Zorro Platform: Analyze
- Develop - Test - Trade
Since the 1990s, private traders can use the Internet to access the financial markets. Several trade platforms offer automated trading functions. But they are,
in fact, mainly designed for manual trading and not well suited for algorithmic strategies (see trading platform comparison). Consequently, private traders are not
very successful with automated trading. Those who are
normally don't use trading platforms, but self-written software.
Zorro is a free tool for executing financial algorithms. It
supports research, exploring, developing, testing, and executing
strategies for stocks, forex, options, futures, bonds, ETFs, CFDs,
or any other financial instruments. The focus is on advanced
statistical or machine learning algorithms. Zorro runs either as a
stand-alone system with direct connection to a broker or exchange,
or as an attachment to a trading platform, such as MTR4/MTR5 or TWS by Interactive Brokers™.
Converting a trade idea into an automated trade system, testing and trading it is a lot easier and faster with Zorro than with conventional platforms. Here's a tiny video course about developing a simple daily trade system:
- C based compiled script language.
- Native string, file, dataset, vector, and matrix functions.
- Event driven: Immediate reaction on price ticks.
- No limits: Full access to the Windows API and external DLLs.
- R bridge: Use R statistics or deep learning packages.
- Sytax highlighting script editor with command help.
- Single step mode for debugging trade strategies.
- Strategy programming course included.
- About 150 traditional and advanced indicators.
- User-defined bar types: Range, Renko, Point-and-Figure, Haiken Ashi...
- Curve pattern detection with Fréchet algorithm.
- Curve spectrum analysis with filter bank and Hilbert transformation.
- Market sentiment analysis.
- Correlation and autocorrelation analysis.
- Fuzzy logic for peak / valley, crossover, and pattern detection.
- Source code of most indicators included. Easy interface to add own indicators.
List of functions by category
Data Mining & Deep Learning
- Algorithm generator with multivarate logistic regression (Perceptron).
- Algorithm generator with pruned decision tree.
- Algorithm generator for price action trading with patterns of up to 20 variables.
- Generated trade algorithms can be exported in C code to other platforms.
- Training / prediction / live trading framework for all R machine learning packages.
- Deep Learning with Keras / Tensorflow, MxNet, H2O,
Historical Data Download
- Download tick* or minute historical price data directly from brokers or price sources.
- Access historical and live data from Google™, Quandl™, AlphaVantage™, Stooq™.
- Convert historical tick-based, M1, or EOD data for stocks,
options, and other instruments from all data providers in CSV or
- Multi-asset, multi-timeframe, multi-algorithm portfolio analysis.
- High backtest speed: up to 100 x faster than MetaTrader4™.
- Single-run and stepwise backtests.
- Precise broker simulation considering fees, margin, spread, swap, slippage.
- Options trading simulation based on historical options
- Out-of-sample testing with several data split methods.
- Anchored and rolling Walk Forward Analysis (WFA).
- Multiple CPU cores for parameter training and machine learning.
- Bootstrap / Monte Carlo analysis with randomized price and equity curves.
- Calculation of Sharpe Ratio, AR, CAGR, R2.
- User-defined optimization target.
- Separate parameter optimization for portfolio components.
- Robustness / consistency analysis through oversampling.
- Adjustable time offsets for feed-independent price data.
- Detrending on price, signal, or trade level.
- Sine, square wave, and noise generators for algorithm testing.
- Optimal capital allocation factor calculation with MVO and OptimalF algorithms.
- Seasonal analysis with day, week, month, or year profiles.
- Tick mode for precision, bar mode for speed.
- Automatic price history download and update.
- Balance curve and trade list import/export for performance analysis.
- Minute- and tick-based price data for major currencies and indices included.
- Detailed performance sheet with portfolio analysis.
- Bar, line, dot, or band chart for indicators or user defined functions.
- Statistics charts for price, profit, season, and parameter profiles.
- Profit and MAE distribution charts.
- Line and symbol drawing functions.
- Detailed trade statistics export to spreadsheet programs.
- User-defined data export to .csv data files.
- All asset types (option chains, stocks, forex, futures, CFDs, ETFs,
- Supports ~1100 brokers and exchanges (IB*, FXCM, Oanda,
- Identical software in test and trading for reproducible results.
- Customized strategy control panels with user-defined buttons and entry fields.
- Native portfolio support with multiple assets, algorithms, and time frames.
- Symbol mapping and broker-dependent parameters.
- Time frames from 1 millisecond* to 1 month.
- Tick-precise trade entry/exit handling with user-supplied execution algorithms.
- Entry limit, stop loss, profit target, profit lock, trailing, timed exit.
- Virtual hedging: Simultaneous long and short virtual positions.
- Full support for NFA and FIFO compliant accounts.
- Money management with OptimalF or Markowitz position sizing.
- Realtime trade simulation ('phantom trades') for equity curve trading.
- Realtime parameter optimization without interrupting the trade process.
- Open API interface for connection of any broker API (source code included).
- Open API interface for connection to external signal services.
- Remote control of external programs by sending key strokes and button clicks.
- Command line parameters for starting Zorro from external programs.
- No high-end hardware required; runs on any old PC or cheap notebook.
- Automatic re-login and seamless restart in case of Internet or PC outages.
- Minimum requirements: Win XP / 7 / 8 / 10, 1 GB RAM, Internet access.
- Runs on cloud servers (VPS) with Windows Server 2003 / 2008 / 2012 / 2016.
- Open concept: all file formats and interface structures are documented.
- New functions can be easily added through user plugins.
- Customized, spreadsheet-based user interface panels*.
- Customized / relabeled Zorro versions available on request.
- Zorro source code licenses available on request.
Included Z strategies
- Ready-to-run autotrading strategies:
- Z1: Trend following system with spectral analysis.
- Z2: Mean reversion system with Fisher transform.
- Z3: Price cluster based trading system.
- Z7*: Price pattern based system with machine learning algorithm.
- Z8, Z9*: Long-term trading systems by portfolio optimization.
- Z10*: Cryptocurrency trading system.
- Z12*: Anticorrelation based trend/countertrend system.
Zorro S version required for some functions.
Zorro version 1.96 was released and is available on the
Download page. This version
comes with an interactive chart viewer/debugger, an options trading workshop,
Visual Studio™ support, a function to record options chains, and many other new features and
improvements. Details on the What's New page.
A comparison table of Zorro vs. TradeStation™ vs. MTrader™ can be found on the FAQ page.