Why do we create a Martingale EA?
Hello, this is Yoko.
Thank you to everyone who read the first installment. As promised, this time I’ll share a little about the roots of our story.
“Why do we build a grid EA (averaging down)?”
Averaging down is scary
First, I’ll be honest. Averaging down is a scary mechanism.
Averaging down is a method that adds to a position with a losing floating P/L, to improve the average entry price. If the market rebounds, you can end up with a profit all at once. It tends to have a high win rate and a smoothly rising equity curve, making it look like a clean rightward trajectory.
But behind thatlies the deferral of lossesas the market doesn’t revert and continues in one direction, the accumulated positions shake the entire account. As the saying goes, “steady gains followed by a big crash” (kotsu kotsu dokan), the danger of averaging down lies in the often unseen “tail” portion.
Kenji always says,“Backtests for averaging down should show the most painful moment, not a perfectly rising curve.”.
Actually, there are broken settings too
This is behind-the-scenes development.
While building Rising Tide, we tried many aggressive settings that could have yielded even more profits. For a period, assets really surged. However,as gold plunged sharply in March 2026, all aggressive settings blew up the accounts.
No matter how good the interim results, if a single crash wipes out the account, it’s pointless. We decided not to commercialize those settings.
The settings we’re delivering now are those that survived 12 years, including the crash in March 2026.Only those that survived. The numbers may look modest, but this is our answer.
Still, why choose averaging down
So, despite being so scary, why use averaging down? There are three reasons.
① Because gold’s market nature suits it
Gold is an asset that trends upward over the long term. In the short term, after sharp declines, it tends to rebound. The idea of “buying near the bottom and waiting for a return” aligns well with gold’s characteristics. Averaging down is not suitable for every currency pair;you must choose the marketas the first step.
② Because risk can be reduced through design
The fear of averaging down cannot be eliminated entirely. But it can be reduced. Our EA includes measures such as the following.
| Widen averaging interval according to market volatility (ATR) During volatility, we spread out the intervals and pick more cautiously. |
| Do not take positions when a sharp rise/fall is detected When prices move significantly in a short time, we wait for the market to settle rather than taking new positions. Averaging down is most dangerous when rushing into one-directional rapid moves. |
| Do not start anew during dangerous times/days Before/after major economic indicators, for example. |
| Have a final brake If floating losses exceed a predetermined amount, the system stops to protect the account. |
③ Because you can verify scary moments repeatedly during testing
We base our approach on 12 years of backtests, including the COVID-19 volatility and the March 2026 crash.
Seeing scary moments again and again with real money at stake is the contract we make to stay with averaging down.
Why we choose a "conservative multiplier"
Averaging down uses a key number called the “multiplier.” It decides how many times the lot size increases with each averaging step.
Increasing the multiplier yields larger profits for modest market rebounds—what people call “big profits.” In gold-focused EAs distributed for free on SNS, you’ll see high multipliers with flashy results. It’s a common industry approach.
But when you expose the underlying data, it looks like this: starting with a 0.01 lot, averaging up to 20 steps.
| Averaging multiplier | Lot at 20th step | Total lots at 20 steps |
|---|---|---|
| 1.0x (even) | 0.01 | 0.20 |
| 1.06x | 0.03 | 0.37 |
| 1.1x | 0.06 | 0.57 |
| 1.3x | 1.46 | 6.30 |
| 1.5x | 22.17 | 66.49 |
| 2.0x | 5,242.88 | 10,485.75 |
The gold row shows the multipliers we use by default for our three EAs.
With a 1.5x multiplier, the 20th-step lot becomes over 2,000 times the initial lot. Once it grows this much, one more reversal can break the account.
We deliberatelyset a conservative default multiplierfor this reason.
| Alchemy Rising Tide | 1.1x |
| Alchemy Flex(Stable type) | 1.06x |
| Alchemy Duo | 1.0x(no increase in lot) |
Even with the same 20-step averaging, 1.1x results in about 6x the 20th-step lot and a total of about 0.57 lots. With Duo, the lot remains the same no matter the number of steps.
When the market rebounds, the profits aren’t as dramatic with high-multiplier EAs. Instead, our EAs have two strengths.
Strength 1: Longevity
Even during drawdowns, the account lasts longer. We measure success by surviving 12 years, including the March 2026 crash.
Strength 2: Can operate with lower leverage
High-multiplier averaging down typically assumes a high-leverage account environment. Because your lots can explode, you need sufficient margin to endure.
Our three EAs are tested to run on low-leverage accounts with default settings (Rising Tide and Duo for 12 years, Alchemy Flex for 7 years). However, in a sharp drop like March 2026, for some modes with 20x leverage, the built-instop-loss brake activated and positions were reset within the expected drawdown. In particular, Duo M5 mode shows that with low leverage, the brake amount remains the default, and multiple activations can occur. We will discuss this in detail next time.
| EA (default settings) | Minimum equity maintenance rate (Leverage 20x) |
|---|---|
| Alchemy Rising Tide | 50.76% |
| Alchemy Flex (Stable) ※7 years of verification |
20.13% |
| Alchemy Duo M15 mode | 196.33% |
| Alchemy Duo M5 mode ※Even with a brake of 9.5 million yen, the March 2026 crash triggered once |
20.50% |
※Duo M5 mode on low-leverage accounts shows the brake activating once at the default brake amount and once more when raised to 9.5 million yen during the March 2026 crash; we will discuss this in detail next time.
However, how far the margin maintenance rate can drop at the most difficult moment depends on the EA and mode. Some domestic brokers will force liquidations at 100% or 50% maintenance. If you use a low-leverage account, check your broker’s liquidation level and either keep extra funds or choose Duo’s M15 mode with more maintenance margin.
Better to survive 12 years than chase big profits.
And to operate without relying on high-leverage accounts.
This is the hallmark of our gold averaging down.
Additionally, Alchemy Flex offers aggressive and ultra-aggressive settings for those who want higher risk for higher profits. These are optional and intended for high-leverage accounts only.Default is the stable 1.06x. We will explain the difference in detail in a future post.
Final brake depends on your fund management
One more important point.
No matter how much the EA designs, the safety of averaging down ultimately depends on thebalance of account funds and lot sizes. Even with the same EA, if you run with larger lots than recommended, a setting that survived 12 years can break quickly.
That’s why on our product pages and manuals we show not only good numbers but also data from the most difficult moments, so you can judge whether your funds can endure it.
For those who still find averaging down unsuitable
If you’ve read this far and feel you can’t come to like averaging down, that is perfectly normal.
For such readers, we also offer“The Golden Guardian”which does not use averaging down. We’ll discuss this character in detail at the end of the series.
Next time
Next time,“Which of the three Alchemy series is right for me?”We’ll compare Rising Tide, Duo, and Alchemy Flex in a comparison chart and tell you which one suits which type of person.
Thank you for reading to the end.
Yoko
※This article presents past verification results and does not guarantee future profits. Investment decisions are at your own risk.