Almost all major investors make their transactions within a certain system. It is the first technical analysis which comes to mind in the financial markets. Principles and rules are determined and the system is developed for basic analysis. As the investments in the Forex market are generally short-term, the system is mostly developed for intra-day transactions.
Systems based on only one indicator are generally unsuccessful. Errors of systems created with a single indicator should be reduced by using other indicators. This is called filtering process.
The relationship between the formations in the charts and the indicators and even the general system should be tested.
TESTING THE SYSTEM
A system without a bunch of testing phases cannot be considered as robust. The improved system must be tested on past prices. This is called a backtesting process. It will be appropriate to cover the test period as long as possible. For example, it would be useful to test the system on an average of five years. It is useful to test the system practically in the demo account after the correct test. Because it is very common for many systems that are successful on paper to be unsuccessful in a real situation. In the demo account also test the system as long as possible. This phase is called forward-test.
The system can also be applied to the actual account if successfully tested the demo account. However, it should not be forgotten that one should not discontinue the process of back and forth testing in order to develop a successful system in the real account.
When developing the system, it should be kept in mind that not all prescriptions apply to each patient. Each pair has its own characteristic of movements. The performance of the systems might give different result in the pairs with higher volatility whilst the less volatility.
Other criteria are used for an objective testing process:
- Total Net Earnings: The total net gain from the total gain in the test process is the total net gain. It is useful in terms of reliability.
- Total number of transactions: The number of procedures in the testing process is also an important criterion. For the success of the system, it is beneficial that profitable positions are more than harmful positions. However, even more, risky systems that are not waiting for losses but waiting in the snow are being developed. The high number of operations means that the system is affected even by small fluctuations. Too much processing increases the expenses of the investor with commissions and spreads.
- Percentage of Earnings / Loss: The percentage of profit or loss in the test process according to the initial capital.
- Annual Earnings / Loss Percentage: The annual percentage of profit obtained during the testing process calculation is important in order to make a comparison with interest, inflation and return on other investment instruments. In case the test period exceeds one year, the value obtained by dividing the number of days in the test period by 365 is the year-to-date value of the profit obtained during the test period. The annual gain is calculated by dividing the total net profit by this value.
- Drawdown: In the Forex market, the loss is a significant risk factor due to the leverage factor. Systems that can jeopardize the margin level and cause a heavy loss must be optimized.
- Total Net Earnings / Risk Ratio: Each investment has a potential risk. The potential risk is the percentage of the maximum reduction in the balance during the testing process. The fact that the total net gain is less than 3 in terms of the gain/risk ratio divided by the potential risk ratio indicates that the system is useless.
- Total Gain: It is the gain or loss that will be obtained if the parity is kept from the beginning of the test period until the end of the test period. When calculating this ratio, it should be taken into account whether the margin problem occurs in the test process due to the leverage factor, unlike other financial markets. At the end of the test period, the system should be questioned if a lower return than the buy-in value is achieved.
- Percentage of Total Earnings: It is the rate obtained by dividing the total earnings by the starting capital.
- Annual Percentage of Total Earnings: The one-year yield of total earnings is the percentage ratio of the initial capital. The system should be questioned if the percentage of return obtained during the testing process is lower than this rate.
- Percentage of Earnings: Percentage of the number of positions closed by the gain in the test process, divided by the number of positions. The longer the duration of the test, the greater the importance of this ratio.
- Total Earnings in Profitable Operations: It is obtained by collecting the yields obtained in the transactions which result in gains in the test process.
- Average earnings per transaction: It is the rate obtained by dividing the total earnings by the number of transactions. The high rate is an important criterion for the success of the system.
- Number of Losses: Number of positions closed during the test period.
- Total Loss in Damage Resulted Processes: The sum of the losses incurred in the loss-closing positions during the testing process. Despite the small number of transactions, the high rate of loss strategy should be reviewed.
- Average Loss in Harmful Processes: The total loss in the resulting transactions is calculated by dividing the total loss by the number of transactions.
- Average Earnings / Average Loss Rate: Average profit in profitable transactions is calculated by dividing the average loss in harmful transactions.
- Earnings / Loss Index: The success rate of the system can be measured by dividing the win ratios in the test phase by the loss rate. However, the profit or loss obtained during the testing process is not real. In terms of opportunity cost, it should be compared with the country’s interest rate. It is taken into account in the open positions. The index value varies between +100 and -100. Since this index expresses the gain or loss in the most accurate way, it is recommended to calculate each test period.
- Yield-Risk Index: As we have already mentioned, the risk must also be calculated when calculating the return in the test process. Yield-risk index is used to measure the risk assumed for the gain in the test process.
OPTIMIZATION OF THE SYSTEM
Applied systems can be optimized to reduce the error rate and adapt to different pairs. For example, a 22-day moving average can be used in a system that produces an early receive signal as a result of the 20-day moving average. Parameters can be tried until the errors are minimized. Optimization should be implemented strictly for every system developed. However, it should be kept in mind that each system can produce false signals. Therefore, it is more useful to try to increase the success of the stop loss strategy instead of trying to optimize the system to give a 0 percent error.
Reducing the parameters applied in the system leads to an increase in the number of signals. The risk is increased in systems where the number of signals is too high. The growth of risk means the potential for growth in earnings, but in the case of such systems, if the stop loss strategy cannot be established, very serious losses may occur. Changing the parameters of all indicators while optimization is not a correct approach. Changing the parameter of a single indicator while optimizing (according to the science of economics, ceteris paribus, ie nothing else) is a more accurate approach. For the Forex trading platforms and other technical analysis software, it is possible to easily change the parameters of the indicators.