backtest-strategy

name	backtest-strategy
description	Guide agents through backtesting strategy ideas with automatic data fetching and performance analysis
metadata	{"tags":"backtesting, strategy, performance, sharpe, drawdown, funding, simulation, yield, lending, carry, delta-neutral"}

When to use

Use this skill when you are:

• Backtesting an existing strategy from wayfinder_paths/strategies/
• Validating a new trading strategy idea before production deployment
• Analyzing historical performance (Sharpe, drawdown, CAGR, funding PnL)
• Testing any strategy type: momentum, delta-neutral, yield rotation, carry trade
• Testing different leverage levels or parameter combinations

How to use

First, determine if you're backtesting an existing strategy or a new idea:

Backtesting an existing strategy (from wayfinder_paths/strategies/)

Load these rules:

1. rules/backtesting.md — Always load first. Config reference, stats format, gotchas.
2. rules/existing-strategies.md — REQUIRED for existing strategies. Workflow for reading strategy source code, extracting parameters, fetching real Delta Lab data, and faithfully reproducing signal logic. Never use generic helpers with default parameters for existing strategies.
3. Load the strategy-type-specific rule if applicable (yield-strategies.md).

Backtesting a new strategy idea

Load these rules in order (most to least specific for your strategy type):

1. rules/backtesting.md — Strategy type → helper mapping, quick start examples, config reference, stats format, gotchas, production path. Always load this first.

2. rules/yield-strategies.md — Detailed patterns for lending/yield strategies: supply rate rotation, leveraged yield loops, carry trade, multi-venue benchmark. Load when the user's strategy involves lending protocols, supply APRs, or borrow rates.

Examples

• examples/basic_momentum.py — Cross-sectional momentum using quick_backtest
• examples/delta_neutral.py — Delta-neutral basis carry using backtest_delta_neutral
• examples/yield_rotation.py — USDC rotation across lending venues using backtest_yield_rotation
• examples/carry_trade.py — Borrow cheap / supply expensive using backtest_carry_trade

Strategy type → helper cheat sheet

Strategy	One-liner
Momentum/trend (perp)	quick_backtest(strategy_fn, symbols, start, end)
Delta-neutral basis carry	backtest_delta_neutral(symbols, start, end)
Yield rotation (lending)	backtest_yield_rotation(symbol, venues, start, end)
Carry trade (borrow/supply spread)	backtest_carry_trade(symbol, start, end)
Active perps / HIP-3 (deployable)	backtest_perps_trigger(signal_fn, decide_fn, symbols, ...)
Full control	run_backtest(prices, target_positions, config)

All helpers are in wayfinder_paths.core.backtesting.

For actively-traded Hyperliquid strategies (perp or HIP-3), use backtest_perps_trigger and the trigger pattern. The same signal.py and decide.py modules run in backtest, live, and reconcile — drift is architecturally impossible. See rules/perps-trigger.md.