VALYNOXSTRATEGY RESEARCH TECHNOLOGY
AlphaPrimeAlphaPrime
Sign inFree 7-day trial
ACADEMY / Markets and trading sessions

Crude oil futures trading hours: CL and MCL in Taiwan time

CL and MCL trade on CME Globex from 6:00 pm to 5:00 pm US Eastern time, Sunday to Friday, with a 60-minute break each day. In Taiwan time the session opens at 7:00 am in winter or 6:00 am in summer and closes one hour before the next open. Settlement is set at 2:30 pm ET, the EIA report lands on Wednesday at 10:30 am ET, and a backtest must fix the session, the daily bar close and the roll rule before the first run.

By · Published · updated · Editorial policy

Contents
  1. What are the crude oil futures trading hours in Taiwan time?
  2. CL vs MCL: same hours, what else is different?
  3. When is the daily settlement price set, and which bar is the daily bar?
  4. When is the EIA inventory report released, and what does it do to volatility?
  5. How do contract expiry and rolls affect an oil futures backtest?
  6. How should you set the session in a backtest for CL and MCL?
  7. Frequently asked questions
  8. References

What are the crude oil futures trading hours in Taiwan time?

Crude oil futures trading hours for CL and MCL run from 6:00 pm to 5:00 pm US Eastern time, Sunday to Friday, with a 60-minute break each day. In Taiwan time that is an open at 7:00 am in winter or 6:00 am in summer, and a close one hour before the next morning's open. Settlement is set at 2:30 pm ET, not at the close.

CL is the WTI crude oil futures contract listed on NYMEX, a CME Group exchange, and MCL is its one-tenth-size Micro contract. WTI is West Texas Intermediate, the US crude grade that both contracts price. Both trade electronically on CME Globex, so "NYMEX trading hours" for oil are the Globex hours above. The daily break is the one hour from 5:00 pm to 6:00 pm ET when no trading takes place and the exchange closes one trading day and opens the next. Because the open is 6:00 pm, the trading day for Monday starts on Sunday evening in the US, which is Monday morning in Taiwan. The free futures market hours tool shows the same session in your own time zone, and the exchange's contract page is the place to check it, because exchanges can change these values.

Taiwan does not use daylight saving time, so the gap to US Eastern time changes twice a year. Daylight saving time is the US practice of moving clocks forward one hour from the second Sunday in March to the first Sunday in November. In winter Taiwan is 13 hours ahead of ET; in summer it is 12 hours ahead. Every oil session time therefore shifts by one hour in Taiwan, and a backtest that uses Taiwan-time session boundaries must shift with it. The same shift applies to other US contracts; the Nasdaq overnight session hours in Taiwan time and the Dow futures night session in Taiwan time move by the same hour on the same dates, which is useful if you trade oil and an index future from one calendar.

EventUS Eastern timeTaiwan time, winter (ET = UTC−5)Taiwan time, summer (ET = UTC−4)
Session open6:00 pm07:00 next day06:00 next day
Settlement window2:28–2:30 pm03:28–03:3002:28–02:30
Session close5:00 pm06:0005:00
Daily break5:00–6:00 pm06:00–07:0005:00–06:00
Week opensSunday 6:00 pmMonday 07:00Monday 06:00
Week closesFriday 5:00 pmSaturday 06:00Saturday 05:00
On a Taiwan clock the CL and MCL daily break falls at 06:00 to 07:00 in winter and 05:00 to 06:00 in summer, with settlement at 03:30 or 02:30, so every session time moves one hour with US daylight saving The break shifts one hour Taiwan clock, 00:00 to 09:00 Settle 03:30 Break 06–07 Winter Trading Summer Settle 02:30 Break 05–06 Summer: every time one hour earlier. So: keep two Taiwan-time templates.
The 23-hour CL and MCL session on a Taiwan clock. US daylight saving moves the open, the break and the close by one hour, so a Taiwan-time session must be defined twice a year.

