Features

StrategyQuant is an unique software that generates, backtests and optimizes strategies automatically and does not require programming skills. The strategy can be easily saved in the source code of your trading platform. It allows you to:

Build unlimited number of trading strategies

Export your strategies to MetaTrader 4/5 or Tradestation with full source code

Develop strategies for any market or timeframe, even multi-market and multi-TF

Improve existing strategies by altering the trading rules

Run automated robustness tests to reduce the risk of curve-fitting

Optimize your strategies with Walk-Forward optimizer

Basic features - Strategies builder

The basic functionality of StrategyQuant is to create, retest and optimize wide range of strategies.

Robustness tests / Overfitting protection

Overfitting (curve-fitting) is a serious problem in all areas  related to machine learning and data mining. 

What exactly is overfitting? See equity charts for two different strategies below.

Left strategy trades well also on the unknown data, while right strategy fails miserably – possibly because it was produced to “fit” the data on which it was generated and it has no real edge.

StrategyQuant X contains a number of features that can help you recognize these bad, overfitted strategies before they cost you money in live trading.

Advanced and unique features

Overfitting (curve-fitting) is a serious problem in all areas  related to machine learning and data mining. 

Very fast backtesting engine with real tick precision

StrategyQuant has its own very fast backtesting engine that can make thousands backtests per second – depending on your data and test precision. This allows SQ X to generate and review tens of thousands strategies per hour.

The backtesting engine was made to match exactly backtests in your trading platform, while being few times faster.

Multi-TF and Multi-symbol strategies

StrategyQuant can generate strategies that look at multiple charts – for example it can trade on H1 timeframe, and look also at signals confirmation on H4 and D1 timeframes.

Or it can trade on EURUSD and check also indicators or price data on GBPUSD and USDCHF charts. You can configure how many signals should be generated for every chart, and what will be its properties.

This allows to create a new class of more robust trading strategies.

No programming knowledge necessary

You don’t need programming skills to start trading algorithmic strategies. With StrategyQuant you can build new algo strategies with a click of a few buttons.

Generate thousands of unique trading strategies in Builder mode

Builder is the main mode of StrategyQuant. It allows you to generate thousands of unique strategies based on your desired properties. 

It is based on machine learning techniques, like random generation and genetic evolution, it combines all available building blocks, entry and exit types to generate strategies that fit the selected properties.

Just configure your input data, selection o ftrading signals and blocks, your desired parameters and len Builder generate new strategies for you.

Avoid overfitting with integrated robustness tests

A big and important part of StrategyQuant features are build-in tools and checks to ensure the generated strategies are not overfitted to existing data – in other words that they have real edge and will continue to work also in the future.

 

Robustness tests (cross checks) have been integrated into the strategy generation process, you can turn them on / off with a click of a button. 

 

Every additional test takes time, so SQ also estimated how much time it takes for evey test to finish per strategy.

Output your strategy to multiple trading platforms

StrategyQuant outputs full source code of your strategy, nothing is hidden.

 

Platforms supported right now: MetaTrader4, MetaTrader 5, Tradestation / MultiCharts.

 

Coming soon: NinjaTrader, Quantopian, QuantConnect, cTrader.

Support for all standard technical indicators

StrategyQuant supports all standard technical indicators (like CCI, RSI, MACD and so on). It also supports different candle formations, 4 types od trade entries and 6 types of exits, and the list will be continuously growing.

 

What’s more – with a little programming you can easily extend StrategyQuant with your own favorite indicator, or ask us to do it for you.

Automatic custom workflow

You can automatize your workflow in Builder, or use Custom projects to build your own unique workflow with as many steps as you like.

 

You can have multiple builds, retests, optimizations and other tasks chained one after each other to achieve exactly what you want – and let it run in a fully automatic mode.

Two types of Monte Carlo tests

Two separate Monte Carlo tests with 8 different simulation types allow you to simulate the behavior of your strategy with different random variations.

 

You can incorporate Monte Carlo tests directly into your building workflow and automatically reject strategies that don’t pass the Monte Carlo tests.

Walk-Forward Optimization and Matrix

Build-in Optimizer allows you to optimize your strategies either in a simple way, or
in Walk-Forward mode.

 

Equity chart is displayed for both original and optimized strategy with al the Walk-Forward periods.

 

With WF Matrix (cluster analysis) you can verify if there are clusters of “optimal” and stable parameters for your strategy.

What's coming in the future?

StrategyQuant X is under continual development. Here are some of the things you can look forward to in the future versions:

Optimization profile, Parameter permutation method

new advanced cross check modules to protect against overfitting

NinjaTrader (C#), Quantopian (Python)

support for generating trading strategies for other platforms

Deep Neural Networks in trading

train and use neural networks as a part of your strategies

Grid support

run SQ on multiple computers for better performance

And Much More…

Get fully functional trial version

Request a free trial and put StrategyQuant X to test.