Do Bitcoin CME Gaps Always Get Filled? BTC Futures Gap Trading Explained

 

The CME gap is one of the most widely repeated technical claims in bitcoin trading, and its central assertion is that a gap eventually gets filled, making it usable as a price target.

 

That claim can now be tested properly, for two reasons. The first is that CME Group moved its crypto futures and options to around-the-clock trading on May 29, 2026, so weekend gaps no longer form and the phenomenon constitutes a closed, complete dataset. The second is that a statistical study covering 427 CME bitcoin futures weekends now exists, spanning December 2017 to May 2026.

 

Judged on fill rates, the claim appears to hold. Per that study, 22% of gaps filled within an hour, 55% within a day, 80% within a week, 89% within a month, and 97% cumulatively as of September 26, 2026.

 

But the study did something most discussions omit: it built a control group, taking an invented price level the same distance from the reopen in the opposite direction and racing the two to see which was reached first. In three of four gap-size bands, the invented level was reached first more often. The 97% figure, in other words, mainly reflects bitcoin's own volatility rather than any special pull exerted by a gap.

 

That distinction is what this article sets out to establish: gaps do get filled, yet the filling itself does not constitute a usable trading edge.

 

 

 

The cause is a difference in trading hours. CME bitcoin futures previously closed over weekends while spot traded continuously, producing a discontinuity on the futures chart at the Sunday reopen.

 

Fill rates are high but carry no time bound. Per the study, 80% filled within a week and 89% within a month, while the 97% figure measures to a cutoff nearly nine years after the earliest sample.

 

The control test does not support the claim. In three of four size bands the equidistant invented level was reached first more often, at 49% against 33%, 62% against 37%, and 67% against 33%.

 

The one exception is mid-sized gaps. Among gaps of 3% to 5%, the gap level was reached first 59% against 41%, with 71% filling within a week against 52% for the invented level.

 

There is a directional asymmetry. Gaps down filled at a 99.4% rate against 94.8% for gaps up.

 

No new gaps form any more. CME has traded continuously since May 29, 2026, so existing gaps remain on charts while no new weekend gaps appear.

 

 

Understanding the cause requires only one fact: two markets kept different hours.

 

CME bitcoin futures launched in December 2017 and previously closed at 4:00 p.m. Chicago time on Friday, reopening at 5:00 p.m. Chicago time on Sunday, leaving roughly 49 hours without trading. Bitcoin's spot market, by contrast, runs continuously and never closes.

 

The mechanism follows directly. After Friday's close, spot kept trading and prices could move substantially over the weekend while the futures chart recorded nothing. At the Sunday reopen, the futures price jumped straight to a level near spot, leaving a blank stretch on the chart. One end of that blank is Friday's close and the other is Sunday's open, and the distance between them is the gap.

 

Worth stating clearly is that a gap is not a price anomaly but a recording absence. The spot market traded continuously throughout; only the futures market did not. A fill therefore means price returning to Friday's closing level, so that subsequent price action covers the blank interval on the chart.

 

That also explains why the phenomenon appears only on futures charts rather than spot ones. Viewed on spot alone, weekend price movement is continuous and no gap exists at all. For bitcoin spot pricing and historical data, see MEXC's BTC price page.

 

The distribution of gap sizes is worth establishing first. Per the study, the median size across 427 weekend gaps was 1.2%, with 186 below 1%, 43 at 5% or more, and the largest at 13.1% in May 2019. Most gaps, in other words, were small, while the conspicuous large ones were a minority.

 

 

Consider first the data that supports the claim.

 

 

At first glance that set looks persuasive, with 97% close to inevitability. But it carries a critical flaw: no time bound. The sample begins in December 2017, so the earliest gaps had nearly nine years in which to fill. For an asset whose intraday moves frequently exceed 2%, given enough time almost any level not far from the current price will eventually be touched.

 

What carries trading meaning is therefore the time-bounded figures. The 80% within a week is the most practical of them, while just 22% within an hour indicates gaps were rarely filled immediately after the reopen.

 

The study's definition of a fill also needs stating, because it affects how strict the results are. The study used Binance BTCUSDT spot prices read at CME's Friday close, 4:00 p.m. Chicago time, and Sunday reopen, 5:00 p.m. Chicago time, defining a fill as bitcoin trading back to the Friday level. Fills were checked on hourly bars, so a level touched at any point within an hour counts as filled even if price left immediately. That is a relatively permissive definition, meaning actual fill rates under stricter definitions would be lower than the figures above.

