The so called Quant or Algo traders prefers to stay mostly away from Traditional trading platforms like Amibroker, Ninjatrader, Metatrader, Metastock, Tradestation etc. Do you know why?
These traditional trading platforms are designed to meet retail level expectation i.e you can connect up to various brokerages, downloading data, charting, custom indicators, backtesting..etc. It is relatively easy to understand their programming language and you can easily automate any simple trading strategies. Retail traders mindset is to create a simple plain vanilla trading algorithm and to find a edge in the trading system.
Retail traders prefer to stay away from trading strategies like Pair Trading, Options Mispricing…etc. Retail traders prefer Data management, coding indicators/signals, charting , backtesting and performance evaluation to be easier, They prefer to concentrate more on trading than their algorithms.
What is a Quant Trader and What Do They Do?
What are the issues with Traditional Trading Platforms?
1)Your trading algorithms are not cross portable. If you are writing a trading algorithm in one trading platform and tomorrow you willing to test another trading platform then you cannot easily port your trading system easily.
2) You cannot code complex mathematical models into traditional trading platforms. For example performing Co-integration between two trading pairs i.e ADF root test or performing complex statical models, implementing feed forward models , machine learning models is quite challenging in traditional trading platforms.
3)Most of the traditional softwares lack multithreaded backtesting capabilities.
4)Performance degradation of your trading system becomes poor when you are planning to test your models on bigger portfolio size.
5)Retail traders are dependent on trading system providers for support, bug fixes and updates.
6)Performing Monte Carlo testing is difficult in most of the trading softwares.
How does the quant traders solve such issues?
Generally most of the quant traders prefer trading programs like C, Python, R, Matlab,FPGA. They build their own data connectors, Order Management, Risk Management, Execution Mechanism, Data managament, backtesting/Optimization/Walkforward testing modules. Programs and modules/packages in python, C are open source and completely free, even for commercial use, as are many of the key scientific libraries. Building complex statistical models are quite easier which the retail level trading platforms are not able to solve. Parallelization and ability to hand huge computational power are another key to depend on such programming models which gives the scalability to your portfolio.
If you ask me to rate the retail level trading platforms for algo trading then multicharts comes at first as it support for backtesting portfolio of strategies, pair trading backtesting, optimization etc which lacks in most of the traditional trading software.