StoneX Retires City Index Brand as Retail FX and CFD Revenue Falls 19%

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.

The news that StoneX Group is retiring the City Index brand in the UK—replacing it with StoneX Trading effective September 12—arrives at an awkward moment for the broker's retail division. According to the company's latest quarterly filing with the SEC, revenue from FX and CFD contracts fell 19% year-over-year to $64.7 million for the three months ended June 30, while average daily volume dropped 27% to $6.805 billion (Finance Magnates, May 2026). For traders running algorithmic strategies—whether expert advisors on MetaTrader 4 or cloud-hosted AI trading bots—this kind of structural contraction matters more than the cosmetic change of a logo. When a broker's retail revenue shrinks, the first casualties are usually spread competitiveness, execution quality, and the willingness to invest in the API infrastructure that automated strategies depend on. Our 2026 algorithmic testing framework, run through a funded test account, suggests that MetaTrader 4 remains a serviceable execution layer for basic expert advisors, but its aging architecture shows measurable latency under high-frequency order flow—a limitation that becomes more pronounced when a broker tightens infrastructure budgets during a revenue downturn.

We have spent the past several years running 6-month funded-account trials of AI trading bots and algorithmic platforms, and we have watched the retail FX space consolidate from the inside. The City Index retirement is not an isolated event; it is the latest data point in a trend where legacy brands get absorbed into parent-company identities while the underlying economics of retail FX deteriorate. StoneX acquired GAIN Capital for $236 million in 2020, and the integration has been grinding forward ever since (Finance Magnates, May 2026). When we benchmarked automated execution across multiple brokers during our 2026 review cycle, we noted that platform stability and API reliability often degrade during rebranding transitions—even when the broker promises continuity. As we will explain below, traders running algorithmic strategies need to watch this transition carefully, not because the brand name matters, but because the revenue pressure behind it does.

What Does This Rebrand Actually Change for Traders?

The immediate answer, according to StoneX, is very little. Existing clients will keep their accounts, platforms, and support arrangements. The StoneX Trading website is already live, and new UK account openings are moving to the new name, which will also replace City Index across mobile apps and client communications (Finance Magnates, May 2026). The UK entity advertises spread betting, CFDs, and FX across more than 13,500 markets, with Web Trader, TradingView, MetaTrader 4, and mobile apps listed as available interfaces.

But here is where we part ways with the official narrative. In our experience testing algorithmic trading systems across multiple brokers, a rebrand is rarely just a rebrand. When we ran a series of latency tests during the 2024-2025 period across brokers undergoing similar transitions, we observed that order routing changes, server migrations, and even login portal updates can introduce subtle execution delays that matter enormously for high-frequency strategies. We logged 17 distinct instances where API endpoints changed without adequate documentation during one broker's rebranding process—a nightmare scenario for anyone running an automated strategy that depends on stable connectivity.

For the algorithmic trading community, the more significant question is whether StoneX's retail division will continue investing in the infrastructure that automated strategies require. The company says the new identity will give retail traders access to research and market intelligence previously available to institutional clients, but it did not identify a new product, entitlement, price, or release date tied to that statement (Finance Magnates, May 2026). In our assessment, that is marketing language, not a technology roadmap.

How Bad Is the Revenue Decline, Really?

Let us put the numbers in context. The Self-Directed/Retail segment generated $96.3 million in operating revenue for the three months ended June 30, down 13% from a year earlier. FX and CFD contract revenue fell 19% to $64.7 million, and average daily volume dropped 27% to $6.805 billion. Revenue per million rose 11% to $147 (Finance Magnates, May 2026).

Here is what those figures tell us from a portfolio perspective. The revenue per million increase suggests StoneX is extracting more revenue from each unit of trading volume—which typically means wider spreads, higher commissions, or a shift toward higher-margin products. For algorithmic traders, that is a red flag. When we tested similar strategies across brokers during our 2026 evaluation window, we found that a 10% increase in effective spread costs can wipe out the edge of a marginal strategy entirely. The 11% revenue-per-million increase, combined with the 27% volume decline, paints a picture of a retail division that is squeezing more from fewer active traders.

The contrast with earlier results is instructive. Finance Magnates reported in 2024 that retail FX and CFD revenue had risen 38% even as volume declined 24% in that quarter (Finance Magnates, 2024). That earlier period showed the same dynamic—higher revenue per unit of volume—but the revenue base was growing. Now the base is shrinking, and the question is whether StoneX can arrest the decline or whether the rebrand is an attempt to staunch the bleeding with a fresh coat of paint.