CL vs MCL: same hours, what else is different?

CL and MCL share one session but differ in size, tick value and last trading day, so they share one session template and need separate cost, tick and roll settings. A tick is the smallest price step the exchange allows, and the tick value is what one tick is worth in dollars for one contract. For both contracts the tick is $0.01 per barrel. For CL, with one thousand barrels per contract, one tick is $10. For MCL, with 100 barrels, one tick is $1. The free futures contract calculator turns a price move in either contract into dollars.

The other difference is the calendar. MCL stops trading one business day before CL for the same contract month. A continuous series built for MCL therefore reaches its roll a day earlier, and a backtest that copies the CL roll dates onto MCL will try to trade a contract that has already stopped. Keep the two roll calendars apart even when the entry and exit rules are the same.

SettingCL (WTI Crude Oil)MCL (Micro WTI Crude Oil)
Trading hours6:00 pm–5:00 pm ET, 60-minute breakSame as CL
Contract sizeOne thousand barrels100 barrels
Minimum tick$0.01 per barrel$0.01 per barrel
Tick value$10$1
A $1.00 moveOne thousand dollars per contract$100 per contract
Last trading dayThree business days before the 25th of the month before the contract month (four when the 25th is not a business day)One business day before CL
CL and MCL share the trading session and the 0.01 dollar tick; they differ in size, with CL ten times the 100 barrels of MCL, in tick value, 10 versus 1 dollar, and in last trading day, three versus four business days before the 25th Same hours, three differences Session and $0.01 tick are shared CL MCL Size 10× MCL 100 bbl Tick value $10 $1 Last day 25th −3 25th −4 One session, separate costs and rolls. So: test MCL as its own strategy.
One session, two contracts. Only the size, the tick value and the last trading day change between CL and MCL.

The practical consequence is in costs. The same commission per contract is a far larger share of a tick in MCL than in CL, so a rule that survives costs on CL can fail on MCL even though every signal is identical. Test the two as separate strategies with their own cost settings.

When is the daily settlement price set, and which bar is the daily bar?

The CL settlement price comes from the 2:28–2:30 pm ET window, not from the 5:00 pm ET Globex close, so a daily bar that ends at the session close and a daily bar that ends at settlement hold different highs, lows and closes. The settlement price is the official end-of-day price the exchange uses to mark positions and margin, and for CL it is based on trading in those two minutes. In Taiwan time the window is 03:28–03:30 in winter and 02:28–02:30 in summer.

This creates two legitimate daily bars for the same date. A Globex-close bar runs from 6:00 pm the previous day to 5:00 pm. A settlement bar runs from 2:30 pm the previous day to 2:30 pm. The stretch from 2:30 pm to 5:00 pm belongs to today in the first definition and to tomorrow in the second. Any rule that reads the daily close, the daily range or a moving average of closes will produce different signals on the two series. Neither is wrong. What matters is that you pick one and keep it for the whole backtest and for live trading, because the signal you tested at a 2:30 pm close does not exist on a chart that closes at 5:00 pm.

Daily bar definitionBar covers (ET)Close usedFits best when
Globex-close bar6:00 pm previous day to 5:00 pmLast trade before the breakYour data vendor and platform build daily bars from the full session
Settlement bar2:30 pm previous day to 2:30 pmExchange settlement priceYou mark P&L at settlement or compare with published settlement data
Two daily bars for the same date in US Eastern time: the Globex bar runs from 6:00 pm the previous day to 5:00 pm, the settlement bar from 2:30 pm the previous day to 2:30 pm, so the 2:30 to 5:00 pm stretch belongs to a different day in each Same date, two daily bars Which day gets 2:30–5:00 pm ET? Previous day Settle 2:30 pm Globex bar Settle bar 2:30–5:00 pm ET Daily closes differ between the two. So: fix one daily close before testing.
Same date, two daily bars. The two and a half hours after settlement are counted in today's bar under one definition and in tomorrow's under the other.

