUSDT price from Binance). With the 1h cooldown applied, 37 of them actually fire, roughly every 17 hours. Forward returns on BTCUSDT:

• Avg 1h return: −0.03%
• Avg 24h return: +0.41%
• 62% of fires positive at 24h, median +0.55%

Compounded across the 13 non-overlapping 24h holds the rule produces, that’s +8.3% over the window, above BTC’s buy-and-hold of +4.9% (a 3.4pp edge).

See every trigger. Numbers are pre-fee, pre-slippage. Subtract about 1pp for fees and slippage and the rule still beats BTC.

The 1h number is indistinguishable from noise on a 28-day window. The trigger is coarse, so the first-hour drift won’t register. The 24h figure is where the bias shows up, and it shows up directionally. We tested the opposite version (negative funding plus a DOWN move) over the same window and it underperformed BTC by 2pp. The edge is one-sided: it lives in the up half of the construction.

Running the backtest first means you see this shape, and the comparison to buy-and-hold, before you turn the rule on.

Backtest your own version

The +8.3% is specific to this rule. Change the threshold or the asset and the result changes. Write your version, run the backtest, see what it does before turning it on.

The same rule alerts you, and trades for you

When the conditions are true, Stingray pings Telegram with the funding snapshot, the price context, and a live-view link. If you’ve connected an exchange, the same rule routes to an execution leg. One construction, no copy-paste, no manual translation between an idea and a position.

Want to try your own strategy?

Open the card and adapt it. Type your version in. Backtest it. Turn it on if you like the shape.