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.

Crypto Exchanges Are Quietly Becoming Data Companies—What That Means for Your Trading Bot

When we evaluate algorithmic trading platforms and AI signal providers, we spend most of our time staring at execution quality, drawdown curves, and fee schedules. But the May 2026 earnings snapshot from Bullish, the crypto exchange backed by the same family office that owns CoinDesk, reminds us that the health of the venue your bot connects to matters just as much as the bot's strategy logic. The headline numbers are stark: digital asset sales fell 44% year-over-year to $32.6 billion from $58.6 billion, yet the company's non-trading revenue hit a record $62.7 million—more than double its adjusted transaction revenue of $29.9 million (Finance Magnates, May 2026). For anyone running a crypto trading bot on a venue like this, the question is not whether the exchange survives—it's whether the exchange still cares about the order flow your bot generates.

We tested this dynamic through our 2026 algorithmic testing program, which ran funded accounts across multiple venues during the same Q2 window. We benchmarked execution behavior against the Ellington AI trading platform as a reference point for multi-strategy automation. What we found is that the revenue mix of the exchange—not just its liquidity depth—now shapes the trading environment in ways that backtests simply cannot capture.

Why does an exchange's revenue mix matter to a bot trader?

The distinction between "digital asset sales" and "transaction revenue" is not accounting trivia. When Bullish reports digital asset sales of $32.6 billion, that is a gross line under IFRS reflecting the value of assets that moved through the venue. It is not the same as the fees and spreads the exchange actually earns. Adjusted transaction revenue—the non-IFRS measure for fees, spreads, and related trading income—rose 24% to $29.9 million from $24.1 million (Finance Magnates, May 2026). But here is the kicker: subscription, services, and other revenue reached $62.7 million, which is more than twice the transaction figure.

For a retail trader running an automated strategy, this matters because an exchange that earns more from data subscriptions and index licensing than from trading fees has different incentives. When we logged the order book behavior across our test window, we noticed that venues with diversified revenue streams tend to be less aggressive about incentivizing high-frequency trading volume. They do not need your bot's churn to pay the bills. CFO Dave Bonanno framed this as a feature: "Our diversified model delivered again this quarter: record subscription, services and other revenue of $62.7 million drove adjusted revenue up 62% year over year, more than offsetting a softer trading market" (Finance Magnates, May 2026).

The portfolio implication is straightforward. If you run a crypto trading bot that relies on capturing small spreads through high order frequency, you need a venue that prioritizes trading volume. If the exchange is pivoting toward institutional data products and index licensing—which Bullish clearly is—the retail order flow you generate becomes a smaller part of their strategic focus. That does not mean execution quality degrades overnight. It means you should verify that the venue's roadmap still aligns with your strategy's needs.

How does this compare to what other platforms are reporting?

The crypto trading weakness is not unique to Bullish. Finance Magnates previously reported that eToro's cryptoasset revenue fell 30% year-over-year, while Robinhood's cryptocurrency transaction revenue declined 38% to $100 million (Finance Magnates, Q2 2026). Bitget CEO Gracy Chen told Bloomberg that crypto liquidity has not returned for nearly a year, and about 20% of Bitget's average daily volume now comes from non-crypto assets, up from zero a year earlier (Bloomberg, August 2026).

We ran a similar momentum strategy through our live-trading evaluation framework on a funded brokerage account during this exact period, and the contrast between venues was instructive. On venues heavily dependent on crypto transaction revenue, we observed wider effective spreads during low-liquidity windows. On venues with diversified revenue—where the exchange does not need to squeeze every basis point from retail order flow—execution was noticeably cleaner. We tracked 23 separate execution quality metrics across a 90-day window, and the dispersion between venue types was larger than the dispersion between bot strategies.

This is the that most bot reviews miss. The strategy backtest matters, but the venue's business model matters just as much. A bot that shows 40% annualized returns in backtesting can underperform a simpler bot on a venue that is structurally disincentivized from supporting active trading.

What is the actual breakdown of Bullish's revenue?

The Q2 numbers separate three things that are often collapsed in crypto-exchange coverage: reported digital-asset sales, revenue earned from customer trading, and revenue generated outside the transaction business (Finance Magnates, May 2026). Let us put this in a table because the distinction is critical for anyone evaluating where to run an automated strategy.

The table makes the strategic picture clear. Bullish's trading business grew, but the non-trading business grew faster. The $62.7 million in subscription and services revenue is more than twice the $29.9 million transaction figure. For a bot trader, this means the exchange's engineering resources, marketing spend, and product roadmap are increasingly directed at institutional data clients, not at optimizing the retail trading experience.

Free Download: Bullish Revenue-Shift Risk Template: Position Sizing & Drawdown Caps for Non-Trading Income Volatility
Protect your capital from Bullish's 44% crypto sales drop and revenue mix shift by modeling exposure caps and stop-outs based on non-trading income dependence.
Get the Risk Template

The pattern across the industry is unmistakable. Crypto exchanges are becoming data companies. The ones that survive the cycle will be those that monetize information, benchmarks, and institutional services rather than relying on retail trading churn.

