XTB Catches Up on Overnight US Stock Trading, Adds Hungarian Tax Account

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.

XTB sits in the algorithmic trading platform bracket, specifically the broker-integration layer that automated strategies actually plug into. That distinction matters more than most retail traders realise. A bot is only as good as the venue it routes orders through, and when a broker changes its session hours, its order-type handling, or its tax wrapper, it changes the economics of every automated strategy running on top of it. So when XTB announced it had started rolling out 24-hour weekday trading in selected US shares and opened a tax-advantaged long-term investment account in Hungary, our review desk treated it as an execution-layer event, not a marketing one (Finance Magnates).

We have benchmarked against Zephyr AI's adaptive engine in our 2026 review cycle, and the single most common failure mode we see in retail algo deployments is not a bad signal, it is a venue mismatch. The strategy assumes one fill model, the broker delivers another, and the equity curve quietly diverges from the backtest. XTB's overnight session is a textbook example of that gap, and the Hungarian TBSZ account is a textbook example of a wrapper that changes holding-period behaviour. Both deserve a careful look.

What XTB Actually Rolled Out

The Warsaw-listed broker said it had started rolling out 24-hour weekday trading in selected US shares, a feature several rivals have offered for two years or more. Extended hours were already on the 2026 product list XTB published with its annual results, which put US markets first and European exchanges later in the year (Finance Magnates). The US names join close to 1,000 European stocks and ETFs that XTB moved to longer hours earlier this year. It also lengthened trading in selected French and Spanish shares to 7:30 a.m. to 10 p.m. CET, in line with German stocks (Finance Magnates).

Clients get access in stages and are notified in the app, where the service can be switched on or off, XTB said. Its updated terms list 582 US stocks, fractional shares included, according to a summary published by Czech personal finance site Finhacker. That is the number our team anchored on, because coverage breadth is the first thing an automated strategy notices. A momentum or mean-reversion model that scans a fixed universe needs to know precisely how many symbols are tradable after hours, and 582 is a narrow slice of the US tape.

Here is the part that should concern anyone running an automated book. Outside regular hours, a market order is executed as a limit order at the best opposing price, and any unfilled part is cancelled, the terms state. XTB warned that liquidity, volatility and order execution can differ from the regular session. In plain English: your bot sends a market order, the venue converts it to a limit order, and whatever does not fill simply disappears. There is no queue, no resting order, no second attempt unless your logic is built to retry.

Why Overnight Sessions Matter for Algo Traders

Overnight access is not a convenience feature for a discretionary trader. For an algorithmic platform it is a structural input. Earnings releases, macro prints and overseas session moves all land outside the regular US cash session, and a strategy that cannot act on them is a strategy with a blind spot. When we mapped XTB's overnight coverage against the 582 US names in its updated terms, the immediate question was universe drift: does the tradable set after hours match the set the backtest assumed during the day?

That question is where most retail algo deployments quietly break. Our backtest harness assumes a stable symbol universe and a single order-fill model. XTB's overnight session introduces two variables at once, a narrower universe and a conditional fill model. Neither is fatal, but both must be coded for explicitly. A bot that assumes market orders always fill will systematically under-execute after hours, and the resulting position sizes will be smaller than the strategy intended. That is a silent drag, not a visible error.

We have not yet run a full six-month live trial on XTB's overnight session inside our 2026 algorithmic testing program, so we are not publishing drawdown or slippage figures here. What we can say is that the fill-certainty gap is the first thing we would instrument, because it corrupts position sizing in a way that is hard to see until you reconcile expected versus actual exposure line by line.

How the Hungarian TBSZ Account Works

The Hungarian product is the TBSZ, a five-year investment account. Gains become tax-free after five years and are taxed at a reduced rate in years four and five, according to XTB's website. Closing the account earlier means the standard 15% personal income tax plus a 13% social contribution. XTB's version opens from HUF 25,000, has no account fee and charges no commission on shares and ETFs up to EUR 100,000 of monthly volume.

The format is growing fast. TBSZ accounts rose by more than a third to over 519,000 at the end of 2025, from about 386,000 a year earlier, according to data from Hungary's central bank, the Magyar Nemzeti Bank, cited by XTB, with assets approaching HUF 7.9 trillion. That is a market moving from niche to mainstream, and it is exactly the kind of wrapper that changes how a systematic investor behaves.

For an algo trader, a five-year tax wrapper is a behavioural constraint disguised as a tax product. Strategies with high turnover, monthly rebalancing, or tactical de-risking do not fit a five-year lock-up cleanly. If your automated system wants to flatten exposure during a volatility spike, the TBSZ early-close penalty of 15% income tax plus 13% social contribution is a real cost of that decision. We cross-referenced the TBSZ terms against our live-trading evaluation framework and the conclusion was straightforward: this wrapper suits buy-and-hold and slow rebalancing, not active rotation.

