Strategy Types

This catalog documents the trading strategies supported by our backtesting system, their intended market structures, technical logic, input assumptions, and application guidance.

Showing 12 of 12 types

directional

Beat baseline using forecast direction

Directional vs baselinePre-gate D-1 forecast vs baseline

1 ex1 params

0 required1 optional

threshold

Filter noise via deviation threshold

Thresholded directionalPre-gate D-1 forecast vs baseline

1 params

1 required0 optional

contrarian

Exploit forecast biases (opposite direction)

ContrarianPre-gate D-1 forecast vs baseline

1 params

0 required1 optional

quantile

Size by percentile of deviation

Quantile-based sizingHistorical quantiles of `forecast-baseline`

1 params

1 required0 optional

quantile_high

Trade high-end forecast error extremes

Tail (high)Rolling quantiles of forecast error

2 params

2 required0 optional

quantile_low

Trade low-end forecast error extremes

Tail (low)Rolling quantiles of forecast error

2 params

2 required0 optional

quantile_extreme

Trade both tails of forecast error

Tails (both)Rolling quantiles of forecast error

3 params

3 required0 optional

auction

Optimize DA bid curves

Dual MILP optimizationPre-gate inputs; asset constraints

10 ex4 params

4 required0 optional

inter_hour_spread

Mean reversion in hour spreads

Spread z-scoreLive intraday prices (hour-pair z-score)

3 params

3 required0 optional

volatility_regime

Capture vol upshifts via straddle-like logic

Vol regime ratioLive rolling vol vs 7-day MA

3 params

3 required0 optional

forecast_anomaly

Contrarian on anomalous forecasts (ML)

IsolationForest anomalyContinuous retraining on forecast errors

3 params

3 required0 optional

calendar_roll

DA–ID premium convergence (calendar/roll)

Roll premium z-scoreNext-day vs current forecast premium (live)

3 params

3 required0 optional