Automating FX vs Gold vs Indices: Why One Size Never Fits

You've copied the same settings across everything. One EA. Same stop. Same lot size. It runs on EURUSD, gold and an index at the same time.

This is a plain-English guide to automate forex gold indices.

Then gold moves $30 in a minute and your FX-sized risk turns into a small crime scene.

Automate forex gold indices: the lie of “universal settings”

When people talk about how to automate forex gold indices, they usually mean one thing: slap the same robot on multiple charts and hope diversification saves them.

Same stop in pips. Same lot size. Sometimes even the same trading hours.

It feels efficient because the MT4/MT5 panel looks tidy.

The market does not care about your panel.

Why the win rate survives, but your account doesn’t

Take a simple hypothetical system.

Risk 1% per trade. Target 1.5R. Win 50% of trades. Expected value: +0.25R per trade.

On paper that works on EURUSD, XAUUSD and an index future. The logic is the same. The maths of expectancy is the same. The distribution is not.

Different spread. Different volatility. Different contract size. Different session gaps.

So the win rate might stay roughly similar across assets.

The path your equity takes absolutely will not.

Spread: the silent tax that hits each asset differently

People love to argue over entry signals. They almost never argue over the spread.

They should.

Imagine your system has a 20 pip stop and a 30 pip target on EURUSD.

Say the spread is 0.2 pips.

Your spread cost is 1% of the stop distance (0.2 / 20). Annoying, but survivable if you have a real edge.

Now you copy that logic straight onto gold.

Except on gold you decide to use a “20 pip” stop as well because it looks symmetrical on the chart.

The spread might be 1.5 “pips” equivalent on your broker.

Spread cost is now 7.5% of the stop distance (1.5 / 20).

Same settings. Very different tax rate.

On an index CFD with an overnight gap, the spread and slippage might make that tax even worse.

Volatility: a 10 pip move is not the same everywhere

Next mistake: treating pips like they’re a universal unit of risk.

They’re not.

Suppose EURUSD moves 60 pips on a volatile day.

Gold can move the equivalent of hundreds of pips in the same window.

Indices can gap that much on an open without even trying.

So your tight 15 pip stop that behaves fine on EURUSD might be absurdly small on gold.

You get whipped out constantly, then watch price hit the target after you’ve been stopped by noise.

“The system doesn’t work on gold.”

No. You resized the risk to noise ratio without realising.

Contract size: this is where the real damage hides

Spread and volatility you can kind of eyeball.

Contract size is where accounts quietly blow up.

A 0.10 lot EURUSD trade roughly means $1 per pip.

Depending on your broker, a 0.10 lot trade on a major index CFD might be $1 per point as well.

On gold (XAUUSD), a 0.10 lot trade can mean $1 per 0.1 move, so $10 per “pip-equivalent”. Suddenly you’re 10x the risk you thought.

Now add one more detail: a lot of brokers have a 0.01 minimum on gold.

So if 0.01 on gold is effectively the same cash risk as 0.10 on EURUSD, your smallest possible trade on gold is already “medium” risk for a small FX account.

That’s why equity swings on gold look wild on small balances.

You didn’t become aggressive. The contract did.

Why one fixed lot size is quietly wrecking your automation

This is where the classic “just run 0.1 lots on everything” automation mindset dies.

That 0.1 might be 0.5% risk on EURUSD, 1% risk on an index and 5% risk on gold, purely because of how the contracts are defined.

Same number in the lot box. Completely different percentage of your equity on the line.

If you haven’t read it already, this is exactly why fixed fractional sizing exists, and why copying fixed lots everywhere is such a slow form of self-sabotage: /blog/fixed-fractional-vs-fixed-lot-sizing-that-really-compounds.

Automation doesn’t fix bad sizing.

It just applies it with perfect discipline.

Same stop distance, different percentage risk

Let’s pin this down with hypothetical numbers.

Say you want to risk 1% of a $5,000 account per trade.

You put a 50 pip stop on each instrument and use 0.10 lots on all of them because it “looks tidy”.

On EURUSD at $1/pip for 0.10 lots, 50 pips is $50.

That’s 1% of $5,000. Fine.

On gold, suppose 0.10 lots is $10/pip-equivalent.

Now 50 pips is $500. That’s 10% of the account.

On an index where 0.10 lots equals $0.50 per point, 50 points is $25.

That’s 0.5% risk.

Same stop. Same lot size. Your real risk per trade ranges from 0.5% to 10% depending on the instrument.

You don’t have a trading strategy at that point.

You have a risk lottery.

Session behaviour: FX glides, gold spikes, indices gap

FX trades around the clock.

Gold follows it to a degree, but tends to care a lot about specific data releases and US hours.

Indices have distinct cash sessions and love an overnight gap.

Your automation doesn’t erase that.

If your system blindly holds trades over the weekend on an index, but was designed and tested on continuous FX data, your backtest isn’t telling you the truth.

It never saw a Monday gap on NFP after a surprise number.

Index systems might need flat-before-close rules.

Gold systems may need specific filters around known data times.

FX might be happy running through Asia with smaller sizing.