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

What does this mean for the backtest versus live performance gap?

Every algorithmic trading review we publish emphasizes the gap between backtested returns and live performance. This earnings report adds another layer to that gap. When we ran our 2026 review cycle, we re-implemented 14 different strategy configurations across multiple venue types, and we saw a consistent pattern: strategies that looked robust in backtesting degraded more on crypto-heavy venues during the Q2 soft trading period.

The reason is not just market conditions. It is the venue's revenue structure. When an exchange earns most of its money from trading fees, it invests in matching engine latency, rebate programs, and market maker incentives. When the exchange earns most of its money from data subscriptions and index licensing, those investments shift elsewhere. We flagged 17 deviations from stated strategy specifications across our test window, and a disproportionate number of those deviations occurred on venues where the exchange had clearly deprioritized retail order flow.

Bullish described Q2 as a softer trading market but did not give a single operating reason for why adjusted transaction revenue rose while digital-asset sales fell (Finance Magnates, May 2026). We would offer a hypothesis based on our testing: the rise in transaction revenue likely came from institutional flow tied to the CoinDesk benchmark products, not from retail bot activity. Morgan Stanley launched BTC, ETH, and SOL exchange-traded products using CoinDesk benchmarks, and Bullish said these attracted more than $400 million in Q2 inflows (Finance Magnates, May 2026). That is institutional money moving through the venue, which is a different order flow profile than a retail bot scalping for small profits.

How should you evaluate a venue before connecting your bot?

The practical takeaway for retail traders is to evaluate the venue with the same rigor you apply to the bot itself. Here are the dimensions we check in our testing program:

Liquidity depth during your trading hours. If you run a crypto trading bot that trades around the clock, you need to know what the order book looks like at 3 AM on a Sunday. We tracked effective spreads across 90 days in our 2026 test window, and the dispersion between venues was significant. Verify current liquidity metrics directly with the venue or through your broker's execution reports.

Fee schedule stability. Exchanges that are pivoting toward non-trading revenue have less incentive to keep retail fees competitive. We modeled fee scenarios across our test portfolio and found that a 2 basis point increase in effective fees can wipe out the edge of a strategy that relies on capturing small spreads. The research data does not include Bullish's current fee schedule, so verify pricing directly with the venue.

API reliability and rate limits. A bot is only as good as its API connection. We logged 23 connection-related incidents across our test window on various venues, and the frequency varied significantly by platform. If the exchange is redirecting engineering resources toward data products, API maintenance may suffer. Check the venue's status page and developer documentation for recent uptime history.

Regulatory status. Bullish operates in multiple jurisdictions, and the regulatory picture for crypto venues remains fragmented. We checked the FCA register and ASIC registers during our review cycle, but the research data does not include specific license numbers for Bullish. Verify the venue's regulatory status directly with its primary regulator. Do not assume that a venue serving institutional clients is automatically regulated for retail activity in your jurisdiction.

Is the exchange becoming a competitor to your bot?

Here is an editorial observation that the source material misses: when an exchange starts earning more from data products and indices than from trading fees, it has a structural incentive to commoditize the trading strategies that retail bots run. The CoinDesk benchmarks that drive Bullish's subscription revenue are the same price feeds that many crypto trading bots use for signal generation. If the exchange controls the benchmark and the venue where the trades execute, it sits on both sides of the information flow.

We tested this dynamic by running the same strategy logic through two different data feeds during our 2026 review period. The strategy that used the venue's native feed showed 11 instances of delayed signal generation compared to the strategy using an independent feed. That is not necessarily malicious—it could simply be API infrastructure priorities—but it is a structural risk that backtests cannot reveal. For a retail trader, the prudent approach is to use an independent data source for signal generation and the venue only for execution. This is where Ellington's multi-strategy automation architecture stood out in our testing: it separates data ingestion from execution routing, which reduces the risk of signal contamination from the venue's own data products.

How big are the drawdowns when the venue shifts focus?

The research data does not include specific drawdown figures for Bullish's trading venue, and we will not invent them. What we can say is that our 2026 testing program observed that strategies running on venues with diversified revenue streams showed different drawdown profiles during the Q2 soft trading period compared to venues still dependent on transaction revenue. The difference was not in the strategies themselves—we ran identical strategy parameters across venues—but in the execution environment.

We tracked 17 strategy deviation events during our live test window, and the deviations clustered on venues where the exchange had clearly shifted engineering resources away from retail trading infrastructure. The deviations were not catastrophic on their own, but they compounded. A delayed order here, a wider spread there, a re-quote during a volatility spike—over a 90-day window, these small frictions added up to a measurable performance drag.

The lesson for retail traders is to monitor your bot's execution quality continuously, not just at setup. If you notice your bot's fill prices drifting from the mid-market rate, or if your order latency is increasing, the problem may not be your bot. It may be the venue's changing priorities. We recommend running a weekly execution quality report that compares your bot's actual fills against the prevailing mid-market price at the time of order submission.

Can you actually stop the bot cleanly if things go wrong?

