CPI Drops at 8:30. Is Your Bot Ready for the Move?
Scheduled macro releases test a bot in three places: the grid range, the DCA ladder and the liquidation price. How to measure each against the expected move, three worked examples, what happened on September 16, and the remaining 2026 CPI and FOMC dates.
Last updated: October 7, 2026.
A CPI release or a Fed decision tests a bot in three places: whether the grid range survives the move, whether a pre-release flush fills the DCA ladder, and how close a leveraged position sits to liquidation. Compare each with the move the options market expects for the day. If the range edges and liquidation price sit more than two expected moves away, leave the bot alone. If they sit closer, widen the range, thin the ladder or cut leverage before the release, or stop the bot with a "keep" option and restart it afterwards.
Macro days rarely break a bot by being volatile. They break bots that were sized for a calm week. The fix is usually a ten-minute check the day before, not a reaction at 8:31.
Key facts at a glance
- US CPI is published at 8:30 a.m. Eastern. The next releases are October 14, November 10 and December 10, 2026.1
- The Fed statement lands at 2:00 p.m. Eastern, with the press conference at 2:30. The remaining 2026 meetings end October 28 and December 9; December includes new projections and the dot plot.2
- September broke the calm. The Fed raised rates by 25 basis points to 3.75%–4.00% on September 16, its first hike since 2023. Bitcoin touched about $75,242 and more than $571 million in longs were liquidated on a day that also brought a Senate setback for the CLARITY Act.23
- A priced-in print needs no reaction. When August's CPI matched forecasts at 3.4%, the options market had priced a move of about 1.3%, and a bot that kept its settings through the release missed nothing.4
- Leveraged bots live through releases. On one multi-exchange platform, the typical COMBO bot in H1 2026 ran about 12 hours at an average leverage of 7.5x.5 A bot launched in the morning is still open when the number comes out.
Q4 2026 release dates, in UTC
US clocks move back on November 1, so the UTC time shifts by an hour after October.12
Step 1: sort the release before you touch the bot
Not every release deserves a change. Three questions sort them.
- Is the forecast stable? If the consensus hasn't moved for days and forecasters agree, the market has mostly priced the outcome. August's CPI is the textbook case.4
- Are expectations split? A Fed meeting where traders disagree between hold and hike, or a CPI print after two surprises in a row, carries real event risk. September's FOMC sat in this group.
- Is something else landing the same day? On September 16, a rate hike and a failed Senate vote on crypto market structure arrived within 24 hours.3 Stacked news widens the move more than either event alone.
A stable forecast with nothing stacked: check the bot, change nothing. Split expectations or stacked news: go to step 2.
Step 2: measure your bot against the expected move
The expected move is the market's own estimate of how far price may travel on the day. Options desks publish it for scheduled releases as a percentage. Before the August CPI, it was about 1.3% for Bitcoin.4 It isn't a ceiling, and real surprises run past it. A working rule:
Worked example 1: GRID on BTC
Bitcoin at $86,000, expected move 1.5%, so roughly $1,300.
A grid doesn't care which way price goes, only whether it stays inside the range. Release minutes add a second cost: spreads widen and the book thins, so a stop loss that triggers then fills by market order below its level.6 Your Bot Didn't Get That Price shows how much that gap can grow.
Worked example 2: DCA on ETH
A long DCA bot with a $200 base order and five $200 safety orders spaced 1% apart. A 3% pre-release flush fills three of the five safety orders before the number is even out.
That's the trap. Leveraged traders often cut positions in the hours before a release, which knocks price down without any new information. The ladder can't tell that flush from a real repricing. In August, a leverage flush pushed Bitcoin toward $63,200 hours before a CPI print that changed nothing.4 If the real move comes after the release, the bot has already spent most of its budget.
The fix: wider steps (2% instead of 1% covers twice the distance with the same number of orders), a cap on active safety orders, or keeping reserve funds unspent until after the release.
Worked example 3: COMBO at 10x
On the long COMBO example in our COMBO explainer, DCA fills pulled the average entry to $75,240 and the liquidation price sat at about $73,080, 2.9% below.7 Against a 1.5% expected move, that's under two expected moves: a normal surprise can reach it.
