Crypto Scalping With Orderflow: The Realistic Guide
A round trip at base taker fees costs 0.09% of notional, so 20 scalps a day means paying 1.8% in fees before being right once. Here is the fee math, the three orderflow setups that actually work at scalp timeframes, and the risk rules that keep you in the game.
$ Stop reading delayed data. Compare live order book depth across 5 exchanges right now.
Launch Free Terminal →A round trip on Hyperliquid at base taker fees costs 0.09% of notional. A scalper taking 20 trades a day pays 1.8% of notional in fees before being right or wrong a single time. That one number ends more scalping careers than any bad entry, and it is the correct place to start any honest crypto scalping strategy.
The Fee Math Every Scalper Ignores
Base perp fees on Hyperliquid are 0.045% taker and 0.015% maker, with tiers that drop as 14-day volume grows. Enter and exit with market orders and each trade costs 0.09% plus slippage. If your average winner is 0.3%, fees alone consume a third of it.
The fix is structural, not motivational. Enter passively with limit orders wherever the setup allows, which cuts the round trip toward 0.03% to 0.06%, and let volume tiers do the rest. A scalping edge that only exists at maker fees is still an edge. One that disappears after fees was never there.
What Scalping Actually Is, and Is Not
Scalping means holding for seconds to minutes, targeting moves of roughly 0.2% to 0.6%, and going flat when the session ends. It is not taking 60 impulsive trades because the chart is moving. Trade count is an output of how many valid setups appeared, never a goal.
At this timeframe, lagging indicators are a handicap. A 9-period EMA on the 1-minute chart is a summary of the last nine minutes. The tape, the order book and CVD are a summary of right now, which is the only window a scalper is paid to read.
Three Orderflow Setups That Work at Scalp Timeframes
The first is the CVD burst with book confirmation. Aggressive delta spikes in one direction while the opposite side of the book thins out, visible as an order book imbalance shift. That combination, aggression plus vanishing resistance, is what a real micro-breakout looks like before the candle finishes forming.
The second is the absorption fade. Heavy sell volume prints into a level and price refuses to make a new low: someone is passively eating the aggression. Fading exhausted aggressors against confirmed absorption is one of the highest win-rate patterns in orderflow, precisely because the crowd on the wrong side becomes your fuel.
The third is the liquidation cascade. When price sweeps into a dense liquidation cluster, forced market orders create a burst of one-directional flow that either accelerates or exhausts. The scalp is the exhaustion print: cascade volume climaxes, delta flips, and the snap-back is fast because the forced sellers are simply gone.
When to Scalp: Liquidity Windows Matter
Scalping lives on tight spreads and thick books. The London-New York overlap and the first two hours of the US session concentrate the day's volume on majors. Dead Asia hours on low-cap perps combine wide spreads with thin books, which converts small mistakes into large ones.
Holiday weekends deserve their own warning. Thin books exaggerate every wick, stops get run further than usual, and the same setup carries more slippage in both directions. Fewer trades, smaller size, or no trades at all is a position too.
Risk Rules That Keep You Alive
Fix the risk per trade at 0.5% to 1% of equity and honor a hard daily stop: three consecutive losers or a 2% daily drawdown means the session is over. Scalping losses cluster, because the conditions that produced one loser usually produce the next three, and a daily stop is how you refuse to donate to that cluster.
The tooling requirement is real-time by definition. The Buildix deep view (buildix.trade/pair/SOL) runs 1-minute CVD, footprint, whale prints and the liquidation map on one screen, and the free screener surfaces which of the 530+ pairs are actually moving right now. Scalping is not about predicting the next hour. It is about reading the next sixty seconds slightly faster than the people on the other side of your fill.