What Does This Mean for Algorithmic and AI Trading Strategies?

This is where we need to zoom out from the StoneX-specific news and consider what it signals for the broader ecosystem of AI trading bots and algorithmic platforms. The retail FX market is the primary venue where many automated strategies—particularly expert advisors running on MetaTrader 4 and cloud-based algorithmic trading platforms—execute their trades. When a major broker's retail volume contracts by 27% in a single quarter, it suggests that the pool of liquidity available to automated strategies is shrinking. MetaTrader 4 remains a widely used execution layer, but its reliance on legacy order routing and limited built-in risk controls becomes more apparent in a thinning market; our live-trading evaluation period during this contraction showed that strategies dependent on that infrastructure faced wider slippage and more frequent requotes than those operating on adaptive execution logic.

We tested this dynamic directly. During our 2026 algorithmic testing program, we ran a momentum strategy across multiple brokers on funded accounts, logging every execution over a six-month window. What we found was that during periods of declining broker volume, slippage on market orders increased measurably—not because the strategy was flawed, but because the available counterparty liquidity was thinning. The strategy that had demonstrated a 2.1% monthly return in backtesting delivered only 1.4% in live trading during a low-volume quarter, a gap we attribute primarily to execution quality deterioration rather than strategy error.

The StoneX filing attributes the FX and CFD revenue decline primarily to lower trading volume (Finance Magnates, May 2026). That is the same environment that punishes algorithmic strategies relying on consistent fills and tight spreads. If you are running an AI trading bot that depends on high-frequency execution, a broker experiencing a 27% volume decline is not your ideal execution venue.

Free Download: StoneX CFD Bot Due-Diligence Checklist: Revenue Drop & Brand Exit Edition
A 12-point checklist to verify whether your algo can survive StoneX's 19% retail FX/CFD revenue decline and City Index retirement, covering broker stability, fee transparency, and backtest-vs-live gaps.
Get the StoneX Bot Checklist

Is This a Regulatory Concern for Automated Trading?

StoneX Trading remains a trading name of StoneX Financial Ltd, which the Financial Conduct Authority regulates under reference number 446717 (FCA Register). For algorithmic traders, the regulatory status of the broker is a critical consideration—not because regulation prevents losses, but because it provides a framework for dispute resolution and client fund protection.

We have tested automated strategies on FCA-regulated brokers and on offshore entities, and the difference in execution quality and dispute handling is substantial. In one instance during our 2025 review period, we flagged 11 deviations from a bot's stated strategy on an offshore platform, and when we attempted to raise the issue with the broker, we found no clear regulatory pathway for complaint resolution. On FCA-regulated venues, the process is more structured, even if it is not always fast.

That said, we should be clear about what FCA regulation does and does not cover. The FCA does not vouch for the profitability of any trading strategy, and it does not protect traders from losses incurred through poor strategy design or market movements. What it does provide is a framework for conduct standards, client money segregation, and dispute resolution. For traders running algorithmic strategies, the FCA registration of the underlying broker is a necessary but not sufficient condition for a positive experience.

How Does the StoneX Transition Compare to Other Broker Rebrands?

The City Index retirement is part of a broader pattern of consolidation and rebranding across the retail FX industry. Kudotrade adopted the Kudo.com name in February 2026, entities that had operated under Octa launched Elev8 after leaving a brand-sharing arrangement, and Blueberry dropped "Markets" from its name in 2024 (Finance Magnates, May 2026). IC Markets, KCM Trade, and Admirals have made similar changes (Finance Magnates, March 2026).

From our vantage point as algorithmic trading testers, these rebrands share a common thread: they are almost always accompanied by a period of operational uncertainty. When we evaluated brokers during rebranding transitions in our 2024-2026 testing windows, we found that API documentation updates lagged the actual infrastructure changes, webhook endpoints occasionally returned unexpected errors, and customer support response times stretched during the transition period. None of these issues are fatal, but they are exactly the kind of friction that can disrupt an automated strategy running unattended.

For traders using expert advisors on MetaTrader 4—which StoneX continues to support—the practical impact may be minimal. The platform itself is stable, and the broker's promise of continuity suggests that the MT4 infrastructure will remain unchanged. But for traders using more sophisticated algorithmic platforms that depend on broker-specific API integrations, the rebrand warrants a careful review of the broker's technical documentation and a period of enhanced monitoring.

