Skip to content
AlgoVerdict
Menu
Ultimate Breakout System product shot

EA review ·

Ultimate Breakout System review

provisional

Ultimate Breakout System

4 signals on real-money accounts (per MQL5 rules), selected by the vendor

The vendor lists 4 public signals. We checked that they exist and can be viewed. The seller still chooses what to show, so such evidence earns at most 8.0 for verifiability.

Checked
Signal exists and is public · real money per the MQL5 signal rules, broker’s live server · trading behaviour matches the described design
Not shown
Return, drawdown, running time. Past figures tempt buyers more than they inform them.
Counts towards
Verifiability (25% of the grade): 6.0 out of 10
Strength of evidence
2 of 3: selected by the seller
Breakout6 marketsMT5

Reviewed by Pitt Petruschke, operator. No stake in Ultimate Breakout System; receives commission from ATS and develops his own EA, Dawnbreak. Disclosure

Cross-checked by an AI agent (different model, blind): 4 of 6 criteria agree. Differences

Verdict

Provisional grade: a flexible toolkit with a stop-loss on every trade and set files that are broadly profitable in the backtest, but a young main signal with a marked decline and several losing months, a test account with a very deep decline by balance, drawdown figures that in many sets are below the decline measured in our backtest, a martingale grid set without a basket stop in the package and a high price.

The Ultimate Breakout System by Profalgo Limited is a breakout toolkit for MT5 with three entry models and about 300 set files, per the vendor the basis of The Gold Reaper. It costs $1,699 (as of 19 Nov 2026). We backtested all 283 set files in the package ourselves: broadly profitable in the tester, but the vendor's live main signal does not match so far; the reasons cannot be separated from the outside. Provisional grade.

Suits you if …

  • Yes: you want an EA with fixed entries and exits
  • Yes: you run it as one building block in a portfolio
  • Caution: you set setfile and risk yourself, without ATS calibration

How the strategy trades

Schematic · no market dataSchematic of the Breakout strategy type: it explains the process, not the outcome. No prices, no real market data, no performance of Ultimate Breakout System.

Example · closed

How it works

  1. In every session a quiet phase marks out a range: its high and its low.
  2. A buy stop sits above the high, a sell stop below the low. Whichever fills first counts; the other is cancelled.
  3. The stop sits on the far side of the range, the target at 1.5× that distance. Because every range has a different width, the levels move each time.

What we check

What we check on real breakout EAs: how often false breakouts hit the stop, and whether the reward-to-risk still holds after spread and slippage.

Schematic with demonstration values: a simulated price path without prices, results in pips weighted to 0.10 lots, deliberately favourable. Not this EA’s market or settings, not a backtest and not investment advice.

Design & risk

Ultimate Breakout System by Profalgo Limited is a Breakout EA on XAUUSD, BTCUSD, NAS100, US30 (MT5), sold on the MQL5 market. Each position has its own exit; losses are not absorbed by averaging in. What matters is whether execution at your broker matches the described process. You buy the vendor default; risk frame and setfile are up to you.

The full review

Overview

The Ultimate Breakout System (UBS for short) by Profalgo Limited (Wim Schrynemakers) is not a finished EA with one strategy but a breakout toolkit for MetaTrader 5. You load set files, each describing one strategy on one market and timeframe, and run them singly or by the dozen from one chart. The package costs $1,699 on the MQL5 Market, or rent from $399 for one month, with a free demo and 10 activations. It was released on 16 May 2025; the current version is 9.8 of 17 November 2026 (read on 19 November 2026). UBS is not available at AlgoTradingSpace.

For this review we did not only read the product page and the signals: we backtested all 283 set files in the package ourselves. The result is mixed: in the tester the package looks strong, but the vendor's live main signal does not match it. Our grade is provisional.

What the Vendor Sells

Per the product page (read 19 Nov 2026):

  • Entry models: breakout of support and resistance with pending orders, volatility breakout after a consolidation, and opening-range breakout. Since version 6.4 there is also an entry after a break and retest.
  • Exit: stop-loss and take-profit on every trade, plus many variants of a trailing stop (time-based, at support and resistance, “MagicTrail”) and a trailing take-profit.
  • What you get: more than 300 set files per the page, future sets, a manual, a video guide, optimisation templates and a private group. The sets name market and timeframe; they come from the developer and by now also from clients.
  • Portfolio mode: since version 5.0 the “OneChartSetup” (autoloader) loads many sets from one chart. The lot size can follow the historical drawdown stored in each set via “Max Risk Per Strategy”.
  • Bonus: per the page one more of the vendor's EAs for free.

