AvaxFundingCarryDeltaNeutral8H

AVAX Positive-Funding Spot-Perp Cash-and-Carry, Single-Asset Delta-Neutral (Cross-Venue: long AVAXUSDT.BINANCE_SPOT + short AVAXUSDT.BINANCE USD-M perp, held for days while funding is richly positive, 3-parameter)

A DELTA-NEUTRAL, cross-venue funding cash-and-carry on a SINGLE high-funding alt, AVAX. Alts like AVAX carry structurally larger and more persistent perpetual funding than BTC/ETH during positive regimes (frequently +0.05% to +0.10% per 8h, i.e. 15-30 bps/day, versus ~1 bp/8h medians on BTC that L13 correctly flags as un-tradeable). When AVAX 8h funding is richly and persistently positive (longs paying shorts), the strategy simultaneously BUYS AVAX on BINANCE_SPOT and SHORTS the equal-notional AVAXUSDT USD-M perpetual on BINANCE, carrying zero net directional (delta) exposure while collecting the funding stream every 8h. It is held for DAYS, not cycled per funding stamp, so the funding accrual dominates the two-leg entry/exit cost. This is deliberately NOT a basket (avoids L12 rank-and-rotate fee death — no per-rebalance N-leg churn) and NOT a thin spread-arb (L1) — it captures a directional funding cash flow whose stated median accrual (15+ bps/day) exceeds the ~0.30% two-leg round trip within ~2 days. It reuses the ONLY carry mechanism in the portfolio that has reached paper (BTC positive-funding spot-perp carry, Sharpe 8.58), but on a materially different instrument with a richer funding profile, and fills the two most under-represented buckets at once: cross_venue (7.1% vs 15% target) and market-neutral direction (diversifying from the 85% long-only concentration).

Implements the hypothesis directly and reuses the proven carry machinery from the SOL/DOGE/AVAX basket (funding parse, searchsorted lookup, timestamp-aligned two-leg management, naked-leg guard) that already reached Layer 4 — so layers 1-3 remain satisfied. The single change that matters is the fix the basket's Layer-4 QA review demanded: the basket gated at 0.0001 (~3 bps/day), 1/5th of the extreme-funding floor, and empirically produced avg_trade_return_pct -0.0337% because ~35 bps gross over an 11.5-day hold could not clear the ~0.30% cross-venue two-leg round trip. Here entry_gate is raised to 0.0005 (0.05%/8h = 15 bps/day), exactly the L13/L1 viability floor the hypothesis is grounded on: a multi-day hold at that rate collects 45-170+ bps, decisively above cost. Scoping to a SINGLE asset (AVAX) also fixes the basket's second flaw — 3 concurrent pairs consumed ~90% of cash and ~1.8x gross; one pair at capital_frac 0.50 keeps spot-long at 50% of cash-account equity and 50% short-perp notional, a bounded, delta-neutral profile. Venue routing is correct: primary is the AVAX USD-M perp (MARGIN account gives a stable equity read and is the funding-accounting leg); the spot long is the cross-venue hedge on BINANCE_SPOT (CASH, long-only — the strategy only ever BUYS spot, never shorts it). leverage is 1.0 and referenced in sizing, so no leverage_set_but_unused gate fires. The signal is continuous (the trailing funding mean, varying every bar) and the entry/exit decisions are thresholded against that same per-8h unit. Fills the under-represented cross_venue and market-neutral buckets with low turnover, avoiding the intraday commission churn and drawdown deaths of directional strategies.

fee/edge failure with no edge at all (L22): Sharpe -0.52, profit_factor 1.0036, win_rate 0.50, avg_trade_return_pct -0.0058% (negative, below the 0.15% futures floor), total_return -1.14%. On top of the negative per-trade edge, impact_cost_pct is 69.3% and capacity_usd is only $208k — market impact consumes ~70% of gross PnL and any edge exists only at toy scale, a hard deployment blocker. Same structural cause as the ETH+SOL sibling: the two-legged spot+perp construction pays ~0.30% cross-venue round-trip to synthesize delta-neutrality, and AVAX positive funding does not net clear it — here it's outright negative. The developer already set entry_gate to the extreme-funding floor (0.0005/8h) per the docstring, so threshold tuning is exhausted, and the tiny capacity shows no scale rescues it. This is NOT a funding-crediting problem (funding is credited; mechanism proven in promoted siblings) — the positive-funding two-legged expression is inherently fee/impact-disadvantaged on this instrument. Abandon at review rather than spend a 2-hour optimization on a negative-Sharpe, zero-PF result.

Single-asset, delta-neutral, cross-venue positive-funding cash-and-carry on AVAX. On 8-hour bars it reads AVAX perp funding from supplementary data (per-instrument series with fallback to the primary accounting series) and, when the trailing 3-point funding mean is richly positive (>= entry_gate = 0.05%/8h ≈ 15 bps/day with >=2 points for persistence), simultaneously BUYS AVAXUSDT.BINANCE_SPOT and SHORTS the equal-notional AVAXUSDT.BINANCE USD-M perp. The short perp collects funding every 8h while the long spot cancels price risk (delta-neutral). The pair is held for days until funding normalizes (trailing mean <= exit_gate = 0.01%/8h) or a 40-day timeout, so multi-day funding accrual dominates the two-leg round-trip cost. A naked-leg guard flattens both legs if one ever exists without the other. Funding is parsed once into sorted numpy arrays; per-bar work is one O(log n) searchsorted plus a small-window mean. Three tunable knobs: entry_gate, exit_gate, capital_frac.

