SolBinanceSpotPerpFundingCarryDeltaNeutral1H

Intra-Binance SOL Spot-Perp Funding Carry, Delta-Neutral Always-On (Long SOL Spot / Short SOLUSDT USD-M Perp, Funding-Gated, Clean Single-Exchange Basis Hedge, Hysteresis Low-Churn)

A MARKET-NEUTRAL, CROSS-VENUE-SCOPE (two venue suffixes on the SAME exchange) funding-carry strategy that harvests Binance perpetual funding while staying delta-neutral. It holds SOL on Binance Spot (BTCUSDT.BINANCE_SPOT-style: SOLUSDT.BINANCE_SPOT) and simultaneously shorts the SOLUSDT.BINANCE USD-M perpetual in equal notional. The perp short collects funding whenever funding is positive (perp at premium). This is deliberately the SINGLE-EXCHANGE form the prior cross-venue (Binance-spot / Hyperliquid-perp) carry failure explicitly recommended: the hedge is a CLEAN intra-Binance basis (the perp is mechanically pegged to spot via the funding mechanism), so there is essentially no inter-venue basis noise — the exact structural defect that killed the cross-exchange version. SOL is chosen over BTC because its perp funding runs structurally higher and more persistently positive than BTC's heavily-arbitraged funding, giving real headroom over fees. The book is NEAR-ALWAYS-ON with hysteresis gating to amortize the ~0.30% full-cycle round trip (spot 0.10%x2 + perp 0.05%x2) over weeks of funding accrual rather than churning entry/exit. Signal cadence on 1H bars + the real Binance funding_rate stream.

Implements the hypothesis exactly: a single-exchange (intra-Binance, two venue suffixes) delta-neutral funding carry that shorts the SOL perp to collect structurally-positive funding and hedges with equal-notional SOL spot. Routing the perp as primary is the critical correctness decision — funding_rates supplementary data keys to the perp symbol, so primary placement guarantees the funding series is loaded and the edge is testable; the spot hedge as a same-timeframe extra leg means calculate_signal reads a contemporaneous spot price (the base barrier defers until both legs have a bar at the same timestamp), eliminating the phantom basis that a stale T-1 leg would create. This is the clean single-exchange form the prior cross-venue (Binance-spot / Hyperliquid-perp) carry failure recommended: the perp is mechanically pegged to spot via the funding mechanism, so inter-venue basis noise — the defect that killed the cross-exchange version — is structurally absent. SOL over BTC because its funding runs higher and more persistently positive, giving headroom over the ~0.30% full-cycle round trip; the wide enter(0.0004 ~ 44% APR)>>exit(0.0) hysteresis keeps the book near-always-on so that round-trip fee amortizes over weeks of accrual rather than churning. Futures venue (BINANCE perp) is required because the funding-collecting leg is a short; leverage=1.0 because both legs are equal-notional and fully funded (delta-neutral carry needs no amplification). All pair entry/exit/sizing is centralized in calculate_signal with the single-instrument hooks inert, and the funding lag tolerance (9h) correctly exceeds the 8h settlement cadence. If funding data is unavailable the book stays flat — no price-only proxy edge swap. Fills the market-neutral and cross-venue-scope buckets on the engine-clean Binance path.

suspected backtest artifact: hedge/accounting non-netting on a 'delta-neutral' SOL spot-perp carry. The metrics are internally contradictory and physically impossible: total_return -1.003 (~-100%, an account blowup) coexists with sharpe +2.27, cagr +132%, all-positive annual_returns, and a max_drawdown of only 1.96% (a real -100% loss cannot show a 2% drawdown), plus end_unrealized_pct -6.60 (a large un-netted open position at the end). The book also posts impossible delta-neutral outlier days (+266% on 2021-02-14, +100% on 2020-10-30) and +-$40k single-trade PnL on a ~7% avg position (kurtosis 54, skew 2.5) while full-sample beta is ~0 (-0.0004) — i.e. the long-SPOT (CASH account) and short-PERP (MARGIN account) legs net on calm bars but break apart on fast-move bars, leaving residual directional exposure and a terminal un-netted loss. This is the SAME hedge/accounting artifact that killed the BTC single-venue cash-and-carry sibling: a high raw Sharpe on a market-neutral carry with a negative total return and an effectively >100% drawdown is an accounting artifact, not an edge. SYMPTOM stated; root cause NOT independently verified - recommend factory-gate review of two-account (CASH spot + MARGIN perp) leg netting on fast-move bars before this instrument/timeframe is retried. Not worth optimizing: the optimizer's Sharpe objective would fit the calm-bar carry while the fast-move artifacts and -100% blowup make the PnL untrustworthy. Not iterate here because the viable expression of this exact mechanism is ALREADY captured by a promoted sibling (EthSpotPerpFundingCarryHedgedNeutral4H, DSR 0.99, holdout ratio 1.198) - the delta-neutral spot-perp funding-carry edge is not lost. This is NOT a funding-uncreditable false negative (funding IS credited/accrued). A clean SOL version, if desired for its higher funding, should be rebuilt on the working ETH-sibling pattern/timeframe with verified leg-netting, not optimized from this hedge-broken 1H instance. FAILURE PATTERN: intra-Binance SOL spot-perp delta-neutral funding carry on 1H whose CASH-spot and MARGIN-perp legs fail to net on fast-move bars produces impossible outlier days (+100% to +266%) and +-$40k carry trades, yielding a positive-Sharpe but -100%-total-return blowup with an understated 1.96% drawdown - a hedge/accounting artifact identical to the BTC single-venue cash-and-carry; the spot-perp carry edge is real and already promoted on ETH, so this broken instance should be retired rather than optimized.

