Is Trading M1 a Good Idea? Why Spread Eats More Profit Than on M15

You can trade M1, but you start at a big cost disadvantage. The spread (the gap between the buy price and the sell price) is a toll you pay in full on every trade, while price on a 1-minute bar only moves a little. So the same cost takes a share of profit several times bigger than on M15 or H1. This article works it out step by step, then looks at gold backtests where the same system made money on M15 but lost on M1.

What a timeframe is, and why it matters for costs

Timeframe (TF) is the length of time one candle covers. M1 = 1 minute, M15 = 15 minutes, H1 = 1 hour. Small timeframes give frequent signals and precise entries, which makes them look attractive.

But trading costs do not shrink with the timeframe. Every trade pays two things.

• Spread. You buy at the Ask price, which is higher than the Bid price shown on the chart. The moment you open a trade, you are down by the spread.
• Slippage. The price you actually get differs from the price you asked for. It usually happens around news or when the market is thin.

Neither of these cares which timeframe you are looking at. If the distance you hope to profit from gets shorter, the same cost becomes a large part of the profit.

Worked example: same spread, different distances

Use a gold spread of 0.27 dollars per ounce (rounded from the 0.272 average we measured, see the next section). The distance price moves per bar is a round number assumed to explain the idea, not a measured value. Real values change with the time of day and volatility. You can check your own with ATR (the average range of a bar).

Assume an M1 bar moves about 2 dollars, M15 about 8 dollars and H1 about 16 dollars.

• M1: 0.27 ÷ 2 = 0.135 = 13.5%
• M15: 0.27 ÷ 8 = 0.034 = 3.4%
• H1: 0.27 ÷ 16 = 0.017 = 1.7%

Now see how this changes the win rate you need. Say you set TP and Stop loss the same distance away, 1 bar (RR 1:1), measured on the Bid chart. You buy at the Ask, which is 0.27 dollars higher.

On M1, with a 2 dollar distance

• The chart moves up 2 dollars to the TP: you actually make 2 − 0.27 = 1.73 dollars per ounce
• The chart moves down 2 dollars to the Stop loss: you actually lose 2 + 0.27 = 2.27 dollars per ounce
• Win rate needed to break even = 2.27 ÷ (1.73 + 2.27) = 2.27 ÷ 4 = 56.75%

Check: over 100 trades, 56.75 wins × 1.73 = 98.18 and 43.25 losses × 2.27 = 98.18. Net exactly zero.

The shortcut: breakeven win rate = 50% + spread ÷ (2 × distance)

TF	TP = SL distance (assumed)	Actual win	Actual loss	Win rate needed
M1	2 dollars	1.73	2.27	56.75%
M15	8 dollars	7.73	8.27	51.69%
H1	16 dollars	15.73	16.27	50.84%

On M1 you need to win almost 57 trades out of 100 just to break even, while on H1 it is barely different from a coin flip. Read more on win rate and breakeven in A high win rate does not mean profit.

Spread does not get more expensive on M1. The distance you trade gets shorter, so the same cost becomes a big chunk of the profit.

What gold spread really is

We measured the gold (XAUUSD) spread from real ticks from January to August 2026 on the account type we use for backtests, 61.3 million ticks in total (Thai time = GMT + 7).

Period	Average spread (dollars per ounce)
Whole period	0.272
22:00 GMT (05:00 Thai time), the widest hour of the day	0.348
Sunday (market opens for the week)	0.329
Friday (narrowest day of the week)	0.254

p95 (the value that 95% of ticks do not exceed) was 0.400 dollars in almost every hour. Around 05:00 Thai time, p95 rose to 0.560 dollars. That is right after the market pauses for about an hour and reopens. See market hours in Forex market hours in Thai time.

These figures come from one account over one period. Other brokers or account types may be much wider or narrower.

Test result: same system, profit on M15, loss on M1

In our backtests, we ran the same systems on several timeframes without changing the rules. Conditions: gold (XAUUSD), real ticks, 5 January to 25 July 2026, starting balance 3,000 dollars, 0.02 lot per trade, one single 7-month test period.

The systems tested

1. Breakout system. Enters when price breaks out of a Donchian channel (the High/Low range of the last N bars), takes profit at 161.8% of the swing.
2. Pullback system. Waits for price to pull back within the swing, then enters, takes profit at 161.8% of the swing.
3. Plain entry at a Fibonacci level. Places a Limit at 61.8% of the latest swing, with no other trigger.

