CopyFX Rating Refresh: Six Strategies Added, Two Falls, and a Score That Reads the Full Curve

Six RoboForex strategies join our CopyFX rating today, nobody leaves, and the two accounts that fell hardest since our last update — Filmont and exactw — stay on the page with the scores their records now earn. One more change is bigger than any single row: from today the score is computed from each strategy's own figures every morning, not typed by hand, and it reads drawdown off the full daily curve rather than the number RoboForex prints. That last part is what reordered the table, so it comes first.

The drawdown figure RoboForex shows is not the fall you would have sat through

Every strategy card on RoboForex carries a "max drawdown", and that CopyFX drawdown figure is the first number most copiers read. We had been reading it as the worst peak-to-trough fall in the account's history. It is not. The platform's own daily drawdown series shows that the figure is the worst single step — the deepest one-day dip — and the strategy's full daily yield curve, which the card loads for its chart, tells a different story about most accounts.

Read 2026-09-09. The last row is the exception that explains the rest: S1's platform figure is larger than its curve fall, because a floating loss inside a single day can exceed any decline in closed results. The two figures measure different things. Our score now takes whichever is worse — and keeps the deepest fall it has ever recorded for an account even after the platform's own trailing windows have forgotten it. The reasoning, the components and their weights are on the methodology page.

Who fell

Filmont — Invictus Maneo was second on this page since we added it in August. Its all-time yield peaked at +741% on 11 August and stood at +337% by 4 September — a fall of 48% from the high in under four weeks, during which the trader also raised the fee to 45% of profit. Today it shows +371% all-time and −35% over the past month, still with 129 copiers. The score goes from 4.55 to 2.2. The account is not disqualified: a real funded account spread across gold, silver and several currency pairs, with a fast first eight months and now a real fall, is exactly the kind of record a rating should keep visible.

exactw — Albert Einstein lost 40% in a single week in January (from +117% on 21 January to +30% on the 27th) and sits 21% below its late-August high today, at +190% all-time. Its platform drawdown figure, 62.6%, was already the deepest on the page. From 3.85 to 2.3.

Vendo DD50 fell from 4.6 to 3.4 without anything happening to it this month — it is up 2.7% over the past thirty days and has 367 copiers, the most of any strategy we cover. The whole move is the table above: in its first summer, between 26 August and 11 September 2024, the account went from +31% to −18%, a 37.5% fall that the platform's 9.05% figure never showed. Our note used to say its drawdown had "stayed under 10%". It had not, and the score now says so. It has been steady since.

S1 AUDCAD/EURGBP moves from 4.1 to 3.3 for a plainer reason: its return over the past twelve months is 18%, and the formula does not give a long, cautious record the benefit of the doubt a hand did.

One footnote to the August collapse: SARUJA AI left RoboForex's public rating altogether on 2 September, not under the platform's own delisting rule but, as far as the data shows, by the trader's hand. The trader's new account under the same name, 10K To 1M JOURNEY, had gathered 1,486 copiers eighteen days after opening on eight closed trades and a 39% drawdown. It stays on our page, at the bottom, with a score of 0.85 and a note that says why.

Who joins

The six were chosen from a screen of RoboForex's entire board — 5,102 strategies across five time windows — for a positive quarter, a real balance of the trader's own, at least fifteen copiers and eight months of history, then read against their full curves. Whether a trader offers a partner programme played no part; the platform lists thousands of strategies with one and we would not have needed a screen. Scores are today's output of the formula and will move with the figures from tomorrow. The worst fall below is the worse of the platform's own figure and the fall on the daily curve — the number the score actually uses.

Each name links to the strategy on RoboForex. Copier counts are as of 9 September.

Short Stop Is Law — new first place, 4.3. A gold-led breakout system (79% XAUUSD, 18% GBP/USD) with a stop on every trade, running since May 2025 at +175% all-time; the deepest fall in its history is 7.8%, in early December 2025. The trader, Kuropatkin, runs the same system on three more accounts — the youngest has 72 copiers and the highest return — and all four have single-digit falls, which is rare evidence that the results come from the method rather than one lucky run. Fee 20%, minimum $300, and the trader's own balance is modest.