Is XTB Regulated for Algo Trading?

XTB is a Warsaw-listed broker with a long-standing European footprint, and its UK, EU and other entity status should be verified directly with the provider primary regulator rather than taken from any marketing page. For UK clients, check the FCA Register for the specific legal entity you are onboarding with. For Australian clients, the ASIC Connect registers are the primary source. We do not assert a licence number we cannot cite against a register entry, and neither should any review you read.

The regulatory point that matters for algo traders is not the headline licence, it is which entity holds your account and what execution venue sits behind it. Regulatory status of the broker is separate from the regulatory status of any bot or signal provider you bolt onto it. XTB is the venue. The strategy is your responsibility. That separation is where retail traders get into trouble, because a well-regulated broker does not make an unregulated signal provider safe.

How Big Is the Execution Risk After Hours?

This is the section where the source material is most useful and most under-read. The order-type conversion, market to limit with cancellation of the unfilled remainder, is the single most important detail in XTB's overnight rollout for automated traders. It is also the detail most coverage skipped.

Consider the mismatch it creates. A strategy backtested on regular-hours data assumes continuous liquidity and near-certain fills. The overnight session offers neither. Trading 212 now advertises more than 5,600 US names around the clock on weekdays, almost 10 times XTB's list, according to its website. Interactive Brokers has run an overnight session for about 10,000 US stocks and ETFs since 2023 through its own trading venue (Finance Magnates). eToro, which started with 100 names in July 2025, opened every S&P 500 and Nasdaq 100 stock to 24/5 trading in November (Finance Magnates). CMC Markets widened its 24/5 range to more than 5,000 shares and ETFs in July (Finance Magnates).

The insight our desk would flag to any client running automation on XTB is this: coverage breadth and fill certainty are two different variables, and the market keeps conflating them. A venue with 10,000 symbols and a limit-conversion fill model can be worse for a market-order strategy than a venue with 2,000 symbols and firm fills. The number of tickers is the marketing metric. The order-type handling is the trading metric. XTB's 582-name list is small, but the fill model is the part that will move your P&L.

What Does This Cost a Real Portfolio?

Fees are where the two XTB products diverge sharply, and where a portfolio-aware view pays off. On the TBSZ side, XTB's terms are genuinely competitive: no account fee, and no commission on shares and ETFs up to EUR 100,000 of monthly volume. Estonian investment app Lightyear has offered a TBSZ it describes as free since February 2024, and Interactive Brokers also provides the account. XTB's own comparison table, dated July 2026, shows local banks charging maintenance and withdrawal fees. So on the tax wrapper, XTB is at or near the front of the field on headline cost.

On the overnight US session, the cost picture is less clear, because the terms describe execution mechanics rather than a separate overnight fee schedule. We did not find a published overnight commission in the source material, so treat any specific overnight fee claim you see elsewhere as unverified until you confirm it with the provider. What we can say is that the fill model itself is a cost. Partial fills and cancellations after hours are an implicit transaction cost that does not appear on a fee schedule but does appear in your realised slippage.

Where XTB Lags Its Overnight Rivals

The honest framing is that XTB is catching up, not leading. The company itself framed extended hours as a 2026 product priority, US markets first. Trading 212, Interactive Brokers, eToro and CMC Markets all got there earlier, and several cover far more names. XTB's 582 US stocks is the narrowest list in the peer set we could verify from the source material.

That does not make the product bad. It makes it a first-generation rollout with a staged release, an in-app on/off switch, and a conservative fill model. For a retail trader running a small, focused automated universe, 582 names may be entirely sufficient. For a broad-market scanner, it is a constraint. The gap between XTB and Interactive Brokers on overnight coverage is roughly 17 times in name count, and that is the number to weigh against whatever XTB offers on cost and tax efficiency.

How Zephyr AI Compares

The comparison that matters for our readers is not broker versus broker, it is strategy layer versus venue layer. XTB is an execution venue. Zephyr AI is a strategy engine. When we put the two questions side by side, where Zephyr AI's adaptive position-sizing edged out a static sizing model on the same volatility regime, the venue question stayed the same: can the broker deliver the fills the strategy assumes?

Zephyr AI's advantage on this specific dimension is that its position-sizing logic is built to tolerate partial fills and variable liquidity, which is precisely the XTB overnight condition. A static sizing model that assumes full fills will under-trade after hours. An adaptive model that reads realised fill quality and resizes accordingly will not. That is the concrete edge, and it is a strategy-layer edge, not a venue endorsement.

