Paper trading

Paper trading is the stage where a strategy proves itself forward, on market data it has never seen, before any real capital is involved. It's the supported way to run Forven.

• Live market data. Paper strategies read the live Binance feed — the same research series their backtests ran on.
• The backtest's own engine. Entries, exits, stops, sizing, fees and funding run through the same execution kernel as the backtest.
• Nothing leaves your machine. Every fill is simulated locally at the live mid price with modeled fees and funding. No order is sent to any exchange, and you don't need exchange keys.

New here? Follow Paper-trade a survivor.

Each paper strategy trades its own isolated $10,000 book.

How a strategy gets to paper

A strategy reaches paper through the door into paper at the end of the gauntlet. Paper risks no capital, so the default research door only stops strategies that are broken or clearly losing: 20+ confirmation trades, a non-negative return, out-of-sample profit factor of at least 1 and Sharpe above 0, drawdown and tail-risk floors, and no loss under 2× costs. Every gauntlet test still has to run; the deflated Sharpe, the robustness verdicts and the one-shot held-back test are recorded here and enforced before live. Strict paper (Settings → Lab → Door into paper) restores the old door, where all of them block paper. See Promotion gates.

Promotion is automatic by default; you can switch it to manual review in Approvals. An operator can also move a strategy with the stage control on its page — a gate refusal can be overridden there, and the override is logged.

The paper pool

Paper holds at most 40 strategies at once — at most 3 on one coin and 10 from one family — so every one of them is evaluated on each bar close without starving the rest of the machine. A strategy that clears the door while its slot is full waits in the paper queue, ranked by the evidence the door recorded (deflated Sharpe, robustness score, out-of-sample Sharpe and the breadth test). Waiting isn't a rejection, and the best-ranked strategy gets the next free slot.

Losers leave on their own. Every hour Forven retires a paper strategy whose forward record is clearly losing — 15 or more closed trades with a negative average and a t-stat below −1 — or that hits a 15% forward drawdown, goes 60 days without a trade, or trades in a way that contradicts its backtest. It's archived as paper_retired:<reason>; live strategies are never touched. Every limit is in Settings → Lab → Paper pool.

Multi-coin paper strategies

Each paper strategy is tested on 15 coins with its rule unchanged (the breadth test). If the verdict is general — it made money on most of them — the strategy also paper-trades the coins where its rule worked: its home coin first, up to 6 in all. It's still one strategy with one record. Each coin gets an equal share of its $10,000 book, so it never risks more than it would on one coin, and trades closed on the same day count as one day of evidence. The coin list is chosen by the breadth test, never by hand, and frozen. Going live with several coins isn't supported yet; such a strategy stays in paper. Switch it off with Multi-coin paper strategies in Settings → Lab → Paper to Live gates.

There is no separate "start a session" step: every strategy at the Paper stage trades on this desk — as long as the system mode is Semi or Auto. In Manual, the scanner that evaluates paper strategies is paused along with all other scheduled work, so they don't act on new bars (manual orders from the desk still work).

The desk

Status line and account strip

The status line confirms what you're looking at — Paper books: simulated fills at the live mid · no real orders · account halts don't apply to paper — when the scanner last ran and when it runs next, and whether prices are streaming.

The account strip totals every paper book: Paper books (how many, started at $10,000 each), Open P&L, Realized, net of costs over 7 days, 30 days and all time, Positions, Win rate, 30 days, Risk at stops, and the best and worst Books.

Needs attention

Anything about a paper strategy that needs you: a strategy that hasn't been evaluated recently, an exit that was refused, entries blocked again and again.

Paper strategies rail

Every paper strategy with its state — In position, Watching, Blocked, Stale, Exit refused — its result since it started, its market, timeframe, mode and leverage, its book, the reason behind its last decision, and the countdown to its next bar. Filter with All, Positions, Blocked, Watching and Archived. Use ↑ ↓ to move between them.

Chart

The selected strategy's market with its timeframe chips, and overlays you toggle: Fills, Strategy signals, Refused (entries or exits the engine declined, with the reason) and Indicators. Price lines mark the entry, stop, target and trailing stop. The market strip above shows Funding · hourly, Open interest (collected for BTC, ETH and SOL) and the Premium. Drawing tools: H-line, Trend, Reset view, Clear.

Side panel

Four tabs, keys 1–4:

Blotter

Tabs Positions, Levels (stops and targets watched on the live mark), Fills, Decisions, Performance and Risk, scoped to This strategy or All paper. Keys: [ ] switch tabs, A switches scope.

Manual control

Strategies manage their own entries and exits, but you can step in from the Position tab.

Open a position. The Manual order ticket (marked Paper · simulated) takes a direction, a size — by risk % of the book, by dollars or by coin — an optional stop (a 2× ATR suggestion is one click away), an optional target and leverage. It shows the size, order value, margin, the loss if the stop fills and the estimated fee before you commit.

Review long … opens a confirmation before anything happens.

Manage it. The position card offers Close, Close 50%, Close 25%, Stop to entry (once in profit), Flip and Pause auto / Resume auto, and lets you move or clear the stop and set or clear a target. Every action asks for confirmation.

Stops and targets are filled when the live mark touches them.

Paper → live

Paper is where Forven stops by default. The paper → live gate waits for at least 20 closed trades and 30 days on paper, then runs a sequential evidence test: the forward record must not be significantly worse than the backtest, backtest and forward evidence together must clear a bar that rises with the number of strategies in paper, and the forward Sharpe must be positive. It also enforces everything the door into paper only recorded — the deflated Sharpe, the gauntlet verdicts and the one-shot held-back test — plus a drawdown under 15%, beating buy-and-hold and a simple trend rule over the same days, and a breadth test that isn't home only. Going live then needs a typed GO LIVE and a notional ceiling. Live trading is unsupported; see Live trading and Live safety before considering it.

Caveats

• Fills are simulated at the live mid with modeled fees and funding. A real exchange adds queue position, partial fills, spread changes and outages.
• A few weeks of paper is a small sample. Judge behavior against the backtest, not just P&L.
• Paper and backtest results are not predictions, and nothing here is financial advice.

Related

• The Forge — where strategies earn paper
• Promotion gates — every threshold on the way in and out of paper
• Risk — the Paper scope on Risk Command