LinkSpotDaily30BarHighBreakoutVolumeConfirmLong

LINK BINANCE_SPOT Daily 30-Bar High Breakout Long-Only with Volume Confirmation (Cross-Venue + Cross-Instrument Port of Paper-Stage Sharpe 4.37 Mechanism)

Direct cross-instrument and cross-venue port of the proven DogeDaily30BarHighBreakoutLong mechanism (currently paper_stage at Sharpe 4.37) applied to LINKUSDT on BINANCE_SPOT venue. This is NOT a new mechanism class — it is a deliberate replication test of an empirically-validated 3-parameter Donchian breakout with volume confirmation, applied to a different instrument on a different venue. Hypothesis: if the 30-bar-high + volume-multiplier mechanism captures a real momentum-confirmation edge on DOGE perp futures, it should generalize to LINK on Binance spot — same mechanism, different participant mix (spot holders vs perp leveraged traders), different fees (0.20% RT spot vs 0.10% RT perp), different asset (mid-cap utility token vs memecoin). LINK is chosen specifically to address the cross-instrument-generalization failures observed earlier this session (ETC 4H Volume Breakout, DOGE 4H Volume Breakout, both abandoned due to lower-liquidity microstructure on alts). LINK's daily spot volume on Binance is ~$50-150M (vs ETC's ~$5-15M and DOGE-spot's parabolic distortion), well above the 'liquidity floor' threshold for breakout viability. LINK is also NOT a memecoin (no narrative-driven parabolic reversals), NOT BTC/ETH (so it's not a re-test of already-covered majors), and has continuous Binance spot data from 2018 (8+ years for robust walk-forward). The strategy is pure OHLCV — NO supplementary data required (avoids the 13+ session abandons from liquidation/funding infrastructure gaps). 3 parameters total: lookback=30, volume_multiplier=1.2, atr_stop_mult=2.0. Volume confirmation requires breakout bar volume > 1.2× trailing 30-bar median to filter out 'thin breakouts' where price grinds above the 30-bar high without conviction. Long-only by venue constraint (BINANCE_SPOT is CASH account — cannot short). Position sizing is volatility-targeted to keep risk-per-trade at ~1.5% of equity. Fills the BINANCE_SPOT venue quota gap (9.9% current vs 15% target).

Implements the hypothesis as a faithful replication test of an empirically-validated mechanism rather than a new edge. The three headline economic parameters are exactly as specified — lookback=30, volume_multiplier=1.2, atr_stop_mult=2.0 — with the volume filter using the trailing 30-bar MEDIAN (robust to single-bar crypto volume spikes that would inflate a mean and weaken the filter, per the hypothesis). Venue routing: long-only with no leverage and a pure-OHLCV edge → BINANCE_SPOT (CASH account, leverage=1), which also fills the stated BINANCE_SPOT quota gap; the strategy never submits a SELL except to close an existing long, respecting the CASH no-short constraint. Sizing is volatility-targeted (risk_dollars / (atr_stop_mult×ATR)) to hold per-trade risk near 1.5% of equity, and is capped at equity/price because a CASH account cannot buy more notional than its cash — without this cap, calm low-ATR regimes would demand >100% notional and clip. The fresh-breakout gate (prev_close ≤ level < cur_close) matters more on spot than on the DOGE perp because spot's ~0.20% round-trip is double the perp's ~0.10%: requiring each entry to be a genuine new breakout (not merely still-extended-above-the-high) avoids re-entry fee churn after a stop-out. Per-trade edge: LINK daily breakouts that clear a 2×ATR trailing stop and run typically capture multi-percent moves, comfortably above the 0.20% spot round-trip and the >0.25% spot viability floor. Verified locally through Layer 1 (static, clean) and Layer 2 (synthetic — all 6 scenarios pass, continuous distance signal varies every bar with no NaN/inf and no frozen-signal warnings).

Failed deflated Sharpe: DSR=0.0625, expected-max=8.62 over 225 trials — the optimized Sharpe of 5.34 is BELOW the best-of-N luck bar and statistically indistinguishable from selection noise (is_significant=false), with PBO 0.8439 indicating near-certain overfitting. Walk-forward is_overfitted=true: avg IS 4.85 collapses to avg OOS 0.614 with windows [0.0, 1.84, 0.0] — the optimizer over-selected (lookback 30->45, volume_multiplier 1.2->1.8, max_hold 30->15) so its chosen config produces ZERO qualifying trades in 2 of 3 out-of-sample windows, and the holdout failed with zero trades (holdout_sharpe 0.0). The favorable full-period sharpe_ci_low (0.59) does not correct for the 225-trial selection that DSR/PBO do and is overwhelmed by the zero-OOS/zero-holdout failures. This is exactly the thin-sample risk flagged at pre-optimization: 54 baseline (33 optimized) trades across a 3-window walk-forward leave too few out-of-sample trades to validate, and the optimizer fits noise. Not a tunable-region case: sensitivity is clean (no cliff to avoid) and the binding constraint is data quantity — a second pass cannot manufacture more daily breakouts and would hit the identical zero-OOS wall. Not revise_hypothesis: the breakout mechanism is already proven on the DOGE sibling at paper_stage (Sharpe 4.37), so the family is covered; LINK daily spot is simply a too-sparse target and the failure is thin-sample overfit (PBO 0.84), not structural decay of a working target — porting to a 4th instrument would repeat the same daily-breakout sparsity. FAILURE PATTERN: cross-instrument port of a proven daily breakout to a lower-frequency target (LINK spot, ~7-8 trades/yr) overfits in optimization — strong in-sample Sharpe but DSR ≈0.06, PBO 0.84, and the selected config produces zero walk-forward OOS trades in 2 of 3 windows plus a zero-trade holdout.