Intraday traders face a smaller version of the same choice. A 60-minute chart that starts at 6:00 pm ET gives bars ending on the hour; a chart that starts at midnight in Taiwan time gives bars that straddle the break. The way a platform anchors bars to the session is explained in the article on US market hours in Central Time, and the same logic applies to oil when the exchange's own clock is Central time.

When is the EIA inventory report released, and what does it do to volatility?

The EIA Weekly Petroleum Status Report is released on Wednesday at 10:30 am ET, which is 23:30 Taiwan time in winter and 22:30 in summer, and the minutes around it bring wide bars and worse fills. The EIA is the US Energy Information Administration, and its weekly report gives the change in US crude oil stocks. Oil traders watch it the way index traders watch an employment report. When a holiday falls on Monday, the release moves to Thursday at the same hour.

For an intraday rule the problem is not the direction of the move but the fill. A market or stop order placed in the first minutes after the release can fill several ticks away from the trigger price, because stop orders become market orders when touched and trade at whatever price is next available. A backtest that assumes one tick of slippage everywhere is too kind to trades taken in that window. There are two honest ways to handle it, and both must be written down before the test:

  1. Charge more slippage in the window. Keep the rule as it is and raise slippage for entries and exits between, say, 10:25 and 11:00 am ET. If the result changes a lot, the rule depended on cheap fills at the report.
  2. Add a time filter. Skip new entries from a few minutes before the release until a fixed time after it. Fix the window once; moving it after seeing the result turns the filter into a tuned parameter.

Hypothetical example

Suppose a CL intraday rule averages $45 of gross profit per trade with one tick ($10) of slippage per side, so $25 net of $20 slippage before commission. Suppose a fifth of its trades enter in the 30 minutes after the EIA release and those trades fill three ticks worse per side. The extra cost is $40 on one trade in five, which is $8 averaged over every trade. The net average trade drops from $25 to $17, a cut of about a third, from one release a week.

Five-minute bars around the Wednesday EIA release at 23:30 Taiwan winter time, 22:30 in summer: the bars in the 30 minutes after it are wider and fill worse, and a filter from 23:25 to 00:00 skips seven bars After EIA: wider bars Taiwan, winter; summer is 22:30 Filter skips Worse fills EIA 23:30 00:00 Fills here cost more than one tick. So: filter the window or add slippage.
The Wednesday release on a Taiwan clock. A time filter removes the shaded bars; a slippage rule keeps them but charges more for them.

Daily-bar strategies are less exposed, because the report lands inside the bar and the daily close is set hours later. They still feel the report through the size of Wednesday's range, which matters for any rule that scales its stop or position with recent range.

How do contract expiry and rolls affect an oil futures backtest?

CL stops trading three business days before the 25th calendar day of the month before the contract month, so a continuous series rolls roughly every month and the roll gap must be adjusted or the backtest trades price jumps that nobody could have captured. A continuous series is a single price history built by joining one contract month to the next. Oil rolls about twelve times a year, which is more than quarterly index futures, so the roll rule matters more in oil than in ES or NQ.

Two months of the same contract rarely trade at the same price. When the next month is priced above the front month, the market is in contango; when it is below, backwardation. At the roll, the continuous series jumps by that difference. If the series is not adjusted, a long position appears to gain or lose the gap overnight, although a real account only changes contracts and sees no such profit or loss. Back-adjusting shifts earlier prices by the gap so the joined series has no jump at the roll date. The full method, and what it does to long histories, is covered in futures backtesting: contract specs, rollover and tick value.

  1. Find the last trading day. Take the 25th of the month before the contract month and count back three business days. If the 25th is not a business day, count back four. MCL's last day is one business day earlier.
  2. Pick the roll date. Most traders roll a few days before the last trading day, often when the next month's volume passes the front month's. Write the rule, not the dates.
  3. Adjust the gap. Shift the earlier contract's prices by the difference between the two contracts at the roll, so the series has no jump.
  4. Apply the same rule to every roll in the history. One rule, applied the same way about twelve times a year, is what makes the series comparable with live trading.

