Every trade idea has to prove itself.
BaroqTech builds automated trading systems for gold. Before a strategy is offered to anyone, it has to survive six years of prices, real trading costs and a test for luck. Every result is published, including the ones that failed.
Nothing to pay. One email when the first three months of results are public. See all 30 test results
The watch so far
Gold, XAUUSD. Updated 8 October 2026.- 30 ideas tested and turned down
- 0 passed so far
- Not tested yet
The five that came closest, and what stopped them
The numbers, as they stand today
Counted from the judge's own journal each time this page is built, so they cannot drift from what was really tested. Updated 8 October 2026.
- Trading ideas tested
- 30
- Passed every test
- 0
- Years of gold prices per test
- 6
- Shuffled copies each result must beat
- 2,000
- Tests each idea has to pass
- 9
- Days of demo forward testing
- 2
Luck or edge? 5,000 other histories of one strategy
ORB - NY opening range breakout, long, the strategy running on our own demo and prop-challenge accounts. Its 367 backtest trades were drawn again at random, thousands of times, to see how much of the result was the order the trades happened to come in. The rules were frozen before the out-of-sample part. Lab numbers, not a promise.
- Win rate
- 41%
- Average trade (USD, test size)
- +2.07
- Out-of-sample profit factor
- 1.19
- Histories ending in a loss
- 15%
- Bad-case drawdown (5 %)
- −1,077
- Real backtest drawdown
- −509
It started with a loss nobody had measured
In nine days, a signal-copying robot placed 259 trades on a demo account and lost 1,390 dollars. It was following fifty Telegram channels, and nothing showed which of them was to blame.
So the first thing built here was the tool that could tell: one that counts a position as one trade, charges every cost, and refuses to call anything good before thirty trades.
The second thing was a rule. Everything is measured the same way, nothing is offered until it has passed, and the measurements are shown to whoever asks. That rule is this company.
Three ways a trading robot fools its buyer, and its builder
Most robots are sold on a chart of the past. That chart is the easiest thing in trading to get wrong, usually without meaning to.
The past is flattering
Any rule can be tuned until history looks good. So each idea is tuned on one stretch of prices and judged on a later stretch it has never seen, eight times over.
Costs are not a detail
Spread, commission, slippage and overnight financing are charged on every test trade. Then the test runs again with the costs doubled. An edge that only survives cheap trading is not an edge.
Luck looks like skill
Test enough ideas and one will shine by accident. So every new test raises the bar for all the others, and each result is compared with two thousand shuffled copies of itself.
The core of it: study the funds that do this best
The most successful quantitative funds keep their strategies secret. How they work is not secret: it is in their papers, their books and their interviews. Studying that is a daily job here, and it shaped every rule on this page.
Many small edges, not one big idea
Its Medallion fund is reported to have been right on barely more than half of its trades, across a very large number of them. So no single strategy is trusted with much here, and a high win rate on its own counts for nothing.
Take published research, then test it yourself
These firms publish their work on trends and momentum. Each published idea is put through the same tests here, on gold, with our costs. Most do not survive, and the record says so.
The loss limit comes before the profit
The large multi-strategy funds are known for giving every team a strict loss limit and cutting what stops working. Here every strategy has its own switch, a daily loss limit and a floor.
Assume it was luck until shown otherwise
Their published test asks how good a result has to be once you count every idea you tried. It is one of the tests each idea has to pass here.
BaroqTech has no connection with any firm named here, and does not have their strategies, their people or their results. The names say whose published work is being studied. Read the playbook
From an idea to a live account
Four steps, in this order. A strategy that stops earning goes back a step. Nothing is deleted.
The lab
The reason is written down before the test: what should happen, and why. Then six years of gold prices decide.
The judge
A locked set of tests that nobody adjusts afterwards, the author included. Most ideas end here.
The demo account
Thirty closed trades over at least three weeks, earning no less than half of what the test promised.
The live account
Small size, a stop on every trade, a daily loss limit, and a floor that switches everything off.
The desk: see what your account is really doing
A private dashboard that reads your MetaTrader account and cannot place a trade. It runs on your own server, so your trading history never leaves it.
- Every source gets a verdict
- Each robot, strategy and signal channel is scored the same way. Before thirty trades over three weeks the verdict reads "too early", whatever the profit says.
- One position is one trade
- A position closed in three parts is not three wins. Commission and swap are inside every number.
- Demo and live never mix
- Two sides, two colours. No total ever adds practice money to real money.