The vendor describes UBS as the basis of his other EAs: “This system has served as the foundation for several top-performing EAs on the MQL5 market, including the acclaimed Gold Reaper EA” (read 19 Nov 2026). The package does contain set files for Gold Reaper, Goldtrade Pro, Goldbot One, Indicement, Daytrade Pro and Bitcoin Reaper strategies. Whether they match the original EAs exactly, we cannot check.

Marketing, verbatim, read on 19 Nov 2026: “This package will only be sold at current price for a very limited number of copies. Price will go to 1999$ soon!” and “+ 300 Strategies included and more coming!”. A buyer wrote in the comments on 13 Jul 2026 that he had paid $1,499; today it is $1,699. The version 9.8 package we tested held 294 files, 11 of them optimisation templates and 22 duplicate sets, so 261 unique strategies. Like the other Profalgo pages, this one explicitly distances itself from “neural network/machine learning AI/ChatGPT” marketing.

Design and Risk

Breakouts with pending orders and a fixed stop on every order limit the single loss. With a toolkit, though, the risk depends mostly on which sets you choose and how you set the risk. Four points stand out.

1. Grid and martingale are built in, but not standard. The version history says in version 8.2 “Fixed — grid distances are now in pips”; the manual documents the grid function with a martingale factor and warns against it. It is off in the standard sets. In one set of the package it is active, more on that below. How such baskets fail is explained in Grid and martingale EAs.

2. Lot size depends on a figure in the set. Under “Max Risk Per Strategy” the EA sizes lots from the stored “Max historical DD”. If that figure is below the actual decline, the EA trades larger lots than the chosen risk promises. In many sets it was lower than the decline measured in our backtest.

3. Many sets, one market. Most sets trade gold, the second-largest group bitcoin. Similar gold sets often lose on the same day. Anyone loading “all sets” gets far less diversification than the number of strategies suggests, and the duplicate sets in the package would trade twice. How that adds up in a shared drawdown is worked through in EA portfolio management.

4. Leverage and account size. The main signal runs at leverage 1:1000, which is why MQL5 does not allow subscriptions to it. In the EU, retail clients get at most 1:30 on major pairs and 1:20 on gold, and far less on crypto. With many sets open at once, margin can then get tight. Deposit load was low on the main signal and at times high on the client account and the scalp accounts. The product page does not state a recommended risk per trade (read 19 Nov 2026). On small accounts the EA rounds many lots up to the minimum, so the real risk per set is above the one set. How to cap risk per trade properly is shown in Position sizing and risk per trade.

The main signal holds positions for several hours on average. The scalp sets, by contrast, hold for seconds to minutes and need tight spreads. We therefore do not list UBS as a scalper, but anyone using the scalp sets is running one.

Our Backtest of All 283 Set Files

Method. We ran each of the 283 set files from the version 9.8 set package provided to buyers on its own in the MT5 strategy tester, three years each from 18 Nov 2023 to 18 Nov 2026. We used 1-minute OHLC data with the historical spread of a raw-spread account, commission and swaps included, without slippage. Lot sizes stayed as in the set; nothing else was changed. The news filter uses the MQL5 calendar and probably does not work in the tester. The sets' optimisation periods are unknown; part of the test period is likely inside them. A backtest is not a live result, and that is exactly what shows here. How to read backtests is explained in Backtesting and forward testing. Test reports and analysis are archived with us.

In the tester the package is broadly profitable. Of 261 unique sets, 250 closed positive over three years and 221 also in the current year 2026. The strongest were support/resistance breakouts on gold and bitcoin. Part of that is tailwind: 2024 to 2026 was a strong gold bull market, and the gold sets' gains came mostly from long trades. The best figures came from bitcoin sets added to the package only one day before the test ended; for them the whole test is in-sample and no proof of quality.

True out-of-sample is weaker. 179 sets carry the file date 7 May 2026, so June to October 2026 is demonstrably after their creation. In that period 139 of 177 unique sets closed positive, but on average with a noticeably lower monthly return than in 2024 and 2025.

Backtest and live do not match so far. The main signal lost in July, September and October 2026. For comparison we drew 44 sets at random 2,000 times from the sets that existed before the signal started and scaled them as the EA does under “Max Risk”. Only about 50 of those 2,000 portfolios were negative in two of these three months in the backtest. Possible reasons: a different selection ran live (the number of trades does not fit a random selection), the small account rounds lots up, and pending orders get filled with slippage live. How much each reason contributes cannot be separated from the outside. More on this gap in Gold EAs: backtest vs. live.

