Quick start
Run your first backtest in the browser — pick a universe, strategy, and optimizer, choose dates, run, read the metrics, and save or publish the result.
This walks you through a full backtest on the Backtest page. It runs entirely in your browser; you can do the whole thing signed out, and only need an account when you want to save a run.
1. Open the Backtest page
Go to /app/backtest/. You’ll see a configuration form on one
side and an empty results area on the other.
2. Pick a universe
Enter a handful of ticker symbols — for example AAPL, MSFT, GOOGL, AMZN, NVDA.
This is the pool the strategy ranks and the optimizer allocates across. Prices
are fetched once and cached in the browser. Not sure which tickers? Start from a
watchlist or the
strategy catalog.
3. Choose a strategy
The strategy decides which assets to hold each rebalance. Three are built in:
| Strategy | Selects | Key params |
|---|---|---|
| Momentum | strongest recent price move | t (lookback) |
| Mean reversion | most oversold vs. its moving average | t / window |
| Smart beta | best risk-adjusted return (mean/std) | window |
Details and formulas: Strategies.
4. Choose an optimizer
The optimizer decides how much of each selected asset to hold:
- MSR — Max Sharpe Ratio
- GMV — Global Minimum Variance
- EW — Equal Weight
See Optimizers.
5. Set the run parameters
- Start / end date — the backtest window.
- Initial capital — starting portfolio value (USD).
- Top N — how many top-ranked assets to hold each rebalance.
- Rebalance frequency — trading days between rebalances.
- Lookback (
t/window) — how far back the strategy looks. - Risk-free rate — annual rate used by Sharpe/Sortino (e.g.
0.04). - Transaction costs (optional) — commission + slippage in basis points.
Every field maps 1:1 to an engine parameter — see the
BacktestConfig reference.
6. Run it
Press Run. The backtest executes in a background Web Worker with a progress indicator; the UI stays responsive. When it finishes you get:
- a daily equity curve overlaid with the buy-and-hold benchmark (default
SPY), - headline metrics (annualized return, volatility, Sharpe, Sortino, max drawdown, Calmar, win rate, profit factor),
- monthly returns and the deepest drawdown episodes.
7. Read the metrics
Skim in this order: annualized return (did it grow?), Sharpe / Sortino (was the growth worth the risk?), max drawdown (worst peak-to-trough pain), then the benchmark line (did you beat buy-and-hold?). Each metric is defined in Metrics, and taught with a live example on Learn.
8. Save or publish
Signed in, click Save to persist the run to your workspace, or mark it public to add it to the community leaderboard. See Publishing & sharing.
Next
- Test more than one idea at once → Backtest Lab
- Save a reusable configuration → Strategies workspace
- Understand the numbers → Metrics