 

The study excluded several special cases, including ordinary holiday closures starting on a Friday such as Good Friday, the Christmas 2020 gap, and early closes other than the day after Thanksgiving. It did not account for trading costs, and it reported a 0.90 correlation in gap size between its spot prices and CME's daily bars.

 

Unfilled gaps are documented too. The study notes 16 weekend gaps remained unfilled when CME moved to continuous trading, two of which have since filled, leaving 14 open, of which 13 sit below the current price with the lowest at $3,198 from December 2018.

 

 

This is the core of the article and the step most discussions of CME gaps leave out.

 

The problem is that a high fill rate alone cannot demonstrate that a gap exerts any special influence. Establishing a gap as a meaningful price target requires answering a stricter question: is the probability of price returning to the gap level higher than the probability of returning to a randomly chosen level the same distance away?

 

That is exactly how the study was designed. It took an invented level the same distance from the reopen in the opposite direction, then compared which of the two was reached first, recording touches within the same hour as ties. The results follow.

 

 

In three of four bands the invented level was reached first more often, and by substantial margins: 62% against 37% in the 1% to 3% band and 67% against 33% in the 5% or more band. For those three categories of gap, a wholly arbitrary equidistant level was more likely to be reached first than the real gap.

 

The sole exception is the 3% to 5% band, where the gap level won 59% to 41%. The study provides one further figure for that band: across 56 such weekends, 71% filled within a week against 52% for the invented level. That is the only evidence among the four bands supporting any additional pull from a gap.

 

The accurate conclusion therefore has two layers. First, gaps do get filled and at a high rate, which is not in dispute. Second, that high fill rate does not beat a random equidistant level across most size ranges, so it reflects bitcoin's volatility characteristics rather than any special property of a gap. Treating "a gap will eventually fill" as a prediction amounts statistically to saying "price will eventually move to some nearby level," which holds for any volatile asset.

 

One limitation of the study belongs here as well. The control design answers whether a gap outperforms a random level, but this is a single study with 427 observations, complete though that sample is, and its fill criterion uses a permissive hourly-touch standard. Both should be stated whenever the conclusion is cited.

 

 

Understanding the underlying reasons helps judge whether the conclusion generalises.

 

The first reason is asymmetry in the time dimension. Once formed, a gap stays on the chart permanently with no expiry, while price moves every day. Given a long enough observation window, the fill rate therefore converges toward 100%. That is not a property of gaps but the inevitable result of combining an unbounded time window with a volatile asset. The 97% figure looks striking precisely because it has no time bound.

 

The second reason is the distribution of gap sizes. With a median of just 1.2% and 186 of 427 below 1%, and for an asset whose daily moves frequently exceed 2%, covering a 1% distance within a day or two is routine. That explains why 77% of sub-1% gaps filled within a day, while also explaining why the invented level won more often in that same band: when the distance is very small, which side gets touched first approaches random.

 

The third reason is direction. The study found gaps down filled at 99.4% against 94.8% for gaps up. That difference is consistent with bitcoin's long-run uptrend over the sample period, since rising prices make gaps below easier to cover on pullbacks while gaps above require price to keep advancing. The asymmetry is therefore more likely a product of trend than of gaps themselves, which also explains why 13 of the 14 currently unfilled gaps sit below the current price.

 

The fourth reason is that the mid-sized exception needs cautious reading. The 3% to 5% band contains only 56 observations, a small sample statistically, so its outperformance could be a genuine effect or an artefact of how the bands were cut. A single study cannot distinguish between the two.

 

 

That the phenomenon has ended is essential to understanding its current status.

 

CME Group announced in February 2026 and implemented from May 29, 2026 the move of its regulated crypto futures and options to continuous trading. Tim McCourt, CME's global head of equities, foreign exchange and alternative products, cited client demand for risk management in digital asset markets being at an all-time high. With futures no longer closing over weekends, the cause of weekend gaps disappeared. Analyst Daan Crypto Trades put it as no new gaps being created going forward, while existing gaps remain on charts.

 

The specifics of maintenance windows differ across reports, so CME's own schedule should govern. One report describes two minutes of maintenance on weekdays and two hours on Saturdays. Another describes a two-hour pause from 03:00 to 05:00 Coordinated Universal Time each Saturday. A third describes a 60-minute window from 22:00 to 23:00 Coordinated Universal Time on Sundays and gives the start date as May 30. Those discrepancies mean anyone relying on precise times should consult the official schedule. Weekend trades settle on the next business day per reporting.

