MACD Settings for One-Minute TradingView Charts: A Testing Guide
There is no universally best MACD setting for a one-minute TradingView chart. Shorter lengths change how much weight recent bars receive; they do not establish better entries or a higher win rate. For scalping, execution costs and the time a signal becomes available can matter as much as the chosen lengths.
Compare 6/13/5, 3/10/3, 5/13/6 and 8/17/9 against a 12/26/9 reference under the same trading rules. The configurations below are candidates for a controlled comparison, not reported profitable systems.
What MACD measures on a one-minute chart
With exponential moving averages selected, MACD is the fast EMA minus the slow EMA. Its signal line is an EMA of that difference, and its histogram is MACD minus the signal line. TradingView's MACD documentation describes the source, lengths and moving-average choices.
A positive MACD means the fast average is above the slow average. A positive histogram means MACD is above its signal line. Histogram height measures that difference in price units; it is not a direct measure of trade probability. A histogram zero crossing and a MACD/signal-line crossover are the same algebraic event.
The EMA calculation carries forward a weighted previous value. A 26-bar EMA does not discard all information older than 26 minutes. Missing bars and session breaks also mean that 26 observations need not represent 26 uninterrupted clock minutes. The default MACD is valid on a one-minute chart; its usefulness requires testing rather than declaring it too slow by definition.
Four parameter candidates and a reference
| Fast / slow / signal | What the configuration changes | Question for evaluation |
|---|---|---|
| 12 / 26 / 9 | Standard reference | Does its timing and turnover remain viable after costs? |
| 6 / 13 / 5 | Shorter fast, slow and signal lengths | Does increased responsiveness compensate for additional reversals? |
| 3 / 10 / 3 | Very short fast and signal lengths | How sensitive are entries to a one-tick price change or delayed fill? |
| 5 / 13 / 6 | Shorter fast average but longer signal smoothing than 6/13/5 | Which differences come from the fast average versus signal smoothing? |
| 8 / 17 / 9 | Fast and slow lengths between 6/13/5 and the reference, with a longer signal length | Does changed timing reduce turnover without removing useful entries? |
These choices do not form a guaranteed ranking from fastest to most reliable: crossover timing depends on the price path and all three inputs. A 3/10/3 MACD does not react to “every tick” in a historical bar-close test, and it should not be presented as an authenticated Linda Raschke configuration without a source for that exact formula.
Avoid assigning one setup to all forex, crypto, news events or sideways markets. Record the actual feed and spread. A spot FX feed may supply tick volume rather than consolidated traded volume; one crypto exchange's activity is not the entire market.
Configure the comparison in TradingView
- Open the exact exchange symbol on a one-minute ordinary-candlestick chart. Record the chart time zone and session settings.
- Add the built-in MACD. Set the price source and both moving-average types before entering the three lengths.
- Save a named template or written configuration for each candidate. Colors help identify lines but do not change the signal rule.
- Inspect several crossings one bar at a time. Record when the condition becomes true and whether it remains true at the bar close.
- Choose one entry/exit specification and use it consistently for every configuration.
Bar Replay is useful for observing a chart without exposing all later bars at once. It is not the same as an automated strategy test, nor does adding the MACD indicator create a performance report. TradingView's strategy documentation explains simulated orders, calculation settings and broker-emulator assumptions.
Pineify can generate an indicator or strategy from configured rules or a written specification. Its MACD configuration exposes source and fast, slow and signal lengths. Inspect the output and run a strategy in TradingView with the intended properties. Generating a script does not validate its signal quality or reproduce live execution automatically.