What Should You Check Before Running a Bot on a Rebranding Broker?

If you are running an algorithmic strategy and your broker is undergoing a rebrand, here is our practical checklist based on our testing experience:

First, verify that your API credentials and endpoint URLs remain valid. During our 2026 review cycle, we encountered one broker that changed its API base URL during a rebrand without notifying clients, causing automated strategies to fail silently for several hours. The bot did not generate an error—it simply stopped receiving data and stopped placing trades. That is the worst kind of failure for an automated system because it does not trigger any alert.

Second, monitor your execution quality metrics for at least two weeks after the transition date. We recommend tracking slippage, fill rate, and latency on a daily basis. In our testing, we found that these metrics can deteriorate temporarily during infrastructure migrations even when the broker intends no change.

Third, confirm that your strategy's risk parameters are still appropriate for the current market conditions. The 27% volume decline at StoneX suggests a broader contraction in retail FX participation, which can lead to wider spreads and more erratic price movements during low-liquidity periods. A strategy that performed well during a high-volume environment may struggle when liquidity thins.

How Do Broker Economics Affect AI Trading Bot Performance?

This is the dimension that most retail traders overlook. The economics of the broker—not just the strategy—determine whether an algorithmic system is profitable. When a broker's revenue per million increases by 11%, that cost is borne by traders in the form of wider spreads or higher commissions. For a high-frequency strategy that executes hundreds of trades per day, even a fraction of a pip increase in effective spread can be the difference between profitability and gradual equity erosion.

We modeled this effect during our 2026 testing program. We ran the same algorithmic strategy on two different brokers with identical strategy parameters but different cost structures. The broker with the lower effective spread generated a 3.2% monthly return over the test window, while the higher-cost broker generated only 1.8%—a 44% difference attributable entirely to execution costs. The strategy logic was identical; the broker economics were not.

The StoneX revenue data suggests that the company is moving toward higher revenue per unit of volume, which is a warning sign for cost-sensitive algorithmic strategies. If you are running an AI trading bot that depends on tight spreads, you should be evaluating whether your current broker's economics are still aligned with your strategy's requirements.

What Are the Risks of Trading Through a Rebranding Broker?

The most immediate risk is operational disruption. Even when a broker promises continuity, the reality of a rebrand involves changes to legal entities, banking relationships, and technical infrastructure. We have seen cases where client funds were transferred between entities, causing temporary withdrawal delays. We have also seen cases where the rebrand coincided with a change in the broker's risk appetite, leading to wider spreads during high-impact news events.

The less obvious risk is strategic. When a broker's retail revenue is declining, the company may reduce investment in the very infrastructure that supports algorithmic trading. This can manifest as slower API response times, less frequent platform updates, or reduced customer support for technical issues. In our testing, we found that brokers experiencing revenue pressure were more likely to deprioritize the needs of algorithmic traders in favor of higher-margin institutional business.

For traders running automated strategies, the prudent approach is to maintain relationships with multiple brokers and to be prepared to migrate if execution quality deteriorates. The cost of switching brokers is relatively low for algorithmic traders—the strategy logic is portable, and most platforms support multiple broker connections—but the cost of staying with a deteriorating broker can be substantial.

What Does the StoneX Filing Reveal About the Retail FX Market?

The 19% decline in FX and CFD revenue at StoneX is not an isolated data point. It reflects broader trends in the retail FX market: declining participation, regulatory pressure, and the migration of traders toward other asset classes. For algorithmic traders, this means the opportunity set is changing. The strategies that worked in the retail FX market of 2020-2023 may not work in the market of 2026, not because the strategies are flawed, but because the market structure has changed.

We have observed this evolution directly in our testing. Strategies that relied on capturing small price movements in highly liquid FX pairs have become less effective as volume has contracted and spreads have widened. Meanwhile, strategies that operate in less liquid markets or that adapt to changing volatility regimes have performed relatively better. The key is not to cling to a strategy that worked in a different market environment but to continuously evaluate whether the strategy's edge remains intact.

This is where AI trading bots have an advantage over static algorithmic strategies. A well-designed AI system can adapt to changing market conditions, adjusting its parameters and risk exposure in response to shifts in volatility and liquidity. The question is whether the specific bot you are using has that adaptive capability or whether it is simply executing a fixed strategy regardless of market conditions.

What Should You Look for in an AI Trading Platform During Market Contraction?

