SolCrossVenuePositiveFundingCarryPersistentDeltaNeutral1H

SOL Cross-Venue Positive-Funding Carry — Long Spot + Short USD-M Perp on Persistently Extreme Funding (Delta-Neutral, Days-Held, Single-Leg-Per-Venue, 3-Parameter)

A DELTA-NEUTRAL, CROSS-VENUE funding-carry book on SOL that transplants the factory's single promoted cross-venue winner ('Cross-Venue BTC Funding Carry: Long Spot + Short Perp', Sharpe 8.5) onto an asset whose funding structurally accrues 2-3x faster than BTC. Mechanism: hold equal-notional long SOLUSDT on BINANCE_SPOT and short SOLUSDT USD-M perp on BINANCE, so price exposure nets to ~0 and PnL is dominated by the funding payments the perp short COLLECTS while positive funding persists. Enter only when 8h funding is BOTH extreme (> +0.05% / 8h, i.e. ~0.15%/day) AND has been positive for the prior 2 consecutive funding stamps (persistence gate, so we hold into an accruing stream rather than chasing a one-print spike). Positions are held for DAYS collecting multiple 8h funding stamps; the ~0.20% round-trip taker cost (0.10% futures + 0.10% spot) is paid ONCE per cycle and amortized across many funding accruals. This is deliberately NOT the abandoned ETH clone: SOL's smaller cap and heavier retail-perp leverage crowding push its positive-funding regimes to median ~0.08-0.15%/8h (24-45 bps/day) during crowding, comfortably clearing L1's 20-bps-per-cycle bar and L13's >15-bps/day accrual bar — the exact conditions BTC met and ETH's thinner funding did not.

Transplants the promoted BTC cross-venue funding-carry structure onto SOL, whose smaller-cap retail-leverage crowding drives structurally higher, more persistent positive funding (median ~8-15 bps/8h in crowded regimes) that clears the fee bar the thinner-funded ETH clone missed. Mechanism: equal base quantity on both legs makes price exposure net ~0, so PnL is dominated by the funding the short perp collects. The DOUBLE gate implements the hypothesis exactly -- (1) extreme +0.05%/8h threshold ensures per-cycle carry beats round-trip cost, (2) 2-prior-stamp persistence avoids chasing one-print spikes and biases toward an accruing stream held for days. Wide enter>>exit hysteresis keeps the book near-always-on so the fee amortizes across many 8h accruals. 3 core parameters (enter threshold, persistence stamps, exit threshold) keep the surface small. 1-HOUR bars act promptly on 8h funding updates; min_bars_required=3 keeps warmup negligible so the entry gate can fire early in the sandbox window. Leverage left at 1.0 (unused in sizing) since the delta-neutral book needs no amplified notional.

NOT a funding-crediting issue (funding is credited and was read: funding_events_available=91). The binding constraint is SOL funding-data coverage: only 91 funding stamps exist across the 2020→2026 backtest window, so the strategy trades only in a single ~30-day window in Oct–Nov 2020 and is flat for the remaining ~5.5 years, yielding just 7 trades. That sample cannot support the 3-phase pipeline (walk-forward's 3 windows and the 15-day holdout would hold ~0 trades), and no parameter tuning creates funding history that isn't collected. Compounding it, the delta-neutral premise is not achieved in the trades that did fire — return_skew -3.30, kurtosis 13.6, largest_loss -$21,042 vs largest_win $19,706, exposure 171%, PF 0.92, Sharpe -4.72, negative expectancy — large asymmetric swings a properly hedged carry should not produce, which my guidance flags as implausible cross-venue behavior to escalate rather than optimize on. The mechanism is already validated on BTC where full funding history exists. Recommend the Research Lead/Data Engineer (a) verify and backfill full SOLUSDT.BINANCE funding history before any SOL carry re-attempt, and (b) confirm the cross-leg hedge actually nets to ~0 delta on a diagnostic run; until SOL funding coverage is complete this asset cannot validate the carry regardless of parameters.

Delta-neutral cross-venue SOL funding carry: SHORT SOLUSDT USD-M perp (BINANCE, collects 8h funding) + LONG equal-notional SOLUSDT spot (BINANCE_SPOT, neutralizes price). Enters only when 8h funding is BOTH extreme (>= +0.05%/8h) AND persistently positive (current stamp + 2 prior stamps all positive), then holds for days collecting multiple funding accruals so the ~0.20% round-trip fee amortizes. Exits when funding turns non-positive, a basis-divergence stop trips, or a time cap is reached. Perp is primary so the funding series (keyed to the perp symbol) is collected; the spot hedge is a same-timeframe extra leg aligned by the base cross-leg barrier for a contemporaneous price. Long-only spot leg respects the CASH-account constraint; net SOL delta ~= 0.

For the analyst/developer: size the per-cycle carry target against ~0.30% round-trip, not 0.20%. Consider raising enter_funding_threshold or enforcing a minimum hold so accrued funding reliably exceeds ~0.30% before the position can be torn down.

FEE ASSUMPTION UNDERSTATED (materially affects the edge thesis, not a code bug). The hypothesis states a '~0.20% round-trip taker cost (0.10% futures + 0.10% spot)'. Real per-cycle taker cost from venue_configs is ~0.30%: futures round-trip = 0.05% x 2 = 0.10% (correct), but SPOT round-trip = 0.10% x 2 = 0.20% (the hypothesis counts only 0.10% for spot, i.e. one-way). So the true amortization bar is 50% higher than assumed. The backtest already charges the real fees (commission_pct_of_gross 1.66%), so the CODE is not wrong — but the economic margin the hypothesis leans on is thinner than claimed, which raises the required per-cycle funding accrual to clear costs.

