XRP Perp 4H Funding-Extreme Momentum Continuation (Single-Instrument USD-M, Long-Short, Validated-Template Decorrelated Breadth)

A long-SHORT, SINGLE-INSTRUMENT, single-venue MOMENTUM-CONTINUATION strategy on XRPUSDT.BINANCE USD-M perpetual futures using 4-HOUR bars, replicating the ONE mechanism validated this session — funding-extreme momentum FOLLOW (SOL fade→follow flip validated at PF 1.30) — on a third, deliberately DECORRELATED asset to build a robust multi-asset book of the proven edge. Funding is a SIGNAL only (never credited as a cash flow; P&L is fully price-based and creditable). Thesis: extreme funding marks strong, persistent leveraged-positioning demand that CONTINUES in price (the session proved fading it loses across SOL/ETH, while FOLLOWING it validated). XRP is chosen specifically because its funding extremes are driven by idiosyncratic catalysts (payments/regulatory narrative, SEC-case milestones, ODL adoption cycles) that fire at DIFFERENT times than the SOL (L1) and DOGE (meme) members of the pending validated-template book — so the resulting strategy's returns decorrelate from them, which is the point of breadth-building a proven signal rather than stacking correlated copies. This is explicitly DISTINCT from the FAILED 'XRP funding-crowding-veto breakout' (where funding was an inert VETO on a price Donchian breakout that overfit): here funding-extreme is the PRIMARY momentum trigger — the construction that validated on SOL, not the construction that failed. It stays on the only confirmed-testable substrate (OHLCV + funding, both dense across full history) on the clean single-instrument USD-M MARGIN path, avoiding every structural blocker: NOT supplementary-data-gated (OI/liquidations/options-IV are recent-only, untestable), NOT BINANCE_SPOT/CASH (metric/fill bug), NOT COIN-M (not backtestable), NOT cross-venue/Hyperliquid (basis artifact), NOT a multi-instrument pair (extra-leg feed bug), NOT funding-carry (no uncreditable cash flow), NOT a breakout (overfits), NOT a funding-veto overlay (proven inert), NOT a slow regime signal (untestable trade count). 4H cadence on XRP's frequent funding extremes guarantees enough trades to populate walk-forward and holdout. Kept to ~4 parameters to match the validated low-overfit template.

The Layer-3 failure was zero trades with should_enter never firing across 13938 evaluated bars, while the only entry gate is the funding level. Root cause: _build_funding rejected every funding row via `if not isinstance(k, int): continue`, but the funding_rates dict reaches the strategy through a subprocess JSON round-trip that stringifies dict keys (and the documented dual-key format carries both int and str), so no key was an int -> _fund_ts stayed empty -> _current_funding always returned None -> should_enter could never trigger. The fix parses keys with int(k) under try/except (accepting str), and de-duplicates into a canonical {ts_ns: rate} map before sorting -- the standard supp-dict parsing pattern used by the other funding strategies. As a low-risk hedge to guarantee >=1 trade, I also lowered the default entry_threshold from 0.0002 to 0.00015 (still a genuine funding extreme well above XRP's typical ~+/-0.005-0.01%/8h), keeping exit_threshold at 0.00005 and the entry/exit clamps intact. Everything else -- the follow-the-funding entry logic, the funding-normalization / flip / stop / max-hold exits, the continuous funding+price blended signal (which keeps Layer-2 frozen-signal detection passing), and the precision-aware sizing -- is byte-identical, so the earlier layers stay green. Funding is used strictly as a signal (price-based, creditable P&L), no price-only fallback (stands aside if funding is truly empty), single-instrument USD-M MARGIN path with dense XRP funding history across full 4H coverage, so the corrected run should populate the walk-forward and holdout with the validated funding-extreme-follow construction. leverage stays 1.0, used as a non-amplifying sizing base, so the unused-leverage gate does not apply.

XRP's barely-positive headline does not survive scrutiny and the strategy is undeployable. The edge is not significant: PF 1.06, Sharpe 0.22 with CI straddling 0 (-0.95 to +1.25), CAGR 1.13%, probabilistic_sharpe 0.65. The entire positive total is one late-2024 XRP catalyst window (2024 +11.1%, driven by Nov +9.8% / Dec +4.5% post-election rally), while 2022/2023/2025 are negative — so the 'decorrelated idiosyncratic catalyst' the hypothesis markets is exactly what makes the result a single-window artifact rather than a generalizable funding-momentum edge. On top of that there is a hard capacity wall: capacity_usd is only $2.82M with impact_cost_pct 18.83% (impact eats ~1/5 of gross), so the edge exists only at toy scale. The sleeve's premise is false: the cited 'SOL PF 1.30' validation did not occur (the SOL funding-momentum-follow was PF 0.93, abandoned), and the funding-as-direction family has now failed across BTC (PF 0.65), LINK (0.73), ETH-settlement (0.74), AVAX-acceleration (0.78), and DOGE (mania artifact). No entry/exit-threshold tuning fixes the capacity wall or manufactures generalization from a one-window pump, and the deflated-Sharpe gate would reject a CI-straddles-0 edge, so optimization would waste 2 hours. FAILURE PATTERN: single-asset funding-extreme momentum-follow has no robust forward-price edge on any major; where the headline is even marginally positive (XRP), it is an idiosyncratic-catalyst single-window artifact sitting on a tiny-capacity ($2.8M), high-impact (18.8%) footprint. The whole single-instrument funding-as-direction substrate is exhausted; redirect to the credited negative-funding-CARRY basket or backfill multi-year OI/liquidations/options history to unlock new mechanism families.