"Users No Longer Think About Their Financial Lives in Separate Boxes": Bybit and Alpaca CEOs on the Convergence of Crypto and Equities

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.

Bybit's expansion into U.S. equities through Alpaca is being filed by most outlets as a product story: crypto exchange adds stocks, users get more choice. We read it as an infrastructure story with direct consequences for anyone running automation across both asset classes. Bybit AI, the conversational layer that CEO Ben Zhou describes, sits inside the crypto trading bot and algorithmic trading platform sub-niche, even though it is not a strategy engine in the way a dedicated bot is. In our 2026 review cycle we benchmarked exchange-native AI layers against the Ellington AI trading platform, and the contrast is the most useful thing we can hand a retail reader: one product is built to help you find a service, the other is built to execute a strategy and keep risk limits intact while you are not watching.

The source interview is a ten question conversation between Zhou and Yoshi Yokokawa, co-founder and CEO of Alpaca, on why the line between a crypto exchange and a traditional brokerage is becoming less meaningful to end users (Finance Magnates, May 2026). Zhou's headline claim is the one in our title: users no longer think about their financial lives in separate boxes. They hold Bitcoin, use stablecoins for payments, trade traditional markets and invest in equities, often across different platforms. Adding stocks to Bybit, he argues, is a natural extension of that behaviour rather than a new business line.

We pulled the full 10-question transcript and mapped it against our platform checklist. The interview is rich on strategy and intent. It is almost entirely empty on the numbers we would need before routing a funded account through any of it. That gap is the review.

Why the convergence story matters to algorithmic traders

Convergence is usually sold as convenience. For a retail account running automation, it is a risk model problem. A single platform holding 24/7 crypto positions and U.S. equity positions that only trade during exchange hours has two very different liquidity profiles sitting inside one login.

Zhou acknowledges the mismatch directly. Crypto operates around the clock and is global by nature, while U.S. equities operate inside a regulated market structure with specific trading hours, settlement processes and corporate actions. His own summary is blunt: you cannot simply take a crypto product and apply the same model to stocks (Finance Magnates, May 2026). We agree, and we would push it further. The bot layer is where that mismatch does the most damage, because an automated strategy does not pause to read a disclosure.

In our 2026 algorithmic testing program we schedule the same strategy class across crypto and equity sleeves, and the friction we hit first was never the signal. It was the calendar. A 24/7 crypto sleeve and a fixed-session equity sleeve produce two different exposure profiles inside one equity curve, and the account holder experiences the sum of both. We flag that as a structural issue rather than a performance issue, and we have no live Bybit AI strategy results to report because the provider publishes none in the source material.

That distinction matters more than it sounds. When we reviewed the transcript against our 7-point automated-strategy control checklist, we found zero of the seven fields disclosed. No entry logic, no exit logic, no position sizing rule, no per-position risk cap, no drawdown limit, no execution throttle, no kill-switch behaviour. That is not a criticism of Bybit AI specifically. It is a description of what an AI assistant is, and is not.

What does Bybit AI actually do?

Zhou describes Bybit AI as an intelligence layer designed to make it easier for users to discover, navigate and interact with financial services through a conversational experience. The stated goal is that a user should not have to understand how the company is organised internally, or which business unit provides a particular service. They should be able to say what they want to achieve and have the platform point them at the appropriate product (Finance Magnates, May 2026).

That is a discovery and navigation layer. It is not a strategy engine, and the interview never claims otherwise. The problem is vocabulary. In retail usage, "AI trading" has collapsed two very different products into one phrase. A dedicated AI trading bot is defined by what it buys, when it exits, how much it risks per position, and what happens when the connection drops. Bybit AI, as described, helps you find the door. It does not walk through it for you, and it does not manage what happens after you do.

We would rather say that plainly than let a reader assume otherwise. If you are evaluating Bybit AI as an automation product, you are evaluating a front end. The execution and custody sit with a partner, and the strategy, if you bring one, is yours.

How does the order chain work, and who holds your assets?

This is the part of the interview with the most practical value, and it is stated cleanly. Orders are routed through Infra Capital Limited, described as Bybit's introducing broker regulated by the Financial Services Commission of Mauritius, and executed by Alpaca, which also holds the underlying securities in custody (Finance Magnates, May 2026).

Yokokawa frames Alpaca's role as pure infrastructure: execution, clearing, custody, middle office, back office, APIs and the operational complexity required to provide access to capital markets. His stated ambition is that the infrastructure becomes almost invisible while remaining fully transparent in how it is disclosed to customers.

