The signal provider's page was beautiful. Eight months of history, 64% gain, max drawdown 12%, a smooth equity curve climbing like a staircase. I subscribed in March 2025, linked my $5,000 MT5 account at a 1:1 copy ratio, and watched the first three weeks add +$450. I remember thinking copy trading was the closest thing to free money I had ever seen. Then came NFP Friday.
The provider was short GBPJPY into the release. Price spiked against the position, and instead of cutting it, the provider doubled down. Then doubled again. My account copied every add dutifully — that is what copy trading does, it copies the mistakes too. By Monday my $5,450 was $4,350. The provider's "max drawdown 12%" was history, not a promise. I unsubscribed at −$1,100 and spent the next month figuring out how copy trading actually works, because the glossy subscription page had taught me nothing.
What I learned: an mt5 copy trading signals guide that only shows you where to click Subscribe is a brochure, not a guide. The real skill is vetting the provider, sizing the copy ratio with actual math, and building kill-switches for the day the provider's discipline breaks. I rebuilt my approach from zero, tested everything on demo first, and now run copied signals as one sleeve of my trading — sized through the DarkWolf MT5 Signal Calculator like any other position. Here is the full playbook.
The signal runs I did before letting anyone else's trades touch my account
After the NFP disaster I made a rule: no provider touches real money until it survives 90 days on my demo copier with full logging. I ran four provider archetypes through that filter. The results cured me of ever trusting a screenshot again.
| Provider style | Demo test result | Hidden behavior found | My verdict |
|---|---|---|---|
| Low-frequency swing trader | +6.2% over 90 days, max DD 8% | None — stops honored every time | Subscribed with real money |
| High-win-rate scalper | +11% over 90 days, max DD 6% | Copied fills averaged 1.8 pips worse — scalping edge died in slippage | Rejected; uncopyable edge |
| Grid/martingale system | +19% over 90 days, max DD 4% | Floating loss hit 31% intraday twice; shown DD hid it | Rejected; the blowup is scheduled, not hypothetical |
| News-event trader | −3% over 90 days, max DD 14% | Honest losing stretch, risk per trade capped at 1% | Watchlisted; honest beats profitable-but-fragile |
The grid system was the scariest find. Its published max drawdown was 4% because it never closed losers during the test window — the floating loss hit 31% intraday twice. Closed-trade drawdown lies; floating drawdown tells the truth. I only trust providers whose published stats include equity-based drawdown now.
How copy trading actually routes a signal into your MT5
Strip it down and copy trading on MT5 is a replication machine with four settings that matter. The provider's terminal publishes every trade action — opens, closes, stop edits. The signal service relays it. Your terminal receives it and opens a proportional copy on your account. The proportion is the copy ratio, computed from the equity ratio between the two accounts, and this is where most subscribers get hurt without knowing it.
Say the provider trades a $50,000 account and risks 2% per trade — $1,000 of heat per position. You subscribe with $5,000 at the default 1:1 ratio and the copier scales by equity: your copy risks $100 per trade, which is also 2%. Fine. But if you bump the ratio to "make it worthwhile," you are not just scaling profits. You are scaling the provider's worst day, the provider's martingale moment, the provider's NFP revenge trade. The CFTC has warned repeatedly that copied forex trading carries the full risk of the underlying strategy plus execution differences — read their advisories before you outsource your risk to a stranger's discipline.
Execution differences are the quiet tax. Your fill will differ from the provider's by your broker's spread, your latency, and your deviation setting — the max price difference you tolerate before the copy is skipped. Set deviation too tight and you miss the best trades; too loose and you buy tops. I run 2–3 pips on majors, wider on gold. And crucially: copying requires your terminal (or VPS) to be online. The provider does not trade for you; your machine replicates their trades. If your VPS drops, you hold unmanaged orphans.
Signal-copying mistakes I watched drain a $5,000 account
Mine, mostly. Mistake one: subscribing at the top. I joined after the provider's best three months — classic performance chasing. Their edge had not changed; my entry timing had. Mistake two: I never checked the provider's open positions before subscribing. They were already carrying a floating −9% the day I joined. I bought into someone else's underwater trade on day one.
Mistake three was the copy ratio. I ran 1:1 on a $5,000 account copying a $50,000 provider, which sounds conservative until you realize the provider's 2% risk was my 2% too — and my account had no cushion for the provider's losing streaks. A 0.5 copy ratio would have halved my NFP damage. Mistake four: no unsubscribe trigger. I had no pre-defined line — no "if equity drops 15%, I walk" — so I rode it down hoping. Hope is not a risk parameter.
The ugliest mistake I see in others: copying multiple providers on one account without realizing they all trade the same London session momentum. Three "diversified" providers, one correlated macro bet, triple size. If you run several signals, the multi-account setup guide shows how I isolate them so one provider's blowup cannot take the others with it.
The copy-ratio math nobody shows you before you subscribe
Every subscriber should compute this before clicking anything. It takes thirty seconds and it is the difference between copying a strategy and copying a disaster at the wrong size.
| Variable | Meaning | Example value |
|---|---|---|
| Provider Volume | Lot size the provider opens per signal | 2.0 lots EURUSD |
| Your Equity | Your account balance at subscription time | $5,000 |
| Provider Equity | Provider's account size (check it is real, not demo-inflated) | $50,000 |
| Copy Ratio | Your chosen multiplier; below 1.0 means fractional copying | 0.5 (conservative) |
Worked example: 2.0 × (5,000 ÷ 50,000) × 0.5 = 0.10 lots per signal on your account. That is the honest size. Now stress it: if the provider's historical max floating drawdown was 20%, your account must survive 20% × your ratio setting without a margin call — and then survive it happening twice in a row. I also run the resulting per-trade dollar risk through the signal calculator before subscribing, because "0.10 lots" means nothing until you see the dollar heat. The risk management settings guide is where I documented the full pre-subscription checklist this formula feeds into.
