Copying Trades Between Your Own MT4 and MT5 Accounts

MT4 shipped in 2005, MT5 in 2010. Both come from MetaQuotes, but the account systems, the programming languages and the order models are all different. So the question "I have MT4 at broker A and MT5 at broker B — can they copy each other?" has a different answer depending on the route you take.

Why the two platforms cannot talk directly

• No shared accounts: MT4 accounts run in the MQL4 ecosystem, MT5 in MQL5 — the credentials are not interchangeable
• Different order models: MT4 is hedging only; MT5 supports both netting and hedging
• Different symbol names: the same EUR/USD may be EURUSD, EURUSD.a, EURUSD.m or EURUSDpro
• Different execution modes: instant and market execution handle slippage differently

Of these four, symbol naming is the one that bites hardest: the software looks up an instrument by name, fails to find it, and places nothing — while the source account shows a filled trade.

Three routes

Route one: copy within one platform. Both accounts on MT4 (or both on MT5), using a local copier EA or the platform's own signal service. Latency is typically tens of milliseconds. This is the most robust setup.

Route two: bridge across platforms. MT4 on one side, MT5 on the other, with a middle layer (a local EA plus a bridge, or a third-party copy service) translating the order from source to target. Latency ranges from hundreds of milliseconds to several seconds, depending on where the middle layer runs.

Route three: manual mirroring. You take the signal and place an order on each platform by hand. Latency measured in minutes; suitable only for swing and position trading.

Six real costs of cross-platform copying

This is the part that matters. People assume copying equals replicating. What actually gets copied is the order intent, not the fill.

• Symbol suffix: the source trades EURUSD, the target has no such symbol, the copy silently fails
• Different spread: source at 0.8 pips, target at 1.6 — identical stop distance, double the real risk
• Different leverage: source at 1:500, target at 1:100 — five times the margin for the same lots, and a possible margin call
• Different contract size: standard, mini and micro lots are not defined identically at every broker
• Execution latency: half a second during a data release can be 10–20 pips of slippage
• Minimum stop distance: some brokers enforce a stop level, and stops placed too close are rejected

What half a second costs during a data release

At the non-farm payrolls print, EUR/USD routinely jumps 30–60 pips in an instant. Suppose the signal you follow opens a position one second before the release:

The source fills at 1.0850. The target, half a second behind, fills at 1.0870 — 20 pips of slippage. On 0.5 lots that is a $100 difference, or 1% of a $10,000 account.

This is why cross-platform copying should either skip signals around data releases or set a slippage cap on the target account so the order is rejected rather than filled badly.

Lot conversion: match risk, never copy the number

Target lots = source lots × (target equity ÷ source equity) × (source value per pip ÷ target value per pip).

Simplified: if both accounts are USD-denominated and both use standard lots, you only scale by equity. A source of $10,000 trading 0.5 lots maps to 0.25 lots on a $5,000 target.

Watch the rounding error. 0.25 lots is fine on an account with a 0.01 minimum. On an account with a 0.1 minimum it becomes 0.2 or 0.3 — a 20% deviation. That is the usual reason a "proportional" copy ends up carrying different risk.

Setting up the symbol map

This is the step most often skipped. Open the target account's instrument list and walk through every symbol the source uses.

Three cases: identical (EURUSD → EURUSD), suffixed (EURUSD → EURUSD.m), and renamed entirely (XAUUSD → GOLD, US30 → DJ30). The first two need one mapping line each; the third requires checking contract specifications individually.

Then verify contract size. "One lot" is 100,000 units on a standard account and 10,000 on some others. If that is off, every converted lot size is off by a factor of ten.

The test is crude but reliable: place 0.01 lots manually on the target and compare margin requirement and per-pip P&L against the source. If they do not scale together, check the contract specification rather than guessing.

When the connection drops

For a locally hosted copier, the real risk is not latency but signals lost while offline. The source opens, the target does not follow, and by the time the program reconnects the price has moved.

Three handling options, each with a cost: catch up (enter at market, possibly at a terrible price), skip (no catch-up, and the two accounts now differ), or close the source (guarantees consistency, but closes a position that may have been correct).

There is no universally right answer — but decide the rule in advance rather than improvising during the outage. Most traders choose skip-and-log, then reconcile weekly.

Pre-launch checklist

• Symbol names match on both sides, or a mapping table is configured
• Spread, commission and swap are in the same range on both accounts
• Target leverage is sufficient, checked against maximum simultaneous positions
• Stops and targets will not be rejected by a minimum stop level
• You know which machine hosts the copier, and what happens if it goes offline
• Run minimum lots for two weeks and compare fills trade by trade

FAQ

Can an MT4 signal be subscribed on MT5? No. MetaQuotes runs separate signal services for MT4 and MT5; crossing platforms requires a bridge.

How much latency is acceptable? Seconds are fine for swing strategies. For scalping and news trading, anything above 200 milliseconds should be treated as unusable.