Strategies

BTC Cash-and-Carry Term-Structure Basis Capture, Market-Neutral (Cross-Venue: Long BTCUSDT.BINANCE_SPOT + Short the Front BTC USD-M QUARTERLY Delivery Future, Harvest the Guaranteed Contango Basis as It Converges to Spot at Expiry, Roll Across Quarterlies, Multi-Week Hold, 3-Parameter)

Outcome: Abandoned

BtcQuarterlyCashCarryTermStructureBasisNeutralRoll

BTC Cash-and-Carry Term-Structure Basis Capture, Market-Neutral (Cross-Venue: Long BTCUSDT.BINANCE_SPOT + Short the Front BTC USD-M QUARTERLY Delivery Future, Harvest the Guaranteed Contango Basis as It Converges to…

Strategy report

Hypothesis

A MARKET-NEUTRAL, cross-venue CASH-AND-CARRY strategy that harvests the single most structurally-robust cashflow in all of finance and one this factory has NEVER tried: the term-structure basis of a DATED delivery future converging to spot at expiry. Long BTCUSDT.BINANCE_SPOT + short an equal-notional BTC USD-M QUARTERLY delivery future. Unlike the deployed funding-carry strategies (which capture the perpetual's 8h funding and are saturated across BTC + 8 majors in the pipeline), the quarterly's premium is NOT funding — it is a fixed calendar basis that is MECHANICALLY GUARANTEED to converge to zero at settlement. That convergence is the edge: it does not depend on any directional forecast, funding sign, order-flow read, or mean-reversion bet — all of which are confirmed-dead families this session (OI reversal -87%, funding-dip accumulation overfit/61%DD, taker-flow no-edge, liquidation feed too-short). It fills THREE of the most under-represented buckets at once: cross-venue (7.1% vs >=15% target), market-neutral / long-short direction (14.4% vs the push toward hedged), and macro_1w_plus horizon (0.6% — nearly empty). Executed on the deepest crypto book (BTC spot + BTC quarterly), so the 88%-impact / $129k-capacity death that killed the thin-alt basket cannot recur. Deliberately 3 parameters. Note: the calendar-spread pattern (long spot / short dated future) is an explicitly sanctioned cross-venue template (researcher.md pattern #6); quarterly klines for delisted contracts are archived on Binance Vision, so the backtest rolls across historical quarterlies via on-demand collection.

Implements the hypothesis exactly: a market-neutral cash-and-carry on the DATED quarterly basis (deterministic convergence at settlement), not perp funding. Fills cross-venue + market-neutral + macro_1w_plus buckets. The edge needs no directional/funding/flow forecast. Executed on the deepest crypto book (BTC spot + BTC quarterly) so thin-alt impact death cannot recur. Rolls across ~20 historical quarterlies (2022-2026) via on-demand Binance Vision collection; the graceful-drop data layer means any missing leg is harmless. Delta-neutral equal-notional sizing (linear USDT-margined; qty=notional/price), spot leg BUY-only (CASH-safe), naked-leg guard, path-independent capped sizing. leverage=1 (two ~45%-notional legs = ~90% gross, net ~0 delta) and is referenced in sizing. Three deliberate tunables: entry_basis_annual, min_entry_days, roll_buffer_days; permissive defaults ensure the roll fires within short sandbox windows given BTC persistent contango.

data_unavailable: backtest data for BTCUSDT.BINANCE_SPOT could not be collected after 5 retries: - bars (bar_type=BTCUSDT_261225.BINANCE-12-HOUR-LAST-EXTERNAL): got 0, needed 50, or 90+ days span The data layer (not the strategy) is the blocker — re-evaluate this hypothesis once collection is fixed / the symbol is backfilled.

Implementation

Market-neutral, cross-venue cash-and-carry that harvests the BTC quarterly delivery-future term-structure basis. Longs BTCUSDT.BINANCE_SPOT and shorts equal-notional the live nearest-expiry BTC USD-M dated quarterly (BTCUSDT_YYMMDD.BINANCE) whenever the annualized basis exceeds a threshold with sufficient convergence runway. Holds the delta-neutral pair as the basis mechanically converges to zero at settlement, closes both legs a couple days before expiry, then rolls into the next quarterly. Primary spot runs on 1-DAY bars; the quarterly chain runs on 12-HOUR bars (a different timeframe) so expired/unlisted contracts never deadlock the cross-leg alignment barrier. Unavailable legs in the chain are dropped by the data layer and self-filtered by a live-contract check.

Verification Results

CLEAN RESTART 2026-09-04 — this run's verdict history and learning records were removed and it was restarted from verification. Its previous abandonment came from the pipeline, not from the market: the Layer-2 harness mis-bound @staticmethod helpers (fixed), QA issued terminal performance verdicts on an unoptimized smoke test (removed — QA now judges correctness only), and sandbox timeouts came from backtest-slot starvation (fixed). The hypothesis and the strategy code are unchanged. Verify the code on its merits; performance is decided later by the full backtest and the optimizer.

Backtest and paper results are hypothetical. Trading involves risk of loss.