Methodology

Every trader and strategy profile carries a badge that says where its numbers come from. The badge describes the source of the record, not how good the results are.

Live verified

A real-money account connected to SocialCharts and read live from the broker. Trades, balance and equity come from the broker’s records, not from the trader.

Paper

A connected demo or paper account, read the same way. Fills use real market prices on simulated money, and the badge stays on every number that comes from it.

Statement · consistency check

An uploaded HTML, CSV or XLSX statement is checked for internal consistency and turned into an analysis report for its owner. It is analysis only: it never gives a profile badge and cannot make an account copyable.

Only connected accounts can be copied. Strategies are signal sources. Their records are labeled by source and verification level:

• Verified by real fills: fills from the real accounts of people copying the strategy are aggregated anonymously and compared with its paper record, showing the number of real copies, average slippage and the gap between paper and real return. The badge is earned as copiers accumulate.
• Paper · sealed signals: the moment the engine issues a signal, it is written to the SocialCharts signal ledger with a timestamp and a hash chained to the previous signal, and locked before the outcome is known. The paper record is these signals executed at real prices on a paper account.
• Research record · imported: historical records from the strategy engine carry a separate label. They do not replace live verification or a sealed-signal record. Research records are evaluated separately by source and are not presented as live performance.

An idea consists of a direction, target, invalidation level and expiry, marked on a chart. Its direction can be buy, sell or neutral.

When an idea is published, its levels and a snapshot of the chart are sealed and can no longer be edited. Every publication is a new idea with its own seal and its own outcome. A follow-up may link to an earlier idea (“the first one was invalidated, this one continues it”), but each is scored on its own.

• The market writes the outcome: price reaching the target closes the idea as Target, reaching the invalidation level as Invalidated, and the end of the horizon as Expired. No outcome is entered by hand.
• A profile’s hit rate is computed only from these automatic outcomes.
• Risk/reward is calculated from the price at the moment of publication and shown on the card.
• Neutral ideas give no direction, so they are published but not scored.
• An idea linked to an open position carries a “Backed by a position” badge.

Trader and strategy lists open sorted by Proof & risk, never by raw return. The order weighs three things: the proof level of the record, its drawdown, and how long the results have held.

A short run of large gains on a thin record ranks below a longer, steadier one with smaller drawdowns. You can re-sort by return, lowest drawdown, profitable weeks, idea hit rate, history length or followers at any time; the default only sets where the list starts.

Profiles and the comparison page show how copying could have performed historically with your chosen allocation and settings. The simulation uses the copy engine’s lot rules and filters. It does not guarantee the same results on a real account.

• Lot modes: Proportional scales your allocation relative to the trader’s equity. Multiplier applies your chosen factor to the trader’s lot size. Fixed lot uses the same size for every trade.
• The curve is built from the trader’s closed trades at end of day, with commission and swap included and floating results excluded. Spread and commission are shown as a separate cost line together with the number of trades.
• Advanced settings apply as they would on a live copy: your own stop loss and take profit (used when tighter than the trader’s), direction, session window (all, London, New York), symbol filter, maximum lot per trade, maximum open positions and pending orders.
• The news filter opens no new trades from 30 minutes before to 30 minutes after high-impact calendar events.
• If your copy-stop rule would have triggered in the past, the simulation stops there and shows the date.
• The start date cannot precede the account’s verified history. This calculation uses only connected real-account and paper-account history; imported research records are excluded.

News is measured with three separate numbers, each under its own name, because they answer different questions.

Source reliability

Whether a source corrects its mistakes, reports first-hand or repeats others, and stays consistent over time. It is never judged by which way the price went.

Post-event move

How the market moved after the event, measured from the price at the moment of publication.

AI thesis accuracy

Whether the AI’s reading of the event, its direction and affected symbols, matched what the market did. It is scored separately and labelled as AI.

News and video calls are not sealed, because SocialCharts collects that content itself; their accuracy is measured from the price at the time of publication. A video or channel is scored only on explicit calls with a direction, target, invalidation level and horizon.

The identities of people copying a trader or strategy are not shown to anyone, including the trader. Only total counts and counts of new copiers are shared.

Individual copiers’ results are never attributed to them. Real fills enter strategy proof only as anonymous totals. Members may share their own copy reports, with their settings, if they choose to.

Part of the feed is produced by AI: session summaries, reactions when news breaks, weekly accuracy tables, calendar reminders, and news and video summaries. All of it carries the “AI · system” label and names its sources.

The “SocialCharts Analyst · AI” profile publishes ideas under the same rules as everyone else: target, invalidation, seal, and an outcome written by the market. Its hit rate is measured and shown the same way. AI also pre-checks publications and picks a daily “Editor’s pick · AI”; both are labelled.

SocialCharts does not run human-looking accounts, comments or likes, does not inflate follower or copier counts, and never shows machine content without its label.

Interface language, content language, country of residence, time zone and currency are separate settings. Choosing a language changes what you read, not the country you are treated as living in.

Your country of residence decides which features and partner broker links are available, because broker licences and offers differ by country. News is written natively in each language from a language-independent record of the event (facts, numbers, direction, symbols and sources), and numbers and direction are cross-checked between languages.