When Is a System Suitable for Automation?

While the average trader can make money using any given toolkit, there are some cases that are not well-suited for complete automation. Some examples include:

Event Driven Systems: Given their spontaneous nature, events require manual entry. However, a computerized system can aid event traders by providing total control following the entry, all the way to the exit of the trade.

Non-Quantitative Systems: Although mechanically driven systems have grown in importance, many traders still rely on fundamental signs that do not willingly lend themselves to mechanization. In some cases these systems utilize automatic entry or exit, and for some the advanced intra-trade tools allow for extremely fine, automated management of trades.

Highly Discretionary Systems: A number of traders prefer to work with their "gut instincts" or use obscure chart patterns to make their trading choices. In these situations some manner of automation will provide them with an edge they might not have had before.

The best way to decide if a system is suitable for complete automation is historical testing. That way traders can construct engines and assessment them historically, then allow the results to speak for themselves. Some engines will demonstrate striking results with no human involvement. Others may have outstanding results but be subjected to "giveback" – which is an indication that some manual adjustment is needed. A swift glance at the real trades in the historical test report will help the trader decide what happened and how he or she might have stepped in to improve the outcome.

Once a trader has customized the parameters to their satisfaction, live testing can begin. These test trades should be identical to live trades, so demo trading with total automation or with some manual operation will duplicate live trading conditions. If these tests prove lucrative, real trading places the trader in a situation where automated trading systems affect the bottom line in a positive manner.