Expert Advisor Review

Dark Nova EA Review MT4/MT5

Not Recommended Early Review MT4 / MT5 AUDCAD, AUDNZD, NZDCAD — counter-trend grid $499 Verified 1 Nov 2026
Early Review: based on a real-money record of nearly 23 months (the account behind MQL5 Signal 2312398 made its first trade on 2 December 2024; MQL5 registered the signal on 30 May 2025). An Early Review is for a real-money record of at least 12 but less than 24 months, with at least 300 trades. It can come out Inconclusive or Not Recommended, but not Recommended, however good the record looks: we want two years before we recommend anything. Both products were published on 12 January 2026, so more than a year of this record predates the listings; the seller links the signal from both pages. On 2 December 2026 the record reaches 24 months: at our next quarterly re-verification after that we will re-read the full history and drop the Early Review label. Time alone does not lift a Not Recommended; what would change it is set out in the disclosures below.
Which product this is: Dark Nova by Marco Solito, sold on the MQL5 Market as Dark Nova MT5 (product 139187) and Dark Nova for MT4 (product 139186), each $499. The live record is an MT4 account.

The Bottom Line

Not Recommended. The Robotic Trader rates Dark Nova Not Recommended because the live record the seller presents for it trades a martingale grid. On that signal, 330 trades were each opened while another trade on the same pair, in the same direction, was still open: 329 after the price had moved further against every trade already open (lower for buys, higher for sells), and all 330 bigger than every trade already open. At its deepest, 13 trades were open on one pair in one direction, at sizes rising from 0.01 to 0.13 lots, 0.91 lots in total, 91 times the size of the first trade (the largest single trade, 0.13 lots, was 13 times the first). The pattern runs through the whole record: 169 of the added trades came before the products were published on 12 January 2026, and 161 after. Our methodology treats any grid, averaging or martingale pattern in the trade history as an automatic Not Recommended, whether or not the seller discloses it.

What "martingale" means here. The classic martingale doubles the size after every loss. Dark Nova's added trades mostly grow by 0.01 lot at each step (0.01, 0.02, 0.03 …), with the first step a doubling. Our definition of martingale is size raised after a loss, or as a position moves against you; this meets the second part. The product's settings include a grid lot method called "Martingale", among others; we do not know which method the signal uses.

What it means in practice. Adding bigger trades after the price moves against a position can let a group of trades close at a profit before the price gets back to the first trade's entry; we did not measure how far the price came back each time. On the closed record, every group of trades on one pair and direction closed at a net profit, although individual trades lost. If the price keeps moving against the position instead, the open loss deepens as the trades get bigger. No stop loss or take profit is recorded at the broker on any of the 1,081 trades; the settings have an optional switch that stops the EA after a set money loss; the product page does not say whether it closes open trades, and we cannot see whether it was used. At our check, open trades were $498.17 in loss: 13.8% of the balance and 16.1% of equity. The balance still includes the $1,200 deposited on 12 October 2026, so both percentages are smaller than they would have been before that deposit. (The AUDNZD group open on that date closed on 16 October; this open loss is a later one.) MQL5 also shows a relative drawdown by equity of 51.10% at some point in the record; it does not say when, or how large the account was then.

The strongest case for it is real. The record is real money and nearly 23 months long, and the seller links it from both product pages. It made a closed profit in every calendar month from December 2024 (trading from the 2nd) through October 2026 (a grid that never closes a group at a loss tends to produce that; open losses are not counted). The grid settings are listed openly in the product description, and one buyer review says that "as the author himself points out" it may open additional positions to cover a previous trade.

What would change this verdict is set out in the disclosures below.

Methodology
ⓘ

Evaluation Methodology

We assessed Dark Nova as an Early Review. Its live record was nearly 23 months old when we read it, short of the 24 months a full review needs; see how we test.

