Overview
Tokyo EA by Vasiliy Strukov is a low-priced EA for MetaTrader 4 and 5 that is sold openly as a grid with martingale. The vendor has sold EAs such as EA Gold Stuff on the MQL5 market for years; we have not tested those. This review is a warning, and the warning follows from documented features of the product, its design and a stop-loss that is switched off in all of the vendor's settings, and from a young vendor signal named "Tokyo" that we attribute to the EA by its name and that shows a very deep decline by equity. It says nothing about the vendor as a person. Our grade is provisional.
What the vendor sells
Per the product page (MT5: 197239, MT4: 197277, read on 17 Nov 2026):
- Price: $50 per version as a purchase. The page shows no rental option. There is a free demo for the strategy tester.
- Published on 22 Sep 2026, both versions at version 1.0.
- Logic: the EA enters at the open of the next candle after the Tokyo Indicator shows a buy or sell arrow, and then trades "in a grid/martingale strategy". It is said to suit any pair and any timeframe, especially strongly trending markets such as gold on M15.
- Settings: the product page states neither a minimum deposit nor a risk per trade. The set files on the blog start at 0.01 lot, on MT5 partly with automatic lot sizing per $1,000 or $3,000 of free margin. The maximum lot in the sets is set so high that it practically imposes no limit. For gold there are sets on M15 and M1.
- Bonus: buyers are asked to get in touch after the purchase and, per the vendor, receive a free indicator. Set files and the manual are on his blog.
The parameter list is detailed. Five items decide the risk:
| Parameter (product page) | Meaning |
|---|---|
| "Use hedge" | The EA may trade both directions at the same time. |
| "Max Orders" | Maximum number of open orders in the basket. |
| "Lot multiplier" | The factor by which each further order grows: this is the martingale. |
| "Real SL points (0 – not use)", "Virtual SL points (0 – not use)" | Stop-loss in points, 0 means off. |
| "DD Reduction Algoritm" | Closes the last order in profit together with the first one in loss to shrink the basket. |
The stop-loss is optional, and off in the vendor's sets
The product page says it itself: for both stop-loss variants it reads "0 – not use". On 17 Nov 2026 we read all 22 set files for Tokyo EA from the vendor's blog, for gold and seven forex pairs on MT4 and MT5. In every one of them the stop-loss in points and the percentage account stop ("Stop_Loss_Percent", an input that does not appear in the parameter list on the product page) are set to 0. All files also have:
- up to 100 orders per basket,
- a lot factor between 1.5 and 1.95 per follow-up order,
- hedging switched on.
What that means is plain arithmetic with these settings: at a factor of 1.65 the tenth order is about 90 times the first. If price runs against the basket for long enough, the open loss grows faster than any single order. Without a stop-loss such a basket only ends when price comes back, the drawdown algorithm dissolves it, or the margin runs out. A stop can be switched on. How the EA behaves then is not documented by any of the sources, and the vendor's sets do not use it.
One buyer writes in his review that a configurable percentage stop-loss hard-protects the account. The input does exist, but it is set to 0 in all of the vendor's sets. Anyone who wants to use it has to set and test it themselves. Why this is the decisive point for grid systems is explained in the guide Grid and martingale EAs: the risks; for gold in particular see Gold EA risk: grid, martingale and stop-loss.
The signal on the seller profile
No signal is linked on the product page. It only says "Real-time results can be viewed here", and the link leads to the seller profile. That profile lists three of the vendor's signals: "EA Gold stuff", "Paris" and "Tokyo" (2395805). We attribute the "Tokyo" signal to the EA because it carries its name and the product page points to this profile for results. It is not explicitly linked, and it has no description of its settings.
What we read on 17 Nov 2026, without figures, as in all our reviews:
- Account type: an MT4 account on the server Weltrade-Live, started on 16 Oct 2026, almost four weeks after launch. Under the MQL5 signal rules (III.12/13) only real-money accounts are allowed, and MQL5 shows the "Real Account" icon.
- Leverage: very high, above the limit at which MQL5 blocks subscriptions (signal rules III.3). The signal therefore cannot be copied. With EU leverage (at most 1:30 on majors, 1:20 on gold) the signal's position sizes can only be reproduced with much smaller lots; a grid then needs even more margin headroom.
- Decline: by balance it looks limited, by equity it is very deep. Open grid baskets at times pushed the account far below its balance before they were closed. According to the MQL5 metric "Max deposit load", the margin tied up at times exceeded the account value.
- Result: very high growth, according to MQL5 overwhelmingly on a single day. An amount equal to the initial deposit was withdrawn (MQL5: "Withdrawals"). The profit comes from gold; most of the forex pairs traded are in the red.
- MQL5 warnings: on 27 Oct 2026 MQL5 reported that most of the growth came from a single day. On 10 Nov 2026 the notices "A large drawdown may occur on the account again" and "Share of days for 80% of growth is too low" were added (read 17 Nov 2026).
