EA Overview
| Martingale | Yes |
|---|---|
| Grid | Yes |
| Scalping | No |
| Timeframes | 15M |
Forward Test Analysis (as of November 28, 2025)
Live Trading on a Darwinex Real Account
Aot EA is being forward-tested as a signal called “Aot Main” on a 100,000 USD Darwinex-Live real account. The fact that the performance is disclosed on a real-money account rather than a demo adds some transparency and credibility to this EA. As of now, the forward period is about 21 weeks, with the balance growing from 100,000 USD to 168,332 USD, which corresponds to a total gain of roughly +69% – a very strong start.
Performance Is Excellent, but Period and Sample Size Are Still Limited
Looking at the stats, there have been 78 trades over 21 weeks, averaging around 11 trades per week. This is not ultra-high-frequency scalping, but more of a swing or medium-term day trading style. The win rate is 94.87% (74 wins, 4 losses), monthly growth is about 20%, and the annualized return is around 250%. On the numbers alone, the EA looks extremely impressive. However, there are only 78 trades over roughly five months, which is far too small a sample to determine whether this performance can be reproduced consistently over multiple years.
Risk Structure and Risk–Reward Revealed in the Statistics
On the risk side, the profit factor is 11.20 and the recovery factor is 12.21, which looks ideal if you only focus on closed trades. However, the average profit per trade is about 1,028 USD, while the average loss is about –1,685 USD, meaning that from a risk–reward perspective, losing trades are larger in size than winning trades. The best trade is +5,069 USD and the worst trade is –5,558 USD, nearly the same magnitude. The maximum deposit load is as high as 84.8%, suggesting that when positions are open, the EA is often using a large portion of the account margin. While this aggressive risk-taking boosts returns in the short term, it also means that when the market trends strongly in one direction, a single move can cause a large drawdown or even blow up the account. Although the maximum balance drawdown is only 4.02%, the maximum relative drawdown on equity reaches 11.23%, indicating double-digit swings in unrealized P/L, which should not be overlooked.
Actual Logic Appears to Be “Stop-Loss Avoidance + Semi-Martingale”
The developer states on the product page that “excessive grid and martingale are not used” and that “risk is strictly controlled.” However, based on my own backtests and chart plots of the trade history, the actual behavior is that of an EA that holds losing positions for a long time without cutting them and waits for the market to come back. Furthermore, when price moves against the position, the EA adds additional trades to improve the average entry price – a classic martingale-style lot control. If you only look at closed trades, it appears to have “few stop losses and a very high win rate,” but behind that is a structure where unrealized drawdown can expand significantly.
Avoid Being Misled by Short-Term Performance
In summary, Aot EA’s forward results are excellent at this point, and the fact that such performance is shown on a real-money account deserves recognition. However, that performance is supported by a combination of a “high win-rate, stop-loss-avoidance logic” and “aggressive use of margin that often deploys most of the account.” The risk–reward profile is far from conservative. As long as the market eventually mean-reverts, the equity curve will look beautifully smooth and upward-sloping. But if the market trends in one direction for an extended period, a single adverse move can cause a large drawdown or, in the worst case, a complete blow-up. There is a high probability that Aot EA is a “high-risk EA” whose risk profile is much greater than what the forward curve suggests. It should never be judged as a “safe winning EA” based only on its short-term forward performance; you must also consider the underlying logic and long-term backtest results discussed below.
Backtest Results Analysis
Conditions of the 20-Year Backtest Run by the Reviewer
On the product page, the developer provides backtests only for the period from 2020 to 2025, which risks overfitting to the “post-COVID” market regime. For this review, I therefore ran my own backtests in MT5 using around 20 years of data to examine long-term behavior. The test conditions were as follows:
- Period: January 1, 2005 – November 15, 2025 (about 20 years of data)
- Currency pairs: 16 pairs
- Lot size: Balance-based dynamic lot size (default settings)
- Other parameters: All default settings
- Spread: Variable spread
Aot EA is a multi-currency EA that controls 16 pairs from a single chart: EURUSD, GBPUSD, USDJPY, USDCHF, USDCAD, AUDUSD, NZDUSD, EURJPY, GBPJPY, EURGBP, EURAUD, GBPAUD, AUDCAD, AUDNZD, EURNZD, and GBPNZD. Thus, when you apply the EA to a single symbol for backtesting, it automatically runs all 16 pairs together. This requires significant time and computing resources for symbol synchronization and calculations. Even so, with this kind of EA, you simply cannot see the real risk from “short, partial tests.” For anyone considering purchase, I strongly recommend running long-term backtests on your own broker environment first.
