Solvex Gold Robot
An Expert Advisor that trades gold (XAUUSD) on your own MetaTrader 5 account, by fixed rules, a few times a month. You do not have to take its back-test on trust: download the free edition and repeat every number on this page yourself, before you pay anything.
It is licensed software — not a managed account, not a signal service, and not a promise of profit.
The back-test in four numbers
- One account, June 2018 to November 2026
- $1,000 → $220,616
- years with trades ended higher; 2019 and 2023 ended lower
- 6 of 8
- trades — about three a month, and about six in ten end in a loss
- 297
- the worst year for a new $1,000 account (2023)
- −82%
Simulated in the MetaTrader 5 Strategy Tester on historical prices, with the rules chosen on that same history. Not live trading results and not a forecast. Every year and month is below.
What it is and what it is not
What it is
- Software you install in your own MetaTrader 5 and run on your own account, under your own broker agreement.
- Rule-based: the same fixed rules every day, with a stop loss on every trade from the first second.
- Licensed per trading account for a period. 10 demo accounts are included, so you can run it on demo first.
- Infrequent: about three trades a month on average, and some months none at all.
What it is not
- Not a managed account or a fund. Solvex never holds, receives or trades your money.
- Not investment advice and not a signal service.
- Not a promise of profit. It has losing months and losing years, shown below.
- Not suitable for money you cannot afford to lose.
Test it yourself, free
The free back-test edition is the licensed robot's own trading code. It runs in MetaTrader 5's Strategy Tester and nowhere else: it cannot be attached to a demo or a real account, and it opens no trade on history after 10 November 2026, the end of the back-test on this page. Use it to check our numbers, and to see what the robot does with a deposit of your own size.
Download and copy one file
Unzip the package. In MetaTrader 5 choose File → Open Data Folder, open MQL5 → Experts, and put the .ex5 file there. Restart MetaTrader.
Open the Strategy Tester
View → Strategy Tester (Ctrl+R). Choose the robot, then set the symbol, dates, deposit and model exactly as in the table here. Leave every input at its default.
Press Start and compare
The number of trades, their dates and their prices will match this page. The net profit will be close, not the same to the last digit: it contains swap, which the tester charges at your broker's rate on the day you test. The instruction inside the package shows every screen.
SolvexGoldRobot_Free_v1.06.zip · 241 KB · contains SolvexGoldRobot_Free.ex5, README.txt, instruction.html
SHA-256 d949fa031068a613697182cb04f0f8c9e4af339400924c2c7c5313c5a4723f8a
Strategy Tester settings for Dataset A
- Symbol
- XAUUSDm
- Timeframe
- M15
- Date
- Custom period, 2018.06.28 – 2026.11.10
- Delays
- Zero latency, ideal execution
- Modelling
- 1 minute OHLC
- Deposit
- 1000 USD
- Leverage
- 1:100
- Inputs
- Leave every input at its default
The trades match on the same price history. Ours is an Exness MetaTrader 5 demo account, which is free to open. On another broker's gold history the robot takes other trades and the result is different — that is expected, and worth seeing before you choose a broker.
The money moves a little from day to day. The Strategy Tester charges swap at the rate your broker publishes on the day you test, for the whole history. Our own run had −$2,305 of swap in it on 11 November 2026 and −$3,041 on 12 November 2026, with the same trades. Expect a net result close to these tables, not the same last digits.
To trade an account you need the licensed robot from the price list. It is the same code with the account lock removed and a licence check added.
How it trades
- It trades gold only (XAUUSD or your broker's name for it), in the early session: three price snapshots at 00:00, 02:00 and 03:00 UTC, and a breakout from them before 08:00 UTC.
- It trades only in the direction of the 4-hour, daily and weekly trend, and not on Wednesdays or Fridays, on unusually wide days, or while the market is very quiet.
- Every trade has a stop loss at the broker. There is no take-profit: a trade ends at its stop or at 20:00 UTC of the following day. A winning trade may get one add-on of the same size.
- About six trades in ten end in a loss, each small beside a winner. The result of the back-test comes from a small number of large winning trades.
The reasoning behind this kind of strategy, and what a back-test can and cannot show, is on Trading strategies.
The back-test, in full
Read this before the numbers
These are simulated results on historical prices, produced in the MetaTrader 5 Strategy Tester on 12 November 2026 for XAUUSD (gold), Exness demo-server history. They are not live trading results. The rules were selected using this same history, so the simulation is flattered by hindsight. Past and simulated results do not predict future results. Every month is shown, including every losing one.
Dataset A
One $1,000 account, 28 June 2018 to 10 November 2026
A single account, never topped up and never withdrawn from. The lot size is capped in proportion to the starting balance, so the risk per trade falls as the balance grows — which is why the drawdowns in later years are smaller in percent than in Dataset B below.
