Why Automate?

With advancements in computer speeds, computers are able to search thousands of stock quotes and moves at once, and make trading decisions quicker than humans.

Computer and Programming Considerations

Because computer speed is the heart of automated trading, trading systems should be run on high-end, fast computers with high speed network connections. Major stock exchanges offer clients to locate their servers at the exchange's data centers, to speed up order execution.

Strategies

Any strategy can be entered into an automated trading system. Popular strategies include arbitrage, index benchmarking, and correlation trading.

Correlation Trading

Correlation trading is based on finding stocks that trade in similar correlations, and locating times when an individual stock or basket of stocks is outperforming another group of stocks. Automated systems will analyze the correlations and short stocks that are outperforming, and buy the laggards, with the goal of the correlation returning to historical averages.

Black Box Trading

Black box trading is a term used to describe a fully functional algorithmic trading system. Programming knowledge needed to run the system is minimal; however, users are limited to minimal changes that they can make to the settings and strategies of the system.

White Box Trading

As opposed to black boxes, white boxes are trading systems that are completely open to manipulation by users. However, advanced programming knowledge is needed to run a white box, as much of the strategies have to be input by the user.

Source:

exchange's data centers