Hypothetical example

Suppose the 25th falls on a Thursday. Counting back three business days gives Wednesday, Tuesday, Monday, so CL stops trading on Monday the 22nd and MCL on Friday the 19th. Suppose at the roll the front month trades at $70.00 and the next month at $70.80. An unadjusted series jumps $0.80, which is $800 per CL contract and $80 per MCL contract. A back-adjusted series adds $0.80 to all earlier prices and shows no jump.

Example month where the 25th is a Thursday: counting back three business days, CL stops on Monday the 22nd, MCL one business day earlier on Friday the 19th, and the roll comes a few days before that on the 16th Count back from the 25th Example month: the 25th is a Thursday Roll MCL stops CL stops 25th 16 17 18 19 20 21 22 23 24 25 −4 Weekend −3 −2 −1 Roll first, then adjust the price gap. So: write the rule, not the dates.
From the 25th back to the last trading day, then the roll. The price gap between the two months is adjusted in the data, never traded by the strategy.

How should you set the session in a backtest for CL and MCL?

Set the full Globex session with the 5:00 pm ET break, state the time zone, fix the daily bar close, charge commission and slippage on every trade and define the roll rule before the first run, because changing any of these afterwards changes the result you are comparing against. The order below is the order the settings depend on each other. A time zone chosen after the session template is set can silently move the break by an hour; a roll rule changed after the costs are set changes how many trades the costs apply to.

  1. Session template. 6:00 pm to 5:00 pm ET, Sunday to Friday, with the 60-minute break. Use the same template for CL and MCL.
  2. Time zone. State whether bar times are in ET, Central (the exchange's own clock) or Taiwan time, and whether the data follows US daylight saving. A Taiwan-time template needs a winter and a summer version.
  3. Daily bar close. 5:00 pm ET Globex close or 2:30 pm ET settlement. One choice, kept for every run and for live trading.
  4. Commission and slippage per trade. Per contract, on entry and exit, with extra slippage at the 6:00 pm ET reopen and in the EIA window if the rule trades there.
  5. Roll rule and gap adjustment. The rule from the previous section, applied the same way at every roll, with MCL's calendar one business day earlier than CL's.
Five backtest settings fixed in order before the first run: session template, time zone, daily bar close, costs per trade and roll rule. In the example the time zone changes, the session template is kept, and the daily bar close, costs and roll rule all shift with it, so the whole test must be rerun Fix five settings first Example: the time zone changes 1 Session template 2 Time zone 3 Daily bar close 4 Costs per trade 5 Roll rule Kept Changed All shift with it Later steps depend on earlier ones. So: change one, rerun the whole test.
Five settings fixed in order before the first run. A change to any one of them sends the test back to the start, because every later result depends on it.

Once these five are fixed, the strategy rules can be tested on top of them without the session itself moving under the result. The rule types that suit a 23-hour market, and how each one is tested, are listed in futures algo trading strategies and how to backtest each type. In AlphaPrime (Valynox), the session, the costs and next-bar execution are part of the research settings that every candidate shares, so its backtesting applies the same commission, slippage and bar timing to each of the strategies it evaluates, and the comparison between candidates stays fair.

The box below shows what a finished daily-bar run with fees, slippage and next-bar execution looks like in a report. It is not a crude oil test. It is a stand-in on NQ daily bars from 2019-01-02 to 2025-06-30, long and short, one contract, using an OSMA rule (the distance between a MACD line and its signal line) as its only entry and exit, with no stop, target or filter. Out of 1,024 candidates the search tried, it kept this one for its return against its drawdown. Picking the kept candidate after the fact flatters it: the best of many tries always looks better than the rule itself would on new data, which is the selection effect Bailey et al. (2014) describe. The numbers show the shape of a report, not what the rule earns or what a reader can expect.

