A trading robot can react quickly without obtaining the price you expected. During a sharp move, the important question is not simply whether a strategy sends an order. It is what happens between the signal, the exchange receiving that order and the trade being filled. Prices, available liquidity and account conditions can change at each step. Understanding that sequence helps you decide which controls to check before leaving automation running.
A market can keep moving while you are unavailable
Crypto markets generally trade throughout the day and night. There is no single global closing bell that puts every exchange, trading pair and automated strategy on hold. Individual services can nevertheless pause for maintenance, suffer an outage or temporarily restrict a particular operation. A market being open does not mean your own route to it is working.
That distinction matters when news arrives while you are asleep or unable to monitor an account. Your strategy may continue submitting orders even though your ability to review them has changed. Check the operating hours and maintenance arrangements of the actual broker or exchange, rather than assuming the whole process is continuously available.
Prepare the human side too. Know where to see open positions, pending orders and recent fills. Keep the official support and service-status routes accessible. Discovering those controls during a fast move leaves less time to understand what the account is doing.
Liquidity changes with the venue and the moment
It is tempting to describe weekends or overnight trading as uniformly thin. That is too broad. Liquidity depends on the asset, venue, trading pair and observation period. The useful check is whether enough buying and selling interest exists near the current price for the order you intend to send.
An order book is a changing collection of offers, not a reservation of prices for your strategy. Orders can be executed or withdrawn before yours arrives. If nearby liquidity disappears, a market order may meet several price levels, or encounter the venue's execution protections. Coinbase's order documentation describes how market orders can fill in portions and how protections can limit execution.
Speed therefore changes the meaning of a displayed quote. The price visible when a signal was calculated may no longer be available when the order reaches the market. Evaluate the strategy's behaviour under that delay, rather than treating the signal price as a promised transaction price.
An order instruction decides which uncertainty you accept
A market order seeks execution against available liquidity. It does not normally set the maximum acceptable purchase price or minimum acceptable sale price in the way a limit order does. A limit order sets that boundary, but it can remain unfilled if the market moves away.
Neither instruction solves every problem. A fast exit may execute at an unfavourable price; a tightly limited exit may not execute at all. Partial fills add another possibility: some exposure closes while the rest remains. The strategy must interpret the actual result, not merely the fact that an order was sent.
Gemini's order documentation also distinguishes immediate-or-cancel and fill-or-kill instructions. Their different handling of incomplete execution illustrates why order settings matter. Before using any such feature, confirm that your actual provider supports it and learn what an incomplete or rejected order means for the next automated action.
A stop trigger is not a guaranteed exit
A stop is a condition for activating an order. It is not a promise that the resulting trade will occur at the trigger price. In a fast market, prices can move through that level before the order finds sufficient liquidity.
A stop-market instruction generally leads to a market order after triggering. A stop-limit instruction leads to a limit order. The second controls the acceptable execution boundary but can leave the position open. The first prioritises execution while retaining price uncertainty and any venue-specific restrictions.
Also check which price activates the stop. Kraken's derivatives documentation distinguishes reference prices such as last traded, mark and index prices. A chart showing one reference may not explain an order triggered by another. Read the product's rule before comparing a visible price with your order history. Several controls with similar names can behave differently across venues.
Leverage can turn a move into forced trading
With leverage, the account's collateral supports a larger exposure. An adverse move can reduce the account's equity until it no longer meets the product's maintenance requirement. The provider may then liquidate positions under its rules. Coinbase's derivatives risk disclosures warn that inadequate margin can lead to forced liquidation and loss of the amount committed.
When many leveraged positions face similar pressure, forced transactions can add to an already fast market. Connections between spot and derivatives trading mean the effect need not remain inside one contract. That is a possible transmission mechanism, not proof that every sharp spot move was caused by liquidations.
Perpetual contracts add funding payments between the two sides of the market. Kraken describes funding as a mechanism helping the contract track its reference market. Rates and settlement arrangements depend on the product. A price move and a funding charge can affect account conditions together, so assess both where your account actually uses perpetuals.
Volatility describes a window, not the next move
Realised volatility measures variation in past returns over a selected period. A short window and a longer window can tell different stories about the same asset. Sampling frequency and calculation method also matter. A number without those details is difficult to compare meaningfully.
Coin Metrics's volatility material illustrates the use of multiple observation windows. The practical lesson is to compare like with like. Do not interpret a change in a displayed measure before checking whether its underlying period or calculation changed.
A quiet historical window does not establish that the next hour will be quiet. A volatile window does not predict the direction of the next move. Use the measure to ask better questions about exposure and operational readiness, rather than treating it as a guarantee or an automatic instruction to trade.
Check what pause and stop actually do
Before leaving a strategy active, review existing positions, unfilled orders, exposure across related assets and the settings that govern new orders. Check whether notifications are supported, enabled and reaching you. A notification is useful only if you understand which event it reports and can access the account afterwards.
XTZ Swap provides a demo, dashboard and controls to pause or stop the robot. Those supplied features do not establish that pausing closes a position or cancels an order already accepted by a broker. Confirm each action's effect with the relevant provider and inspect the resulting account state.
If an order appears delayed, check its status before submitting another one. Repeated instructions can create exposure you did not intend. Use a demo to learn the interface, while recognising that a demonstration cannot establish how a live order will execute under stress. Crypto trading can lose the entire amount committed.