For a martingale EA, exactly this gap between balance and equity is the core. Balance only shows closed baskets; equity shows how deep the account stood in between. An account can close in profit for a long time and still have run close to the edge. How to read these metrics is covered in Trading metrics: drawdown and profit factor.
What buyers report
For MT5, MQL5 shows 60 reviews on 17 Nov 2026, the first from 25 Sep 2026 and the last from 16 Nov 2026, so all from the roughly eight weeks since launch. 48 of them count towards the star rating, average 5 of 5; the remaining 12 are entries without text for which the page shows no stars. The MT4 version has 20 reviews, 16 of them rated, also averaging 5 of 5.
What these numbers mean and what they do not:
- Who may review: under the MQL5 market rules (section III) only users who bought the product, or, for free products, downloaded the full version.
- "Activations: 5" is not a number of users or past installations. Per MQL5 it is the number of computers on which a buyer may activate the EA; 5 is the minimum the market rules (section IV) allow. MQL5 does not show how many copies were sold, so no ratio of reviews to buyers can be read from it.
- Many reviews in a short time mainly describe the first weeks. A martingale grid shows its weakness in a long move against it, and that did not happen on the signal during these weeks. Buyer reviews do not count towards our grade.
On the content side, several buyers address the design themselves; we do not judge their statements:
- Several write that it is a grid/martingale EA that needs risk management and enough capital. One buyer says fully automatic use needs a cent account; another advises trading semi-manually with small capital.
- Several describe setting the trading direction themselves according to the trend, or switching buying and selling on and off depending on the market phase. The vendor provides a guide to semi-manual trading for this. The result then also depends on the user's judgement.
- Praise goes mainly to the vendor's quick answers and to the option of letting the EA manage one's own trades. Several buyers report large profits after a few days; we cannot verify that.
- In the comments, several buyers ask why the EA closes positions very quickly or does not trade on a given day. In his answers the vendor refers them to private messages.
How the provisional grade is set
We grade Tokyo EA on the same six criteria as every EA (calculation):
- Risk control: martingale without a proven hard stop has a cap. Below it, the worst figure from any source counts, here the decline by equity on the signal we attribute to the EA by its name. It leads to the lowest step of the table.
- Evidence: one public account would score more under the formula. We lower it because the account can only be attributed by its name, has run for a few weeks, started after launch, and its settings are unknown.
- Transparency: design, parameters and set files are open, which counts in favour. A deduction because no signal is linked on the product and the percentage stop is missing from the parameter list.
- Price and value: $50 sits in the cheapest band for paid EAs.
- Usability: neutral mark; we have not tested the installation ourselves.
- Broker independence: one broker documented; a deduction because the account runs with very high leverage and a grid needs margin headroom.
Martingale also caps the overall grade. The traffic light is red because the formula says so: the decline on the signal sits in the step where yellow is no longer possible without a long, calm account (methodology). The traffic light therefore also rests on the signal we attribute to the EA only by its name. Independently of that, the warning about the design stands: a martingale grid with the stop-loss off in all of the vendor's sets. As with every EA, an AI from a different model family also grades Tokyo EA blind; where it differs strongly, the mean counts until a human has re-checked. For Tokyo EA that applies to risk control, evidence and broker independence: the second review was stricter there, and the mean counts.
Who it is (not) for
If you want to test it anyway: first the demo in the strategy tester over several years and explicitly including a long trending phase against the basket, then a separate small account, with the stop-loss or the percentage account stop switched on, a low lot factor, a capped number of orders and only money you can afford to lose. On gold the cost per lot varies a lot between brokers, as the best broker for gold EAs shows; a grid that holds baskets for days needs a VPS anyway (VPS for EAs).
It is not suitable for prop-firm accounts, for anyone who needs a limited loss per trade, as a second gold EA next to another gold system, or for anyone who wants to run the vendor's sets unchanged and fully automatically. Gold EAs with a stop-loss on every position are compared in Best gold EA 2026.
Verdict
Tokyo EA is a low-priced grid/martingale EA that names its design openly. The warning follows from documented features: the stop-loss is optional and off in all of the vendor's sets, the only signal we attribute to the EA by its name has been running since mid-October 2026 and shows a very deep decline by equity, and MQL5 itself warns there that a large drawdown may occur again. The many top ratings come from the first eight weeks and say little about how the EA survives a long move against it. Our grade is provisional. We grade Tokyo EA with the same formula as every EA. We earn nothing from the EA itself; we link the vendor and MQL5 only as sources, never as partner links. Links marked “Ad” on this page (VPS, ATS-listed alternatives) are partner links. Past results are no guarantee of future ones, and trading leveraged products can lead to the loss of the capital invested.