What we read:

  • The signal and its full history: the live signal both product pages link (MQL5 Signal 2312398, "Dark Nova 3 Pairs"), and its complete trade history exported from MQL5: 1,081 closed trades and 5 deposits, to 30 October 2026.
  • Both product pages (MT5 and MT4) and their input lists.
  • Reviews: MQL5 buyer reviews on both product pages, and a web search for other reviews.

How we tested for a grid: we looked for trades opened on the same pair, in the same direction, while an earlier trade was still open, more than 60 seconds after it (no trade in this record opened within 60 seconds of another on the same pair and direction). For each, we checked whether it opened at a price below (for a buy) or above (for a sell) the opening price of every trade already open, meaning all of them were then in loss, and whether it was larger. We call that "after the price had moved further against" the open trades. A group is a chain of same-pair, same-direction trades in which each trade after the first opened while an earlier trade in the chain was still open; a single trade is a group of one.

Prop-firm test. Both product pages say "FTMO compatible", so our prop-firm test applies. We have not run it yet; we will add the result to this review as a dated entry.

Note. Figures were read on 1 November 2026. They are a dated snapshot, not a live feed.
  • Open trades: MQL5 hides open trades from non-subscribers. We can see their total open loss, but not how many there are or how long they have been open.
  • The latest day: our history runs to closes on 30 October 2026.
  • Drawdown: MQL5 checks open losses at intervals, so a brief deeper dip between checks may not appear in its figures.
  • Our figures: figures marked "our calculation" come from the exported history. Profit figures are after swap (−$110.23 in total); the account charged no commission.
  • What MQL5 tracking shows: that the trades came from this account. It does not show that this EA placed them, or which version or settings it ran. Both products were published on 12 January 2026, so more than a year of the record predates them; the seller links the signal from both pages, and the grid pattern appears both before and after that date.
  • Testing: we did not test the EA ourselves.

Live Performance Data

The record that counts: "Dark Nova 3 Pairs" (MQL5 Signal 2312398).

Real account — linked from both product pages

Dark Nova 3 Pairs

Recorded: 2 Dec 2024 – 30 Oct 2026; our calculation from the exported history unless marked (MQL5 Signal 2312398)
1,081
Total Trades
75.5%
Win Rate (MQL5; excludes one trade closed at zero)
2.48
Profit Factor
+$3.19 / −$3.96
Average Win / Loss
+$1,549.69
Net Profit
51.10%
Drawdown by Equity (MQL5, relative)
Win rate, profit factor, net profit and the monthly results count closed trades only. Open losses are outside them, which is why the closed record looks smooth while the drawdown by equity reached 51.10%.
AccountReal money, USD, Headway, MetaTrader 4, leverage 1:500. A paid MQL5 signal (999 USD a month).
DatesMQL5 shows "Started 2025.05.30", which is when the signal was registered; the earlier trades come from the account's own history. The history begins with a $100 deposit and a first trade on 2 December 2024.
Cash flowsOpening deposit $100; four further deposits totalling $1,950 (see "Deposits" below); no withdrawals. With the closed profit, that comes to $3,599.69, which is the balance MQL5 showed. Equity was $3,101.52: open trades $498.17 in loss.
Growth figureMQL5 shows 1,166.09%, which we do not use as our measure. It chains percentage returns between deposits, so it is not profit divided by money paid in, and should not be read as the return earned on all $2,050. Closed profit was $1,549.69; counting the open trades (equity), the gain was $1,051.52. Neither is a percentage return: $1,950 of the $2,050 paid in arrived in 2026, $1,200 of it on 12 October 2026.
Drawdown — use 51.10% for the grid's riskMQL5's relative drawdown by equity, 51.10%, counts open losses; MQL5 does not say when it occurred or how large the account was then. The balance figures leave open losses out, so they look smaller:
  • MQL5's relative drawdown by balance, 14.94%;
  • MQL5's maximal drawdown by balance, 8.42% ("relative" is the largest percentage fall; "maximal" is the largest fall in money, shown as a percentage);
  • ours, 13.86%: the largest fall in the closed balance from a previous high to a later low, from $249.18 to $214.64, both on 7 April 2025 (no deposits were made in 2025, so including or leaving them out makes no difference).
We have not reconciled our 13.86% with MQL5's 14.94%; MQL5's 8.42% uses a different definition.
Per tradeBest +$126.12; worst −$45.32. 815 won, 265 lost, 1 closed at zero. Each win averaged $3.19 and each loss $3.96: a payoff ratio of 0.81, which a high win rate has to carry.
ActivityAverage holding time 58.6 hours (our calculation); up to 22 trades open at once across the three pairs.
By year (closed profit)2024 (from 2 December) +$33.81; 2025 +$623.39; 2026 to 30 October +$892.49. Every calendar month made a closed profit.
By pairAUDCAD 447 trades, +$688.34; NZDCAD 378, +$502.84; AUDNZD 256, +$358.51.
Lot sizes0.01 to 0.13 (our calculation).
ConcentrationThe best trade made 8.1% of net profit, the best 5 trades 24.9% and the best 10 37.4% (our calculation). The other trades made the remaining 62.6%.
SettingsThe signal does not state its settings, and the product pages do not say whether the grid is on by default.
View the signal →

