Features
Indicators and signals
Compute indicators, build your own, stream updates, and turn them into signals
Compute indicators, build your own, update them bar by bar, and turn them into signals or learning targets. This section is for anyone who builds the inputs of a strategy, from standard indicators to their own formulas or a model's predictions. The pages below go from built-in and custom indicators to streaming updates, trading signals, and labels for machine learning.
Start here
| What do you want to build? | Where to start |
|---|---|
| RSI, moving averages, bands, or another existing study across symbols and settings | Technical indicators |
| Your own formula or a reusable version of a pandas calculation | Custom indicators |
| Calculations that continue as new bars arrive | Streaming indicators |
| Entry and exit conditions from crossings, levels, or calendar rules | Trading signals |
| Return, breakout, or trend targets for a prediction model | Machine learning labels |
Indicators and signals pages
Technical indicators
Run TA-Lib, pandas-ta, ta, SMC, WorldQuant Alphas, and native indicators at scale
Custom indicators
Turn Python, pandas, or Numba calculations into reusable indicators with parameter grids
Streaming indicators
Update indicators bar by bar with stateful accumulators that match batch results
Trading signals
Generate, combine, clean, and analyze entry and exit signals before backtesting
Machine learning labels
Create triple-barrier, trend, and pivot labels, and backtest model predictions
From a calculation to a strategy
An indicator gives you values, such as RSI or distance from a moving average. A trading rule turns those values into signals, such as buying when RSI crosses back above a threshold. You can compare windows and thresholds across several symbols, inspect how often the signals fire, and then pass them to Signal backtesting to measure the trades.
Start with an integrated library or bring your own Python, pandas, or Numba calculation. Custom indicators use the same labeled outputs and comparison helpers, so you can combine a familiar indicator with your own research feature. For larger searches across settings, continue to Parameter optimization.
If you use machine learning, indicators can supply model features and labels define the future outcome the model should learn to predict. The model's predictions then become trading inputs. The labels page shows that full workflow, including a comparison with a simple baseline.
See the Indicators documentation for the development pipeline, or follow the Signal development tutorial for a worked strategy.
Copyright © 2021–2026 Oleg Polakow. All rights reserved.
Site content and documentation are provided for using and evaluating VectorBT PRO and for educational purposes. Any other use, including building or supporting competing products or services, requires prior written consent.