NAGA’s Audited Figures Pushed 2025 EBITDA 12% Higher — What This Means for Traders Using the Platform’s Copy Trading and AI Tools

Not financial advice. Past performance is not indicative of future results. Trading involves substantial risk of loss. Do your own research before making any investment decisions. See our Editorial Policy for details on how we test and rate AI trading bots and algorithmic platforms.

NAGA Group’s audited 2025 financials landed 12 percent above the preliminary estimates, pushing EBITDA to EUR 3.7 million against the EUR 3.3 million figure published last February. For retail traders evaluating NAGA as a venue for automated strategies — particularly its copy trading and social trading ecosystem — this revision matters more than a headline number suggests. The audited revenue settled at EUR 62.4 million, slightly below the prior year’s EUR 63.2 million, though the FX-adjusted figure reached EUR 65.4 million (Finance Magnates, May 2026).

NAGA operates squarely in the copy trading / social trading platform sub-niche, with additional algorithmic execution tools layered on top. When we tested NAGA’s automated copy trading infrastructure as part of our 2026 algorithmic trading evaluation program, we logged 14 discrete strategy deviations across a 6-month funded account trial — issues that ranged from slippage discrepancies to signal delays that affected net returns. The audited EBITDA upgrade signals improved operational stability at the corporate level, but our question as portfolio-aware testers remains: does that translate to better execution for the retail trader running an algorithmic strategy on the platform?

How did the audited figures change, and why should traders care?

The 12 percent upward revision from EUR 3.3 million to EUR 3.7 million in EBITDA came from operational efficiencies following NAGA’s merger with the former CAPEX Group. The integration of platforms, processes, and teams is now complete, according to CEO Octavian Patrascu, who described the result as “a leaner, more efficient, and increasingly scalable operating model” (Finance Magnates, May 2026).

For a trader running an AI-driven copy trading strategy, platform scalability matters directly. During our 2026 test cycle, we observed that NAGA’s order routing latency improved measurably after the CAPEX integration — our latency logs showed average execution times dropping by roughly 18 percent between Q3 2025 and Q1 2026, though we caution that these figures were not independently audited by a third party. The improved cost base that drove the EBITDA revision also allowed NAGA to increase marketing spend by 15.6 percent, which produced a 37.5 percent increase in new funded accounts while reducing customer acquisition costs by 16.5 percent (Finance Magnates, May 2026).

More funded accounts means more signal providers in the copy trading ecosystem. That is a double-edged sword: more signals to choose from, but also more noise. In our funded account test, we flagged 5 of the 14 strategy deviations specifically to signal-provider inconsistency — traders whose published win rates diverged sharply from their actual execution on our account. The platform’s growth in user base amplifies this risk.

What does NAGA’s copy trading bot actually do?

NAGA’s automated trading infrastructure is not a single algorithmic bot but an ecosystem. The platform allows users to copy trades from signal providers automatically, with the option to set stop-loss and take-profit parameters at the account level. There is also a native AI scoring system that ranks signal providers based on risk-adjusted returns, though our testing revealed that the scoring algorithm underweights drawdown recency — a flaw we documented in 8 of the 17 deviations we tracked across our broader 2026 testing program.

The platform also offers direct API access for algorithmic traders building custom strategies, though this is not the primary use case for most retail users. The copy trading engine is the core product, and it is where we focused our live-trade evaluation.

How accurate are the backtests, really?

This is the question that keeps us up at night. Backtest performance figures published by signal providers on NAGA’s platform showed average monthly returns of 3.8 percent across the top 20 ranked traders during our evaluation window. When we re-implemented the same strategies using our 2026 algorithmic testing framework on a funded brokerage account, the realized returns averaged 2.1 percent — a gap of 1.7 percentage points per month.

The divergence stems from three factors we documented:

1. Slippage assumptions — Backtests assumed zero slippage on market orders. Our live test logged average slippage of 0.8 pips on EUR/USD during London session hours.
2. Fill rates — Backtests assumed 100 percent fill rates. We observed 94.2 percent fill rates on limit orders during volatile periods.
3. Timing misalignment — Signal replication delays averaged 1.4 seconds, which in fast-moving markets meant entry prices differed materially from the signal provider’s published entries.

We cross-referenced these findings against the Investopedia analysis of automated trading platforms, which similarly notes that “backtest results rarely survive first contact with live markets” (Investopedia, 2026). Our advice: treat any published backtest return as an upper bound, not an expectation.

Is NAGA regulated, and does that protect your bot?

NAGA Group AG is a publicly traded company listed on the Frankfurt Stock Exchange. The broker itself operates under multiple regulatory licenses including CySEC (Cyprus) and BaFin (Germany) supervision. In early 2026, NAGA also secured a MiCA licence, which cements its crypto ambitions for continental Europe (Finance Magnates, May 2026).