Keeping a copied account alive through a drawdown month
Here is the part nobody posts about: even good providers have losing months, and your job during those months is to not make it worse. My rules now. One: an equity stop on the copied account — if it drops 15% from its peak, copying pauses automatically and I review. Not the provider's drawdown. My account's. Two: I never add funds to a copied account mid-drawdown to "average down" the subscription. That turns a 15% planned stop into a 30% hope trade.
Three: I re-vet the provider monthly, not yearly. Has their trading frequency changed? Are stops getting wider? Did they start holding through news they used to avoid? Style drift kills more subscriptions than bad months do. Four: I keep copied risk capped at 30% of my total trading capital — the rest runs my own systems. If every provider I copy blows up simultaneously, I am hurt, not dead. Five: during the provider's historically worst calendar month (most signal pages show monthly breakdowns), I halve the copy ratio preemptively. It costs a little upside. It has saved me twice.
The psychological trap is the subscription fee effect: once you pay $50/month, you feel compelled to "get your money's worth" by running max ratio. I lost $1,100 learning that the fee is sunk and the ratio is a risk decision. Unsubscribing from a bad provider is a profitable trade. The equity curve tracker is what finally made this mechanical for me — it plots the copied account's equity separately, so I see the 15% line approaching instead of feeling it.
Before I let a copied signal run with real size, I run it through the DarkWolf MT5 Signal Calculator to sanity-check the provider's stop and target against my own risk. These are the inputs I use when vetting a provider:
| Input Field | What it means | Sample value |
|---|---|---|
| Currency Pair | The pair the copied signal is on — must match the provider's symbol | EURUSD |
| Current / Close Price | Live price when I review the signal, not the provider's delayed quote | 1.0842 |
| Last Candle High | Completed candle high for the signal's timeframe | 1.0895 |
| Last Candle Low | Completed candle low for the signal's timeframe | 1.0810 |
| Account Balance ($) | My copy account equity, so lot math uses my dollars | 10000 |
| Risk Per Trade (%) | My risk cap per copied trade, regardless of the provider's aggression | 1.5 |
| ATR (optional) | I fill this in to check whether the provider's stop is at least 1 ATR | 0.0048 |
| Timeframe | The timeframe the copied signal was generated on | H4 |
Six weeks of signal copying on a $100,000 account
In July 2025, after the demo vetting, I allocated $100,000 across two vetted providers — $60,000 to the swing trader, $40,000 to a second intraday provider that passed the same 90-day filter. Copy ratios set by the formula above at 0.5 and 0.4 respectively. Equity stops armed at −15% per sleeve. Here is the honest log.
| Week | What happened | Account result |
|---|---|---|
| Week 1–2 | 27 signals copied, fills within 1.2 pips of provider on average | +$1,840 |
| Week 3 | Intraday provider hit a 6-trade losing streak; my sleeve −4.1% | −$2,260 net week |
| Week 4 | Swing provider caught a 190-pip EURUSD move; deviation skipped 1 entry (too fast) | +$3,920 |
| Week 5–6 | Quiet markets; 19 signals, provider discipline held, no style drift | +$1,150 |
| Six-week total | 84 signals copied, worst sleeve DD −6.8%, zero margin scares | +$4,650 (+4.65%) |
Respectable, not miraculous — and that is the point. Copy trading done right is a boring, diversified return stream, not a lottery ticket. The FCA's guidance on copy and mirror trading makes the same point in regulator language: understand the strategy, understand the risks, and never copy what you could not explain. Six weeks in, both providers are still subscribed. The grid-trading "superstars" from my demo tests are not.
Copy trading questions I wish someone had answered honestly
How much money do I need to start copy trading on MT5?
Enough that the minimum copy size does not over-risk you. Most signal services suggest $500 to $1,000 minimum, but the real constraint is the formula above: your equity divided by the provider's equity, times their typical volume, must keep per-trade risk under 1–2% of your account. I started with $5,000 and still got hurt by ratio mistakes, so treat $2,000–$5,000 as a sensible floor for copying a serious provider.
Can I lose more than the provider loses?
Yes, easily. Slippage on your fills, wider spreads at your broker, copying at a higher ratio, and subscribing mid-drawdown all make your results worse than the provider's published numbers. In my NFP incident the provider lost 24% while I lost 20% — I got off lightly because my ratio was 1:1 and my broker's spreads held. Never assume your curve will match theirs; assume it will lag by the execution tax.
Should I copy a provider with no losing months?
No — that is the biggest red flag in copy trading. A strategy with zero losing months over a long history is either martingale/grid hiding floating losses, or the history is too short to matter. I want to see at least one honest 8–12% drawdown month handled with discipline: stops honored, size unchanged, no revenge trading. Recovery from a real loss tells you more than a year of smooth gains.
What is the deviation setting and what should I set it to?
Deviation is the maximum price difference you accept between the provider's fill and yours before your terminal skips the copy. Too tight and you miss fast-moving winners; too loose and you chase entries at terrible prices. I run 2–3 pips on majors and 30–50 points on gold. During news weeks I widen it slightly rather than disabling it — skipping a trade beats buying the top of a spike.
Can I run copied signals and my own EAs on the same account?
You can, but I stopped doing it. Copied trades and EA trades share margin, and a provider's surprise position can trigger your EA's risk limits or vice versa. Worse, closing trades manually gets confusing fast. I now isolate copied signals on their own account per the multi-account approach, with the EA running separately. Clean separation means clean attribution when something goes wrong.