| Backtesting Software for Forex Traders |
Backtesting software is a complicated subject so we have split this lesson into two sections. In this first section, we will look at the pros and cons of software applications like those provided by your dealer, and specifically the kind of work software does well and doesn’t do well.
This is an important discussion because, like many products in the forex, software and system results are often marketed without your best interests in mind. There is often intentionally missing or misleading information included with software or purchased systems. As usual, information and education is your most effective weapon against this sort of marketing.
In the second section of this lesson we will look at a demonstration of a backtesting software package to illustrate the benefits of software.
We will also demonstrate the use of a spreadsheet to capture many of the benefits of software without the expense. Backtesting Software Pros
- Impartial A computer is impartial and is not fooled by bias. A computer will test exactly what you ask it to. If a particular formation or trading signal appeared in the past, it will find it and identify it for you. The results from a computer will almost always be more complete than testing by hand from a chart.
- Efficient Backtesting software can do in seconds what would take someone hours to do by hand. This is a great advantage when developing system signals. Software can find your trading signals and then plot them on the chart for you to examine in greater detail.
- “What if” analysis Developing appropriate stop losses and profit targets is an important component of testing any system. Software can retest the same system with new “what-ifs” to adjust for different stops, limits, costs, etc. with just a few key strokes.
Backtesting Software Cons
- Computers do not understand trading Backtesting software does not understand market dynamics. It does not adjust for vacation time, distractions, market crises or other issues that you will deal with everyday. It cannot be subjective in its analysis. If you regularly make subjective “judgment calls” during your trading, your actual results will not look like what the computer produced.
- Missing Variables Backtesting software, especially those available from dealers, will often have limited capability to adjust for trading costs. Typically, sample systems or systems available for sale will have ignored spreads, costs, slippage and other issues that will occur in real life. Sometimes this is intentional to make the results look better and to encourage frequent trading and big commissions or sales.
- Past Data Reliability Past data is usually full of erroneous ticks and price problems. Sometimes these price anomalies can be significant, and the computer cannot tell the difference between good data and bad data. You always have to assume that there is some bad data in your historical sample. Never accept a system’s backtesting results as a perfect indication of what you will see in the future.
Recommendations for Traders
- Don’t get your backtesting software or systems from your dealer. They may have ulterior motives and may bias the software or systems to make short term, frequent trades look much better than they are. To borrow a term from the X-Files, “trust no one.”
- Do your own work. It is well worth the effort to learn how to create and test your own systems. Do not buy them from other people and don’t follow someone else’s ideas without doing your own research. Getting trading and system ideas from your peers or mentors is a great idea; just don’t rely on that information without your own objective analysis.
- Revisit your backtesting once forward testing has been done. Backtesting is good and can provide a lot of detail about your ideas that would be difficult to determine otherwise. However, testing into the future by trading (with a small investment) or paper trading is critical.
You will learn a lot about factors you may have overlooked when you originally built the system itself. You may also find that you have overstated the risks and costs of a system and your performance is better than you thought it would be.
Comments deemed inappropriate will be removed
3.25 Copyright (C) 2007 Alain Georgette / Copyright (C) 2006 Frantisek Hliva. All rights reserved." |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Learning Markets Partners More Partners | Become a Partner |
 |  |  |  |  |
 |
| Learn to Invest | Reviews of Stock Brokers | Stock Picks | Technical Analysis | Broker News | Investor Education | What to Invest In | Live Market Analysis | Should I Invest? |