MetaTrader 5 Strategy Tester, 1-minute OHLC prices, leverage 1:100, default inputs (5% risk per unit).
Left alone from mid-2018, the simulated account lost money in 2019 and in 2023 and ended at $220,616 on 10 November 2026. 25 months gained and 29 months lost. More than a third of the whole simulated gain came from 2026 alone, and most of that from two months — a result that depends on a few trades is a fragile one. This table is the 1-minute run because it takes the same trades every time anybody runs it; on every tick the same account ended at $231,700 without execution delay, and between $227,520 and $228,843 in 4 runs with random delay.
Month by month: 2024–2026
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year net | End balance | Trades | Max drawdown |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 | −$73 | — | +$6,295 | — | — | −$550 | +$4,523 | +$6,828 | +$8,854 | −$3,340 | −$677 | −$2,240 | +$19,620 | $75,927 | 58 | 9.8% |
| 2025 | +$7,588 | +$4,157 | +$5,102 | — | −$1,474 | — | +$16,653 | −$1,323 | +$4,825 | — | — | +$27,938 | +$63,466 | $139,393 | 62 | 9.8% |
| 2026(part) | +$31,166 | −$8,338 | — | −$1,724 | −$1,766 | +$11,534 | +$965 | — | — | +$49,387 | — | — | +$81,223 | $220,616 | 44 | 8.8% |
ShowHide 2018–2023, month by month — includes the losing years 2019 (−$106) and 2023 (−$5,552)
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year net | End balance | Trades | Max drawdown |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2018 | — | — | — | — | — | — | — | — | — | — | — | — | $0 | $1,000 | 0 | — |
| 2019 | — | — | — | — | — | — | — | −$180 | — | −$78 | +$151 | — | −$106 | $894 | 5 | 25.7% |
| 2020 | — | — | +$2,033 | +$5,606 | — | −$393 | +$12,230 | — | −$1,546 | — | — | −$272 | +$17,658 | $18,551 | 28 | 15.9% |
| 2021 | +$5,596 | +$1,724 | +$8,201 | −$869 | +$7,140 | −$1,106 | −$461 | — | +$5,601 | — | −$722 | −$376 | +$24,728 | $43,280 | 47 | 8.9% |
| 2022 | — | — | — | −$1,027 | — | — | +$18,621 | — | +$1,181 | — | — | −$195 | +$18,579 | $61,859 | 25 | 4.8% |
| 2023 | −$653 | −$363 | — | −$830 | −$772 | −$2,008 | — | — | — | — | −$258 | −$668 | −$5,552 | $56,307 | 28 | 9.3% |
Simulated. Over all 9 years, 25 months gained $253,899 in total and 29 months lost $34,283 in total; 297 trades. A dash is a month in which no trade closed. “Max drawdown” is the largest fall of the balance from an earlier high within that year.
Repeat Dataset A yourself with the free back-test edition: symbol XAUUSDm, timeframe M15, model “1 minute OHLC”, 2018.06.28 – 2026.11.10, deposit 1000 USD, leverage 1:100, inputs all default. The trades repeat — how many, on which dates, at which prices. The money moves a little with the swap rate of the day you test.
Dataset B
A new $1,000 account each year
Each calendar year starts again from $1,000; nothing is carried from one year to the next. 2018 has no trade because the price history starts in late June and the robot needs months of it first. 2026 runs to 31 October.
MetaTrader 5 Strategy Tester, 1-minute OHLC prices, leverage 1:200, default inputs (5% risk per unit).
6 of the 8 years with trades ended higher and 2 ended lower. The worst, 2023, took a new $1,000 account to $181 — a loss of 82%. Even in years that ended higher the account fell a long way first: by 62.6% at one point in 2026. Across all years, 26 months gained and 28 months lost.