Large Total Profit Over 20 Years, but Maximum Drawdown Around 90%
Over the full 20-year backtest, total net profit reached about +15.26 million USD from an initial deposit of 10,000 USD, and the equity curve ends up strongly upward-sloping. There were 4,084 trades, with a win rate of 87.17% (3,560 wins, 524 losses), a profit factor of 2.20, and a recovery factor of 11.18 – metrics that suggest a “high win-rate, positive-expectancy EA.”
However, the impression changes completely when you look at the risk metrics. The relative maximum drawdown on balance is 90.40%, and on equity it is 91.85%, meaning that at some point the account temporarily lost nearly 90% of its value. In addition, average profit is about 7,868 USD while average loss is about –24,062 USD, so each losing trade wipes out roughly three winning trades’ worth of gains. The largest profit is +210,773 USD versus the largest loss of –385,168 USD, again showing much larger losses than gains – a classic small-profit/large-loss risk–reward profile.
Looking at the entire equity curve, from 2005 to around 2019 the performance is mostly flat to mildly upward, and the curve only really “takes off” after 2020, forming a typical hockey-stick shape. This indicates that the EA is heavily optimized for recent market conditions (especially after 2020), while earlier periods experienced severe drawdowns and stagnation.
Details for 2005–2012: Small Profits, Occasional Large Losses
When we zoom in on the 2005–2012 period, the EA’s fundamental risk characteristics become much clearer. The balance drops sharply in the early years, then over a long stretch of time it repeatedly accumulates small profits steadily and then gives back a large chunk of them in occasional big drawdowns. This is exactly the behavior of a “small-profit, big-loss EA.”
Because the win rate is high, the balance line does creep upward during normal conditions. However, in trending markets or when volatility regimes shift significantly, unrealized losses build up in positions that are never cut, and a single event can wipe out most of the accumulated gains. In fact, maximum drawdown in this backtest exceeded 90%, so it is entirely reasonable to say the EA “just barely avoided a complete blow-up by luck.”
Gap Between the Developer’s 2020–2025 Backtest and Long-Term Reality
The backtests published by the developer on the product page cover only the most recent five years (2020–2025), where the equity curve appears as a very clean, smooth uptrend. These years have been highly favorable for Aot EA’s logic, so it naturally shows “high win rates with relatively small drawdowns and attractive returns.” However, when you overlay the 20-year backtest, it becomes obvious that in earlier periods the EA experienced large drawdowns and long stagnation, and it is certainly not an EA that “wins consistently in all market conditions over 20 years.”
In other words, it is dangerous to look only at the strong performance after 2020 and assume that “the EA has been stably rising for 20 years.” Long-term backtests are essential to understand the blow-up risk inherent in a small-profit, high win-rate strategy.
Why You Must Run Long-Term Tests on Your Own Environment Before Buying
Because Aot EA is a multi-currency system that avoids cutting losses and aims for a high win rate, its short-term performance can look extremely attractive, but over the long run it carries a non-trivial risk of large drawdowns and account blow-ups. The default configuration automatically runs all 16 pairs from a single chart, so backtesting requires time and PC resources. Even so, simply trusting the developer’s selective backtests over favorable periods is not recommended.
As this review shows, running a 20-year backtest reveals a small-profit/large-loss profile, a tendency toward “small gains, occasional crashes,” and drawdowns approaching 90%. Anyone considering buying or using Aot EA should, without fail, run sufficiently long backtests and forward tests on their own broker and account type before making a decision.
Trading Logic and Risk Profile
Official Logic as Described by the Developer
According to the developer, Aot EA is a “next-generation multi-currency EA that combines AI sentiment analysis with adaptive optimization.” It is said to obtain market sentiment via the Claude API and use that information as a filter, entering only when the statistical edge is strong. The product description also states that “excessive grid and martingale are not used” and that “Smart Loss Reduction (SLR) strictly controls drawdowns,” which, on the surface, makes it sound like a relatively safe, high win-rate EA combining AI-based entry precision with robust risk management. In addition, the EA is designed to control 16 currency pairs from a single chart, and its multi-currency structure is promoted as a way to diversify and enhance overall portfolio performance.
Actual Behavior Observed in Forward Trade History
In this review, I plotted Aot EA’s actual forward trade history on MT5 charts and examined the entry and exit behavior in detail. White arrows mark buy entries and red arrows mark sell entries. The real behavior is quite different from what the official description suggests.
For example, on AUDCAD, after opening a sell position near a high, the market moves upward against the trade, but the EA does not stop out and instead holds the floating loss for an extended period. During that time, it opens additional sell positions, averaging up to pull the overall entry price closer to the current market price – a classic grid + martingale-style lot control. Once the market finally reverses and returns toward the initial level, the EA closes the basket of positions at an overall profit.
