Strategies & Builder¶
Built-in strategy catalog¶
| Strategy | Type | Best for |
|---|---|---|
| MACD Momentum | Trend-following | Single-stock daily swings |
| Mean Reversion | Counter-trend | Range-bound stocks, RSI signals |
| Trend Following | Trend | EMA crossover with ADX filter |
| Stochastic Crossover | Counter-trend | Overbought/oversold reversals |
| Volume Breakout | Trend | Confirms direction via volume spikes |
| VWAP Trend | Trend | Intraday/multi-day VWAP-anchored |
| Covered Call (F&O) | Income | Premium collection on held stock |
| Calendar Spread (F&O) | Volatility | Time-decay capture |
Each strategy exposes a small set of parameters (EMA periods, stop-loss %, etc.) you can tweak before running. Defaults are honest, not curve-fit to a specific stock.
Strategy Builder (no-code)¶
The Strategy Builder lets you compose a custom strategy from indicator building blocks. You pick:
- Entry condition — e.g.
RSI < 30 AND MACD > Signal - Exit condition — e.g.
RSI > 70 OR stop-loss hit - Position sizing — fixed quantity or % of capital
- Risk — stop-loss %, trailing stop, max position size
Save it; it now shows up in the strategy dropdown on the Backtest page next to the built-ins.
Start simple
First-time builder users often combine 4-5 conditions and end up with a strategy that fits the past but won't generalise. Start with one indicator pair; only add more if you can explain why each one helps and the out-of-sample performance survives adding it.
Optimizer¶
The Optimizer runs a grid search across a parameter range and ranks the runs. Pick the strategy, pick which parameters to vary (e.g. EMA fast 5-15, EMA slow 20-50), and the grid is materialised.
The result table is sortable by total return, Sharpe, drawdown, win rate. Look at all four, not just total return — the strategy with the highest return often has the worst drawdown and is the most curve-fit.
Overfitting is the enemy
Grid search will always find some parameter combination that worked well in the past. That doesn't mean it'll work in the future. Best practice:
- Out-of-sample test: optimize on 2020-2023, then run the best parameters on 2024-2025 untouched. If it falls apart, your strategy was fitting noise.
- Pick a robust region, not the single best cell. If only one cell in the grid looks good and everything around it is bad, that cell is luck.
Scanner¶
The Scanner runs a quick condition (e.g. "RSI < 30 today") across a list of symbols and shows the matches. Useful for finding setups manually before building them into a strategy.