 

The count of unfilled gaps also differs by basis. Per the study, 16 weekend gaps were unfilled at the switch, two have since filled, and 14 remain. Per CoinDesk's reporting before the switch, three gaps were unfilled at that time, two above the roughly $73,000 spot price at around $80,000 formed in late January and around $78,500, and one just below $70,000. The difference comes from scope, with the former covering all weekend gaps since 2017 and the latter focusing on recent gaps near the price at that time. Citations should specify which.

 

One piece of scale context is also worth knowing. Per Cole Kennelly, founder of Volmex Labs, open interest in CME bitcoin futures options stood near $800 million to $900 million against roughly $27 billion to $30 billion for BlackRock's IBIT exchange-traded fund options. That comparison indicates CME holds a limited share of crypto derivatives, which in turn weakens another common narrative, namely that gaps fill because institutional money deliberately moves to close them.

 

 

Translating the findings into practice yields four points.

 

First, treat a gap as a reference level rather than a prediction. A gap marks a price at which trading once occurred, and like any technical level it can serve as a point of attention, but the control test shows it is no more likely to be reached first than a random equidistant level. Using it to form a directional view therefore lacks statistical grounding.

 

Second, if using it at all, focus on the 3% to 5% range. That is the only band to win the control test, with 71% filling within a week against 52% for the invented level. But bear in mind the band contains only 56 observations, making it small-sample evidence.

 

Third, always attach a time window. The 97% figure carries no trading value because it is unbounded. The usable figures are 80% within a week and 89% within a month, while just 22% within an hour indicates immediate fills after the reopen were uncommon. Any gap-based plan should specify a holding period.

 

Fourth, note that the phenomenon has stopped generating new samples. No new weekend gaps have formed since May 29, 2026, so every gap remaining on charts is a historical residue. Whether they still influence price cannot be tested with new data, and under continuous trading the original cause no longer exists.

 

For readers looking to participate in the bitcoin spot market, introductory material sits in MEXC's complete guide to buying bitcoin, with practical methods covered in how to buy BTC. Related platform activity can be found on the BTC Carnival page.

 

Data beats narrative. Rather than waiting for a gap to be filled, look at what price is doing right now: open the BTC spot order book on MEXC and judge from real trades and real depth, where the pair trades as BTC against USDT. As of 9:13 p.m. Eastern on October 22, bitcoin traded near $81,397, down 2.36% over twenty four hours, for a market capitalisation around $1.64 trillion.

 

 

For James Mitchell, the CME gap is a textbook case of how a factually correct statement produces a wrong conclusion. The 97% fill rate is real, yet it contains almost no information, because the statistical window is unbounded while bitcoin moves every day. The question with discriminating power is not whether a gap will fill but whether returning to a gap is more probable than returning to an arbitrarily chosen equidistant level. The study ran that test, and the answer is negative in three of four size bands by clear margins, at 62% against 37% in the 1% to 3% band and 67% against 33% above 5%.

 

The significance of that result extends beyond CME gaps. Any technical claim should face the same test of whether it beats a random benchmark. Absent a control group, a high hit rate easily turns out to be an artefact of construction rather than a signal. The gap case is unusually clean because its control can be defined precisely as the equidistant level in the opposite direction. Most technical indicators cannot be tested so neatly, but this one can, and the test does not support the popular claim.

 

The directional asymmetry deserves recording too. Gaps down filled at 99.4% against 94.8% for gaps up, while 13 of the 14 currently unfilled gaps sit below the current price. That distribution aligns exactly with bitcoin's long-run appreciation over the sample, so it more likely reflects trend than any pull from gaps. Misreading it as gaps exerting downward attraction would produce systematically wrong judgements through an uptrend.

 

The piece of evidence worth preserving is the 3% to 5% band. It is the only group to pass the control test, at 71% against 52% within a week. One possible explanation is that this size is large enough to draw attention yet close enough to be covered within a reasonable period, producing something self-fulfilling. But 56 observations is a small sample, so that explanation can only stand as a hypothesis. The honest formulation is that if any real CME gap effect exists, it is probably confined to mid-sized gaps and modest in magnitude.

 

One final point concerns the phenomenon's end. CME trading continuously since May 29, 2026 closes this dataset completely. That is a rare benefit in empirical work, since the sample no longer grows and the conclusion cannot be overturned by later data. But it also means the 14 gaps remaining on charts have no mechanism sustaining them. With weekend closure gone as a premise, continuing to treat those historical gaps as price targets requires an entirely new rationale, and no data currently supplies one.

 

 

 

It is a discontinuity in price on the CME bitcoin futures chart. Those futures previously closed at 4:00 p.m. Chicago time on Friday and reopened at 5:00 p.m. Chicago time on Sunday, leaving roughly 49 hours without trading, while bitcoin spot ran continuously. Weekend spot price movement went unrecorded by futures, so at the Sunday reopen the futures price jumped straight to a level near spot, leaving a blank on the chart whose two ends define the gap.

