Range Trading: A Strategy for a Sideways Market

Range trading means working off the edges of a price channel: buying near the lower zone, selling near the upper one, and exiting around the middle or at the opposite edge. You don't need to guess direction here. You need to know where the boundary is, where the stop goes, and at what point the range counts as broken.

Picture a training scenario. Bitcoin has spent three weeks moving between two prices, breakout traders keep getting caught in fakeouts (false pushes beyond a level with a quick return), and trend strategies bleed the deposit through fees. A sideways trend in crypto is the normal state of the market between impulses, and flat-market trading there comes down to discipline.

The examples below are for training, prices are hypothetical, fees and slippage are built into the math, and thresholds like "three touches" are starting parameters for your own testing.

What a Range (Sideways Market) Is

A range (sideways market, flat) is a section of the chart where price reverses several times from the same zones above and below and can't hold beyond them. There's no directional move, just swings inside the channel.

Formally, a channel is drawn from two touches of each boundary, but that's only a hypothesis. A third touch is already an argument that there's interest at the edges: a seller defends the top, a buyer absorbs the bottom. An argument, not proof. The fourth touch may well turn into a breakout.

Where does a sideways market come from? Most often it's a pause after an impulse: some traders lock in profit, others aren't ready to pay more yet.

In a range, the rules of the game change.

The same "buy off the level" trade can have a different expected value in a trend and in a flat market, so first we define the market regime and only then pick a strategy.

A quick regime filter is Bollinger Bands (a moving average and two bands, by default two standard deviations away from it). The bands run almost horizontally, price moves from one to the other and doesn't close beyond them for a series of candles - a typical sideways picture. Indicator settings are covered in the article “Bollinger Bands: How to Use Them in Crypto”.

How to Define the Boundaries

A range boundary is a zone, not a line. Price rarely reverses to the exact tick: it pierces the level, collects stops and comes back, and the depth of the pierce depends on the coin and on volatility.

We mark it up in two ways and look for where they line up.

Horizontal Levels

The classic approach. Find the extremes price reversed from and draw horizontal zones through them. The logic of drawing them is covered in the article “Support and Resistance Levels: How to Draw and Trade Them”.

Markup order.

• Take a timeframe noticeably higher than your working one. For scalping on the 1-minute chart, that's usually 15m or 1h
• Mark the highs and lows that produced a reversal with a meaningful move. Training benchmark - at least a third of the width of the future channel
• Build the zones from candle bodies to the outermost wicks of past touches
• Count the touches. No boundary has held three times? The range is still a hypothesis
• Check the width in percent. A narrow channel may not cover costs, the math is below

Pay separate attention to the pierce. Price went beyond the boundary with a wick and came back fast? The boundary is still holding, and some of the stops behind it have probably been taken out already. A candle closed beyond the zone, and the next one did too - that's a different story.

Volume Profile (POC, VAH, VAL)

Volume Profile shows how much volume was executed at each price over a chosen period. Executed, specifically: it's the sum of trades that already happened, not the orders sitting in the order book right now. The profile tells you about the past, the order book about current intentions, which can change in a second.

Profile markers for a range:

• POC (Point of Control) - the price with the highest executed volume for the period. Inside a range, price often returns to this level
• VAH and VAL (Value Area High and Low) - the upper and lower boundaries of the value area

The value area is the range around the POC where most of the period's volume traded. The default is most often 70%, but that's a convention, and the percentage can be changed in the settings.

Next, build a fixed range profile from the start of the range and compare it with the horizontal markup.

A caveat. VAH and VAL aren't guaranteed range boundaries, they're statistics on how volume is distributed. Price can pass through them with no reaction, especially when the profile is built on a short stretch.

A fixed range profile isn't available on every platform. An additional tool is clusters, meaning the distribution of volume by price inside a single candle. The POC of each bar shows the price with the highest volume inside that candle. Reading the profile is covered in the article “Volume Profile: How to Use It in Crypto”.

If the order book, clusters and tape still blur into one big noise, start with the free lesson from the crypto trading course on the Secret Terminal YouTube channel. The lesson is part of a full five-part course for beginners.

Strategy: Bouncing Off the Boundaries

Range trading comes down to simple logic: buy near the lower boundary, sell near the upper one, and once price holds beyond a boundary, admit that the range is gone.

Sounds basic. People blow up on the details.

Entry, Stop, Target

Entry. Not a limit order placed at the boundary in advance, but after a reaction: price has stalled at the zone, aggressive selling on the tape (the real-time stream of executed trades) has dried up, and buying has appeared. A limit order "on the touch" gets a better price, but in a real breakout it opens your position exactly when the seller is stronger.

