Start from two timeframes, not five
The fastest way to ruin a day trading setup is hopping through six timeframes until one of them agrees with you. The working setup is two: a context timeframe that tells you the trend and where the levels sit, and an execution timeframe where the decision and the stop live. Pick the pair, then stay on it for a full week before changing anything.
Scalping
- Context timeframe
- 15 minutes
- Execution timeframe
- 1 minute
- Useful candle count
- 150 – 250
Day trading
- Context timeframe
- 4 hours
- Execution timeframe
- 15 minutes
- Useful candle count
- 200 – 300
Multi-day swing
- Context timeframe
- Daily
- Execution timeframe
- 1 hour
- Useful candle count
- 250 – 400
That last column matters more than it looks. A chart crammed with a thousand candles hides the structure of the session you are in; a chart showing forty makes every pullback look like a reversal. The range buttons above the chart (1D, 5D, 1M, 3M) move you between those windows in one click without changing the timeframe.
Studies: two on the candles, one below
Every study you add asks for attention, and attention is finite. The working rule: one or two averages on the candles to read trend, and a single study in the lower pane to read momentum. Anything beyond that hands you contradictory signals and delays the decision.
On the candles
- Suggested setting
- EMA 21 (+ EMA 50 when needed)
- What you read from it
- Whether price sits above or below the short trend
- When to switch it off
- In a tight range where they cross every second candle
Session reference
- Suggested setting
- Session VWAP
- What you read from it
- Volume-weighted average price since the day opened
- When to switch it off
- On the daily timeframe — one candle is already a whole day
Lower pane
- Suggested setting
- RSI 14 or MACD
- What you read from it
- Strength of the move, or its acceleration
- When to switch it off
- When both are open and you catch yourself reading the study instead of price
Market structure
- Suggested setting
- Support & resistance + patterns
- What you read from it
- Levels where past reversals clustered
- When to switch it off
- Leave these on — they sit closest to price itself
Levels: draw few, and leave them alone
- 1Open the context timeframe (four hours, say) and find the highest high and lowest low of the last fortnight.
- 2Draw a horizontal line at each with the horizontal-line tool in the side rail.
- 3Add a third level only if price has already reacted to it at least three times.
- 4Switch to the execution timeframe — the lines stay put, because they are anchored to price, not to the timeframe.
- 5On the level that matters most, press the bell button to turn it into an alert that reaches your phone and e-mail when it is crossed.
Three clear levels are read in a second; twelve levels make every price “near a line”. If you have to work out which line matters right now, there are too many.
A setup that still works on the phone
A small screen cannot carry three panes under price. On the phone keep exactly one lower pane (RSI, MACD, or volume) and lean on the range buttons instead of repeatedly pinching out. One finger drags the chart, two fingers zoom, and a long press on a drawing opens its options.
desktop phone ┌──────────────────────────┐ ┌──────────────┐ │ price + EMA21 + levels │ │ price + EMA21│ │ │ │ + levels │ ├──────────────────────────┤ ├──────────────┤ │ RSI 14 │ │ RSI 14 │ ├──────────────────────────┤ └──────────────┘ │ volume │ one pane only └──────────────────────────┘
Review the setup once a week
- ▸Did every open study actually feed one decision this week? If not, switch it off.
- ▸Is any drawn level still untouched by price after ten days? Delete it.
- ▸Was an alert dismissed after it fired, or is it still armed and repeating? Check your alert list.
- ▸Has your style shifted (day trading into multi-day)? Move the timeframe pair with it.