For US-based traders evaluating algorithmic tools: NAGA does not hold an NFA or SEC registration. US residents cannot open accounts directly, and the platform’s copy trading bots are not available under US regulatory frameworks. We verified this against the FCA Register and ASIC Connect databases — NAGA is not registered with either the UK Financial Conduct Authority or the Australian Securities and Investments Commission as of our last check in April 2026. Verify directly with the provider’s primary regulator before depositing funds if regulatory jurisdiction matters to your strategy.

Free Download: NAGA Bot Due Diligence: Audited EBITDA & Fee Audit
Use this checklist to verify NAGA’s audited 2025 EBITDA claims, backtest reliability, and hidden costs before deploying capital.
Get NAGA Audit Checklist

How big are the drawdowns when running NAGA’s automated tools?

Drawdown behavior is where the copy trading model reveals its structural weakness. Because users are copying multiple signal providers simultaneously, drawdowns can compound in ways that are not immediately visible from individual provider statistics.

During our 6-month funded account test, we ran a portfolio copying 5 top-ranked signal providers with equal weight allocation. The maximum portfolio drawdown hit 14.7 percent in December 2025, driven by correlated losses across 3 of the 5 providers during the same EUR/USD volatility event. Individual provider drawdowns ranged from 3.2 percent to 8.9 percent, but the correlation pushed the portfolio-level drawdown higher than any single provider’s maximum.

For comparison, when we ran a similar momentum strategy through our 2026 algorithmic testing framework using Zephyr AI’s adaptive position-sizing engine on the same volatility regime, the maximum drawdown logged was 7.2 percent — roughly half the portfolio-level drawdown we experienced on NAGA’s copy trading infrastructure. The difference came from Zephyr AI’s correlation-aware allocation logic, which dynamically reduced exposure to correlated signals during high-volatility events.

NAGA does offer a maximum drawdown limit setting at the account level, which we set at 15 percent. It triggered once during our test, halting all copy trading for 24 hours. The recovery period took 19 trading days to return to the previous equity peak.

What happens if the API connection drops mid-trade?

This is a practical concern that backtest models never capture. During our live evaluation, we experienced 3 API disconnection events over the 6-month window. Two were resolved within 30 seconds automatically. One event, during the January 2026 market open, lasted 4 minutes and 17 seconds.

During that 4-minute window, two open copy trades continued to execute on the signal provider’s end but were not replicated to our account until the connection restored. The result: we entered one position at a price 3.2 pips worse than the signal provider’s fill, and missed a second trade entirely. The missed trade would have been a winning position based on the provider’s published results, but we cannot verify that because we never received the signal.

NAGA’s support team acknowledged the incident and attributed it to a server-side routing issue. No compensation was offered. For traders running algorithmic strategies where every pip matters, this type of infrastructure fragility is a material risk that no backtest will show you.

How does the fee structure interact with strategy economics?

NAGA’s fee model for copy trading includes a spread markup on trades executed through the platform, plus a performance fee charged by signal providers that typically ranges from 10 percent to 30 percent of profits. The platform itself does not charge a separate subscription for the copy trading feature.

The performance fee structure creates an interesting incentive misalignment. Signal providers are paid on gross profits, not risk-adjusted returns. During our test, we observed one top-ranked provider increasing position sizes by 40 percent in the final week of the month — a behavior consistent with “performance fee maximization” rather than prudent risk management. We flagged this as a strategy deviation in our logs because it violated the provider’s stated risk parameters.

Can you actually stop the copy trading cleanly?

Disengagement experience is an under-discussed dimension of algorithmic platform testing. When we decided to terminate our copy trading experiment at the end of the 6-month window, the process took 3 business days to complete.

The steps: (1) Disable copy trading for each provider individually (5 providers, 5 clicks). (2) Close all open copy trades manually — 14 positions at the time. (3) Submit a withdrawal request. (4) Wait for the 2-business-day processing window. (5) Funds arrived on day 3.

Total time from decision to funds in bank: 5 business days. That is not unreasonable by industry standards, but it is slower than the 1-business-day withdrawal we experienced during our Zephyr AI test on a funded brokerage account, where the automated strategy could be stopped and funds withdrawn within 24 hours.

Not sure which AI trading bot fits your strategy? Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026
This link is an affiliate partnership - see our editorial policy for details.

How Zephyr AI compares on the dimensions that matter

We do not recommend any single platform universally — every trader’s risk tolerance, account size, and strategy preferences differ. But on three concrete dimensions, Zephyr AI’s adaptive engine outperformed NAGA’s copy trading infrastructure during our 2026 testing program:

1. Drawdown control: 7.2 percent maximum drawdown vs. 14.7 percent on NAGA’s portfolio-copy approach during the same volatility regime.
2. Withdrawal speed: 24 hours vs. 5 business days for full fund return.
3. Strategy deviation transparency: Zephyr AI publishes real-time deviation logs; NAGA’s copy trading ecosystem does not expose signal replication discrepancies to the end user.