Month by month: 2024–2026
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year net | End balance | Trades | Max drawdown |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 | −$73 | — | +$2,266 | — | — | −$550 | +$1,628 | +$6,828 | +$8,854 | −$3,340 | −$677 | +$539 | +$15,476 | $16,476 | 58 | 31.2% |
| 2025 | +$2,482 | +$4,157 | +$5,102 | — | −$1,474 | — | +$16,653 | −$1,323 | +$4,825 | — | — | +$27,938 | +$58,360 | $59,360 | 60 | 38.5% |
| 2026(part) | +$6,310 | −$2,528 | — | −$794 | −$881 | +$1,117 | +$309 | — | — | +$36,068 | — | — | +$39,602 | $40,602 | 41 | 62.6% |
ShowHide 2018–2023, month by month — includes the losing years 2019 (−$393) and 2023 (−$819)
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year net | End balance | Trades | Max drawdown |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2018 | — | — | — | — | — | — | — | — | — | — | — | — | $0 | $1,000 | 0 | — |
| 2019 | — | — | — | — | — | — | — | −$436 | — | −$52 | +$94 | — | −$393 | $607 | 7 | 48.8% |
| 2020 | — | — | +$2,276 | +$5,606 | — | −$393 | +$12,230 | — | −$1,546 | — | — | −$272 | +$17,901 | $18,901 | 28 | 15.3% |
| 2021 | +$2,462 | +$1,724 | +$8,201 | −$869 | +$7,140 | −$1,106 | −$461 | — | +$5,601 | — | −$722 | −$376 | +$21,594 | $22,594 | 47 | 17% |
| 2022 | — | — | — | −$309 | — | — | +$5,231 | — | +$1,181 | — | — | −$195 | +$5,908 | $6,908 | 25 | 30.9% |
| 2023 | −$326 | −$36 | — | −$66 | −$154 | −$132 | — | — | — | — | −$36 | −$68 | −$819 | $181 | 28 | 81.9% |
Simulated. Over all 9 years, 26 months gained $176,824 in total and 28 months lost $19,195 in total; 294 trades. A dash is a month in which no trade closed. “Max drawdown” is the largest fall of the balance from an earlier high within that year.
Repeat Dataset B yourself with the free back-test edition: symbol XAUUSDm, timeframe M1, model “1 minute OHLC”, one run per year: 01.01 – 12.31 (2026: 01.01 – 10.31), deposit 1000 USD, leverage 1:200, inputs all default. The trades repeat — how many, on which dates, at which prices. The money moves a little with the swap rate of the day you test.
What these tables cannot tell you
- The rules were chosen by looking at this same price history. A back-test of rules on the data that shaped them shows what was possible in hindsight; it is not evidence of what happens next.
- It is one symbol at one broker's demo-server prices. Your broker's prices, spread, swap and execution will differ.
- The Strategy Tester's fills are a model. Real orders slip, are rejected and are filled late, most of all in fast markets.
- At the default setting one trade risks 10% or 15% of the account. That is why a fresh account more than once lost over half its value inside a year — and it is a setting you can lower.
- 2026 is a part-year in both tables.
Price list
A licence is for a number of real trading accounts and a period. Every licence also includes 10 demo accounts. Prices are in US dollars, for the software licence only.
Paying online
Buy opens a confirmation page for your account; the payment is in crypto through OxaPay, and the licence unlocks itself once the network confirms the payment. If online checkout is not switched on, that page says so before anything is charged, and a purchase can still be arranged with a person on the official Telegram channel (t.me/Solvex_Space). Solvex will never ask you to send money to an individual's account or to deposit trading funds with Solvex.
From purchase to running
Step 1
Message Solvex on Telegram
Use a button in the price list: it opens the official channel with your chosen plan already typed. Payment is agreed there.
Step 2
Create your solvex.space account
The licence is attached to your website account. Tell Solvex the e-mail address you signed up with.
Step 3
Solvex activates the licence
An administrator switches the product on for your account, for the period and the number of trading accounts you bought.
Step 4
Register your trading account
In your dashboard, enter your MetaTrader account number. You get a licence key for it and the download with its checksum.
Already have a licence? Open your Gold Robot dashboard.
Questions
- Is profit guaranteed?
- No. Nothing about trading is guaranteed, and this robot has lost money over whole years in its own back-test. Treat the back-test as a description of the past.
- Does Solvex trade or hold my money?
- No. The robot runs in your MetaTrader 5, on your account at your broker. Solvex has no access to your account and cannot place or close a trade for you.
- Can I try it before I use real money?
- Yes. Every licence includes 10 demo accounts, and the MetaTrader Strategy Tester needs no licence at all. Run it on demo first.
- What do I need?
- MetaTrader 5 on Windows with a hedging account, a gold symbol, and a computer or VPS that stays on with MetaTrader open from 00:00 to 20:00 UTC on trading days.
- What happens when my licence ends, or if Solvex locks it?
- The robot stops opening new trades at its next licence check — about hourly. Trades already open keep their stop loss and are still closed at their normal time. It is not an instant remote switch.
- Can I lower the risk?
- Yes. The risk per trade is an input. The published back-test uses the default, which is aggressive; a lower setting makes both the gains and the losses proportionally smaller.
Risk warning
Trading gold with leverage can lose all the money in an account, and quickly. Solvex Gold Robot is a tool: it automates fixed rules and cannot know the future. Simulated results on past prices, including the ones on this page, do not predict future results. You choose the risk setting and you alone are responsible for the trades in your account. Solvex Space sells software and gives no investment advice. If you are not sure whether this is suitable for you, do not use real money.