Trade with stop loss — 4.1. Gold, GBP/USD and USD/JPY with a fixed stop on every position, as the name says, on a real account since December 2024: +153% all-time, +108% over the past year, 55 copiers, deepest fall 13.7%. The trader's description talks about neural networks; nothing public substantiates that, and the curve does not need it.

WALLSTREET GOLD 2 — 4.1. A gold-only robot that has run since March 2024, opening a few trades at a time with a stop and a target on each: +533% all-time, +277% over the past twelve months, 127 copiers, fee 25%. Its one deep fall — the 28.9% in the table — came in its first year, when it gave back most of its early gain over five months; the two years since have been a staircase rather than a spike.

CAZADOR DE PIPS — 3.9. Gold only, from a Spanish-speaking trader who publishes setup instructions for copiers and trades in measured bursts: +220% since May 2025, 109 copiers, a 16.5% setback last autumn recovered within weeks. Fee 20%; the minimum to copy is $1,200, higher than most.

SteadyWinner I — 3.9. The most diversified account we have found on the platform: a portfolio of forty breakout systems across gold (41% of trades), the US tech and S&P indices, the German index, currencies and oil, on a real ECN account with a $31,000 balance of the trader's own. +1,137% since November 2025, deepest fall 29.7% in its first weeks before the portfolio filled out, 18 copiers, fee 25%, minimum $2,000. It is under a year old, and the score says so.

Capital Logic — 3.5, the high-risk seat. A quantitative strategy across gold and the major currency pairs from a trader who describes themselves as a professional actuary, with $139,000 of their own money on the account and 347 copiers, 69 of them recent. +896% since January 2026, a 36.6% fall in five days in March, and the maximum fee the platform allows, 50% of profit. Fast, popular and volatile — the sort of account our position-sizing guide was written for.

Checked and left out: Diamondhill, which is the same expert advisor as Short Stop Is Law under another name; Lucky Cat, a cent account with a few hundred dollars behind a +371% year; and HONEY MAKER, the most-copied strategy on the whole platform with 804 followers and a total return of 16.5% in twenty months — a reminder that the crowd is not a signal.

What "computed daily" means, and what it does not

Every morning at 03:40 UTC the site reads each RoboForex strategy on the page off the platform's own data — all seven windows, the full daily yield curve, the balance, the copiers, the fee, the instruments — and re-runs one formula over it. The score, the risk and return gauges and the fee shown in the table come from that read. Our own record of each strategy is kept alongside, so a fall that rolls out of RoboForex's trailing windows still counts here, and a strategy that is renamed, changes its fee or disappears from the platform's rating is flagged the same day rather than noticed by accident weeks later, which is how every one of the changes above was found until now.

What stays editorial is the list. The platform carries more than five thousand strategies, and a rating is only useful when it is short. Which accounts are on the page is still a decision; how they score is not. The full weights are on the methodology page, and the diversification guide covers how to hold several of these at once without holding the same bet six times.

FAQ

Why did Vendo DD50's score drop if the account did nothing wrong? Because the score now reads the whole curve. The 37.5% fall in 2024 was always there; the platform's drawdown figure did not show it and our earlier score was typed from that figure. The account has been steady for two years since, which the score also reflects — it is still comfortably in the upper half of the page.

Filmont and exactw fell hard. Why keep them? A fall that has already happened is the most useful thing a rating can show. Both are real funded accounts with long, visible records; the reader who wants a fast, volatile strategy should see exactly what that has meant in practice. Removal is for accounts whose fall has gone past the point where recovery would need a return larger than anything in their own history, and neither is there.

Does the new score change the risk of copying? No. It changes what you know before you copy. Every figure here moves daily and every strategy on RoboForex is a leveraged CFD account; check the live card, size any single strategy against its worst fall rather than its best month, and do not hold six gold robots believing you are diversified.

Bottom line

Six strategies added, none removed, two big falls scored honestly, and a rating that from today re-reads every account each morning from the platform's own data and measures drawdown the way a copier would have felt it. The first place changes hands to a gold breakout system with a single-digit worst fall; the former leader keeps a good score and loses its halo; and the number RoboForex prints as "max drawdown" goes back to being what it is — one bad day, not the whole story.