Market-neutral, intra-Binance SOL spot-perp funding carry on 1H bars. The perp SOLUSDT.BINANCE is the primary instrument (so the Binance funding_rates series keys correctly and is collected); SOLUSDT.BINANCE_SPOT is a same-timeframe extra leg whose contemporaneous close is delivered by the base cross-leg alignment barrier. When the perp's 8h funding is meaningfully positive (>= enter_funding_threshold), the strategy SHORTS the perp (to collect funding) and BUYS equal base quantity of spot (to hedge price), giving net SOL delta ~0. It harvests funding every 8h while the spot long cancels direction. The book is held near-always-on via wide enter>>exit hysteresis and torn down only when funding goes non-positive (exit_funding_threshold), a net adverse pair-PnL basis stop trips (max_basis_stop_pct), or a 20-day time cap is hit. A naked-leg guard flattens any one-sided book. The spot leg is only ever bought then closed, respecting BINANCE_SPOT's long-only CASH constraint. Three core parameters plus the hysteresis gap; per-leg sizing is capital-relative and equal across legs for exact delta neutrality.

Sound, well-motivated hypothesis: single-exchange spot-perp basis is cleaner than cross-venue; SOL funding runs richer than BTC; genuinely low-parameter; the mechanism is real and has a PROMOTED sibling (ETH 4H spot-perp carry)

Full-sample beta ~0 (-0.0004) and benchmark_meaningful=false confirm the intent is market-neutral; funding is read and credited

Low commission drag (commission_pct_of_gross 0.94%) — the carry itself clears fees on calm bars

Internally inconsistent metrics: total_return -100.3% (account blowup) vs sharpe +2.27, cagr +132%, all annual_returns positive, and max_drawdown only 1.96% (cannot coexist with -100% total return); end_unrealized_pct -6.60 shows a large un-netted open position at the end

Implausible delta-neutral outlier days: +266% (2021-02-14), +100% (2020-10-30), and +-$40k single-trade PnL on a ~7% avg position (kurtosis 54) — legs net on calm bars but break apart on fast-move bars

Matches the documented CASH-spot + MARGIN-perp non-netting hedge/accounting artifact that killed the BTC single-venue cash-and-carry sibling; not optimizable until netting is verified

total_return -100% vs CAGR +132%, DD 1.96%

internally consistent

+266% on delta-neutral pair

physically plausible

SolBinanceSpotPerpFundingCarryDeltaNeutral1H tried the cleanest form of the spot-perp funding carry — a single-exchange Binance SOL basis where the perp is mechanically pegged to spot — chosen for SOL's richer funding and to dodge the inter-venue basis noise that sank the cross-exchange carries. The hypothesis was sound, but the backtest returned physically impossible metrics: a -100% account blowup reported alongside a +2.27 Sharpe, +132% CAGR, and a 1.96% max drawdown, with +266% delta-neutral outlier days and ±$40k carry trades. The reviewer abandoned it as the same CASH-spot/MARGIN-perp non-netting artifact that killed the BTC cash-and-carry sibling, recommending a factory-gate review of two-account leg netting rather than a guessed fix. Critically, the edge is not lost — it is already promoted on EthSpotPerpFundingCarryHedgedNeutral4H — so this hedge-broken 1H instance should be retired and any SOL version rebuilt on the working ETH pattern with verified leg-netting.

Two-account delta-neutral carries (CASH spot + MARGIN perp) can fail to net on fast-move bars, producing a positive-Sharpe but -100%-total-return blowup with an understated drawdown — an accounting artifact, not an edge; the spot-perp carry edge itself is real and already promoted on an ETH 4H sibling, so a SOL version should be rebuilt on that working pattern with verified netting rather than optimized from this broken instance.

The analyst abandoned it at the backtest-review gate as a suspected hedge/accounting artifact: the CASH-spot and MARGIN-perp legs net on calm bars but break apart on fast-move bars, leaving residual directional exposure and a terminal un-netted loss — the same two-account non-netting artifact that killed the BTC single-venue cash-and-carry sibling — making the PnL untrustworthy and not optimizable until leg-netting is verified.

A market-neutral, single-exchange spot-perp funding carry on Binance SOL — long SOLUSDT.BINANCE_SPOT and short the SOLUSDT.BINANCE USD-M perp in equal notional — harvesting the perp's positive funding while the spot leg cancels price risk, deliberately the clean intra-Binance basis form recommended after the cross-venue carry failures, with near-always-on hysteresis gating and only 3 core parameters.

Its metrics were internally contradictory and physically impossible: total_return -100.3% (an account blowup) alongside Sharpe +2.27, CAGR +132%, all-positive annual returns, max_drawdown of only 1.96%, end_unrealized_pct -6.60, and implausible delta-neutral outlier days (+266%, +100%) with ±$40k single-trade PnL on a ~7% average position (kurtosis 54) — even though full-sample beta was ~0 and commission drag was low (0.94%).