Stop. Beyond the zone: the extreme of the pierce plus a buffer. A large density level in the order book (a big limit order or a cluster of orders at one price) is good cover for a stop, but it can be pulled at any moment. When triggered, a stop-market sends a market order; the fill price and completeness depend on available liquidity. In a thin order book, slippage (the difference between the stop price and the actual fill) will increase the loss, and a stop-limit may not fill at all.

Target. The first part of the position is closed near the POC or the middle of the channel, the second near the opposite boundary with some room to spare, not right at the edge where the other side's orders are already waiting.

Training example. BTC/USDT, perpetual futures, hypothetical prices.

Range 64,000 - 66,000, width 2,000 (about 3.1%), POC at 65,000. The last pierce of the lower boundary reached 63,900. Deposit 10,000 USDT, risk per trade 0.5%, i.e. 50 USDT including costs.

• Long entry at market after the reaction, estimated fill price 64,150
• Stop-market at 63,780, below the extreme of the pierce. Slippage allowance 20 USDT per 1 BTC, estimated exit 63,760. This is an allowance for the calculation, there are no statistics behind it
• Targets as limit orders: half at 65,000 (POC), half at 65,800, 200 below the upper boundary
• Fees are hypothetical, taker 0.05% and maker 0.02%. Real rates depend on the exchange and account tier, check them on your own platform, for example on Binance

I size the position from the full cost of the stop: distance, both fees and the slippage allowance.

Risk per 1 BTC = (64,150 - 63,760) + 64,150 × 0.05% + 63,760 × 0.05% = 390 + 32.08 + 31.88 = 453.96 USDT.

Size = 50 / 453.96 = 0.1101, rounded down to 0.110 BTC. Position notional 0.110 × 64,150 ≈ 7,057 USDT, about 0.7 of the deposit, so the calculation doesn't require a notional above the deposit. Funding isn't included here; actual slippage may exceed the built-in allowance.

The "straight-line" calculation, based only on the 370 distance to the stop, gives 50 / 370 = 0.135 BTC. If stopped out, that position would lose 0.135 × 453.96 ≈ 61.3 USDT, 23% more than planned.

Net profit per 1 BTC at the first target: 850 of price movement - 32.08 for the entry - 13.00 maker fee = 804.92 USDT. At the second: 1,650 - 32.08 - 13.16 = 1,604.76 USDT. I build in the maker rate with a caveat: it only applies if the limit order filled passively.

With both targets hit, reward to risk after costs is about 2.7 to 1. On paper, without fees and slippage, it would be almost 3.4 to 1.

A separate cost line on perpetual futures is funding, a periodic payment between longs and shorts determined by the funding rate. Only those holding an open position at settlement time pay or receive it, and the amount is calculated from the position value at the mark price. At a +0.01% rate, a long with a notional of about 7,057 USDT will pay roughly 0.71 USDT, at +0.1% already about 7 USDT, which is 14% of the planned risk.

On Binance the base interval is 8 hours, but some contracts have a shorter one. On Bybit the interval also depends on the pair, so check the settlement time in the contract specification.

Why does flat-market trading in a narrow channel often not pay off? In the table I take the target conditionally as 60% of the width: entries and exits rarely land at the very edges.

In a 0.4% channel, more than 40% of the target goes to the exchange before slippage even kicks in. A limit entry after the reaction, filled passively, lowers costs, but in a channel like that the stop beyond the zone is often comparable to the target.

Order Book Confirmation

The chart shows where the boundary is. The order book and the tape show whether someone is defending it right now.

What we look for at the lower boundary before a long:

• A density level in the order book at the boundary or slightly below it, not flickering but sitting there for a long time. Order lifetime gives context but doesn't prove intent: even an order that sat for half an hour can be pulled
• The tape's reaction. Aggressive selling hits the density level, and it doesn't get eaten or it refills. Then the tape goes quiet and market buys appear
• A footprint in the cluster: large volume traded on the touch, yet price didn't go lower. It looks like absorption, when a limit buyer takes on the selling

And what should make you cautious.

• The density level at the boundary is being eaten in chunks, the order size shrinks and doesn't recover
• The order book is thin right past the boundary. There are no visible orders there capable of slowing a pierce, and the stop may slip badly
• The order is moved lower or pulled as price approaches, without being filled. Whether it's spoofing (an order that was never meant to fill) or the owner changed their mind doesn't matter: the support is gone

A POC or a large volume node near the lower boundary says there was a lot of trading there in the past. A density level at the same price shows that orders are sitting there right now, though they can be pulled. The overlap adds context, but the edge of this combination has to be tested on data; a bounce isn't guaranteed.