Analyst: if the full-history backtest confirms avg_trade_return_pct < 0.15%, abandon as fee/edge (SOL funding did not accrue fast/persistently enough to clear the corrected ~0.30% round-trip over held cycles). If it clears on fuller data, proceed.

NEGATIVE SANDBOX EDGE — flag for BACKTEST_REVIEW (analyst gate, not a code defect). The Layer-3 smoke shows avg_trade_return_pct = -0.234% (below the +0.15% viability floor), total_return -1.72% over 25 days, Sharpe -4.94, on 7 trades / 65 funding events. This is consistent with the corrected ~0.30% round-trip fee out-running the funding collected over this window, i.e. a genuine fee/edge outcome. It is NOT an L17 code-defect signature: win_rate 0.43 (not 0), profit_factor 0.91 (not 0), |Sharpe| 4.94 (< 5), and the large per-leg avg_win/avg_loss (10.7k / 8.8k) are the two hedge legs correctly offsetting and netting to the small -0.234% — normal for a delta-neutral pair, not leaked directional exposure. On a 7-trade smoke sample this is too noisy for QA to abandon on; the full backtest + analyst review is the correct gate.

Optional robustness: verify spot-leg market readiness (_market_ready / _latest_spot_close not None) BEFORE submitting the perp leg, so the pair is opened atomically and never runs naked even for one bar.

TRANSIENT NAKED-LEG IMBALANCE (guarded, minor). long_trades=3 vs short_trades=4 means at least one cycle opened the perp SELL but the matching spot BUY did not become an open position (base template drops an extra-leg entry when its market isn't ready, and _submit_entry_instrument blocks CASH shorts). The strategy's own naked-leg guard (bool(perp)!=bool(spot) -> _close_pair) flattens the one-sided book on the next bar, so directional exposure is bounded to ~1 bar plus an extra entry/exit fee pair. Not a correctness failure, but it adds fee drag and a bar of unhedged risk on affected cycles.

Mechanism is proven: the BTC cross-venue funding carry sibling reached promotion at Sharpe 8.5, and funding is now credited by the engine so the carry is backtestable in principle

Design is reasonable: perp-primary so funding keys resolve, same-timeframe spot hedge on the alignment barrier, long-only spot leg (CASH-compliant), 3 core params

Only 91 funding stamps available across a 2020→2026 window (SOL should have thousands) — the SOLUSDT funding series loaded is nearly empty, so the strategy trades only in a single ~30-day window in Oct–Nov 2020 and stays flat thereafter

Just 7 trades total — far too few for walk-forward (3 windows) or the 15-day holdout; optimization would fit noise

Negative and non-neutral: total_return -1.72%, PF 0.92, Sharpe -4.72, expectancy -$425/trade, avg_trade_return_pct -0.23%

Delta-neutral premise not achieved in practice: skew -3.30, kurtosis 13.6, largest_loss -$21k vs largest_win $19.7k, exposure 171% — large asymmetric swings a working hedged carry should not show

SolCrossVenuePositiveFundingCarryPersistentDeltaNeutral1H transplanted the factory's one promoted cross-venue winner — the BTC spot-vs-perp funding carry (Sharpe 8.5) — onto SOL, whose smaller cap and heavier retail leverage should make positive funding accrue faster and clear the fee bars that the abandoned ETH clone missed. The design was reasonable (perp-primary so funding keys resolve, same-timeframe long-only spot hedge, 3 core params), but the SOLUSDT funding series loaded nearly empty at just 91 stamps, so the book traded in only one ~30-day window in late 2020 and produced 7 losing, non-neutral trades with a Sharpe of -4.72 and a ±$20k win/loss asymmetry no working hedge should show. The reviewer abandoned it as a data-coverage wall rather than a funding-crediting bug, recommending the funding history be backfilled and the hedge's zero-delta confirmed before any SOL carry re-attempt. The mechanism remains validated on BTC where full funding history exists; SOL simply could not test it.

Verify that the underlying funding history is actually collected before transplanting a data-dependent carry to a new asset — even a proven mechanism can't validate on 91 funding stamps, and a supposedly hedged book showing skew -3.30 and a ±$20k win/loss asymmetry signals the delta-neutral construction should be diagnostically confirmed to net ~0.

The backtest reviewer returned 'abandon' at the pre-optimization gate: the binding constraint was SOL funding-data coverage (only 91 stamps, when SOL should have thousands), leaving too few trades for walk-forward or holdout, and the delta-neutral premise was not achieved in the trades that did fire — large asymmetric swings the reviewer flagged as implausible hedged behavior to escalate rather than optimize.

A delta-neutral, cross-venue funding-carry book on SOL that held equal-notional long SOLUSDT.BINANCE_SPOT against a short SOLUSDT USD-M perp — a transplant of the promoted BTC carry — entering only when 8h funding was both extreme (>= +0.05%/8h) and had been positive for the prior 2 stamps, holding for days to amortize the round-trip fee across many funding accruals.

The backtest fired only 7 trades because just 91 SOLUSDT funding stamps were available across the 2020-2026 window, so it traded only in a single ~30-day window in Oct-Nov 2020 and stayed flat afterward, producing a negative, non-neutral result: total return -1.72%, profit factor 0.92, Sharpe -4.72, expectancy -$425/trade, with skew -3.30, kurtosis 13.6, exposure 171%, and a largest loss of -$21k against a largest win of $19.7k. No optimization stage ran.