Portfolio123

A ranking system scores the research universe; portfolio rules determine what a simulation buys, holds and sells. Portfolio123 lets you work on both layers. Its formula editor uses a spreadsheet-like approach, with predefined factors and custom calculations, so a research model does not have to begin with Python code.

Choose whether a factor ranks companies across the full universe, within a sector or within an industry. A strategy can then combine a custom universe, buy and sell rules, position sizing and hedging. Commission and slippage assumptions belong in the comparison because apparently similar models can produce different results after trading costs. The historical datasets include point-in-time fundamentals, estimates and corporate actions.

Book of Strategies brings several models together with chosen weights and correlation analysis. That is useful when the question concerns the relationship between strategies rather than the best result from one ranking. Broker-connected management is a separate stage from researching a historical simulation.

The homepage offers a 30-day Screener and Backtest trial without a credit card. Registration is required to build and retain research. Check the current subscription options before committing to a particular data region or research allowance.