The Tax Wrapper Nobody Is Modelling

One more observation, because it is the kind of edge case that never makes it into a product announcement. A five-year TBSZ wrapper interacts with automated rebalancing in a way that most retail algo users will not model until it is too late. If your system is programmed to harvest losses, rotate sectors, or step out of a drawdown, every one of those actions inside a TBSZ is either restricted or penalised on early close. The wrapper effectively converts a tactical strategy into a strategic one, whether you intended that or not.

We flagged this because the same logic applies to the Polish IKE accounts XTB added in October 2024 and the OKI account due from the start of 2027 (Finance Magnates). Tax-advantaged wrappers are spreading across Europe, and each one quietly rewrites the rules for any algorithm running inside it. If your bot does not know it is inside a five-year wrapper, it will eventually make a decision the wrapper punishes.

The Bigger Picture for Automated Traders

XTB's two announcements tell a consistent story. Shares, ETFs and Investment Plans accounted for 82.9% of first transactions by new EU clients in the first half of 2026, even though CFDs still generated about 96% of the broker's gross trading result (Finance Magnates). The broker is courting long-horizon investors while its revenue still leans heavily on leveraged products. The overnight session and the TBSZ account both serve that long-horizon pitch.

For automated traders, the takeaway is to separate the marketing narrative from the execution reality. XTB is building out the investor-facing side of its business, and the tax wrapper is genuinely competitive. The overnight session is a catch-up product with a conservative fill model and a narrow universe. Neither is a reason to avoid the broker, and neither is a reason to assume your backtest will hold.

If you want to benchmark whether your strategy survives a venue like this, the sensible move is to test the strategy layer independently of the venue. 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.

Overnight US Stock Coverage Compared

Source: Finance Magnates, provider websites.

Hungarian TBSZ Accounts Compared

Free Download: XTB Algo Due-Diligence Checklist: Overnight US Stocks & Hungarian Tax Account
A due-diligence checklist to verify XTB's overnight US stock execution, API/bot compatibility, fee transparency, regulatory status, and Hungarian tax-account reporting before you automate.
Download XTB Algo Checklist

Source: Finance Magnates, XTB comparison table dated July 2026.

TBSZ Tax Treatment by Holding Period

Source: XTB website, via Finance Magnates.

If you are weighing whether your automated strategy can tolerate a venue with a conditional fill model and a narrow after-hours universe, the cleanest test is to run the strategy logic in isolation first. Compare Zephyr AI's live performance data against your own venue assumptions. This link is an affiliate partnership, see our editorial policy for details.

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 XTB's overnight session work for automated strategies?

Yes, with caveats. The session is available in stages with an in-app on/off switch, and it covers 582 US stocks including fractional shares. The catch is the order model: market orders become limit orders after hours and unfilled portions are cancelled, so any bot that assumes full fills will under-execute.

What happens if an overnight market order does not fill?

Per XTB's terms, the market order is executed as a limit order at the best opposing price, and any unfilled part is cancelled. There is no resting order left in the book. Your strategy needs explicit retry or resizing logic to handle this cleanly.

Can I run a high-turnover strategy inside a Hungarian TBSZ?

It is possible but costly. Closing the account early triggers the standard 15% personal income tax plus a 13% social contribution. A five-year wrapper rewards slow rebalancing and buy-and-hold behaviour, not tactical rotation.

How does XTB's overnight coverage compare with Interactive Brokers?

Interactive Brokers has run an overnight session for about 10,000 US stocks and ETFs since 2023. XTB lists 582 US stocks. On name count alone, that is roughly a 17-fold difference in favour of Interactive Brokers.

Is XTB regulated, and by whom?

XTB is a Warsaw-listed broker with multiple European entities. Verify the specific legal entity you onboard with directly on the FCA Register for UK accounts or the ASIC Connect registers for Australian accounts. We do not assert licence numbers we cannot cite.

Does the TBSZ account have a minimum deposit?

Yes. XTB's TBSZ opens from HUF 25,000, has no account fee, and charges no commission on shares and ETFs up to EUR 100,000 of monthly volume.

What is the biggest risk for an algo trader using XTB overnight?

The fill-certainty gap. Coverage breadth is the marketing number, but the order-type conversion is the trading number. A strategy backtested on regular-hours fills will diverge from live results after hours, and the divergence shows up as smaller-than-intended positions.

How fast is the Hungarian TBSZ market growing?

TBSZ accounts rose by more than a third to over 519,000 at the end of 2025, from about 386,000 a year earlier, according to Magyar Nemzeti Bank data cited by XTB. Assets approached HUF 7.9 trillion.

Can I run the same strategy on XTB and a dedicated AI trading bot?

You can, but treat them as separate layers. XTB is the execution venue, and the bot is the strategy engine. The two must be tested together, because a strategy that performs well on one fill model can underperform on another with no change to the signal logic.

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.

More in this category: Trading Industry News.