The best indicators for scalping are three: the 9 and 21 EMA as a trend filter, RSI 14 as a momentum trigger, and session VWAP as a volume confirmation. More than three on a 1-minute or 5-minute chart adds delay, and delay kills a scalp.

One setup you can paper-trade tomorrow: exact M1/M5 settings, cost math, numbered rules, and a worked example.

In this guide:

  • Cost math: why a 1-pip spread eats 20% of a 5-pip target
  • The Core 3: what each indicator does, its exact settings, and what it cannot do
  • M1 vs M5 settings and the rule for removing an indicator
  • A worked example, when not to trade, risk rules, and five mistakes

The best indicators for scalping are the 9 and 21 EMA, RSI 14, and session VWAP. The EMAs filter trend direction, RSI 14 confirms momentum through its 50-cross, and VWAP confirms volume: long only above it, short below it. Use M5 for structure and M1 for entries, with 5–6 pip stops and targets of at least 1.5 times the stop.

Key takeaways

  • Three indicators only: EMA 9/21 (trend), RSI 14 (momentum), session VWAP (volume). One per job.
  • On a 5-pip target, 1 pip of cost is 20% before slippage. Above 1 pip, do not scalp 5-pip targets.
  • At 1:1.5 gross with 1 pip of cost, break-even is about 48%. At 1:1, it is 60%.
  • RSI is a momentum filter: trade the 50-cross with the trend, ignore 70/30.
  • Session VWAP is the veto: no long below it, no short above it.

What scalping indicators are — and what they are NOT

A scalping indicator is a filter: direction, momentum, volume. Indicators do not predict price; all are computed from past prices, so signals arrive after the move starts.

What they are NOT: a strategy, a win-rate guarantee, or a substitute for cost control. No indicator combination makes a 5-pip target profitable at a 2-pip spread. ESMA disclosures show 70–85% of retail CFD accounts lose money, usually from costs and overtrading. This article gives you no win rate — no honest source can give you one for your broker and execution. It gives you the formula for the win rate you would need, so you can measure your own results against it.

The top guides fail here: Piprider lists 12 indicators with no cost table or worked example; MQL5 has no cost math or VWAP; Headway gives no testable rule; WR Trading lists five bullets with no settings.

The cost math first

Total cost = spread + commission + slippage, in pips. On EUR/USD in the London–New York overlap (12:00–16:00 UTC), a raw-spread account costs about 0.5–1.0 pips; a standard account, 1.0–1.5 pips. Outside liquid hours it can be 2 pips or more. Measure yours over 20 trades.

TargetCost 0.5 pipCost 1.0 pipCost 1.5 pips
5 pips10%20%30%
8 pips6%13%19%

Rule: never take a scalp whose target is less than five times your total cost. A 5-pip target needs cost at or below 1 pip.

Cost is paid win or lose: net win = target − cost, net loss = stop + cost, break-even = 1 ÷ (1 + net win ÷ net loss).

Gross setupStop / targetCostNet win / net lossNet RBreak-even win rate
1:15 / 5 pips1.0 pip4.0 / 6.00.6760%
1:1.55 / 7.5 pips1.0 pip6.5 / 6.01.0848%
1:1.55 / 7.5 pips1.5 pips6.0 / 6.50.9252%

A 1:1 scalp at 1 pip of cost needs 60% just to break even — which is why this guide targets 1:1.5. Raising the target from 5 to 7.5 pips drops it to 48%: the target is the lever, not the entry.

The Core 3 setup: timeframe hierarchy and exact settings

One indicator per job. No two indicators may do the same job — that is how you get conflicting signals and freeze on M1.

Timeframe hierarchy: M15 sets the bias, M5 the structure, M1 the trigger. Trade only the M15 direction, on the right side of VWAP, entering on M1. Never take an M1 signal against the M15 bias or VWAP.

Pairs and sessions: EUR/USD and GBP/USD, in London (07:00–10:00 UTC) or the London–New York overlap (12:00–16:00 UTC). Outside those windows, spreads widen and the Core 3 generates noise. XAU/USD needs wider stops and smaller size — see XAUUSD scalping strategy.

Indicator 1: EMA 9 and EMA 21 — the trend filter

Settings: 9 and 21 EMA on M15, M5, and M1, on close prices.