Three ways to buy room before the release:
- Lower leverage. At lower leverage the same position needs a bigger move to liquidate.
- Add margin. COMBO lets you change margin on a running bot.8
- Close and relaunch after. For a bot that runs about 12 hours on average, skipping one release costs little.
Liquidation Cascades Explained shows how one forced close drags price to the next cluster, which is exactly what happened to $571 million of longs on September 16.
Step 3: pick the action
Pause, Edit or Close explains what each button does to orders and coins.
Two things to avoid. Don't move a grid range after the candle that broke it; you'd be rebuilding a range strategy inside a trend. And don't add leverage before a release because "it'll go my way." Being right on direction doesn't help if a wick liquidates the position first.
Release day, hour by hour
| When | What to do |
|---|---|
| T−24h | Sort the release (step 1). Check consensus and anything else scheduled the same day |
| T−2h | Measure every bot on BTC, ETH and correlated pairs against the expected move (step 2). Edit or pause what fails |
| T−15m | Don't touch anything. Changes made now fill into the thinnest book of the day |
| T+0 to T+15m | Watch. Stop losses may fill by market below their trigger; that's the cost of the release minute |
| T+1h | Spreads usually normalize. Restart paused bots from History if the range still makes sense at the new price |
| T+1 day | Review which bots survived and why. Adjust default ranges and leverage for the next release |
What September 16 looked like for each bot
Bitcoin slid into the decision, touched about $75,242, slipped below $75,000 and recovered toward $76,000.3
- A grid with its lower edge around $76,000 left the range on the way down, and a grid with an edge at $74,000 kept trading the swing.
- A long DCA bot filled safety orders on the slide; the ones with reserve left caught the recovery toward $76,000.
- A long COMBO with liquidation inside the drop was closed by it. That's where the $571 million came from.
None of these outcomes depended on predicting the hike. They depended on where the edges and the liquidation price sat relative to a move the market had signalled in advance.
The 10-minute pre-release check
- Which of my bots run on BTC, ETH or pairs that move with them?
- Is the forecast stable or split? Is other news stacked on the same day?
- For each grid: are the range edges more than two expected moves away?
- For each DCA bot: how many safety orders would a 3% flush fill?
- For each futures bot: how far is liquidation, in expected moves?
- Do my stop losses sit where a thin book could fill them badly?
- If I pause, when exactly will I restart, and what has to be true first?
Run your setup through the October 14 CPI in demo first. Launch the same grid, DCA or COMBO settings on a demo account before the release and watch what it does to the range, the ladder and the liquidation price. The 7-day Bitsgap PRO trial needs no card, and demo bots trade live prices with no funds at risk.
Frequently asked questions
Should I stop my trading bot before CPI or FOMC? Not by default. If your grid edges and liquidation price sit more than two expected moves away and the forecast is stable, the bot can run through the release. Adjust or pause when they sit closer, expectations are split, or other news lands the same day.
What time is the US CPI release in UTC? CPI is published at 8:30 a.m. Eastern. That's 12:30 UTC for the October 14, 2026 release and 13:30 UTC for November 10 and December 10, after US clocks change on November 1.
When are the next FOMC decisions in 2026? October 28 and December 9, 2026, with the statement at 2:00 p.m. Eastern (18:00 and 19:00 UTC) and a press conference 30 minutes later. The December meeting includes updated economic projections and the dot plot.
What is the expected move? The options market's estimate of how far an asset may move on a given day, published as a percentage before scheduled releases. Before the August 2026 CPI print, it was about 1.3% for Bitcoin. Real moves can exceed it.
How do macro releases affect a grid bot? A large move can push price out of the grid range in one candle, after which the bot stops trading on one side. Spreads also widen around the release, so a stop loss filled by market order can execute below its trigger price.
Why do DCA bots fill safety orders before news? Traders often cut leverage in the hours before a release, which pushes price down without new information. A DCA bot can't tell that flush from a real repricing and fills safety orders into it, using capital it might need if the release moves price further.
Is it safe to run a leveraged bot through FOMC? Only if the liquidation price sits well beyond the expected move. On September 16, 2026, more than $571 million in long positions were liquidated around the Fed's rate hike. Lower leverage, more margin or closing before the release all buy room.
