A backtest system for stock trading is something resembling a growth hack for startups. The intended version of the strategy that you decide to apply to make money is passed through a set of quantitative and qualitative data. In other words, its effectiveness is checked on historical quotes with various instruments and the selection of certain settings. As practice shows, initially no one has hundreds of options for trading strategies, but for a start and two or three ideas are quite enough.
Testing of trading systems is like preparing football players before a match, or crashing a car in anticipation of its announced release. Without the initially established, on the points of the scheduled actions, it will not be possible to win against competitors or create a modern car that will be in demand from the buyer.
Currency exchanges "force" some traders and depositors to act irrationally. This not entirely correct behavior includes the use of strategies that have not been tested before the "match". The probability of obtaining any elementary income with such systems is zero.
Pros of Checking Systems for Trading
The test differs from an intuitive choice in that it shows an objective result about the trading strategy and its use. Even if it is demonstrated on historical indicators.
Testing the chosen concept of trading, you can immediately determine whether the "holy grail" has been developed or all the stated advantages of the strategy are only in the words of the developers.
A profitable system generates income in all directions of trends, since it is never known in advance when the price will change. Price, as you know, is formed by the expectations of market participants, and it is impossible for an artificial concept to consistently reproduce changes in these expectations.
Possible backtest problems
One of the problematic points when testing systems and, accordingly, their acquisition, is that there are ways that help to distort the data: the strategy looks good on history, and in real life it is impossible to make money with it.
Any programmer knows how to correct the indicators that are used to demonstrate profitability in backtest. For example, for a trend-following strategy, the system is revised on quotes that have been in effect for the previous two years.
Once an option is selected that looks suitable, the developer looks at whether the system is working for a longer period. In most cases, the results will not be very impressive, which the customer of the strategy needs to know about.