PAPER · READ ONLY
AUTONOMOUS ALL-MODE PAPER PORTFOLIO
Every market and every mode is now on the paper book.
WAIT
The router is ranking every market and mode after costs.
CORE DIRECTIONALFLAT
WAITING FOR POSITIVE NET EDGEValidated routes trade normally; exploratory routes use one quarter of the risk budget and a 10× cap.
RANGE GRID · SIMULATEDFLAT
WAITING FOR A PAYING RANGEA bounded long-only ladder of limit fills, opened only when the step clears its own fees. Fills are reconstructed from five-minute samples, so it under-counts rather than invents round trips.
ARBITRAGE & OPTIONSSIGNAL ONLY
RESEARCHED, NEVER TRADEDBoth engines are scanned, costed and displayed, and neither ever opens a paper position. They stay signal-only until an atomic (arbitrage) and defined-risk (options) paper lifecycle exists.
VERLAUF · 14 TAGE0 ABSCHLÜSSE · 0 GEWINNER
Tageswerte aus dem Readiness-Report, Trades aus dem Paper-Buch. Ein Tag ohne Snapshots zeigt Abschlüsse, aber keine Equity — er wird deshalb nicht als Nullpunkt gezeichnet.
TIEF —HOCH —
NETTO0,00 €
GEBÜHREN0,00 €
BRUTTO VOR GEBÜHREN0,00 €
TREFFERQUOTE—
TAGABSCHLÜSSEGEWINNERNETTOGEBÜHRENEQUITY
—00———
ZULETZT GESCHLOSSENPAARSEITEERGEBNISGEBÜHRGRUND
NOCH KEINE TRADES—————
PAPER OPPORTUNITY BOOK0 ROUTES
Every mode is retained even when it is blocked, so WATCH, SIGNAL and COST VETO are visible instead of disappearing. SIGNAL marks a research-only route that will never be traded.
MODEPAIRSIGNAL / STRUCTURESTATUSNET EDGE
Adaptive validation: only a robust holdout can pass strictly and only an adequate one can pass on probation — a thin holdout is reported as THIN SAMPLE and never as a validated edge. Invalidated edges exit after 15 minutes.
ROLLING WALK-FORWARD EDGE
Strict passes and positive probation signals may enter.
BTC/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
ETH/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
XRP/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
SOL/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
ADA/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
DOGE/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
LINK/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
AVAX/EURLOADING
NO DATA- MODEL / HORIZON
- —
- WIN RATE
- —
- HOLDOUT / TRAIN
- —
- SAMPLE QUALITY
- —
- PROFIT FACTOR
- —
- STOP / TARGET
- —
Models are selected only on the older training window, then judged on unseen recent holdout trades. Fees, spread and a post-only execution buffer are deducted before the route can pass. A holdout of fewer than 10 trades cannot pass at all: it is reported as THIN SAMPLE, and such a route may only be taken from the exploration budget at a quarter of the normal risk.
CONTROLLED MARKET UNIVERSE
Only liquid EUR pairs enter the race.
BTC/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
ETH/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
XRP/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
SOL/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
ADA/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
DOGE/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
LINK/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
AVAX/EURFILTERED
NO DATA- BEST MODE
- —
- SIGNAL
- —
- SPREAD
- —
- NET EDGE
- —
A market is eligible only with at least €400k rolling execution-feed volume, a spread at or below 6 bps, a complete five-minute history and agreement with an independent venue. The cost model still charges the full live spread before any paper entry.
ALL TRADING MODES
Six engines are researched. Four of them trade the paper book.
SPOTREADY
Spot
HOLD- MARKET
- ANALYZING
- LEVERAGE
- 1×
- NET EDGE
- 0.0 bps
- ENGINE
- EXECUTABLE · PAPER
Long only · no liquidation
ISOLATED MARGINREADY
Isolated margin
HOLD- MARKET
- ANALYZING
- LEVERAGE
- 30×
- NET EDGE
- 0.0 bps
- ENGINE
- EXECUTABLE · PAPER
Long/short · dynamic 10×/30× cap · borrow reserve
PERPETUALREADY
Perpetual futures
HOLD- MARKET
- ANALYZING
- LEVERAGE
- 30×
- NET EDGE
- 0.0 bps
- ENGINE
- EXECUTABLE · PAPER
Long/short · isolated dynamic 10×/30× · funding reserve
GRIDSTANDBY
Grid
HOLD- MARKET
- ANALYZING
- LEVERAGE
- 1×
- NET EDGE
- 0.0 bps
- ENGINE
- SIMULATED · PAPER
Range only · bounded simulated ladder
ARBITRAGEMONITORING
Arbitrage
HOLD- MARKET
- ANALYZING
- LEVERAGE
- 1×
- NET EDGE
- 0.0 bps
- ENGINE
- SIGNAL ONLY
Research only · never opens a paper position
OPTIONSSIGNAL ONLY
Options
HOLD- MARKET
- ANALYZING
- LEVERAGE
- 1×
- NET EDGE
- 0.0 bps
- ENGINE
- SIGNAL ONLY
Research only · never opens a paper position
Spot, isolated margin and perpetuals compete for the single core paper position. Range-grid runs its own sleeve and holds one bounded, simulated ladder. Arbitrage and options are signal-only: they are scanned, costed and published, and neither has ever opened a paper position — an atomic arbitrage fill and a defined-risk option lifecycle are not implemented, so pretending to execute them would be fiction.