For a retail trader, the practical question is not which company is cleverest. It is who is accountable at each step, and to which regulator. Three named entities sit between your USDC and a share of stock, and the source material gives us a role for each but not a licence number for any.

One more detail worth flagging for anyone planning to build automation on top of this. The only API mentioned anywhere in the source material belongs to Alpaca, as part of its infrastructure stack. Bybit AI is not described as exposing a strategy API, an order-management API or a webhook layer for third-party bots. If you intend to run a bot against this flow, you are integrating against the infrastructure provider, not the platform, and you should confirm that directly with both parties before you write a line of code.

Crypto and equities still run on different clocks

Zhou's most useful line in the whole interview is that the opportunity is to simplify the experience without simplifying away the rules. The rules, as described, differ in almost every operational dimension.

Free Download: Bybit × Alpaca Crypto-Equities Convergence Due-Diligence Checklist
A due-diligence checklist for algo traders evaluating Bybit and Alpaca's crypto-equities convergence, covering broker compatibility, regulatory status, fee transparency, and withdrawal flows.
Get the convergence checklist

The USDC funding line is the one retail readers will notice first, and it is also the one where the source material stops short. The interview confirms that eligible users can fund supported stock purchases with USDC. It does not state conversion fees, FX spreads, funding costs or any other charge attached to that path. We will not estimate them. Verify with the provider.

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 we could and could not verify

We keep a verification matrix for every platform we cover, and this one is unusually sparse. That is a finding, not a formatting problem.

We checked the FCA Register and the ASIC Connect register against the entities named in the interview. Neither search returned a firm-specific record in the material we reviewed, and we are not going to assert a licence number we cannot cite. That is not evidence of anything improper. It is evidence that the disclosure in the source article is directional rather than complete.

Does Bybit AI publish backtest or live performance data?

No. Nothing in the source material includes backtest results, live results, drawdown bands, win rates or risk-adjusted return figures. The interview is a product convergence discussion, not a performance report.

That silence is worth treating as information. Every automated strategy we have evaluated shows some gap between simulated and live results, and the usual drivers are well understood: slippage, latency, available liquidity at the moment of the fill, and the difference between a backtest's assumed fill price and what the order book actually offers. We are not going to put a percentage on that gap for Bybit's equities flow, because no such figure exists in the source material.

Here is the contrast that adds something. Yokokawa describes infrastructure that should be almost invisible while remaining fully transparent in how it is disclosed to customers (Finance Magnates, May 2026). Invisible execution and transparent disclosure are two different promises. Only the second one is verifiable by a retail trader, and only if the provider publishes. When we asked the same question of every platform in our 2026 review cycle, the pattern was consistent: exchange-native AI layers publish adoption metrics, dedicated bot vendors publish strategy metrics, and almost nobody publishes the gap between the two.

What are the real risks for a retail portfolio?

With no drawdown data available, we will describe the structural risks the source material does support rather than inventing numbers.

Corporate actions are the first. Zhou notes that stocks have dividends and stock splits which need to be handled appropriately. An automated strategy that ignores a split will misread its own position size, and an equity sleeve funded from stablecoins adds a conversion step to every dividend event. We logged corporate-action handling as unverified for this flow.

Regulatory layering is the second. The user-facing promise is one platform. Underneath it sit at least two regulatory regimes, one for the introducing broker and one for the execution and custody provider, plus whatever applies in the user's own jurisdiction.

The third risk is the one the convergence narrative tends to skip. When crypto and equities sit in one account, the platform's risk engine sees two books, but the retail trader experiences one portfolio. A crypto position that gaps over a weekend and an equity position that gaps on an earnings print are correlated in the only way that matters to the account holder: both land on the same equity curve. Neither the exchange's risk engine nor the broker's margin engine is described in the source material as producing a consolidated, portfolio-level drawdown view. That is the gap we would want closed before running automation across both sleeves inside a single login, and it is a gap that a dedicated platform with portfolio-level controls is structurally better positioned to close.

Is Bybit AI regulated, and by whom?

The regulatory picture here is layered, and the layers are not interchangeable. The source material states that Infra Capital Limited, Bybit's introducing broker for the U.S. equities service, is regulated by the Financial Services Commission of Mauritius, and that Alpaca executes the orders and holds the underlying securities in custody (Finance Magnates, May 2026). You can check the Financial Services Commission of Mauritius register directly, and for the U.S. side we would point readers at FINRA BrokerCheck and SEC EDGAR rather than repeat a status we cannot confirm.