The stored drawdowns are often below the measured decline. In 71 of 258 sets with a figure, the measured decline by equity exceeded the set's “Max historical DD”; by balance it was 41 sets. The manual does not say which one is meant. Many client sets are among the largest deviations; in two client sets for USDJPY the figure is below the loss of single trades in our backtest and would lead to very large lots under “Max Risk”. Anyone using “Max Risk Per Strategy” should check the figure per set and rather set it too high.

One set is a martingale grid without a basket stop. The “Grid” folder holds a gold set with martingale factor 1.9, up to ten levels, without a basket stop in the set (a group loss limit is possible in the autoloader but is not configured) and a stored historical drawdown of zero. In the backtest it was one of the most profitable sets, but with a marked open loss, although the starting lot was only a tenth of the usual one; at the usual starting lot the open loss would have exceeded the test account. We would not include it in a portfolio.

Not all sets hold up in the backtest. The Daytrade Pro sets are the outlier on the downside: 7 of 13 closed negative over three years, 8 of 13 in 2026. The ORB Master sets were all positive over three years, but only 3 of 12 in 2026. The gold volatility sets on M5 all lost in 2026. In October 2026, the weakest month, 106 of the 261 sets lost at the same time.

What follows: start with a small, deliberate selection, not all sets; leave out the grid set, the Daytrade Pro sets, the gold M5 volatility sets, the two USDJPY client sets and the duplicates; cap the risk per market; test the selection as one combined run with the autoloader, and before real money let it run on demo for a few weeks next to the backtest.

The Vendor's Signals

The product page links three MQL5 signals. What we read on 19 November 2026, without figures, as in all our reviews:

  • Main signal “Ultimate Breakout System 44 strategies” (2373798), since 16 May 2026, leverage 1:1000. Per its description 44 strategies run across gold, forex, indices and bitcoin from one chart; until the end of July there were 30. The account is above its starting capital; the largest decline by balance is marked. June and August brought most of the growth; July, September and October were losing months. MQL5 noted in August that most of the growth came within a few days. By market, gold was positive and bitcoin negative, although the bitcoin sets were among the strongest in the backtest. On 15 Jun 2026 the developer reported strong growth with a very small decline for this portfolio in the comments; the decline since then is considerably larger.
  • “UBS Scalp Set C” (2385027), since 4 August 2026 on a small account, a gold scalp set with holding times under a minute. Most of the growth came in the first month; MQL5 warns of a high risk of negative slippage when copying.
  • “UBS Christmas” (2385751), a client account, not run by the vendor, since 9 August 2026. Here too most of the growth came in the first month, and October was a marked losing month. MQL5 shows that not all trades were algorithmic.

Not linked is a fourth UBS account of the vendor, the test account “UBS Scalp Test …” (2383386), since 23 July 2026, per its description first with seven gold scalp sets from UBS and since 20 August with Set C only. It is in profit, but had a very deep decline by balance in its first month. We count it anyway: what matters to us is the worst value from all sources, not only the linked ones, because the vendor chooses which accounts appear on the product page. Per its description the account clearly trades UBS, and since 20 August it has run the same Set C as the linked signal 2385027. For context: it is a small account, where lot rounding and single losses weigh more heavily. How we classify it is in the calculation method.

Real money or demo: per MQL5 rules only real-money accounts are allowed for signals, MQL5 shows the “Real Account” icon on all four, and the accounts run on the brokers' live servers. Which sets and risk values exactly run is only roughly known for the main signal. The older signals of the individual Profalgo EAs, whose strategies UBS builds on per the vendor, run longer; but they evidence those EAs, not this toolkit. We do not run UBS on an account in our own EA park.

What Buyers Report

On the MQL5 Market UBS shows 5 out of 5, the “Reviews” tab lists 64 entries, plus 108 comments (as of 19 Nov 2026). Buyer reviews do not count towards our grade.

Praise goes mainly to the flexibility, the bundled sets and the support: “Wim keeps updating and improving it, and his support is top-notch.” (review of 27 Oct 2026, read 19 Nov 2026). Another buyer writes that he is still getting to grips with the risk settings and the number of sets (review of 21 Oct 2026). A renter found the configurations confusing and asked which set to use for gold (comment of 23 Mar 2026). One buyer reported that the autoloader could not be migrated to the MQL5 VPS; the developer confirmed: “the autoloader doesn't work on MQL5 VPS” (comment of 30 Jun 2026, read 19 Nov 2026). Clients share their own sets and analyses in the comments, for example on the correlation of index sets.

