How reliable is backtesting as a way to measure the success and profitability of a trading system?
Backtesting is very popular these days as traders try to find ways to ensure the profitability of a system before they test it in actual market conditions and take risk while doing so. Trading forex online, as everyone knows, is a risky endeavor, and any offer or claim that promises to reduce this risk is greeted with a lot of enthusiasm by the trader community, sometimes without sufficient scrutiny of the claims. In this article we’ll discuss some of the important disadvantages of backtesting which make it completely inefficient as a substitute for actual trading and testing with a mini account.
1. Price action is chaotic
Let’s first of all understand and remind ourselves that the price action is chaotic. It is a chaotic process where the rules, as well as the patterns created by them change all the time. Yes we observe trends, ranges, patterns on the charts all the time., but the order, duration, and size of these patterns makes all the distinction between profit or loss. And since it is safe to say that these patterns are almost never repeated, it is futile to apply the same strategy repeatedly to all market conditions and to expect it to be profitable at all times. In other words, backtesting is built on faulty expectations and principles. It will not deliver satisfactory results.
2. Backtesting cannot account for non-market risks
There are many factors unrelated to strategies and analysis which may complicate the trading process, even if we were in possession of that elusive perfect trading methodology. Connection problems, server failure, computer crashes, software glitches all contribute to minimizing the value of backtesting.
3. Backtesting is unrealistic, and leads to complacency
Let’s also remember that backtesting leads to false confidence. Many traders use backtesting-optimized trading software or packages, and based on the past performance of these tools, they take excessive risks in order to profit from the purported efficiency of robots, plans, systems etc. That attitude of course leads to nothing but losses in the best case scenario.
4. Backtesting is not a scientific method, it is a marketing tool
Finally, backtesting has no scientific justification for its claims. Scientists are in consensus that past results have no bearing on the future performance of a technical strategy, and as such, the results generated by backtesting are only useful to marketers and peddlers, not to traders.
In short, one of the first things that you should be learning as part of your forex education is that backtesting is useless as a way to determine the validity of a scenario. It can be useful for entertainment value, for understanding how technical indicators work, and perhaps helping you get used to long-term trading, since you can observe through backtesting that a position held over the longer term can be as profitable as any short-term strategy. But backtesting surely will not eliminate the curtain hiding the future. If that were so easy, all the online peddlers of magical trading systems would be enjoying their luxury resorts at touristic places instead of advertising their products all the time.