- It adds up to the cent
- Every page can be checked against MetaTrader's own list of deals.
- A meter for prop challenges
- Profit against the target, and the room left before the daily limit and the account floor.
- Made for the phone too
- The same pages, laid out for one thumb.
The record, as it stands
No strategy has met the bar yet. Until one does, the trading system is not for sale.
When there are three months of results on a public, verified account, they go on this site as they are, losing days included. You will not find a promised monthly return here, now or then.
- Trading ideas tested
- 30
- Passed every test
- 0
- Gold price history used
- 6 years
- Demo forward test running since
- 7 October 2026
Made for gold traders who want to know, not hope
You copy signal channels
Find out which channels earn and which only look busy. A channel that wins nine trades in ten can still lose money.
You are in a prop challenge
See the distance to the target and to each loss limit at a glance, open losses included.
You have bought robots before
You have seen the perfect chart and what came after it. Here you see the tests a strategy had to pass, and the ones it failed.
Price
Nothing is charged today. Join the waiting list and the lower price is yours for as long as you stay.
The desk
$12 a month on the waiting list
$24 a month for everyone else
- Statistics, calendar and every chart on your own accounts
- A verdict for every robot, strategy and signal channel
- The meter for prop challenges
- Cancel any month
The trading system, with the desk
$39 a month on the waiting list
$79 a month for everyone else
- The gold trading system for MetaTrader 5
- The desk is included
- On sale only after a strategy has passed every test and three months of results are public
- Cancel any month
| Included | The desk | The trading system, with the desk |
|---|---|---|
| Statistics, calendar and charts on your own MetaTrader accounts | Yes | Yes |
| A verdict for every robot, strategy and signal channel | Yes | Yes |
| Prop challenge meter: target, daily limit, floor | Yes | Yes |
| Demo and live accounts kept apart | Yes | Yes |
| The gold trading system for MetaTrader 5 | No | Yes, once a strategy has passed |
| Daily loss limit and account floor built into the system | No | Yes |
| Runs on your own computer or server | Yes | Yes |
| Cancel any month | Yes | Yes |
Planned prices in US dollars. They are fixed on the day each product goes on sale, and the waiting-list price will not be higher than shown here. No payment details are asked for now. Refunds and cancelling
Why there are no testimonials on this page
Because nobody has used the trading system yet, and a quote from a customer who does not exist is the first lie a trading robot tells. Here is what you get instead.
Every result, with its reason
All 30 ideas tested so far are on the record, with their score and the test each one failed. Nothing is left out because it looked bad.
Numbers you can trace
Every figure on this site comes from a file that can be shown: the judge's journal, a dated report, or MetaTrader's own list of deals.
Real accounts, when they exist
Three months of results on a public, verified account come before any sale. Reviews will come from real customers, after they have used it, and will be shown unedited.
Questions
Does it make money?
That is not proven yet. 30 ideas have been tested and none has passed yet. A forward test on a demo account began on 7 October 2026. This site will say when that changes.
Do you trade my money?
No. BaroqTech makes software that runs on your own MetaTrader account. Nobody here can touch your funds, and nobody gives you personal investment advice.
Can I buy it now?
Not yet, and nothing is charged for joining the list. The trading system goes on sale once a strategy has passed every test and three months of results are public. People on the waiting list hear first and keep the lower price.
What will I need?
MetaTrader 5, a broker that offers gold (XAUUSD), and a Windows computer or server that stays on.
Can I use it on a prop firm challenge?
The desk reads any MetaTrader 5 account, challenge accounts included, and its meter shows the distance to the target and to each loss limit. Whether a firm allows automated trading is that firm's rule, and some forbid it. Read your firm's rules before you attach any robot.
Can I cancel, and are there refunds?
Nothing is sold today. When the products go on sale, every plan can be cancelled any month, and the first payment is refunded in full if you ask within 14 days. The full refund policy
What happens to my email address?
It is kept for the one email you asked for, or for the answer to your question, and for nothing else. Ask and it is deleted. Privacy
Is there a guarantee?
No. Trading gold with leverage can lose money quickly, including all of it. A test, or a demo account, does not tell you what a real account will do. Read the risk warning
Ask something else
Write your question and you will get an answer by email, from a person.
Your address is used to answer this question and for nothing else.
Hear about it when there is something to show
One email when the first three months of results are public. No newsletter and no countdown timers.
Your address is kept for that one email and nothing else. Ask, and it is deleted. Privacy