Difficult to answer to this problem which animates many debates for several years. This is why we will support the different points of view in an objective way and thus be able to draw good conclusions.
We all already have a subjective opinion on the issue, the supporters of automatic trading will tell us that through a program, there is no psychological bias to take into account. Supporters of discretionary trading will argue that an expert advisor cannot easily adapt to different market conditions. The two opinions do not contradict each other, they are complementary. As a result, there is no point in putting one side or the other aside as both support their theory with broadly valid arguments.
In recent years and more and more currently, we find a plethora of expert advisors for sale on the market. Nevertheless behind sumptuous backtests lies often deception. What many beginners frustrated with discretionary trading are unaware of is that it is possible to skew a backtest using optimization. In addition, the backtests offer only a limited vision of the performances of an EA. To validate these performances, it is necessary to analyze a live test (forward test).
Moreover, the automaton confers numerous advantages, in particular, that of not demoralizing itself in front of losses and to keep a constant behavior. It is also not subject to fatigue or absence, ... In short, if it is well built and regular, it can become a very relevant tool for people who do not fully enjoy their free time ( employees, students, ...). Notwithstanding, this constancy in the actions of the operator has a cost, that of not necessarily adapting well to unfavorable trading conditions. Clearly, if an EA is specifically designed to trade trend, it will be subject to strong successions losses in a market range.
The best way to ensure that an EA follows a strategy that is considered valid is to program it yourself. The task, although seeming obvious to the regulars, is no less important. Indeed there are quite a few tutorials on the net, the programming is not necessarily simple. The creation of a PLC also requires programming skills, sound knowledge of the financial markets, technical analysis and risk management.
In addition to the work of creation, the longest surely remains to test the automaton, first through the backtests and their analyzes, then by a long live tracking (live or forward) on different assets simultaneously.
Much more discussed in the discussions and affordable in practice, discretionary trading involves trading "manually". In terms of knowledge, it is not necessary to have a lot to get started. You just have to know how to manipulate a platform and some notions about the financial markets. Nevertheless, it takes a lot more to win, and especially experience. Trading is not just a matter of knowledge, it is a practice that gives rise to success.
Moreover, as we mentioned earlier, the trader will necessarily be confronted with cognitive factors, for example, loss aversion or risk aversion which are the two main problems of the novice trader or not. It is said everywhere that you have to "cut your losses quickly and let your gains run" is an important rule, but is rarely applied rigorously by traders. Many will tell you, that they trade with a big ratio, that is to say, that their earnings are always much higher than their losses. Nevertheless, the ratio remains constant, that is to say, that it certainly cuts its losses as it should, but we do not let fly the gains. "Let your gains run" means that the
Despite the clichés, discretionary trading is not easy to understand. In short, it is necessary to know how to predict, to invest intelligently and to control oneself in order to hope to realize long-term capital gains.
In conclusion, we will compare two points of view. In summary, automatic trading is particularly well suited to people who do not have much free time during the day, or who do not want to hang around the computer all day. In return, you have to make the effort to learn to programme, then spend hours programming and testing strategies. It's a very long and often painful job but it can be very rewarding.
On the other hand, discretionary trading can also pay off. But in the same way, it is long to master and requires a lot of work on oneself that many traders do not do. So many of them find themselves quickly on the straw. In the end, it is very difficult to favor one approach over the other, because in all cases the work is necessary.
Although some predict that the machine will sooner or later overcome man, given the results of many commercial programs, this does not seem to be the case at this time. Nevertheless, some seriously constructed EAs produce very good performances. Difficult to decide, usually, trading enthusiasts try to master both. The perspectives offered by each are impressive.