Developer interview for the EA "Legato USDJPY" with a logic close to discretionary trading

Ebi-sawa, this time we would like to interview about the EA "Legato USDJPY." Before asking in detail,Keisan'sFXcareer, could you tell us?

KeiAt first, I traded with discretion without studying anything. Looking back now, I was taking on positions with a lot of lots that I can hardly believe myself, and if you keep doing that you’ll go bankrupt, right? After that I reflected, and then when I was studying a bit, I learned about Himawari Securities' automated trading. When I checked the performance of automated trading, it was a straight upward trend and looked amazing, so I thought, if I do this I’ll win, a simplistic idea. At that time I didn’t think about lot size or portfolio at all, and I did a high-risk method and failed, losing my assets again.

Ebi-sawa2 times, you were ousted, right?

KeiYes. I hadn’t thought about anything. Then,MT4 I learned that I could create automated trading systems myself. With Himawari, someone else built the system, and I didn’t know the logic behind it, and I invested money and failed. So I thought it would be better to do it myself, and during my commuting time I read programming books and learned to create my own automated trading. After learning MT4, I thought, then I should create my ownEA, and the process of learning programming on the train is something not easily imitated. People say you became able to program well.

KeiOriginally I studied technical analysis quite a bit, so while learning the program, I felt the tangible sense of creating anEA.

Ebi-sawa, among the variousEAs I created, my current favorite is “Legato USDJPY.” What kind ofEAis this?

KeiI wanted to create anEAcloser to discretionary trading. It is designed with emphasis on taking profits and stop losses, focusing on risk-reward, and additionally prioritizes volatility so that settlement happens as much as possible by the logic. The logic is trend-following, and the take-profit to stop-loss is1対1 or more. It does not aim for high win rate, and the win rate is around5割, a system that accumulates profits little by little while going through wins and losses.

Everyone, when choosing anEAI think people emphasize items like profit factor (PF) or revenue, win rate, maximum drawdown, etc., but I believe the ratio of average profit to average loss is important. When this average profit to average loss is1対1 or more, and the win rate is5割 or more, the equity curve will be upward. By the way, thePF is1.2～1.4, and I keep the maximum drawdown below15％.

Ebi-sawa, many people valuePF, but why do you not place importance onPF?

KeiFrom trial and error, I concluded thatPF around1.2～1.4 is the best for now. Statistically, if you keep running the EA,PF converges toward 1.0. Therefore, a value a little above 1.0 allows for stable continued profits. If you try to create a PF above 2, you would need to widen the stop losses or create periods with no trading, which is quite difficult to sustain.

Ebi-sawa, how long do you hold positions?

KeiThe time frame I trade is5 minutes, and the image is scalping to day trading. If the trend continues, I hold it longer, and if it’s a tight range I close it quickly.

Ebi-sawa, are there any cautions in operation?

KeiBecause it is a trend-following logic, it performs well when a trend exists, but when there is no trend it tends to stagnate. It’s not good to just look at the stagnation period and judge that it can’t be used.

Ebi-sawa, then what about stopping the EA during ranges, or applying discretionary judgments—is that acceptable?

KeiIf you can judge trend vs range, you should stop, I think. However, making that judgment is quite difficult, isn’t it?

Ebi-sawa, how much margin should be prepared so that you can operate safely without worrying?

KeiInitial amount100 million yen,1 standard lots trading as the design for forward testing. It is a very safe design. The maximum stop loss is30pips, so if you lose once it would be1 time maximum loss of3000 yen.

Ebi-sawa, I see, a very safe design. Finally, for those interested inLegato USDJPY, please share a message.

KeiManyEAdevelopers tend to add various indicators, adjust parameters, and look for a holy grail. However, I do not do that, and instead pursue how close I can get to discretionary trading. Discretion is about how much profit you can take and how much you can prevent losses, you know. With that aim in mind, I’m building it. While ranges can lead to stagnation periods, it steadily profits in trending markets, so I hope you understand that and use it accordingly.