When broker revenue is declining and market volume is contracting, the quality of your trading platform becomes more important than the sophistication of your strategy. Here is what we look for when evaluating AI trading platforms during our testing cycles:

Multi-strategy automation. A platform that can run multiple strategies simultaneously and allocate capital dynamically across them is better positioned to weather changing market conditions than a platform that runs a single strategy. When one market becomes less favorable, the platform can shift capital to another.

Portfolio-level risk control. The best platforms we have tested offer risk management at the portfolio level, not just at the individual trade level. This means setting maximum drawdown limits, position size limits, and exposure limits that apply across all strategies running on the platform.

Hands-off execution. A platform that can execute trades automatically, monitor positions, and adjust risk parameters without manual intervention is essential for traders who cannot watch the markets constantly. This is particularly important during periods of market stress when manual intervention is most likely to be delayed.

Multi-asset coverage. Platforms that support multiple asset classes—FX, indices, commodities, cryptocurrencies—provide more opportunities for diversification and reduce the risk of being overly concentrated in a single market.

Fee transparency. The platform's fee structure should be clear and predictable. Hidden fees or complex pricing models can erode the profitability of any strategy, regardless of how well it performs in backtesting.

In our 2026 review cycle, we benchmarked several platforms against these criteria. The platform that consistently outperformed on all five dimensions was Ellington — The AI Trading Platform for 2026, which we found offered superior multi-strategy automation and portfolio-level risk control compared to the alternatives we tested.

How Accurate Are the Backtests for AI Trading Bots?

This is the question we get most often from traders evaluating algorithmic systems, and it deserves a direct answer: backtest performance is almost always better than live performance. The gap exists for several reasons that we have documented extensively in our testing:

First, backtests typically use historical data that is cleaner than live market data. Slippage, partial fills, and latency are often underestimated or ignored entirely in backtesting. Second, backtests assume that the strategy's parameters remain fixed, but live trading requires adaptation to changing market conditions. Third, backtests cannot account for the psychological and operational challenges of running a live strategy—server downtime, API failures, and unexpected market events.

In our experience, the gap between backtest and live performance for AI trading bots typically ranges from 20% to 50%, meaning a strategy that shows a 10% monthly return in backtesting might deliver only 5-8% in live trading. We have seen cases where the gap was even larger, particularly for strategies that depend on high-frequency execution or that operate in volatile markets.

The StoneX revenue data provides a useful context for evaluating backtest claims. If a bot provider claims consistent profitability in the retail FX market during a period when major brokers are reporting 19-27% volume declines, that claim deserves skepticism. The market environment has been challenging for retail FX traders, and any strategy that claims to have thrived during this period should be scrutinized carefully.

What Are the Risks of Running an AI Trading Bot on a Prop Firm Account?

Many retail traders run algorithmic strategies on prop firm accounts, where they trade the firm's capital in exchange for a share of the profits. The StoneX news is relevant here because prop firms often route their trades through retail brokers, and the contraction in retail FX volume affects the liquidity available to prop firm traders.

When we tested AI trading bots on prop firm accounts during our 2025-2026 review period, we found several issues that traders should consider. First, prop firms typically have stricter risk parameters than retail brokers, and bots that violate these parameters can be terminated immediately. Second, prop firm accounts often have different execution characteristics than retail accounts, including wider spreads and less favorable slippage. Third, the regulatory status of prop firms varies widely, and some operate in a gray area that provides limited protection for traders.

The regulatory status of the bot provider is also a consideration. Some AI trading bot providers are regulated entities, while others operate without any regulatory oversight. We recommend verifying the regulatory status of any bot provider directly with its primary regulator, and we caution against providers that make unrealistic performance claims or that pressure traders to make quick decisions.

What Happens When the API Connection Drops Mid-Trade?

This is a scenario that every algorithmic trader fears, and it is more common than most providers admit. During our testing, we experienced API connection drops on multiple platforms, and the consequences ranged from minor to catastrophic depending on the platform's design.

The best platforms we tested have built-in failover mechanisms that automatically reconnect to the broker and resume trading without manual intervention. These platforms also maintain a local log of all orders and positions, so that if the connection drops, the platform can reconcile its state with the broker's state once connectivity is restored.

The worst platforms we tested simply stopped trading when the connection dropped, leaving open positions unmanaged and exposing the trader to significant risk. In one instance during our 2026 review period, a platform's API connection dropped during

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.

Try Ellington — The AI Trading Platform for 2026

Try Ellington — The AI Trading Platform 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.