Why Your Hyperliquid Backtest Doesn't Match Your Live Results
If you backtest a Hyperliquid strategy using price-time priority — the way a centralized exchange matches orders — your live results won't match your backtest. Hyperliquid's HyperBFT consensus doesn't work that way.
HyperCore's real execution order
There are three block-priority tiers, in this exact order:
- Non-GTC/IOC orders
- Cancels
- GTC/IOC orders
Plus hourly funding applied on its own schedule. None of this matches a CEX-style continuous price-time model, and most generic backtest engines silently assume the CEX model because that's what they were built for.
What the HyperBFT Backtest Ordering Engine does
A Python/Polars backtest engine that replicates HyperCore's actual 3-tier priority and hourly funding — so your backtest results actually mean something when you go live.
Requirements: Python 3.x + Polars. Bring your own input data (Parquet).
Status: 3/3 tests passing.
Honest note: this is a simulation engine, not a bot. It doesn't trade or send orders — it just makes sure your backtest logic matches how Hyperliquid actually executes.
