How to build a trading system

Rules are specific terms applied for a specific operations. So far, nobody found the best possible number of rules for building trading system. Each trader determines the number rules to build a trading system on their own. The system built on the basis of the sole rule won’t be successful. On the other hand, a great number of rules will consume your time and confuse you.

To explain this principle, let’s turn to Tushar Chand’s conclusions. He made extremely complicated and ambitious research concerning the principles of building of trading systems.

Results of his research show that profitability of system grows as the number of rules increases (on conditions that rules are sensible), but it falls down as the number of trades goes down. The reason is that only few situations satisfy the constantly changing combination of rules. Every new rule is like a filter through which only certain transactions can pass. Besides,you need to have more data to work with these rules.

One more interesting fact refers to the parameter of any trading system. It is a maximal drawdown or maximum intraday drawdown (MIDD). It points out the maximum number of continuous loss for a certain time period. So, increase in the number of rules makes the MIDD level go up also, which is probably related to drop of forecast's reliability. It happens because the number of variables grows. Then, along with reduced number of trades, rising loss starts to drop down too. But it drops slower than the common profit.

Conclusion: if to introduce the rules only to exclude unprofitable trades, you can prevent profitable transactions too.

Thus, an excessive increase of the number of rules (complication of system) won't lead to greater performance of trading system.

An important feature of every system is its sustainability. Sustainability is the inalterability of rules, in particular, open and close terms. Especially they shouldn’t change during long-term periods, if there are open positions. Their variation is possible only for objective reasons such as modification of market conditions, which makes current system less efficient (inoperative) and makes  improve its terms or create a better system.

Indicators, stop limit orders, interval charts can be parameters of this optimization. It is very hard to comply with the rules. But if a trader decides not to implement them, all results will turn to be poor. On the first stage it’s necessary to trade according to the rules, e.i. to rely on system. The principle of system is following: collect data, make decision and take actions.

Mechanical means unequivocal, strict and understandable. There is no place for ambiguity. Signals of the system also should be well-defined without any misunderstanding. A good system can be represented in the form. It is an important condition to test the system. Otherwise, the system doesn’t exist at all.

Modern programs of technical analysis usually include the function of an automatic test of trading systems (for example, MetaTrader4).  All important indicators ( number of trades, net profit for a certain period, profit rate, MIDD etc.) are calculated by computer quickly and accurately. Also these programs allow to improve parameters of system. Their serious drawback is that you need to write a special algorithm in a certain programming language. Programming skills  are required (for a person who knows the basics of programming, it won’t be a serious task). If you have no desire to write a program,  test systems manually. In this process you need to account for the data of previous operations. Thus, the result of future transactions is rather hypothetical. To find out the potential of trading system in real conditions, you have to trade on demo account (virtual practice account). It takes from 1 to 3 months which is a quite long period. But results will be more reliable. And remember: trade in real conditions is much harder, because it’s almost impossible to monitor market conditions all the time.

A correctly conducted test will identify positive and negative features of trading system and find out its efficiency. Time and money save is an undoubted advantage of the test made with historical data. Moreover, testing will approach you to the market and your own analytical qualities will become much deeper.

Risk management usually means a percentage of a capital, which is exposed to risk in a certain trade. Risks should be limited by a stop order (stop loss order). Also, in the process of design of trading system it’s important to take into consideration the volume of capital. The recommended margin must be rom 2 to 33% (not more) of the deposit.

Besides, you should understand that there is no linear dependence between profits and losses which is typical for trade. 10% of loss demands 11% of profit to make trade profitable. But if your loss is 50%, then you need 100% of profit. Analytics agree that maximum loss rate should be 33%. With such a rate, you need to gain 50% of profit, which is quite achievable. If loss is more than 50%, be sure that you will lose the deposit.

First of all, you should ask yourself “what it was created for”. Trading  system is supposed to be used for those conditions and currency pairs for which it was designed. Besides, it must be the system, which provided good results during the test. In other words, the system designed for Swiss franc hour bars is applicable neither for Swiss franc day bars nor for Japanese yen hour bars. In this case, the system should be modified.

Conclusion: implementation of the above-stated principles helps to create efficent trading systems and apply them in an appropriate way. Of course, you may rely on your own principles in design of system. But at the same time, none of the listed principles is unnecessary.