A spike with no news is almost always a liquidity event rather than an information event. A single order larger than the resting depth walks the book, stops trigger into the thin space behind it, and price travels much further than the size alone justifies. The news, if any, comes later.
The instinct when a chart moves two percent in nine seconds is to go looking for the reason. Most of the time there is no reason in the sense you are hunting for. There is a mechanism, and the mechanism is more useful than the headline would have been.
The default explanation: the book was thinner than the order
An order book is a ladder of resting limit orders. At any given moment there is a finite quantity available at the best bid, a finite quantity one tick lower, and so on down. Those quantities are small — usually far smaller than traders assume from looking at a daily volume figure.
When an order arrives that is larger than the quantity resting at the touch, it does not wait. It fills at the best price, then the next, then the next, consuming each level until it is done. That is the entire spike: one participant needing size in a book that did not have it.
Two things then amplify it. Stop orders resting in the space the price just crossed convert into market orders and push in the same direction. And the firms quoting the market widen or pull their quotes, because a sudden one-sided move is exactly the signal that somebody might know something — the adverse-selection defence described in how market makers make money. Depth disappears at the moment it is most needed, so the same size moves price further than it would have ten seconds earlier.
Six real causes, in rough order of how often you will meet them
- A large order meeting a thin book. The default. A fund needs to be out of a position by the close, or an algorithm miscalibrates its participation rate for the current volume.
- A stop cascade. The initial push reaches a price where a cluster of stops sits — under a round number, under yesterday’s low, under the session low. Those stops become market orders, which trigger the next cluster. This is the honest, unglamorous version of the story told in liquidity grabs and stop hunts.
- Index and ETF mechanics. A basket trade in an ETF has to be hedged in the underlying constituents. A single large ETF order can therefore print a simultaneous, newsless move in dozens of individual names at once.
- Correlation, not causation. Nothing happened to your instrument. Something happened to the dollar, to oil, to the index, or to a peer in the same sector, and your chart is reflecting it. Always check the sector and the index before you check the newswire — see correlation risk in trading.
- News you have not seen yet. An SEC filing, a regulatory decision, a headline on a terminal you do not pay for, or a story breaking in another language. The market reacts in seconds; the consumer-grade news feed catches up in minutes. Absence of news on your screen is not absence of news.
- An error. A fat finger, a misconfigured algorithm, a bad price feed. Rare, and usually followed by a broken-trade notice or a halt, which is its own kind of confirmation — see trading halts explained.
Notice that only one of those six involves information about the company. Five are structural. That ratio is the reason the reflex to hunt for a headline is usually wasted effort.
What the regulator’s own guardrail tells you
US equity markets have a formal mechanism built specifically because sudden, unexplained moves happen often enough to need one. Under the Limit Up-Limit Down plan, trades cannot occur outside a price band set around a rolling average of the security’s recent price. In the SEC announcement of the proposal, the Commission describes bands of 5 percent for the more heavily traded securities and 10 percent for the rest, states that “the percentage bands would be doubled during the opening and closing periods,” and specifies a five-minute trading pause if trading cannot occur within the band for more than 15 seconds.
Three things follow from that for a trader, and none of them are obvious:
- The bands are wider at the open and the close. The regulator itself expects more violent, less explicable movement in those windows. A spike at 9:33 a.m. deserves less of your attention than the same spike at 11:30 a.m.
- Fifteen seconds is the threshold. The plan is built around the idea that a genuine repricing re-establishes a two-sided market almost immediately, while a mechanical one does not. That is a useful instinct to borrow.
- A pause is information. If the move was large enough to trip a pause, it was large enough that the exchange itself could not find a balancing side. Reopening after a pause is a fresh auction, not a continuation.
How to tell a liquidity spike from a real move
| Check | Liquidity event | Information event |
|---|---|---|
| Breadth | The name moved alone; peers and the index are flat | The sector or the whole tape moved with it |
| Volume after the spike | Collapses back to normal within a minute or two | Stays elevated for the rest of the session |
| The candle close | Long wick, close back near the origin | Closes at or near the extreme and holds there |
| The retest | Price returns through the whole move without resistance | The new level is defended when price comes back to it |
| The spread | Normalises almost immediately | Stays wider than usual as risk is repriced |
None of these is decisive alone. Together they are usually enough to place a move in one bucket or the other within about two minutes — which is exactly the amount of time most traders do not give themselves.
What to do in the sixty seconds after
The worst moment to act is the moment it happens. During the spike the spread is at its widest, depth is at its lowest and your slippage is at its worst; the identical trade idea costs meaningfully more to express now than it will shortly. A useful sequence:
- Check breadth first, news second. One glance at the index and a peer answers the question faster than a newswire will.
- Do nothing until the candle closes. The close is the cheapest available filter between a mechanical wick and a real repricing.
- If you are already in the trade, let the plan work. A spike is precisely the scenario your stop was defined for. Widening it mid-move is how a managed loss becomes an unmanaged one.
- If you are flat, wait for the retest. A level created by a spike has to prove it exists. If it is real, it will still be there in five minutes and it will hold when price returns to it — the logic behind what a retest actually is.
Frequently Asked Questions
Why did the price spike when there is no news?
Usually because an order arrived that was larger than the orders resting in the book. It consumes every available price level on its way through, and stops sitting in the thin space behind it trigger as market orders, extending the move. No information changed hands. Only the balance of size did.
How can I tell a liquidity spike from a real move?
Check three things. Did the rest of the sector or index move with it, or was it alone? Did volume stay elevated after the spike, or collapse straight back? And did price still hold the new level when the candle closed? A move that is alone, unaccompanied by follow-on volume, and does not hold is almost always mechanical.
What is a trading pause and when does it trigger?
Under the Limit Up-Limit Down plan, US exchanges set price bands around a rolling reference price and trades cannot occur outside them. If the market cannot trade back inside the band for more than 15 seconds, a five-minute trading pause is triggered. The bands are wider during the opening and closing periods.
Should I trade a spike I cannot explain?
Not at the moment it happens. During the spike the spread is at its widest, the book is at its thinnest and your fill quality is at its worst, so the same idea costs more and risks more than it will in five minutes. Waiting for the move to be confirmed or rejected costs you the worst part of the move and very little else.
Bottom line
Price does not need news to move. It needs an imbalance between an arriving order and a resting book, and that imbalance occurs many times a day in every liquid market. The regulatory architecture assumes as much: the Limit Up-Limit Down bands exist precisely because sudden moves happen, and they are deliberately doubled at the open and the close where the book is least reliable. For a trader the practical residue is short. Check breadth before you check headlines, treat the size of the candle as a reading of liquidity rather than conviction, and let the candle close before you decide anything. Our FAQ explains how that discipline shows up in the room day to day.