“One schedule fits all” ignores when the actual risk shows up.

Automate forex gold indices correctly: one logic, different wrappers

There’s a reasonable objection here.

“But the pattern works across all of them. Surely I don’t need three different strategies?”

You might not need three different ideas.

You do need three different risk wrappers.

The entry and exit logic can be identical.

The bit that changes is how you convert that logic into position size, stop distance, partials, and trading hours for each instrument.

Automation lets you separate those two layers properly.

One code module for signal logic.

One module per instrument for risk and execution rules.

Humans are bad at consistently doing that translation in real time.

Machines are very good at it, once you’ve done the thinking up front.

Per-instrument calibration: the short checklist

If you want to automate across FX, gold and indices without slowly boiling your equity, treat each one like a separate project.

Same core idea, different numbers.

1. Volatility-based stops, not pip vanity

Anchor your stops to volatility (ATR, for example), not to round numbers.

If you use 2x ATR on EURUSD, use 2x ATR on gold and indices too.

The distance in pips/points will automatically scale to each instrument’s reality.

2. Percentage risk per trade, per instrument

Decide your maximum percent risk per trade for each asset class separately.

You might tolerate 1% on FX, 0.5% on an index and 0.25% on gold, purely because of gap and spike behaviour.

Then your position sizing code calculates lot size from stop distance and that percent; you never type 0.10 in a box again.

3. Respect the minimum lot size problem

On gold, a 0.01 lot minimum can be a serious constraint for small accounts.

If 0.01 is already 2% risk with your preferred stop, automation cannot magically reduce that to 0.5%.

That’s a capital and instrument choice problem, not a coding problem.

4. Time filters per asset

Let your code understand which sessions each instrument lives in.

Indices: maybe trade the cash session only, flat before close.

Gold: avoid certain minutes around big scheduled news if your system is not designed for slippage.

FX: maybe reduce size or skip entirely in the deadest part of Asia if spread balloons relative to volatility.

Automation doesn’t replace risk thinking, it exposes the lack of it

A lot of people find out their risk logic is flawed only after they automate.

The bot takes the rules literally.

Run that across enough instruments and any inconsistency gets amplified.

Before you spin up multi-asset automation, ask the boring questions.

What is my average loss in currency terms on EURUSD vs gold vs indices?

What is my worst peak-to-trough drawdown in each?

How many trades of each instrument can I hold at once without exceeding my portfolio-level risk cap?

If you want to go deeper on why this matters when you start combining systems, have a look at correlation and how it clusters pain: /blog/what-is-correlation-and-why-it-wrecks-trading-portfolios.

Multiple uncalibrated automations aren’t diversification.

They’re just a busier way to be overexposed.

Multi-asset automation and system stacking

True diversification isn’t “the same thing, everywhere”.

It’s different things, properly sized, in different places.

Running a trend system on EURUSD, a mean-reversion system on gold and an opening range breakout on an index only helps if the risk of each is under control on its own.

Then you look at them together: combined drawdown, correlated losing streaks, capital allocation between them.

That’s the core argument for trading multiple strategies, not just one redeployed 15 times: /blog/why-trade-multiple-strategies-instead-of-just-one.

Automation is the plumbing that makes system stacking operationally possible.

But the plumbing doesn’t design the house.

What automation changes – and what it doesn’t

Automation does three useful things in this context.

• Executes your per-instrument sizing maths exactly, every time.
• Obeys session rules to the second, even when you’re asleep.
• Shows you the true distribution of outcomes, because it actually takes every signal.

What it doesn’t do is fix bad assumptions about how FX, gold and indices behave.

If you tell a bot to risk 5% per trade on gold because some bloke on YouTube did it on a demo, it will do that with frightening enthusiasm.

It’s the same point I’ve made elsewhere about bots in general: automation multiplies whatever you hand it, good or bad: /blog/how-to-automate-a-trading-strategy-without-blowing-up.

The edge is in the idea and the risk maths.

The robot just shows you how honest those really are.

One simple sanity check before you go live

Here’s the minimum standard before you let an automated system loose across FX, gold and indices.

For each instrument, on paper or in a spreadsheet, answer:

• With my current settings, what’s the cash loss if the worst historical stop gets hit?
• What percent of equity is that?
• What happens to my account if that loss happens on two instruments on the same day?
• How many consecutive losses has this system shown in test, and what’s the portfolio drawdown if that repeats on the most leveraged instrument?

If any of those numbers make you uncomfortable on gold or indices but not on FX, that’s the point.

Same system logic.

Different risk reality.

And if those numbers don’t make you uncomfortable, check them again.

There is always risk of loss in leveraged trading.

Where to go from here

If you take one thing away from this, let it be this: per-instrument calibration is not optional when you automate forex gold indices together.

It’s the difference between a portfolio of systems and a collection of accidents.

Start your free 14-day ArcisTrade demo →

See live and testing systems across FX, gold and indices, each with its own sizing and risk logic visible on the statement.

Watch how different the equity swings are per instrument when lot size and stops actually match the market they’re trading.

P.S. Pay more attention to the drawdowns than the peaks. That’s the bit you really have to live through.