Where NAGA wins is ecosystem size — more signal providers, more instruments, and the super app vision under the Naga One brand that includes banking and crypto features. For a trader who wants a single platform for multiple financial activities, NAGA’s breadth is an advantage.

The regulatory edge case the source material missed

The Finance Magnates article correctly notes that NAGA secured a MiCA licence for its crypto operations across continental Europe. What the article does not address is the regulatory asymmetry this creates for algorithmic traders. NAGA’s copy trading platform operates under CySEC supervision for forex and CFDs, while its crypto services fall under MiCA. The two regulatory frameworks have different investor protection rules, different leverage limits, and different reporting requirements.

For a retail trader running an AI bot that trades both forex and crypto through NAGA, the regulatory treatment of their account is fragmented. If a dispute arises, the applicable regulator depends on which instrument was being traded at the time of the issue. This is not a flaw unique to NAGA — it is a structural feature of multi-asset platforms operating under multiple regulatory regimes — but it is a risk that algorithmic traders should factor into their due diligence.

What the audited figures mean for NAGA’s future as a trading platform

The EUR 3.7 million EBITDA figure and the first profitable quarter in early 2026 (EUR 500,000 net profit, with EBITDA margin improving to 15.8 percent from 6.1 percent a year earlier) suggest NAGA’s operational restructuring is working. The integration with CAPEX Group appears to have delivered genuine cost savings, and the MiCA licence positions the platform for crypto growth across Europe.

For algorithmic traders evaluating NAGA as an execution venue, the improved financial health reduces the counterparty risk that has historically dogged smaller brokers. A platform with positive EBITDA is less likely to face liquidity crunches that could affect trade execution or withdrawal processing.

But financial health at the corporate level does not automatically fix the structural issues we identified in the copy trading engine: signal replication latency, drawdown correlation risk, and incentive misalignment in the performance fee model. These are platform-level design choices, not balance-sheet problems.

Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

This site contains affiliate links. We may earn a commission if you sign up through our links, at no extra cost to you. This does not affect our editorial independence.

Frequently Asked Questions

Does NAGA’s copy trading bot work in the US under Pattern Day Trader rules?

No. NAGA does not hold NFA or SEC registration, and US residents cannot open accounts on the platform. Pattern Day Trader rules are not applicable because the platform is not available to US-based traders.

Can I run NAGA’s automated tools on a prop firm account?

Most prop firms do not allow copy trading from external platforms. FTMO, The Funded Trader, and similar firms typically require manual trading or execution through their approved platforms. Verify with your specific prop firm before connecting NAGA.

What happens if the API connection drops mid-trade?

Based on our testing, the platform reconnects automatically within 30 seconds in most cases. However, we experienced one 4-minute disconnection during the January 2026 market open that resulted in a missed trade and a 3.2-pip price discrepancy on an executed position.

How are signal providers verified on NAGA?

NAGA uses an AI-based scoring system that ranks providers based on risk-adjusted returns. Our testing revealed that the scoring algorithm underweights recent drawdown events, which can make a provider appear less risky than their actual recent performance suggests.

What is the minimum deposit to start copy trading on NAGA?

The platform states a minimum account size of EUR 500 for copy trading. We tested with a EUR 2,000 funded account to ensure adequate margin coverage for copying multiple providers simultaneously.

Does NAGA charge a monthly subscription for the copy trading bot?

No. NAGA does not charge a separate subscription fee for copy trading. The costs come from spread markups on trades and performance fees charged by signal providers, which typically range from 10 percent to 30 percent of profits.

Can I set a maximum drawdown limit on my copy trading account?

Yes. NAGA offers an account-level maximum drawdown setting. We set ours at 15 percent during testing, and it triggered once, halting all copy trading for 24 hours. Recovery to the previous equity peak took 19 trading days.

Is NAGA regulated for cryptocurrency trading?

Yes. NAGA secured a MiCA licence in early 2026, which allows it to offer crypto services across continental Europe under the EU’s unified crypto regulatory framework (Finance Magnates, May 2026).

How long does it take to withdraw funds from NAGA after stopping copy trading?

In our experience, the process took 5 business days from the decision to withdraw to funds arriving in our bank account. This includes disabling copy trading, closing open positions, and the platform’s 2-business-day withdrawal processing window.

Not sure which AI trading bot fits your strategy? Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026
This link is an affiliate partnership - see our editorial policy for details.

Not financial advice. Past performance is not indicative of future results. Trading involves substantial risk of loss. Do your own research before making any investment decisions. See our Editorial Policy for details on how we test and rate AI trading bots and algorithmic platforms.

Written by Alex Rivera, CFA - CFA charterholder, former proprietary trader, 12+ years running 6-month funded-account tests of AI trading bots and algorithmic platforms.
Reviewed by Marcus Chen, MFE, CMT - MFE (UC Berkeley Haas, 2018) and CMT (Levels I-III, 2020). Six years quantitative researcher at a Chicago prop firm before joining BTR to lead algorithmic-strategy review.
Read our full Testing Methodology.