MACD settings on M1: the default 12,26,9 vs faster sets

[Fairman]

MACD settings on M1: the default 12,26,9 vs faster sets

MACD's default settings, 12, 26 and 9, come from daily charts. On M1, they cover roughly the last half hour of price. Some scalpers use them as is, others switch to faster sets like 5, 13, 1 or 3, 10, 16, which you'll see mentioned in older trading books.

What changes with faster settings: the MACD line reacts more quickly to price, crosses happen more often, and the histogram swings more. That can give earlier signals. It also gives far more false ones in choppy conditions.

Slower settings smooth things out, but on M1 they can lag noticeably. By the time a cross appears, the move might be mostly done.

The signal line setting matters too. A signal line of 1 essentially removes it, leaving the MACD line crossing zero as the main event. That simplifies the indicator to a momentum gauge.

I think the most useful way to use MACD on M1 isn't for crosses at all, but for divergence and the histogram's size. A shrinking histogram while price pushes higher suggests weakening momentum, whatever the settings.

If you want to test settings, compare them on the same few sessions and count signals that led somewhere versus ones that didn't. Settings that only look good on one day usually aren't telling you much.

[Fairman]

It's also common to use the MACD from a higher timeframe as a filter while trading on M1. For example, only take long M1 entries when the M5 or M15 MACD is above zero. That uses the default settings on a timeframe closer to where they make sense, and keeps the M1 chart cleaner. It's a filter rather than a signal, and it avoids the problem of fast MACD settings on M1 throwing out constant crosses. It also means you only need to look at M1 for timing, which keeps things simple during fast moves.

[Fairman]

Divergence on MACD is easier to read with the histogram than with the lines, at least for me. Comparing the height of two histogram peaks is straightforward, while comparing line peaks can be ambiguous when the lines are close together. Whatever settings you use, divergence works better when it happens at a level that matters, like a session high, than in the middle of nowhere. Divergence alone happens constantly on M1 and isn't much of a reason to trade. A divergence at the session high is worth far more than one in the middle of the range.

[FXS Prop Desk]

On M1, the faster settings react better to short moves but produce many more signals. The 3, 10, 16 set is often used for momentum shifts rather than crossovers.

A common approach is to use the histogram rather than the lines. A shrinking histogram can show momentum fading before the lines cross, which can be useful for timing exits on short trades.

[Fairman]

Zero line crosses on MACD are often more meaningful than signal line crosses on M1. The signal line cross happens frequently and catches a lot of noise. The zero line cross means the fast average has crossed the slow one, which is a bigger shift in momentum. Using zero line crosses as a filter, and signal line crosses only for timing within that direction, cuts down on the number of signals. On faster settings, zero line crosses come more often, so the filter is less strict, which is a trade-off to keep in mind.

[Fairman]

How do faster MACD settings handle news spikes? A sudden spike on a data release can push the fast MACD to an extreme, and the indicator takes a while to come back. During that time, signals might be distorted. I'd ignore MACD signals for a few minutes after major news, at least until the spike has worked its way through the averages. With the default slower settings, the effect lasts longer. With faster settings, it's shorter but sharper. Either way, MACD readings right after a spike aren't very useful. A short pause after big releases avoids most of the problem.