How the EA Trades — What the History Shows

What the seller says:

Strategy"a counter-trend approach that identifies oversold and overbought market conditions using Bollinger Bands, then confirms entries with multiple filters", on AUDCAD, AUDNZD and NZDCAD, M15.
Grid settings (from the input list)"Enable Grid"; "Grid Management: Choose grid lot sizing method (LotsSum, Martingale, AllLotsSum, Fix)"; "Min Distance … between grid orders"; "Min Distance Multiplier: Progressive distance multiplier for each grid level".
Stop settingsA money-loss setting only ("Enable Monetary Loss", "Monetary Loss Amount", "Stop EA After Money Loss"); no per-trade stop-loss input is listed, and the page does not say whether the money-loss setting closes open trades.
Set-up"You can start by using Default Settings, no specific set files are required"; minimum leverage 1:20 and "minimum deposit 1000 USD".

What the history shows (all 1,081 trades, our calculation):

Adding to losing positions330 trades were opened on the same pair and in the same direction while an earlier trade was still open, more than 60 seconds after it. 329 of them were opened after the price had moved further against every trade already open. The one exception was an AUDNZD buy on 26 December 2024.
Bigger at each stepAll 330 were larger than every trade already open. 319 steps added 0.01 lot and 11 added 0.02. 191 groups had added trades; in 185 of them the first step was from 0.01 to 0.02, a doubling (the other 6 started at 0.02 and stepped to 0.04).
How deepUp to 13 trades open on one pair and direction, reaching 0.13 lots (AUDNZD, from 21 August 2026). At 13 deep, the open trades totalled 0.91 lots, 91 times the first trade. The record shows 13 levels; it does not show a limit.
No group closed at a lossGrouping trades into chains as described in "How we tested" gives 751 groups. None closed at a net loss: the worst-performing group still closed $0.35 in profit. Individual trades did lose (265 of them), but the rest of each group made up the loss. (The one trade that closed at exactly zero was in a three-trade group, so it does not count as a group closing flat.)
StopsThe exported history records no stop loss or take profit on any trade. The export does not show whether the optional money-loss switch was enabled.
Buys and sells together126 trades were opened while a trade in the opposite direction on the same pair was open.
Close order90 trades closed while an earlier trade on the same pair, in the same direction, was still open (not first-in, first-out).
Before and after the listings169 of the 330 added trades came before 12 January 2026, when both products were published, and 161 after.

That is a grid with rising sizes: what our methodology calls an averaging and martingale pattern.

