All paper strategies · click a row for a price chart
Per-stack equity CSV (e.g. alts_paper_equity.csv, mr_paper_equity.csv, ma_trading_paper_equity.csv) → last snapshot each UTC day: realized+unrealized when both columns exist (else total_equity_usd), summed for the book. Daily book Δ = this row combined Book EOD minus prior row combined Book EOD; the first row uses inception Δ = that EOD (prior implicit $0). Sharpe / Sortino on consecutive daily simple returns of combined EOD equity, annualized with ×√365.25 (calendar days; not 252 trading days). Sharpe / Sortino are — until there are ≥2 return observations (≥3 EOD rows). Sortino uses downside deviation vs 0 MAR: sqrt(mean(min(0,r)²)). Sortino or Calmar show inf when the denominator is zero (no down days / no drawdown). Calmar = compound CAGR from first to last row’s date (calendar span) divided by peak-to-trough max drawdown on the EOD series. History file: results_paper/combined_book_daily.json. Do not add Funding + Realized + Unrealized for a total: funding is already inside cumulative realized for these stacks; total equity ≈ realized + unrealized only.
Shared results_risk/risk_state.json (manual risk) and results_risk/cross_algo_safety.json (auto: post-close tighten + SL halve — all combined Alts WS stacks read/write the same file). Alts WS bots honor risk when started with
--risk-strategy-key. Set RISK_CONTROL_TOKEN on paper_api and paste the same value below if POST is locked.
Loading…
Live snapshot from /combined/metrics · same risk file for all stacks in COMBINED_CORE_KEYS (see paper_api).
—
Newest first · Prev/Next for older trades
alts_paper_state.json / strategy state on disk (not a live socket from this page). Unrealized per leg is rounded to cents in state. Open table includes Mark = last mid/book price written into state.