Stochastic Crossover Settings for MT4: 1 of 144 Made Money on M5, 39 of 126 on H1, and the Exit Decided It

Most Stochastic guides give you settings without telling you what they did. This one starts from the numbers. In my Stochastic buy and sell signals test, 4,263 %K/%D crosses on the EURUSD 5-minute chart made nothing before costs. The question here is simpler: is there any combination of settings, chart and exit where the cross is worth trading, and how do you set it up in MetaTrader 4?

The short answer: not on M5, whatever the numbers. On H1, with the zone filter and a patient exit, about three setups in five made money in this sample. Below are the settings, the exact MT4 steps and the warnings that come with a small sample.

What do the Stochastic settings in MT4 mean?

MT4 shows five inputs when you add the Stochastic Oscillator. I wrote about the fade use of them in my Stochastic settings for MT4 guide. For crossovers, three of them matter:

• %K period (default 5 in MT4, 14 in most books): how many bars the high-low range covers.
• Slowing (default 3): an average of the raw %K. 1 means no smoothing, the “fast” Stochastic.
• %D period (default 3): the average of %K. The cross is %K crossing this line.
• Price field: Low/High (the standard) or Close/Close. I used Low/High.
• MA method: Simple, which is what I tested.

In this article I write settings in MT4’s order: %K / %D / Slowing. The MT4 default is 5/3/3. The textbook one is 14/3/3.

How was the grid tested?

• Data: 16,925 EURUSD M5 bars, June 3 to August 26, 2026, split at July 15. M15 (5,642 bars) and H1 (1,417 bars) were built from the same M5 data.
• %K: 5, 8, 14, 21. %D: 3, 5. Slowing: 1, 3, 5.
• Zones: every cross; buys only below 20 and sells above 80; the same with 30/70.
• Exits: a short fixed hold (12 bars on M5 and 4 on M15, one hour; 2 bars on H1, two hours), or at the next opposite signal of the same kind.
• Entry: the open of the bar after the cross. The cross is only certain once the bar has closed.
• Cost: 1 pip per trade, from my spread and swap test. Swap is not included, which matters for the H1 trades below.

Which timeframe worked for Stochastic crossovers?

Pips per trade, EURUSD. 18 H1 setups had fewer than 30 trades and are left out.

On M5 the settings made almost no difference. All 144 setups sat between -2.57 and +0.24 pips per trade, and the only positive one, 5/5/5 in the 20/80 zone, made +1.07 in the first half and lost 0.91 in the second. It is the same lesson as my optimized vs default settings test: when the signal has no edge, tuning its numbers finds noise.

Why did the exit matter so much on H1?

An hourly cross held for two hours is a short trade on a bigger chart, and it lost like the short trades on M5. Held until the opposite signal, the same cross becomes a swing trade lasting a day or two. A buy below 20 stays open until the next sell above 80. That gives the move time to be larger than the 1-pip cost, which on M5 it almost never was.

What did the default 14/3/3 do on H1?

The first row is the result most people would want to trade, and it needs three warnings. It is 36 trades, only just over my 30-trade minimum. The average winner was 26.92 pips and the average loser 29.63, so a few bad trades would flip it. And widening the zone to 30/70 turned the same default into a loser. It is a promising result to forward test, not a proven one.

Which settings did best on H1?

All exit at the opposite signal. Pips per trade after 1 pip of cost, swap not included.

No single %K stood out. Among the 56 H1 setups with the patient exit, %K 8 made money in 10 of 15, %K 21 in 9 of 12, %K 5 in 9 of 14 and %K 14 in 6 of 15. Slowing 5 was the most consistent, 14 of 18 positive. The top of the table is a group of neighbours rather than one magic setting, which is what you want to see. Still, choosing the best of 56 on three months of data flatters the winner. I would forward test the middle of the group, such as 8/5/1 or 5/3/5 in 30/70, not the top line.

How do you set up a Stochastic crossover in MT4?

1. Open an H1 chart of the pair. If you are new to MT4, my first chart setup covers the basics.
2. Go to Insert → Indicators → Oscillators → Stochastic Oscillator.
3. On the Parameters tab, set %K period, %D period and Slowing. For the default in this test: 14, 3, 3. Leave Price field on Low/High and MA method on Simple.
4. On the Levels tab, keep 20 and 80, or change them to 30 and 70 if you use that zone.
5. Only act on a cross after the hourly bar has closed. Rushing did not pay: entering at the close of the signal bar instead of the next open gave -1.10 pips per trade on M5 against -1.00. And a cross on a bar that has not closed yet can still disappear.
6. Buy only when %K crosses above %D below 20. Sell only when it crosses below %D above 80. Ignore everything in between.
7. Close at the next opposite zone signal, not after a fixed time.
8. Put a hard stop on every trade. The worst single trades in the table lost 80 to 140 pips while waiting for the opposite signal. My stop-loss distance test covers how far to place it.

The built-in Stochastic has no alert for crosses. You either watch the hourly close, or use an arrow indicator that marks the cross on the closed bar only. If you do, check that it does not repaint: my arrow repaint test shows how to check.

What should you watch out for?

• Swap. The H1 trades lasted 25 to 50 hours on average, so most were held overnight at least once. Swap is not in these numbers, and it can remove a few pips per trade depending on direction and broker.
• Sample size. H1 gave 30 to 70 trades per setup over three months. That is enough to rank setups, not enough to trust one.
• One pair. All of this is EURUSD. Gold behaved differently from EURUSD under the same rules in my other tests.
• Losses are large. In the default 14/3/3 setup the average loser, 29.63 pips, was bigger than the average winner, 26.92. A few in a row would erase months.

Frequently asked questions

What are the best Stochastic settings for crossovers?

On EURUSD, the timeframe and exit mattered more than the numbers. On H1, buying crosses below 20 and selling crosses above 80, held until the opposite signal, the default 14/3/3 made +4.93 pips per trade over 36 trades. Nearby settings such as 8/5/1 and 5/3/5 with 30/70 levels also made money in both halves.

Does the Stochastic crossover work on the 5-minute chart?

Not in this test. Only 1 of 144 M5 setups made money after a 1-pip cost, and that one lost in the second half of the data.

What is the default Stochastic setting in MT4?

MT4 opens the Stochastic Oscillator with %K 5, %D 3 and Slowing 3, Price field Low/High and a Simple moving average. Many books use 14/3/3 instead.

Should I use 20/80 or 30/70 levels?

Both beat trading every cross. Across all 414 setups, no zone averaged -1.01 pips per trade, 20/80 -0.85 and 30/70 -0.26. For the default 14/3/3 on H1, 20/80 was clearly better.

Should I close a Stochastic trade after a fixed time?

Not on H1. Hourly crosses closed after two hours made money in 5 of 70 setups. Held until the opposite zone signal, 34 of 56 did.