Backtest engineer: (1) confirm end_unrealized sums BOTH legs; (2) assert funding booked to the short perp is POSITIVE while funding>0. Resolve both before judging the edge — blocks the whole carry family.

CROSS-CUTTING, NOW SEEN IN 4 INSTANCES: every version of this spot-perp positive-funding carry (SOL/DOGE/AVAX basket, ETH+SOL, and now single-asset AVAX) reports a NEGATIVE total_return driven by a large NEGATIVE end_unrealized_pct (-6.46% here; -9.34% and -10.30% on siblings) on a supposedly delta-neutral book. A truly hedged pair should mark near zero. Two candidates the analyst/backtest engineer MUST resolve: (a) metric artifact — end_unrealized captures only the perp MARGIN leg's unrealized loss while the offsetting spot CASH gain sits elsewhere, understating total_return; or (b) real — the engine isn't crediting positive funding to the held SHORT perp, so the carry cash flow that is the entire edge never lands. Realized expectancy is positive (+$220/trade, PF 1.069, Sharpe 5.16) with near-symmetric largest_win/loss (+$48.6k/-$47.7k), consistent with (a), but the consistency across all four instances demands proof.

Evaluate on total_return / Sharpe / funding over the full window; abandon at BACKTEST_REVIEW if total_return stays negative once end_unrealized is correctly attributed.

avg_trade_return_pct (-0.00785%) is a per-LEG artifact for this delta-neutral pair and must not be the fee gate — its sign disagrees with the positive avg_trade_pnl_usd (+$220). Economic edge = funding captured net of fees (total_return / Sharpe / funding). Sample small (46 trades / 205 days).

Proven mechanism family (funding carry) with promoted siblings; funding is credited so the carry is real in-backtest

Exits/hedging function: avg_holding 7.7d within the 40d cap, max_drawdown 1.14%, commissions only 1.09% of gross

benchmark_meaningful=false correctly flags market-neutral

Negative risk-adjusted result: Sharpe -0.52, sortino -0.25, omega 0.82 (<1), total_return -1.14%

profit_factor 1.0036 (zero edge) and win_rate 0.50 (coin flip)

avg_trade_return_pct -0.0058% — negative and below the 0.15% futures floor (L22)

impact_cost_pct 69.3% and capacity_usd only $208k — impact eats ~70% of gross and the edge exists only at toy scale; hard deployment blocker

Entry gate already at the extreme-funding floor (0.0005/8h) — threshold tuning exhausted; two-legged construction structurally cannot clear the ~0.30% round-trip on AVAX funding

AvaxFundingCarryDeltaNeutral8H narrowed the failed ETH+SOL positive-carry basket to a single high-funding alt on the thesis that AVAX's structurally hotter, more persistent funding would clear the two-leg cost, gating entry at the extreme-funding floor and reusing the mechanism behind the promoted BTC carry. The hedge and funding accrual worked exactly as designed — perfectly delta-neutral, tidy 7.7-day holds, tiny drawdown and low commissions — but the economics were worse, not better: zero edge (profit factor 1.0036, 50% win rate), a negative per-trade return, and a compounding liquidity problem where impact consumed ~70% of gross at a toy $208k capacity. The reviewer abandoned it on the first iteration as a fee/edge failure with a hard capacity blocker, noting the same structural cause as the ETH+SOL sibling and that both threshold tuning and asset choice were exhausted. The funding-carry mechanism is proven for single-leg capture, but its cross-venue delta-neutral positive-funding form simply cannot clear its own costs on these alts.

The positive-funding two-legged delta-neutral expression is inherently fee- and impact-disadvantaged versus single-leg funding capture, and a richer-funding alt does not fix it: AVAX's carry failed to clear the ~0.30% round-trip and its thin liquidity made impact eat ~70% of gross at a $208k capacity — the same structural cause that killed the ETH+SOL sibling, so no asset swap or threshold tuning rescues the construction.

The backtest reviewer returned 'abandon' at the pre-optimization gate for a fee/edge failure with a hard deployment blocker: the two-legged spot+perp construction pays ~0.30% cross-venue round-trip to synthesize neutrality and AVAX positive funding does not clear it (here outright negative per trade), the entry gate was already at the extreme-funding floor so threshold tuning was exhausted, and market impact consumed ~70% of gross with any edge existing only at toy $208k scale.

A delta-neutral, cross-venue positive-funding cash-and-carry on a single high-funding alt (AVAX): when AVAX 8h perp funding ran richly and persistently positive, it bought AVAX on BINANCE_SPOT and shorted the equal-notional USD-M perp on BINANCE to collect funding with zero net delta, holding for days with the entry gate set at the extreme-funding floor (0.05%/8h) — a single-asset variant of the promoted BTC positive-funding carry, 3 parameters.

With full data (40 trades, funding credited, metrics_reliable=true) the hedge worked (beta -0.0001, avg holding ~7.7 days within the 40-day cap, max drawdown 1.14%, commissions only 1.09% of gross) but the result had no edge at all: Sharpe -0.52, profit factor 1.0036, win rate 0.50, avg_trade_return_pct -0.0058% (negative, below the 0.15% floor), total return -1.14%, with impact_cost_pct of 69.3% and capacity of only $208k. No optimization stage ran.