Deposits

The history records every deposit (our calculation):

DateEntryTrades open at the time
2 Dec 2024+$100.00 (opening deposit)0
27 Jan 2026+$300.009
27 Jan 2026+$250.0010
27 Jan 2026+$200.0010
12 Oct 2026+$1,200.0014 (across the three pairs)

In total: $2,050 in, no withdrawals. All four top-ups, on two dates, were made while trades were open. The $1,200 on 12 October 2026 came while 13 of the 14 open trades belonged to the record's longest and deepest group, the AUDNZD buys opened from 21 August 2026; that group closed four days later, on 16 October 2026, with a net profit of $72.37. We cannot see the floating loss or margin position when each deposit was made, so we draw no conclusion about why they were made. MQL5 reports a maximum deposit load (margin as a share of the account) of 15.2%; that does not establish the position at the deposit times. Deposits matter here because they make MQL5's growth figure unsuitable as a measure of what the money paid in earned (see "Growth figure" above). The signal began with $100, a tenth of the "minimum deposit 1000 USD" the product page recommends.

Time Below a High, and How Long Groups Stayed Open

We publish the longest, because averages hide them; see our methodology.

Longest closed group55.8 days (AUDNZD, opened 21 August 2026, 13 trades deep).
Longest time below a previous high25 days, from 6 February to 3 March 2026, measured on closed profit with deposits left out (our calculation). A closed-profit measure misses open losses. The longest group stayed open 55.8 days; that is an upper bound on how long a loss in that group could have lasted, not a measured 55.8-day loss.
Trades open at our check$498.17 in loss. We cannot see when they were opened.
Against the 90-day ruleNo closed group stayed open for 90 days. The rule cannot be checked for the trades open now. The verdict rests on the grid.

What the Sales Page Says, and What the Record Shows

"Live from more then 1 Year … Growth 681%, Win Rate 77%, Profit Factor 2.93"These appear to be an earlier snapshot of the signal, which now shows 1,166.09% growth, 75.5% winning trades and a profit factor of 2.48. Not necessarily false; they appear out of date.
"Scalping Trading"The average hold across all 1,081 trades was 58.6 hours (about 2.4 days), and the longest group stayed open 55.8 days.
"FTMO compatible"Not yet tested. Prop-firm accounts limit daily and overall losses, counting open losses; a grid's open losses are what such limits measure. Our prop-firm test will be added here when it has run.
"minimum deposit 1000 USD"The live signal started with $100.
Grid settingsThe product pages list them openly. Whether the grid is on in the default settings is not stated on the pages.

Price and Licence Terms

Price$499 on the MQL5 Market, MT5 and MT4 alike ("Last copy at 499$ -> next price 599$"). Version 1.10, 10 activations. Both published 12 January 2026. A free demo is available on MQL5.
Offer on the product page"If you Buy this Expert Advisor you can write a feedback at market and get a second EA for Free" (see "Independent reviews" below).
Recommended setup (product page)M15; AUDCAD, AUDNZD and NZDCAD; an ECN broker; a VPS; minimum leverage 1:20; minimum deposit $1,000.
US brokers (FIFO, no hedging)No, as run. On the signal, 126 trades were opened while an opposite-direction trade on the same pair was open, and 90 trades closed while an earlier trade on the same pair, in the same direction, was still open, out of first-in, first-out order. The input list includes "Fifo Closing" and an option to stop buys and sells at the same time; we have not established how the EA trades with them set, and no record shows it.
Prices recorded1 November 2026.

Key Evaluation Criteria

  • ❌ No grid, averaging, martingale or recovery strategy: failed. 329 of 330 added trades were opened after the price had moved further against every trade already open, and all 330 were larger.
  • ⚠️ Stop losses: none recorded at the broker on any trade. The settings have an optional switch that stops the EA after a money loss; the page does not say it closes open trades, and we cannot see whether it was used.
  • ⚠️ Drawdown in line with the marketing: the product page gives no drawdown figure; MQL5 shows 51.10% relative drawdown by equity.
  • ⚠️ Edge: payoff: a 75.5% win rate with a payoff ratio of 0.81. The best 10 trades made 37.4% of net profit and the rest 62.6%.
  • ⚠️ Claims match the record: the sales page's figures appear out of date; "Scalping" is hard to reconcile with an average hold of 58.6 hours (about 2.4 days) across all 1,081 trades, and groups open up to 55.8 days; "FTMO compatible" not yet tested.
  • ⚠️ No wipeout or near-wipeout: none on closed balance; but MQL5 recorded a 51.10% relative drawdown by equity, and a deeper dip between its checks cannot be ruled out.
  • ⚠️ Real-money record: nearly 23 months and 1,081 trades. That qualifies for an Early Review but is short of the 24 months a Recommended verdict needs.
  • ⚠️ Third-party tracked record: MQL5 has tracked the signal only since it was registered on 30 May 2025; the first six months, from 2 December 2024, come from the account's history imported at registration. The seller links it. MQL5 shows the trades came from this account, not that Dark Nova placed them; the products were published on 12 January 2026, 13 months after the record began.
  • ⚠️ No loss held open beyond 90 days: none among closed groups (longest 55.8 days); not checkable for the trades open now.