What it does: long bias only when price is above both EMAs and the 9 is above the 21; short bias only on the reverse. A pullback to the 9 EMA that holds is the entry zone.

What it cannot do: tell you when a trend ends: in a range the EMAs tangle and every crossover loses. Flat EMAs on M5 mean "no trade".

Indicator 2: RSI 14 — the momentum trigger

Settings: RSI 14 on M1 and M5. Not RSI 7 — it doubles the noise on M1.

What it does: for a long, RSI on M1 must cross back above 50 after dipping toward 40–50; for a short, cross back below 50 after rising toward 50–60. The 50 level is the trigger; 70 and 30 are not entries.

What it cannot do: call tops and bottoms: in a trend RSI can sit above 70 for a dozen candles. Above 70 in a long bias means "do not short", never "sell now".

Indicator 3: Session VWAP — the volume confirmation and veto

Settings: session VWAP, resetting at the start of the trading day, one line only — no bands. Plot on M5, read from M1.

What it does: the volume-weighted average price paid so far today — the closest thing scalping has to a fair-value line. Above VWAP, long bias only; below, short bias only. A pullback to VWAP that holds is the best entry here.

What it cannot do: help in thin liquidity: in the Asian session VWAP flattens and price crosses it constantly. In spot forex it uses tick volume as a proxy — treat it as a bias line, never place a stop exactly on it.

M1 vs M5 settings

SettingM1 (trigger chart)M5 (structure chart)
Typical stop4–6 pips beyond structure6–10 pips beyond structure
Typical target5–8 pips (min 1:1.5 gross)8–12 pips (min 1:1.5 gross)

Beginners start on M5: fewer signals, wider stops, far fewer false entries. If you cannot follow the rules on M5 for 20 journaled trades, M1 will charge you faster. See Forex scalping for beginners.

The rest of the toolbox, compared

Bollinger Bands, MACD, and Stochastic each do one job well. None earns a fourth slot — each duplicates a Core 3 job.

IndicatorSettingsJobBest forVerdict
Bollinger Bands20, 2Volatility / range edgesRange scalpingUse instead of RSI only in a proven range; never fade a band without VWAP agreeing
MACD12, 26, 9 (or 5, 13, 3 fast)MomentumTrend scalpingA slower RSI: useful on M5, too lagging for M1; a backup, not a fourth line
Stochastic14, 3, 3Momentum / timing in rangesRange scalpingFaster than RSI and noisier; crosses fail in trends like RSI 70/30

Verdict: trend scalping, keep the Core 3. Range scalping, swap RSI for Bollinger Bands plus Stochastic — only after M5 confirms a range. Breakout scalping, keep the Core 3. Beginners, change nothing for 20 trades.

Entry, stop, and target rules

Every rule below is testable: after the trade, you can say yes or no to whether it fired. If any rule did not fire, journal it as a rule break, not a setup loss.

Long setup (short is the exact mirror):

  1. Bias (M15): price above the 21 EMA and the 9 EMA above the 21 EMA. If not, no long.
  2. Structure (M5): price above VWAP, last pullback held above the 9 EMA. Below VWAP, no long.
  3. Cost check: total cost at or below 1 pip, target (step 6) at least five times that cost.
  4. Pullback (M1): price pulls back to the 9 EMA or VWAP, RSI dipping toward 40–50, no close below the M5 21 EMA.
  5. Trigger (M1): enter long on the first M1 candle closing back above the 9 EMA with RSI crossing back above 50. Enter on the close, never mid-candle.
  6. Stop: 1–2 pips beyond the pullback low or the M5 21 EMA. Typical: 5–6 pips. Wider than 10, skip.
  7. Target: 1.5 times the stop, at the next M5 level. Typical: 7.5–9 pips. Take half at 1R, move the stop to breakeven minus cost. Never move the target further away.
  8. Invalidation: exit before the stop if M1 closes back below VWAP or RSI closes below 40. A setup that loses confirmation is closed, not hoped back.

The removal rule: after every 20 journaled trades, ask of each indicator: did it change a decision? If not, remove it. Never add a fourth indicator to fix a losing week — the fix is in the cost table, the session filter, or the journal.

Worked example (example — not live prices)

Assumptions: EUR/USD in the London–New York overlap, $5,000 account, 0.5% risk, 1.0 pip cost, M5 bias long. Prices are illustrative, not live.