 

 

Fill rates are high in proportional terms, but that does not constitute a usable prediction. Per the study of 427 weekends, 80% filled within a week, 89% within a month, and 97% cumulatively as of September 26, 2026. But the study's control test showed that in three of four gap-size bands an invented equidistant level in the opposite direction was reached first more often, so the high fill rate mainly reflects bitcoin's volatility rather than any special pull from a gap.

 

 

It took an invented level the same distance from the reopen in the opposite direction, then compared which of the two was reached first, counting touches within the same hour as ties. The results were 33% for the gap level against 49% for the invented level in the sub-1% band, 37% against 62% in the 1% to 3% band, 59% against 41% in the 3% to 5% band, and 33% against 67% above 5%. Only the 3% to 5% band favoured the gap level.

 

 

Because it carries no time bound. The sample begins in December 2017, so the earliest gaps had nearly nine years in which to fill, while bitcoin's intraday moves frequently exceed 2%. Over a long enough window almost any level not far from the current price will be touched, making 97% closer to a mathematical certainty than a trading signal. The figures with trading value are the bounded ones, such as 80% within a week and 22% within an hour.

 

 

Per the study, the median across 427 weekend gaps was 1.2%, with 186 below 1%, 43 at 5% or more, and the largest at 13.1% in May 2019. Most gaps were therefore small while conspicuous large ones were a minority. Sub-1% gaps filled within a day 77% of the time, yet in that same band the invented level still won the race more often.

 

 

No. CME Group announced in February 2026 and implemented from May 29, 2026 the move of crypto futures and options to continuous trading, removing weekend closure as a cause, with analysts describing it as no new gaps being created going forward. Existing gaps remain on charts, and per the study 16 were unfilled at the switch, two have since filled and 14 remain, of which 13 sit below the current price with the lowest at $3,198 from December 2018.

 

 

They do, though probably not because of the gaps themselves. The study found gaps down filled at 99.4% against 94.8% for gaps up. That difference is consistent with bitcoin's long-run uptrend over the sample, since rising prices make gaps below easier to cover on pullbacks while gaps above require further advances. Reading it as gaps exerting downward attraction would produce systematically wrong judgements during an uptrend.

 

 

The information above is provided for general market information and educational analysis only and does not constitute investment advice, financial advice, legal advice, tax advice or a recommendation to trade. Digital asset prices can fluctuate sharply, historical statistical relationships do not guarantee future results, and any inference drawn from historical fill rates may fail. The statistics cited here come from a single third-party study whose sample comprises 427 weekends from December 2017 to May 2026, whose fill criterion uses the relatively permissive standard of an hourly price touch without accounting for trading costs, and some of whose size bands contain small samples, so its conclusions should be understood within those methodological limits. The trading hours, rule changes, counts of unfilled gaps and market data cited reflect publicly available information at the time of publication, different sources use different bases, and precise trading hours and maintenance windows should be taken from CME Group's official schedule. This article states explicitly that the phenomenon has generated no new samples since May 29, 2026 and that the relevance of historical gaps under continuous trading cannot be tested with new data. Readers should conduct their own research and make decisions based on their own financial circumstances, investment objectives and risk tolerance, consulting a qualified professional where appropriate. The MEXC Crypto Pulse team accepts no liability for any direct or indirect loss arising from the use of this information.

 

 

James Mitchell specializes in technical analysis, market trends, and trading strategies for both Bitcoin and altcoins. Based in London, he has over 10 years of experience in financial markets. Before joining MEXC Learn, James worked as a senior analyst at a leading European investment firm, where he developed expertise in risk management and quantitative trading. His transition to cryptocurrency markets began in 2017, and he has since become recognized for his data-driven approach. He holds a Master's degree in Financial Economics from the London School of Economics. His analytical approach combines traditional technical analysis with on-chain metrics to provide readers with actionable insights.

 

Areas of Expertise: Technical Analysis, Market Trends and Cycles, Trading Strategies, Bitcoin and Altcoin Analysis, Risk Management.

 

 

 

Want the fastest access to MEXC's latest updates? Join our official Telegram group now!

MEXC Platform: Download App | Zero Fee Trading | Buy USDT Now

Join MEXC Community: X (Twitter) | Telegram | Discord

Get Insights: MEXC Learn | Terms of Service | MEXC Blog | MEXC News

Account Verification: Understand KYC | How to Complete KYC

External Content Platforms: Substack | Medium | Paragraph | LinkedIn | X(News)