Suitability & Risks

What's Genuinely Positive
  • ✅A real-money record of nearly 23 months, linked by the seller.
  • ✅A closed profit in every calendar month of the record (closed trades only; open losses are not counted).
  • ✅The grid settings are listed openly on the product pages.
  • ✅All three pairs made a profit; the best 10 trades made 37.4% of net profit.
Key Risks
  • ⚠️Martingale grid: losing positions are added to at bigger sizes. A steady trend against the open trades is the scenario that makes the losses grow. At 13 deep, the open trades totalled 0.91 lots; the record shows no limit on depth.
  • ⚠️No recorded stops: no stop loss at the broker on any trade (the settings include an optional switch that stops the EA after a money loss; the page does not say it closes open trades, and we cannot see whether it was used).
  • ⚠️Open losses: at our check, open trades were $498.17 in loss, and MQL5 shows a 51.10% relative drawdown by equity at some point in the record.
  • ⚠️Payoff: the average loss is larger than the average win; the high win rate carries the record.
  • ⚠️US traders: not usable as run (hedged positions; 90 trades closed before an earlier same-pair, same-direction trade).
  • ⚠️Prop firms: unproven until our test has run.

Important Disclosures

Verification History
1 November 2026 — Published as an Early Review (record of nearly 23 months), Not Recommended.
  • What we read: MQL5 Signal 2312398 and its full exported trade history; both product pages and their input lists; MQL5 buyer reviews; and a web search for other reviews.
  • Not yet done: our prop-firm test. It will be added here as a dated entry.
  • Our policy: we do not edit published figures to track current values. We add dated entries here instead.
What would change this verdict.
  • The strategy: the verdict rests on the grid. A real-money record of at least 24 months and 300 trades from a clearly documented configuration that shows no grid, averaging, martingale or recovery pattern would be assessed on its own merits.
  • The Early Review label: the record reaches 24 months on 2 December 2026. At our next quarterly re-verification after that date we will re-read the full history and remove the label. Time alone does not lift a Not Recommended for a grid.
Independent reviews
  • MQL5 buyer reviews (read 1 November 2026): 4 reviews on the MT5 page and none on the MT4 page; the written ones are positive ("Great EA trading bot, low risk and steady profits"; "I recommend an excellently performing EA"). One reviewer adds that, "as the author himself points out", it is "a counter-trend bot that may open additional positions to cover the previous trade".
  • Incentive: the product page offers a second EA free to buyers who leave a review. We cannot tell which reviews, if any, were left under that offer.
  • Not found: we found no Forex Peace Army listing or other independent review.
  • How we use reviews: individual reviews are users' accounts, not facts we have verified, and they play no part in the verdict.
Affiliate Disclosure: We have no commercial relationship with the seller or with MQL5, and no link on this page is an affiliate link. We found no affiliate programme for this EA. We did not contact the seller before publication; this assessment rests on material the seller and MQL5 publish.
Corrections. If you are the vendor and believe a figure here is wrong, contact us at [email protected] with the evidence; we will check it and record any change in the verification history above.
Risk warning. Trading foreign exchange and contracts for difference carries a high level of risk and can result in the loss of all invested capital. Automated trading does not remove that risk; it removes the delay between a decision and its execution. Past performance, including verified past performance, is not a reliable indicator of future results. Nothing in this review is financial advice, a personal recommendation, or an invitation to trade. We are an independent publisher, not a regulated financial adviser.