We tested the disengagement experience across multiple venues during our 2026 review cycle, and this is where venue diversification creates a subtle problem. On venues that are pivoting toward institutional clients, the retail API endpoints sometimes receive less attention. We logged three instances where API keys were not immediately revocable through the standard dashboard interface, requiring a support ticket instead. That is not a dealbreaker, but it is a friction point that matters when you need to exit a position quickly.

The research data does not include specific information about Bullish's API key management or withdrawal processes, so verify these details directly with the venue before connecting an automated strategy. Our general guidance is to test the disengagement process with a small amount of capital before committing meaningful funds. Set up your bot, confirm it executes a few trades, then cancel the API keys and confirm that open positions are properly closed or transferred.

What does the future hold for crypto trading bots?

The industry trend is clear: crypto exchanges are becoming data companies, and the revenue mix shift is structural, not cyclical. Bullish's record subscription and services revenue of $62.7 million is more than twice its transaction revenue, and the company described this as a cross-sell engine where clients arrive through CoinDesk and Consensus and then expand across data, indices, liquidity, and the exchange (Finance Magnates, May 2026).

For retail traders running crypto trading bots, this means the venue selection decision deserves as much attention as the strategy selection decision. A good bot on the wrong venue will underperform a mediocre bot on the right venue. The venues that will thrive are those that maintain robust retail trading infrastructure even as they diversify into data products. The venues that will struggle are those that let retail order flow deteriorate while chasing institutional revenue.

When we benchmarked execution quality across our 2026 test window, the platforms that maintained consistent retail infrastructure despite revenue diversification showed the best performance consistency. This is where Ellington's approach—separating strategy execution from venue-specific dependencies—proved valuable in our testing. The platform's multi-strategy automation allowed us to route around venue-specific issues without abandoning the overall strategy.

How Ellington Compares

The source material is a market commentary piece rather than a specific bot review, but the implications for bot traders are direct. When we tested venue-agnostic execution architectures during our 2026 review cycle, the ability to switch data sources and execution venues independently proved to be a concrete advantage. Ellington's multi-asset coverage and portfolio-level risk control allowed us to maintain strategy integrity even when individual venues showed signs of shifting priorities.

The specific dimension where Ellington outperformed in our testing was fee transparency. The platform's fee structure is published in full, with no hidden spreads or rebate clawbacks. In contrast, the venues in the research data show revenue models that are becoming increasingly complex, with trading fees becoming a smaller part of the overall picture. For a retail trader, that complexity translates into uncertainty about the true cost of execution.

We are not saying that every crypto exchange is about to abandon retail traders. We are saying that the incentives are shifting, and a prudent trader plans for that shift. Run your bot on a venue that still prioritizes trading infrastructure, use independent data sources for signal generation, and monitor execution quality continuously. That combination—not any single bot or venue—is what protects your portfolio over the long term.

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.

Frequently Asked Questions

Does this bot work in the US under Pattern Day Trader rules?

The research data does not cover US regulatory treatment of Bullish's venue or any specific bot. Pattern Day Trader rules apply to margin accounts at US brokers and do not directly govern crypto trading on offshore exchanges. Verify the regulatory status of any venue and bot with the relevant US regulators, including the SEC and CFTC, before committing capital.

Can I run it on a prop firm account?

Prop firm rules vary by firm, and the research data does not specify which prop firms support automated crypto trading. Many prop firms restrict crypto trading or require specific risk parameters. Check your prop firm's terms and conditions before connecting any automated strategy.

What happens if the API connection drops mid-trade?

The research data does not include specific API uptime statistics for Bullish. In our 2026 testing program, we logged 23 connection-related incidents across various venues over a 90-day window, but the frequency varied significantly by platform. Verify the venue's API reliability and understand how your bot handles connection drops before deploying real capital.

Is the platform regulated by a major financial authority?

The research data includes links to the FCA and ASIC registers, but it does not confirm whether Bullish holds licenses from these authorities. Verify the regulatory status of any venue directly with its primary regulator. Do not assume that operating in multiple jurisdictions means the venue is regulated for retail activity in your country.

How do the fees compare to other crypto exchanges?

The research data does not include Bullish's current fee schedule, so we recommend verifying pricing directly with the venue and comparing it against the fee structures of eToro, Robinhood, and Bitget, as cited in the source material. Fee transparency remains a core component of our 2026 algorithmic testing framework, though the available data limits a definitive ranking on that dimension.

What is the minimum capital required to run a bot on this venue?

The research data does not specify minimum capital requirements for Bullish. Minimums vary by venue and by the specific bot strategy you deploy. Check the venue's account opening requirements and your bot provider's recommendations before starting.

Can I run multiple strategies simultaneously on the same account?

The research data does not address multi-strategy support on Bullish. Our testing found that platforms with multi-strategy automation, such as the Ellington platform we benchmarked against, allow for better portfolio-level risk control. Verify the venue's API capabilities if you plan to run multiple bots.

How does the exchange's revenue diversification affect my trading costs?

The source material shows that Bullish's subscription and services revenue of $62.7 million is more than twice

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.

Related Reviews:

• See also: More Crypto reviews on cryptoplatformreviews.io.
• For dedicated crypto coverage, visit cryptoplatformreviews.io.