Source: AlphaPrime product engine (CPU), run 2026-10-31. NQ daily bars, 2019-01-02 to 2025-06-30, long and short, 1 contract, $20 commission and $10 slippage per trade.

NumberResult
Net profit (closed trades)$133,935
Max drawdown$46,600
Trades213
Win rate54.93%
Profit factor1.418
Average trade$628.8
Time in market22.69%
The equity curve and drawdown of the stand-in NQ daily run, with the out-of-sample stretch marked. Look at how the curve behaves after the marked line compared with before it, and at the depth of the drawdown relative to the climb.
The equity curve and drawdown of the stand-in NQ daily run, with the out-of-sample stretch marked. Look at how the curve behaves after the marked line compared with before it, and at the depth of the drawdown relative to the climb. NQ · daily · 2019-01-02 to 2025-06-30 · CPU engine backtest

Read the box the way you would read your own oil run. Net profit (closed trades) of $133,935 against a max drawdown of $46,600 over 213 trades says how much the account would have moved against the trader on the way; the average trade of $628.8 is the number to compare with the $30 of commission and slippage charged per trade. For a CL daily rule, the same report would be built from bars that close at the time you chose in step 3, with rolls handled by the rule in step 5. If the oil result changes a lot when you switch between the Globex-close bar and the settlement bar, the rule is sensitive to the bar definition, and that is worth knowing before any live order.

A next step you can take today: load your CL or MCL daily data twice, once with bars ending at 5:00 pm ET and once with bars ending at 2:30 pm ET, and count how many daily closes differ by more than one tick. That count tells you how much the bar definition alone can move any daily rule you test.

Frequently asked questions

How long is the daily break in crude oil futures?

CL and MCL pause for 60 minutes each day, from 5:00 pm to 6:00 pm US Eastern time. In Taiwan time the break is 06:00 to 07:00 in winter and 05:00 to 06:00 in summer. The break marks the end of one trading day and the start of the next, so the trading day for Monday begins on Sunday evening in the US.

How is the crude oil futures settlement price calculated?

The CL settlement price is set by the exchange from trading on CME Globex between 2:28:00 and 2:30:00 pm ET, not from the last trade of the session at 5:00 pm. In Taiwan time the window is 03:28 to 03:30 in winter and 02:28 to 02:30 in summer. A daily bar that closes at settlement therefore has a different close from a daily bar that closes at the end of the Globex session. The exchange can change the procedure, so check its settlement page before relying on a specific time.

Is the oil futures market open on Sunday?

Yes. CL and MCL open on Sunday at 6:00 pm ET, which is Monday at 07:00 in Taiwan in winter and 06:00 in summer. The Sunday evening session is the first part of Monday's trading day. Trading then continues through the week, with a one-hour break each day, until Friday at 5:00 pm ET.

Do CL and MCL have the same trading hours?

Yes, both trade from 6:00 pm to 5:00 pm ET, Sunday to Friday, with the same 60-minute break. The differences are in size and expiry: CL is one thousand barrels with a $10 tick, MCL is 100 barrels with a $1 tick, and MCL stops trading one business day before CL for the same contract month. A backtest can share one session template but needs separate cost and roll settings for the two.

References

  • Bailey, D. H., Borwein, J., López de Prado, M., & Zhu, Q. J. (2014). Pseudo-Mathematics and Financial Charlatanism: The Effects of Backtest Overfitting on Out-of-Sample Performance. Notices of the American Mathematical Society, 61(5), 458–471.
  • Pardo, R. (2008). The Evaluation and Optimization of Trading Strategies (2nd ed.). John Wiley & Sons.
  • U.S. Securities and Exchange Commission. Stop, Stop-Limit, and Trailing Stop Orders (Investor Bulletin). https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins-15