  • Entry: 1.0850, M1 close back above the 9 EMA, RSI crossing above 50 after a pullback to 1.0846.
  • Stop: 1.0844, 6 pips from entry.
  • Target: 1.0859, 9 pips, at the prior M5 high. Gross R 1.5.
  • Net math: win = 9 − 1 = 8; loss = 6 + 1 = 7; net R = 8 ÷ 7 = 1.14; break-even 47%.
  • Position size: budget = $5,000 × 0.5% = $25. Lots = $25 ÷ (6 × $10) = 0.41.
  • Outcome: win +$32.80, loss −$28.70. The stop set the size.

At 1.5 pips of spread, break-even would be 50% on the same trade. That is why the cost check comes before the entry.

When NOT to trade

  • High-impact news: 2 minutes before to 5 minutes after CPI, NFP, and FOMC. Spreads widen and indicators lag the move.
  • Thin liquidity: Asian-only hours, holidays, 30 minutes before London. VWAP is flat, EMA crossovers are noise.
  • Spread filter breached: cost above 1 pip for a 5–8 pip target, or above 1.5 pips for 8–12 pips.
  • After the daily loss limit: two losses in a row is a pause; three ends the session.
  • Bias and VWAP disagree: M15 says long but price is below VWAP. No decision from the market means no trade from you.

Risk management

  • Risk per trade: 0.25–0.5% of equity. Never 1%: at 20 trades a day that is 20% of the account in one session.
  • Daily loss limit: 2–3% of equity, hard stop. Hit it, close the platform.
  • Three-loss pause: after three straight losses, stop 30 minutes and review the journal — conditions usually changed, not the setup.
  • Maximum 3–5 trades per session. M1 offers 30 signals a day; 30 trades at 1 pip is 30 pips of cost before a win or loss.
  • One pair, one session. EUR/USD in the overlap until 20 journaled Core 3 trades. See EUR/USD scalping strategy: London open.

Common mistakes

  1. Stacking six or more indicators. Every extra indicator adds a veto; conflicting vetoes mean you never trade. Fix: run the Core 3 only.
  2. Selling RSI above 70 in a trend. Extremes measure momentum strength, not reversal timing. Fix: 70/30 are never entries; only the 50-cross counts.
  3. Ignoring the spread filter. A 5-pip target at 1.5 pips gives up 30% before slippage. Fix: check cost first (rule 3).
  4. Entering mid-candle. A perfect RSI cross can vanish by the close. Fix: enter on the close only.
  5. Widening the stop. Moving a stop turns a 6-pip loss into a 15-pip loss. Fix: never move the stop away from entry — only toward breakeven after half is taken at 1R.

FAQ

What is the single best indicator for scalping? There is none. A trend filter alone gets chopped in ranges, a momentum trigger alone trades against trends, and a volume line alone says nothing about timing. The smallest stack covering all three jobs is EMA for direction, RSI for timing, VWAP for confirmation.

What are the best indicator settings for 1-minute scalping? EMA 9 and 21, RSI 14, session VWAP. On M1, trigger on the RSI 50-cross, stops 4–6 pips, targets 5–8 pips. Do not shorten RSI to 7 — it adds noise faster than entries. M15 bias, M5 structure, M1 trigger.

Can I scalp with RSI alone? You can, but you will trade every overbought and oversold reading, including inside strong trends where RSI stays extreme. RSI alone cannot tell a pullback from a reversal. Add the EMA 9/21 bias filter first.

Do these indicators work on gold (XAU/USD)? The logic transfers, the numbers do not. Gold moves in dollars per ounce: a 6-pip stop is meaningless, spreads are wider, and the London sweep matters more than any setting. Use the same Core 3 roles with wider stops, smaller size, and the rules in XAUUSD scalping strategy.

Why do my scalping indicators work in backtesting and fail live? Usually costs and execution. Backtests often assume the mid-price; live, you pay spread plus commission plus slippage on every trade — 20–30% of a 5-pip target. Re-run any backtest with your measured cost added, and demo-trade for two weeks first.

Sources consulted

Educational material only — not financial advice. Forex and gold trading with leverage can produce substantial losses. Backtest and demo-trade any setup before risking real capital.