Systematic trading researchWho survives in public copy trading? Evidence from long-lived forex signals on eight platforms
Huyen X. Vu · Quantex Research, AlgoTrade Pte. Ltd., Singapore · ORCID 0000-0002-2209-842X
Abstract
Copy-trading platforms let retail investors follow other traders' accounts. Followers mostly see the records that are still listed, so the survivors set what looks normal. On one day, 6 November 2026, we read the public lists of eight platforms by ordinary browsing: MQL5 Signals, SignalStart, ZuluTrade, Collective2, Neotech (formerly FXCE), eToro, FundSeeder and Myfxbook. We applied one filter, fixed before reading: real money, at least three years of history, maximum drawdown of 40% or less, a trade in the last 30 days, and no more than 12 months of history imported from before the listing. Three facts stand out. First, attrition is visible even among the oldest records: of SignalStart's 100 oldest signals, 27 have no current rank, and 11 of these show a loss of 90% or more. Second, age is often imported: 58 of those 100 SignalStart signals, and 19 of 28 long-lived low-drawdown MQL5 signals, carry more than 12 months of history from before they were listed. Third, the top of a leaderboard is often not retail real money: of 37 long-lived Neotech accounts, 26 are firm-funded and 2 are demo. After filtering, 31 records remain on four platforms (29 real accounts and 2 hypothetical Collective2 systems). Most show modest drawdown, and almost none has subscribers. At typical fees, a follower needs roughly US$2,250 to US$8,600 of capital just to cover the fees. We register a six-month forward watch of the 31 records to test whether margin use, lot growth, symbol count and stop-loss use predict which records stop.
Keywords: copy trading; social trading; survivorship bias; signal providers; retail forex; back-filled history
Full paper · PDF, 7 pages · v1.1
Free account: sign in with Google, LinkedIn, ORCID or email to download.
Download the full paper (PDF)Key findings
- Losses stay hidden. Of SignalStart's 100 oldest signals, 27 are no longer ranked; 11 of those show a loss of 90% or more.
- Age is often imported. 58 of those 100 SignalStart signals and 19 of 28 long-lived, low-drawdown MQL5 signals carry over a year of pre-listing history.
- The top is often not retail money. Of 37 Neotech accounts with three or more years of history, 26 are firm-funded and 2 are demos; the other 9, the ones that pass, are real cent accounts with very small balances.
- Few pass, and fewer get copied. Only 31 records pass a single plain filter (29 real, 2 hypothetical). Most have low drawdown and almost no subscribers.
- Fees demand capital. At typical fees, followers need roughly US$2,250 to US$8,600 before growth covers the cost.
| Platform | Records | Max drawdown, median (range) |
|---|---|---|
| SignalStart | 11 | 31.4% (1.9–38.7) |
| MQL5 | 9 | 21% (8–30) |
| Neotech | 9 | 9.4% (1.8–16.0); real cent accounts, very small balances |
| Collective2 | 2 | 19.2% and 23.8% (hypothetical) |
In the full paper
- Related research on copy trading and fund survival
- Data, the filter and the back-fill check, platform by platform
- Results: silent losses, imported history, who sits at the top, the 32 records, fee drag
- A registered six-month forward watch (hypotheses H1 to H4); first update after 6 Dec 2026
- Discussion, limitations and references
- Appendix A: all 31 records, each with a link to its public page on the platform
Related Quantex reports
- R1: How long do funds and trading signals survive?
- R2: Copy-trading signals across 8 platforms (the same 6 Nov reading, short version)
Declarations
- Type: working paper; not peer reviewed.
- Version 1.1 (13 Nov 2026): one SignalStart record was removed after a check showed no trade in the 30 days before 6 Nov; counts and medians updated.
- Use of AI tools: data collection, analysis and drafting were assisted by AI tools (Claude, Anthropic). The author checked the results and is responsible for the content.
- Conflict of interest: the author is the founder of AlgoTrade Pte. Ltd., which develops systematic trading models. No platform named here paid for or reviewed this paper.
- Funding: none.
- Comments: welcome at [email protected].
- How to cite: Vu, H. X. (2026). Who survives in public copy trading? Evidence from long-lived forex signals on eight platforms. Quantex Research working paper R3, v1.1. https://quantex.com.sg/research/who-survives-copy-trading/
Quantex Research is the research arm of AlgoTrade. Build a trading CV from your trade history with AlgoTrade Connect, free at algotrade.ltd.