How the Provisional Grade Comes About

The grade follows the same thresholds as for every EA (calculation method), with deliberate deviations downwards where the data formula would be too kind:

  • Risk control: the worst decline across all four signals falls into the “very deep” band. It comes from the unlinked test account; without it, the main signal's decline would be “marked”. Even then we would deduct: a martingale grid set without a basket stop in the package, drawdown figures that in many sets are below the measured decline, and lot rounding on small accounts. We do not apply the martingale cap because the function is off in the standard sets.
  • Verifiability: four public accounts count, but vendor accounts are capped with us. We lower further, because of the quality of the evidence: MQL5 notes on the main signal that most of the growth came within a few days; one of the linked accounts is a client account with partly manual trading; the two scalp accounts have only run for a few months. Losing months and the gap between backtest and live feed into the risk assessment and the text, not into verifiability.
  • Transparency: credit for the bundled sets, the manual, the shared live portfolio sets and a detailed version history. Deductions because the product page does not link the test account with a very deep decline, many drawdown figures are below the decline measured in our backtest, the manual leaves open which kind of drawdown is meant, and the page uses scarcity marketing.
  • Value for money: $1,699 falls into the “over $1,000” band.
  • Usability: we ran the EA ourselves in the tester with all sets. Deductions for the entry hurdle of a toolkit, the autoloader not running on the MQL5 VPS, and numerous fixes to lot size and risk calculation in autumn 2026.
  • Broker independence: evidenced at two brokers; a small deduction for the main signal's high leverage and the scalp sets' dependence on spreads.

As for every EA, an AI from a different model family also rates the EA blind; where it deviates strongly on a criterion, the mean applies there until a human has checked. For the Ultimate Breakout System that is the case for value for money and broker independence. We will review the grade when the main signal reaches one year or the drawdown figures in the sets are corrected.

Who It Is (Not) For

Suitable for experienced traders who want to build, optimise and test strategies themselves and have the time to check every set individually. Anyone already running The Gold Reaper or another Profalgo EA gets the same strategies to assemble themselves with UBS, but no extra diversification.

Not suitable for anyone expecting a finished system, for small accounts with many sets, for traders with EU leverage and a large portfolio, and for anyone who wants to use “Max Risk Per Strategy” without checking the stored drawdowns. Why backtests often disappoint live is described in Why EAs fail live.

Verdict

The Ultimate Breakout System is, per the vendor, a breakout toolkit for MT5 with a stop-loss on every trade and about 300 set files, for $1,699. In its favour are its flexibility, the openly bundled sets and a backtest in which almost all sets close positive. Against it stand a young main signal with a marked decline and losing months the backtest hardly led one to expect, an unlinked test account with a very deep decline by balance, drawdown figures that in many sets are below the measured decline, a martingale grid set without a basket stop in the package and a high price. Our grade is therefore provisional. Use it only after your own test and only with money you can afford to lose. 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. The operator of AlgoVerdict sells its own EAs on the MQL5 Market, including Dawnbreak, a breakout EA, and therefore competes with Profalgo. We do not grade or compare our own products. Past results are no guarantee of future ones, and trading leveraged products can lead to the loss of the capital invested.

What you need to run it

  • Broker with low spreads on XAUUSD, BTCUSD, NAS100, US30Raw or ECN account, no EA restrictionsrequired
  • VPS close to the brokerThe EA has to run around the clockrequired
  • Fix the risk per tradeChoose a lot size that fits the accountset it up yourself

For and against

In favour

  • Per the vendor a stop-loss and take-profit on every trade; the standard sets use no grid and no martingale
  • In our own backtest (no live evidence) almost all of the 261 unique sets close positive over three years, most of them in 2026 as well
  • Set files, manual, video guides and even the live portfolio's sets ship with the purchase
  • Flexible: build your own strategies on any market, run many sets from one chart
  • Very active development with a detailed version history

Against

  • Backtest and live main signal do not match so far: three losing months since launch, a pattern random backtest portfolios show very rarely
  • Marked decline on the main signal, very deep decline by balance on a vendor test account that the product page does not link
  • In many sets the stored drawdown figures are below the decline our backtest measured, so “Max Risk Per Strategy” can understate the risk
  • One set in the package is a martingale grid without a basket stop in the set (a group loss limit is possible in the autoloader but is not configured)
  • Weaker modules in the backtest: Daytrade Pro sets weak, ORB Master sets mostly negative in 2026 (positive over three years), the gold M5 volatility sets all negative in 2026
  • High price and scarcity marketing (“Price will go to 1999$ soon!”, read 19 Nov 2026)
  • A toolkit: you have to choose sets, check each set's risk and test it yourself

Buying and running

Our interests: We earn no commission on Ultimate Breakout System; we link the vendor and MQL5 only as sources, never as partner links. ATS pays us a commission when you buy through the link to the alternatives. We develop our own EA, Dawnbreak. We have no stake in Ultimate Breakout System. The grade is set beforehand and independently of that.