A similar pattern appears in other pairs: the EA opens buy positions during a sharp drop, then holds them even as price continues to fall, accumulating deep floating losses until the market eventually recovers and the trades are closed. Again, we see additional buy entries used to average down, implying a typical grid/martingale behavior of “adding at intervals and waiting for the market to come back.” The closed trade history shows very few stop losses and a high win rate, but the charts make it clear that the underlying design is to avoid realizing losses and endure drawdown until price retraces.
The Risk Hidden Behind the High Win Rate
This type of logic performs well in ranging markets or when price frequently mean-reverts. It can hold floating losses while ultimately closing in profit on pullbacks, which is why the forward and partial backtests for Aot EA display such smooth, upward-sloping equity curves. However, from a risk–reward perspective, a single losing sequence can wipe out the gains from many small winners – a textbook small-profit, large-loss pattern. In the 20-year backtest, the maximum drawdown reached nearly 90% of the account balance, so it is reasonable to say the EA “has just been lucky not to completely blow up so far.”
Moreover, the multi-currency design, with 16 pairs traded simultaneously, may appear to lower risk through diversification, but in practice it can work in the opposite direction. When correlated pairs are held at the same time, a strong directional trend can cause floating losses to expand concurrently across many pairs, leading to a very rapid and deep portfolio-level drawdown. For an EA like Aot that tries to avoid cutting losses and keeps positions open, multi-currency trading can turn into “multiple simultaneous blow-up risks” rather than diversification.
Role and Limitations of the AI Filter
Aot EA is marketed on its AI sentiment analysis, using the Claude API to read news and market sentiment and filter entries. This can indeed help reduce unnecessary entries. However, judging from the actual trade history, the AI filter seems limited to entry timing, while the core risk management (stop loss, lot sizing, position exit rules) still relies on a grid + martingale structure. No matter how sophisticated the AI, as long as the fundamental design is “do not cut losses, just average down and wait for a retracement,” the risk of blowing up in a strong trend remains.
Overall, while Aot EA’s trading logic looks attractive on the surface as a “next-generation EA combining AI and multi-currency trading,” its real behavior is that of a typical high win-rate, high-risk EA that combines no stop loss, grid, and martingale. The short-term forward performance and beautiful backtest curves are certainly appealing, but you must fully understand the underlying risk structure and calmly consider “how far the account might fall when it inevitably gets caught in a large trend.”
Overall Evaluation and Conclusion
At first glance, Aot EA looks like a “next-generation multi-currency EA powered by AI” based on its forward performance and the developer’s description. However, when you combine the 20-year backtest with the actual trade history, it becomes clear that its true nature is a high win-rate, high-risk EA that rarely cuts losses and waits for retracements using grid and martingale. The key points of this review are:
- Current forward performance is excellent: The 100,000 USD Darwinex-Live real account has achieved around +69% over about five months, making short-term performance very attractive.
- But period and sample size are insufficient: Forward results cover only 21 weeks and 78 trades, clearly not enough data to judge multi-year reproducibility.
- 20-year backtest shows maximum drawdown around 90%: Total profit is huge, but there are periods where nearly 90% of the account balance is lost, a result that can also be read as “barely avoiding a blow-up.”
- Classic small-profit, big-loss behavior: The EA has a high win rate but poor risk–reward, with a single loss capable of erasing multiple winning trades’ profits.
- No stop loss, grid, and martingale in live trading: Chart plots of forward history clearly show that when trades move against the position, the EA avoids cutting losses, averages down, and exits only on a retracement.
- AI filter likely only adjusts entries: While sentiment analysis may trim entries, it does not change the fundamental risk structure of “averaging down instead of cutting losses.”
- Multi-currency design is a double-edged sword: Simultaneously running 16 pairs can cause floating losses to balloon across the entire portfolio when correlations align, accelerating drawdowns instead of diversifying them.
Overall, Aot EA is an EA that can easily attract attention with “short-term forward performance” and the “AI” buzzword, but in reality it is a high-risk strategy that trades high win rate for the possibility of a single catastrophic loss. If you are considering using it, do not decide based on the equity curve alone. Carefully compare it with your own risk tolerance and investment objectives, and only proceed after thorough testing and sober analysis.
In actual behavior, Aot EA uses a high win-rate logic that combines grid and martingale with almost no stop loss. In the 20-year backtest, the maximum drawdown reached around 90%, meaning that once it gets caught in a strong trend, you can lose most of the account. It is essentially a “high-risk EA” for traders who can tolerate the possibility of a blow-up over the medium to long term.