Define an actual one-minute experiment
The following is a hypothetical comparison design, not a trading recommendation or an observed result:
| Item | Example definition to record before testing |
|---|---|
| Signal | Bullish MACD/signal crossover evaluated at a confirmed one-minute close |
| Entry | First eligible simulated execution after that decision, with the order-timing setting recorded |
| Exit | Bearish crossover or a separately specified protective stop; define which takes precedence |
| Position | One position at a time with a fixed stated sizing rule |
| Session | A named time zone and explicit start/end times; session-end handling stated |
| Costs | Spread, commissions and slippage for the instrument and order type |
| Comparison | Same dates and rules for each parameter set, followed by a later period not used to select settings |
The design still needs instrument-specific values before execution. A universal five-pip stop is unsuitable as a default across currencies, futures, equities and crypto. Stop distance must be translated through tick size, contract value, quantity and any currency conversion. A stop order can fill beyond its trigger in a gap or fast market.
For a hypothetical cost example, suppose average gross wins and losses are each 5 pips and round-trip costs average 1 pip. With win probability p, average net outcome is 5p minus 5(1−p) minus 1, or 10p−6 pips. The break-even win rate is therefore 60%. A 55% win rate would lose 0.5 pip per trade under those assumptions. These are illustrative inputs, not measurements of any MACD configuration.
Three rules that can be tested separately
Signal-line crossover
Use the crossover event rather than repeatedly entering while MACD remains above its signal line. Decide whether intrabar signals are permitted. A signal that appears and disappears before the close cannot be evaluated fairly using only closed-bar markers.
A 9/21 EMA filter or RSI above 50 adds another price-based condition. Compare the unfiltered rule and filtered rule on the same dates. The filter may remove winners as well as losers, and there is no measured 30% reduction in false signals established here.
Divergence reversal
Specify the price and MACD pivot rules, their matching window, and the number of later bars needed to confirm them. Entering at a pivot that was only identified afterwards introduces future information. Divergence can remain present while price continues in the same direction; a fixed 10–15 pip retracement does not follow from the pattern.
Zero-line filter
Test whether requiring positive MACD for long entries changes net outcomes. This condition says the fast average exceeds the slow average; it does not identify a safe side of the market. If adding a Bollinger Band squeeze, define the band calculation, width units and comparison period. Reduced width does not guarantee an imminent expansion or its direction, and an arbitrary 5% threshold does not generalize across instruments.
Timing and validation limits
A five-minute filter used on a one-minute chart needs a clear rule about confirmed versus developing five-minute values. Its final value is unavailable to earlier one-minute decisions. TradingView's repainting documentation explains why reloaded historical bars can differ from realtime observations.
Session labels also need care. London and New York change daylight-saving time on different dates, so a fixed GMT overlap does not describe the whole year. News can increase activity and also spread, slippage and order rejection risk. “High volume” does not confirm that a move must continue, and lower activity does not guarantee a reversal.
Do not treat 100 trades or two weeks of paper trading as automatic validation. Assess dependence between trades, coverage of different conditions, realistic costs and the number of parameter choices tried. Keep a later evaluation window untouched while selecting settings. Overfitting raises the risk that a result will not generalize; it does not logically guarantee that every fitted system will fail.
Track signal timestamp, confirmed-bar status, intended and actual execution prices, fees, exit reason and adverse movement. Review whether performance depends on a few unusual trades. If using a trailing stop, define activation, offset and update timing; it cannot guarantee that displayed gains are locked in.
Frequently asked questions
Is 6/13/5 the best one-minute MACD setting?
It is a shorter-length candidate, not an established winner. Compare it with the reference and other stated candidates under the same costs and rules.
Can I use 12/26/9 on a one-minute chart?
Yes. The calculation works on that interval. Whether its timing suits the strategy is an empirical question, and the 26-bar EMA includes decaying influence from earlier data.
Does adding RSI remove false signals?
It changes which trades qualify. RSI and MACD both use price, so agreement is not independent proof. Measure net results and rejected winners as well as rejected losers.
Should forex and crypto use different settings?
Differences in trading hours, spreads, contracts and feeds justify separate evaluation. They do not establish an automatic mapping from asset class to one parameter set.
How much testing is enough?
There is no fixed count or duration that proves reliability. Use a documented comparison, an untouched later period and realistic execution assumptions, then monitor behavior against those assumptions.