Man who grew 5,000,000 yen in EA operations to up to 18,000,000 yen A 5,000-character interview: "If you want to turn a small amount of capital into tens of millions, start with an initial capital you can afford to lose."

Masuyoshi’s Profile

A full-time trader. Investment targets are stocks, futures, FX, and Bitcoin. Started system trading in stocks and commodity futures about 20 years ago. Achieved 100 million yen from 2 million yen in Nikkei 225 futures.

Twitter:https://twitter.com/stock_fx_trader

※This article is a republication and editorial rewrite of an article from FX攻略.com, August 2019 issue. The market information in the main text differs from current market conditions, so please be aware of that.

●Interviewer: Michihiko Ebisawa (Editorial Department)

Performance improves after building a portfolio

Ebisawa: I understand Masuyoshi has a track record of increasing initial capital from 500,000 yen to 18,000,000 yen with EA. First, please tell us what prompted you to start using EA.

MasuyoshiI’ve been involved in stocks and commodity futures for about 20 years, and my entire investing history has been system trading. So, for FX, I decided to apply system trading and do it with EA as well.

Ebisawa: You’ve been doing system trading for 20 years! In other words, the concept of system trading existed. What are its魅力 (merits) or advantages?

MasuyoshiTo win, you need a method, risk management, and mindset. However, 20 years ago or 10 years ago, information about methods wasn’t readily available. Now there are many EAs, and you don’t have to come up with your own methods anymore—that’s the biggest appeal.

Ebisawa: When did you first start using EA in FX?

MasuyoshiI started using EA more than 5 years ago, but back then I would buy an EA, it wouldn’t work well, and I’d stop. I also tried creating my own EA, but since I can’t code, I would have a logic prepared and have a broker build it, which still didn’t go well. The real success came when I started full-scale operation around late 2017.

Ebisawa: You say performance improved since late 2017. What do you think is the reason?

MasuyoshiI began operating with portfolios of 10 or 20 EAs rather than one or two. Building a portfolio helps reduce total drawdown while increasing profits. I think that’s the reason.

Ebisawa: If you run 20 EAs, choosing them must be challenging. What criteria do you use to buy EAs?

MasuyoshiThere are many opinions, but I place importance on the Recovery Factor (RF). RF is calculated as net profit divided by the maximum drawdown. It indicates how much return you can expect relative to risk. RF increases as net profit grows and decreases with larger drawdown. I wouldn’t say I only look at that, but that’s the guiding sense I use.

Ebisawa: When building a portfolio, some people check correlations between EAs; how about that?

MasuyoshiI do examine correlations, but the RF is the primary criterion for selecting EAs.

Ebisawa: The idea of focusing on RF—did you come up with it yourself, or is it a theory from 20 years ago?

MasuyoshiI created it myself. Since I was involved in commodity futures 20 years ago, there wasn’t even a term RF back then.

Ebisawa: What are the benefits of using RF as the main axis?

MasuyoshiSimply, to increase RF you need to increase net profits and keep drawdown down. So a good EA is one with large net profit and small drawdown, as indicated by a high RF.

Ebisawa: I see. Indeed, RF alone can help judge strong EAs. Are there other cautions when selecting EAs?

MasuyoshiRecently, some backtests run for 20 years or longer, but I don’t rely on such long histories. Back then, around 20 years ago, Japanese retail FX brokers had wider spreads, and the market has changed since moving to a world where spreads can be as low as 0.5 pips. If backtests are too long, the EA may fail to adapt to current market moves. So I think excessively long backtesting periods may hurt accuracy for recent markets.

Ebisawa: Then, what backtesting period do you find useful?

MasuyoshiAbout 5 years, maybe 10 is long. Looking ahead, as AI participants increase in the market, market moves may change, so we will need to find EAs that adapt accordingly.