In Secret Terminal, the order lifetime timer sits in the order book next to a large density level, and the density levels themselves are drawn on the chart as horizontal lines.

When the Range Ends

Every range breaks sooner or later. The bounce strategy only makes sense while the range is alive, and the biggest losses usually come on the exit from it.

Signs the sideways market is weakening:

• Touches of one boundary become more frequent, and pullbacks from it get shorter. Price is pressing into one edge
• Lows are rising or highs are falling, and price is squeezing up against one boundary
• The density level that held the boundary for a long time shrinks or gets pulled
• Bollinger Bands are tighter than at any point during the range. A squeeze often precedes a volatility expansion but doesn't show direction
• The tape speeds up on the approach to the boundary, although it used to fade

Two or three signs at once are a reason not to open new trades off that edge. The threshold is conditional, and no single sign proves a breakout on its own.

How do you tell a fakeout from a real exit? In advance and with certainty - you can't. A fakeout usually pierces the boundary with a wick and brings price back into the channel within one or two candles. Candle closes beyond the zone, volume above the range average and a sluggish retest from the other side of the boundary point to a real exit, but even after that set, price can return to the channel.

When it doesn't work. A hypothetical continuation of the example. The fourth approach to 64,000 within a day, pullbacks keep getting shorter, the density level at the boundary is being eaten by market selling and doesn't recover. A 15m candle closes at 63,850, the next one is also below the zone. The scenario is cancelled, we close the long manually around 63,850: (300 + 32.08 + 31.93) × 0.110 ≈ 40 USDT loss. Had price reached 63,780 first, the stop-market would have triggered, about 49.9 USDT at the estimated slippage and more in a thin order book. Both outcomes are part of the plan. The mistake would be to re-enter the long "at a good price" or to average down.

Price has held beyond the boundary and there's no stop? Close manually. Flipping in the direction of the breakout is a different strategy with its own entry rules, covered in the article on the breakout strategy

The Mistakes That Cost the Most

Trading a range that isn't there. Two touches and an urge to trade don't make a range. Confirmation first, trades second.

Placing the stop right behind the line. A stop like that can get hit on a small pierce, after which price often heads to your target without you.

Ignoring costs. In the example above, fees and slippage turn 3.4 to 1 on paper into 2.7 to 1 in the account, and in a 0.4% channel two taker fees eat more than 40% of the target.

Averaging down on a breakout. "It'll come back into the range any second" is the most expensive phrase in a flat market. Holding beyond the boundary cancels the scenario entirely.

How to read limit orders in the order book and choose entry points and stop levels is covered in the free lesson of the same course.

FAQ

What is range trading in simple terms?

It's trading off the boundaries of a price channel, betting on price returning inside the range. A long is opened near the lower zone, a short near the upper one, the stop goes beyond the zone, and the target is around the middle or at the opposite edge. Price holding beyond a boundary cancels the scenario.

What timeframe is best for range trading?

There's no universal one. In a training setup, the boundaries are marked on 15m-1h and the entry is found on 1m-5m: the higher timeframe filters out noisy mini-ranges, the lower one gives the entry point and a tight stop. Test the combination on your own instrument.

How many touches does it take to consider a sideways market confirmed?

In the training framework, at least three touches of one boundary and two of the other. Fewer than that is a hypothesis, and the chance that you're looking at a pause in a trend is still significant. Even five touches don't guarantee a sixth, and each new one can become a breakout.

Is a sideways trend in crypto easier to trade on Bitcoin or on altcoins?

Ranges form on any coin, the difference is in execution. On liquid pairs like BTC/USDT and ETH/USDT the order book is deeper, and a stop usually fills closer to the estimated price. On low-volume altcoins the order book is thin, pierces are deeper and slippage on the stop is bigger.

Can you trade a range using Volume Profile alone?

You can, but it's a weak version of the strategy. The profile only shows executed volume from a past period, current orders aren't in it. Without confirmation from the order book and the tape, buying at VAL can easily turn out to be buying at the start of a breakout.

The range boundary and its defense on one screen. Secret Terminal plots auto-levels and untested "old levels" on the chart, projects order book density levels onto the same chart, shows a lifetime timer for large orders, a tape with a small-trade filter and clusters with POC highlighting. The funding rate with a countdown to settlement shows when a charge or credit may hit. The terminal is free, works with Binance, Bybit, OKX, MEXC and WhiteBIT, and data is stored locally.