What the article does not provide is a licence number, a licence status, active or suspended, or effective dates for any entity in the chain. We searched the FCA Register and the ASIC Connect register and neither returned a firm-specific record in the material we reviewed. Verify directly with the provider and with the primary regulator in your own jurisdiction before you size a position.

Zhou's own framing is the right one. As platforms expand across financial products, licensing, compliance, custody and local regulatory relationships become even more important, not less. The companies that succeed, in his words, will be the ones that combine breadth with strong regulatory and operational foundations.

Can you actually switch it off cleanly?

Disengagement is the least glamorous dimension of any automated system, and it is the one that costs real money when it fails. For a crypto-only bot, stopping means closing the position and revoking the API key. For a converged platform, stopping means closing two books on two calendars. Equity positions cannot be liquidated at 3 a.m. on a Sunday. Crypto positions can.

We test kill-switch behaviour explicitly in our 2026 framework, and we could not test it here. The source material does not describe a user-facing kill switch, an API revocation flow, a strategy-level stop, or what happens to an open equity position if the platform connection drops outside market hours. That is a straightforward question to ask the provider before funding anything, and we would ask it before we asked about fees.

How Ellington compares

On the dimensions that matter to a retail account running automation, the difference between a convergence layer and a dedicated platform is concrete. Bybit's described setup spreads execution and custody across three named entities and two regulatory regimes, with no consolidated drawdown view described and no strategy API disclosed. Ellington's portfolio-level risk controls keep execution, risk limits and reporting inside one account, with the same rule set applied across sleeves rather than per book. That is the dimension we would weigh first: not how many asset classes a platform can list, but whether one set of risk rules governs all of them.

Alpaca's utility framing is the right instinct at the infrastructure layer (Finance Magnates, May 2026). We would simply want the same utility-grade consolidation at the risk layer before handing it a funded account.

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.

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

What is Bybit AI and is it a trading bot?

Bybit AI is described in the source interview as a conversational intelligence layer that helps users discover, navigate and interact with financial services. It is not described as a strategy engine with entry, exit, sizing or risk rules, so it does not meet the definition we use for an AI trading bot.

Does Bybit AI work in the US under pattern day trader rules?

The source material does not address U.S. pattern day trader rules for this service. U.S. equity trading is described as operating inside a regulated market structure with its own trading hours and settlement processes, and the introducing broker is described as Mauritius-regulated. Confirm margin and day-trading requirements with the provider and with FINRA Rule 4210 before you trade.

Can I run Bybit AI on a prop firm account?

Nothing in the source material describes prop firm compatibility, funded-account rules or third-party account support. Prop firms typically restrict automated execution and cross-asset exposure, so verify directly with the provider and with any prop firm before you connect anything.

What happens if the API connection drops mid-trade?

The interview does not describe failover behaviour, order-state recovery or what happens to an open equity position if connectivity drops outside market hours. This is a question to put to the provider in writing, ideally before funding.

Who holds my shares if I buy stocks through Bybit?

The source material states that Alpaca executes the orders and holds the underlying securities in custody, while Infra Capital Limited acts as the introducing broker routing orders (Finance Magnates, May 2026). Confirm current custody arrangements with the provider, since these relationships can change.

What fees does Bybit charge for stock trading?

No commission, spread, FX or withdrawal fee figures appear in the source material. The interview confirms that eligible users can fund supported stock purchases with USDC but does not state the conversion cost. Verify with the provider.

Can I fund stock purchases with USDC?

Yes, as described. The interview states that eligible users can access supported U.S. stocks and ETFs using USDC and manage eligible crypto and equity positions within the Bybit ecosystem. The conversion mechanics and any associated cost are not disclosed.

Does Bybit AI publish backtest data or win rates?

No. No backtest results, live results, drawdown bands or win rates appear in the source material. If a provider does not publish strategy-level performance, treat that as a data point rather than a neutral omission.

How do I stop the automation cleanly?

The source material does not describe a kill switch, API revocation flow or strategy-level stop. Because crypto trades around the clock and U.S. equities do not, a single stop command will not behave symmetrically across both sleeves. Ask the provider how each book is closed and in what order.

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.

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.

More in this category: Trading Industry News.