Long-only daily 30-bar-high breakout on LINKUSDT Binance SPOT with volume confirmation — a direct cross-venue/cross-instrument port of the paper-stage (Sharpe ~4.37) DogeDaily30BarHighBreakoutLong mechanism. Enters BUY when the current daily close makes a FRESH break above the prior 30-bar high (prev close was at/below the level, current close strictly above) AND the breakout bar's volume exceeds 1.2× the trailing 30-bar MEDIAN volume, while flat. Exits the long on the first of: an ATR(14) Chandelier trailing stop at 2.0×ATR below the highest high since entry (the core risk control, also serving as the initial stop), a breakout-failure (close back below the broken level), or a 30-bar time stop. Position sizing is volatility-targeted to ~1.5% equity risk per trade (stop distance = 2.0×ATR), hard-capped at 95% of cash since spot has no leverage. Pure OHLCV — no supplementary data.

Mechanism matches hypothesis exactly: fresh 30-bar-high breakout + volume-median confirmation + ATR chandelier exit, long-only spot; clean entry diagnostics (0 dropped)

Plausible realized metrics: Sharpe 2.21, max DD 19%, end_unrealized 0.36%, profit factor 1.57 — no artifact inflation

Less outlier/bull-regime-fragile than the abandoned ETC/DOGE alt breakouts: modest positive returns across most years; LINK is liquid (capacity $238M)

Direct replication of a proven paper-stage (Sharpe ~4.37) DOGE sibling mechanism

Only 54 trades over ~7 years (~7.7/yr) — thin for a 3-window walk-forward; optimizer may fit noise and DSR will be demanding

Edge concentrated in a few big winners (30% win rate, avg win ~3.7x avg loss, 15 max consecutive losses) — OOS/holdout could miss the trades that matter

sharpe_ci_low -2.59 straddles 0 (significance uncertain pre-deflation); total return modest (17.6% over 7y, CAGR ~2.3%)

Clean execution and sensitivity (0 cliffs); realized metrics (end_unrealized 0.19%), liquid instrument (capacity $2.1B optimized)

Mechanism is genuinely proven on the DOGE sibling (paper_stage Sharpe 4.37); LINK default-param baseline had a real (if thin) Sharpe 2.21

Failed deflated Sharpe: DSR=0.0625, optimized Sharpe 5.34 BELOW the expected-max luck bar 8.62 over 225 trials (is_significant=false)

PBO=0.8439 — near-certain overfitting of the parameter selection

Walk-forward is_overfitted=true: avg IS 4.85 vs avg OOS 0.614 with windows [0.0, 1.84, 0.0] — 2 of 3 OOS windows have ZERO qualifying trades after the optimizer over-selected (lookback 30->45, vol_mult 1.2->1.8)

Holdout failed: zero trades (holdout_sharpe 0.0)

Root cause is thin sample (54 baseline / 33 optimized trades) — exactly the risk flagged at pre-optimization; too few OOS trades to validate

LinkSpotDaily30BarHighBreakoutVolumeConfirmLong was a deliberate generalization test — porting the paper-stage DOGE daily breakout mechanism to higher-liquidity LINK spot to dodge the lower-liquidity microstructure that sank the ETC/DOGE 4H alt breakouts. It cleared the pre-optimization gate on a plausible if thin Sharpe 2.21, but reviewers flagged just 54 trades over 7 years as too sparse for a 3-window walk-forward. Optimization confirmed it: deflated Sharpe 0.06 far below the luck bar, PBO 0.84, an in-sample 4.85 collapsing to OOS windows with zero trades in two of three, and a zero-trade holdout. The analyst judged this thin-sample overfit rather than a tunable or structural case — the mechanism is already covered by its DOGE sibling, LINK daily spot just too sparse — and abandoned it; it reached optimization and analysis but was abandoned after a single iteration, never advancing to risk review.

A cross-instrument port of a proven daily breakout to a lower-frequency target like LINK spot (~7-8 trades/year) overfits in optimization: the binding constraint is data quantity, not a tunable region — too few out-of-sample trades to validate, so the optimizer fits noise (PBO 0.84) and its over-selected config produces zero OOS/holdout trades; a second pass cannot manufacture more daily breakouts.

After passing the backtest-review gate (verdict: optimize), the analyst abandoned it at the post-optimization stage: it failed deflated Sharpe (DSR 0.06 with optimized Sharpe 5.34 below the expected-max luck bar 8.62 over 225 trials, is_significant=false, PBO 0.84), the walk-forward was overfit (avg IS 4.85 collapsing to avg OOS 0.61 with the over-selective config producing zero qualifying trades in 2 of 3 OOS windows), and the holdout failed with zero trades.

A long-only 30-bar-high Donchian breakout strategy on LINKUSDT.BINANCE_SPOT daily bars (OHLCV only) requiring volume confirmation (breakout-bar volume > 1.2× the trailing 30-bar median) and exiting on a 2× ATR chandelier trailing stop, breakout failure, or a 30-bar time stop — a deliberate cross-venue/cross-instrument port of the paper-stage DOGE daily breakout mechanism (Sharpe ~4.37), chosen on LINK for its higher spot liquidity.

The base backtest over 2019-01-16 to 2026-05-18 showed a plausible but thin result across only 54 trades (~7.7/yr) — total return +17.6%, CAGR 2.26%, profit factor 1.57, Sharpe 2.21 (CI [-2.59, 4.90]), a 19.1% max drawdown, 30% win rate with ~3.7:1 win/loss and 15 max consecutive losses; optimization produced a high in-sample headline (Sharpe 5.34, +28.5% over 33 trades) that did not generalize.