Frequently Asked Questions

Is Dark Nova EA a scam?
We have seen no evidence of fraud, and we do not call it one. The seller links a real-money signal that MQL5 tracks, and lists the grid settings openly. Our verdict is about the strategy's risk, not about honesty.
Why is Dark Nova EA rated Not Recommended?
Because the live record the seller presents for Dark Nova shows a martingale grid. On the seller's signal, 330 trades were each opened while another trade on the same pair, in the same direction, was still open: 329 after the price had moved further against every trade already open, and all 330 bigger. Our methodology treats any grid, averaging or martingale pattern in the trade history as an automatic Not Recommended.
Does Dark Nova use martingale?
By our definition, yes. Its added trades mostly grow by 0.01 lot at each step, with the first added trade a doubling in every group that added one, so size rises as a position moves against it. That is not the classic double-after-every-loss martingale, but it meets our definition. The product's settings include several grid lot methods, one of them named "Martingale"; we do not know which the signal uses.
Why is this an Early Review?
Because the record was nearly 23 months old when we read it. The account behind MQL5 Signal 2312398 made its first trade on 2 December 2024. An Early Review is for a real-money record of at least 12 but less than 24 months, with at least 300 trades; it can come out Inconclusive or Not Recommended, but not Recommended. The record reaches 24 months on 2 December 2026; at our next quarterly re-verification after that date we will re-read the full history and drop the label. Time alone does not lift a Not Recommended.
Is the Dark Nova live signal real?
Yes. MQL5 Signal 2312398 is a real-money account at Headway on MetaTrader 4, linked from both product pages. MQL5 registered it on 30 May 2025, so the earlier trades come from the account's own history. MQL5 shows the trades came from this account, not that Dark Nova placed them or which version or settings were used; both products were published on 12 January 2026, 13 months after the record began. The grid pattern appears both before and after that date.
Is the MT5 version covered?
Yes. Both are rated Not Recommended. We cover both listings because the seller links the same MT4 signal from both. The verdict assesses the record offered as evidence for them; MQL5 does not show which EA, version or settings placed its trades. We found no separate MT5 live record. Both input lists include the same grid options, but settings alone do not show whether the grid is on.
Is Dark Nova good for prop firm challenges?
Unproven. The product pages say "FTMO compatible", but a grid holds its losses open, and prop firms count open losses against their daily and overall limits. We have not yet run our own prop-firm test; the result will be added to this review.
Can I use Dark Nova with a US broker?
Not as run on the signal. It held buys and sells on the same pair at the same time (126 trades were opened while an opposite-direction trade on the same pair was open), which US no-hedging rules do not allow; and 90 trades closed while an earlier trade on the same pair, in the same direction, was still open, which breaks first-in, first-out order; how a particular US broker would treat those closes we have not established. The settings include a FIFO option; we have not established how it trades with it.
Review Our Complete Testing Methodology

Discover More EAs

For another grid EA we rate Not Recommended, see our Quantum King EA review; for why grids look safe until they are not, our guide to martingale, grid and scalping EAs explains the trade-off. Most of the EAs we review are not Recommended.

Back to Best Expert Advisors →
Evidence-led, independent reviews
The Robotic Trader Logo
Risk Warning & Affiliate Disclosure: Trading foreign exchange and CFDs on margin carries a high level of risk and may not be suitable for all investors. Past performance, including a verified live track record, is not a guarantee of future results. The Robotic Trader may receive compensation from partner brokers, VPS providers and prop firms through affiliate links, at no extra cost to you. This supports our research and does not constitute personal financial advice. Commercial relationships never affect our verdicts, and every one is disclosed on the page it affects. Always trade with money you can afford to lose.

© 2026 The Robotic Trader. All rights reserved.