LIVE OPTIONS INTELLIGENCE
Expiry, strike, volatility and Greeks—not guesses.
The engine evaluates long call, long put, bull call spread, bear put spread, long straddle and iron condor. Greeks are model-derived from live underlying price, strike, time to expiry and mark IV. Only structures with bounded maximum loss are considered. This sleeve is signal-only: nothing here is ever opened on the paper book.
CROSS-VENUE ARBITRAGE
Every spread gets a direction and a full cost veto.
BTC/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
ETH/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
XRP/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
SOL/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
ADA/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
DOGE/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
LINK/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
AVAX/EUREDGE
NO ROUTE- GROSS SPREAD
- —
- ALL-IN COST
- —
- NET EDGE
- —
Binance, Kraken and Coinbase are scanned for the cheapest ask and highest independent bid. Fees, both spreads, the execution buffer and inventory rebalancing are all costed. This sleeve is signal-only: a cross-venue leg pair cannot be filled atomically here, so no paper position is opened on it.
SELECTED-MARKET INTELLIGENCE
Five specialists score ANALYZING.
CHAMPION VS CHALLENGER
Four shadow books, one snapshot, no shared cash.
BOOKLEVERAGE / REGIME GATEWINDOWNET P&LEXPECTANCYPFHITCOSTSMAX DDTRADESSAMPLE95% INTERVAL
CHAMPIONlive paper book · no regime gateALLNO DATA—UNMEASURED—————NOT ESTIMABLE
NO SHADOW BOOKS YETThe experiment tables are seeded on the first authorised tick after the migration runs.——————————
DECISION LOGNO DATA
NOTHING DECIDED YETThe decision log fills on the first authorised tick after the migration runs.
PROMOTIONINSUFFICIENT EVIDENCE
NO OUT-OF-SAMPLE RESULT YETA recommendation needs holdout trades. Until then the answer is that nothing is known, which is not the same as nothing working.
Leakage controls: one shared confirmed snapshot per tick, no same-tick entry and exit, decisions labelled with their walk-forward window before they are made, holdout-only promotion gates, an embargo between windows, and Šidák-corrected intervals. Survivorship: every pair the scanner offers is decided on and logged, including the ones that reject, so the decision log is not a record of winners only.
EXECUTION GATEWAY
Binance multi-product gateway with resilient market data and a cost veto.
Binance is preferred; Kraken takes over book and history when the hosting region cannot reach Binance, while Coinbase confirms the price
Perpetual paper routes use a conservative 2 / 5 bps fee model plus funding reserve
Spot baseline: fees + live spread + post-only execution buffer
COST VETO ACTIVE
Private Binance trading APIs, borrowing and real orders remain disabled. The current system executes simulations only; no website credential can withdraw funds.
SYSTEM MAP
Operational layers
01Market intelligenceCONNECTING
02Opportunity routerWaiting for feeds
03Risk governorIsolated paper risk · dynamic 10×/30×
04ExecutionPaper only · 0/2 sleeves used · no real orders
05Scheduled tickWaiting for health report
RISK CONSTITUTION
Every mode is visible. Every sleeve remains isolated.
0/2PAPER POSITIONS
- Core risk per entry is capped at 0.1% of cash over the stop distance
- Exploratory entries take a quarter of that risk and a 10× leverage cap
- One grid ladder at a time, deploying at most 2% of cash
- Arbitrage and options are signal-only and hold no capital
- No live credentials, withdrawals or borrowing
OPERATIONAL HEALTH
A stale system reads as stale, not as calm.
Waiting for the first health report.
Realized P&L booked by routes that passed walk-forward validation.
Exploratory entries are booked under their own strategy keys and never counted inside the validated statistics above.
Every closed paper trade is attributed to the mode and execution class that produced it. A grid ladder closes several fills from one entry, so both counts are reported.
Health is derived from the last completed paper tick, the age and status of the market scan, per-venue coverage and whether an open position can be marked. It reports statuses and ages only — never a credential, a connection string or a stack trace.