System	M15 net	M15 DD	M1 net	M1 DD
Breakout	+3,938	18.0%	−2,922	97.8%
Pullback	+1,649	9.7%	−2,282	82.5%
Entry at Fibonacci level	+1,123	–	+79	–

DD (Drawdown) is how far the account falls from its peak along the way. The breakout system on M1 lost almost the entire account at one point. An earlier version of the pullback system using an RR 1:2 TP also lost money on M1 (−1,302 dollars, DD 62%). The third system on M1 ended up roughly flat: the profit it had on M15 was almost all eaten by spread.

All of these are backtest results. Past results do not predict future results, and they are not returns you should expect.

A bigger timeframe is not always better

Looking at every timeframe in the same test set, the picture is not a straight line of “bigger is better”.

TF	Breakout net	Pullback net
M1	−2,922	−2,282
M5	−713	+914
M15	+3,938	+1,649
M30	+910	+1,494
H1	+785	−54
H4	+1,972	−918

M15 is the only timeframe where both systems were clearly profitable. On H4, the breakout system took only 33 trades in 7 months, too few to trust the number. The pullback system on H1 and H4 got so few signals that it lost money. So the lesson is not “run to a bigger timeframe”. It is that the cost must be small compared with the distance you trade, and you need enough trades to draw a conclusion. A single backtest period does not predict the future.

How much of the edge does spread eat?

Another way to look at it is to separate out the “profit before spread”. Here is an example from our 8-bar box breakout system on M1, backtested on 2023 (simulated ticks, not real ticks), at 0.01 lot.

• Average profit per trade after spread = 0.225 dollars
• Spread already paid per trade = 0.272 dollars
• Profit before spread = 0.225 + 0.272 = 0.497 dollars
• Spread ate 0.272 ÷ 0.497 = 54.7%, about 55%

Check: 0.497 − 0.272 = 0.225, which matches the profit left over.

That means if the spread were just 0.225 dollars per trade wider, that year would come out at zero. So we set a rule: in the worst year, profit per trade must still be at least 0.20 dollars. If it is thinner than that, the system only works on paper. Backtest results do not predict future results.

Always look at profit per trade alongside total profit. A system that trades often with thin profit per trade is more likely to die from costs than from its strategy.

So can M1 work?

Yes, if the distance you trade is big enough. Some of our systems that have survived across several years do look at the M1 chart, but they trade wide price ranges of 1% of price or more, with wide Stop losses and far profit targets.

Think it through. Say gold is at 3,000 dollars. A 1% range is 30 dollars, and a 0.27 dollar spread is only 0.27 ÷ 30 = 0.9% of the range, compared with 13.5% of a single M1 bar in the first example.

The timeframe is just the lens you look through. What really sets your cost is the TP and Stop loss distance. Read more on choosing distances in Take profit near or far? and Stop loss: wide or tight?

Reading spread in MT5: points vs price

MT5 shows spread in points, which is the last decimal place of the price. It is not dollars and it is not pips.

• To see it: in the Market Watch window, right-click, choose Columns, then tick Spread.
• To check the number of decimals: right-click the symbol, choose Specification, and look at the Digits field.

Convert to price: spread in price = number of points × size of 1 point

Symbol	Price format	1 point	Spread shown in MT5	In price
Gold, 2 decimals	2650.25	0.01	27	0.27 dollars
Gold, 3 decimals	2650.250	0.001	270	0.27 dollars
EURUSD, 5 decimals	1.08505	0.00001	12	0.00012 = 1.2 pips

27 and 270 are the same spread. The only difference is how many decimals the broker uses. If you do not check Digits first, you can be off by a factor of ten.

Convert to money: 1 lot of gold is usually 100 ounces. A spread of 0.27 dollars = 27 dollars per 1 lot, or 0.27 dollars per 0.01 lot. For pips and lots, see Pip vs lot: what is the difference?

Common mistakes

• Only looking at how many points the spread is without comparing it to the TP distance. The same number can be tiny on H1 and huge on M1.
• Thinking trading often = profit fast. Every trade pays the spread, so the number of trades is a cost, not a freebie.
• Backtesting with a tighter spread than your real account. Small-timeframe systems are the most sensitive to this. Test with data as close to reality as you can. See Tick data quality in MT5.
• Opening trades right after the market reopens. Around 05:00 Thai time (UTC+7) the spread is wider than usual, which hits systems with thin profit per trade hard.
• Concluding that M1 never works. The problem is a small distance compared with the cost, not the timeframe itself.