Terms

Ad
Price
$1,699
Sold
MQL5 market only
Alternatives at ATS

We do not link to Ultimate Breakout System itself. The affiliate link changes neither grade nor text.

What we do not know

  • Which sets exactly run on the main signal.The vendor shared set packages for the live portfolio in the comments (most recently on 30 Jul 2026, when it grew from 30 to 44 strategies). Whether further sets or other risk values were added since, the signal page does not show.
  • How much of the gap comes from execution.Our backtest has no slippage. Breakouts with pending orders get worse fills live in fast markets; without the main signal's trade list we cannot separate how much of the gap that explains.
  • Whether another gold scalp test account uses UBS.Whether another gold scalp test account of the vendor (2380747, since 5 Jul 2026) uses UBS is open; the signal description does not name the EA. We therefore do not count it.
  • Whether price- and volatility-based lot sizing applies under “Max Risk”.Many sets scale the lot size with price or ATR. Whether the EA also applies that factor under “Max Risk Per Strategy” could not be checked without the source code.

Frequently asked questions

What is the Ultimate Breakout System?

An Expert Advisor for MetaTrader 5 by Profalgo Limited (Wim Schrynemakers) for building and running breakout strategies on any market. It has three entry models: support and resistance breakout, volatility breakout and opening-range breakout. It ships with about 300 set files, per the vendor including the strategies of his other EAs such as The Gold Reaper. It costs $1,699 on the MQL5 Market (as of 19 Nov 2026).

Does the Ultimate Breakout System have a verified live signal?

Yes. The product page links three MQL5 signals: the main signal with 44 strategies since May 2026, a gold scalp signal since August 2026 and a client account. There is also a vendor test account that is not linked. Per MQL5 rules only real-money accounts are allowed there, and MQL5 shows the “Real Account” icon on each. We do not run the EA on an account in our EA park.

What does your backtest of the set files show?

We ran all 283 set files of the version 9.8 package over three years in the MT5 tester, with 1-minute data and no slippage. In the tester the package is broadly profitable. The vendor's live main signal, however, had three losing months, which random portfolios built from the same sets show only very rarely in the backtest. In many sets the stored drawdown figures are also lower than the decline measured in our backtest, and one set is a martingale grid without a basket stop. A backtest is not a live result.

Does the Ultimate Breakout System use grid or martingale?

The standard sets do not: they place a stop-loss on every trade. The EA does have a grid function with a martingale factor, mentioned in the version history (version 8.2) and the manual. In the package we tested it is active in exactly one set, with martingale levels and without a basket stop in the set (a group loss limit is possible in the autoloader but is not configured). We would not use that set.

Why is the grade provisional?

Because the main signal has only run for about half a year and the scalp accounts for a few months, and because backtest and live trading do not match so far. We will review the grade when the main signal reaches one year or the vendor corrects the drawdown figures in the sets.

Is there a verified live track for Ultimate Breakout System?

Yes, 4 public signals on real-money accounts (per MQL5 rules), selected by the vendor. Which settings run there is not disclosed. Because the vendor chooses what to show, such accounts earn at most 8.0 out of 10 for evidence. We deliberately do not show returns or drawdowns.

Changes to this review

  1. Upgraded from the MQL5 list entry to a full review with a provisional grade. Product page, version history, comments and four MQL5 signals read on 19 Nov 2026; our own backtest of all 283 set files on 18 Nov 2026.
  2. Second review by an AI agent (different model, blind): 4 of 6 criteria agree. Differences:
  3. Preis-Leistung: Zweitprüfung 4,2, unsere Bewertung 5,6. Der hohe Kaufpreis steht zwar einem umfangreichen Paket mit Sets, Anleitung, Vorlagen und Zusatzleistungen gegenüber, ist angesichts der erheblichen Auswahl-, Prüf- und Risikomanagementarbeit für den Käufer aber schwach gerechtfertigt.
  4. Broker-Unabhängigkeit: Zweitprüfung 6,0, unsere Bewertung 7,2. Live-Nachweise bei unterschiedlichen Brokern sprechen dafür, aber die Bindung an MT5, die hohe Hebelabhängigkeit und die Spread- sowie Slippage-Empfindlichkeit der Scalp-Sets begrenzen die Übertragbarkeit.

More commodity EAs we reviewed

Compare gold EAs

Compare MQL5 EAs