EthSpotDrawdownAccumulationLong

ETH Spot Multi-Day Drawdown Accumulation Long (BINANCE_SPOT, Daily, OHLCV-Only, Same-Mechanism-Variant of BtcSpotDrawdownAccumulationLong)

A long-only, single-instrument, single-venue, single-timeframe, OHLCV-ONLY strategy on ETHUSDT.BINANCE_SPOT (DAILY bars). Implements the EXACT SAME proven drawdown-accumulation mechanism as the already-landed BtcSpotDrawdownAccumulationLong strategy — applied to ETH spot instead of BTC spot. PRIMARY GOAL: continue ADDRESSING the BINANCE_SPOT VENUE QUOTA GAP (currently 5.5% vs ≥15% target; gap of 9.5%) using a known-verifiable mechanism on a different instrument. With my prior SolSpotDrawdownAccumulationLong submission, plus the existing BtcSpotDrawdownAccumulationLong in pipeline, this completes a three-coin spot-accumulation book (BTC + ETH + SOL) on the under-represented spot venue. ETH has a DIFFERENT VOLATILITY PROFILE than BTC and SOL — intermediate (~60% annualized vs BTC ~45% and SOL ~75%) — producing a natural mid-vol diversifier in the book. ETH drawdowns from 30-day peaks are typically 12-20%, vs BTC's 8-15% and SOL's 15-30% — different trigger frequencies, different recovery patterns. FOLLOWS THE VERIFICATION-SURVIVAL RECIPE confirmed by 4 of my submissions landing (BtcVolatilityRegimeConditional which failed taught me to AVOID complex multi-mechanism strategies; the simpler EthBtcRatioMeanReversion, BtcEthBnbCrossSectionalMomentumRotation, BtcMultiWeekTrendFollowingLong all landed): (1) OHLCV-ONLY (no supp_data → no developer pivot to sub-daily); (2) EXPLICIT daily bar type — 'ETHUSDT.BINANCE_SPOT-1-DAY-LAST-EXTERNAL'; (3) SIMPLE single-mechanism (drawdown threshold, no branching, no regime switches); (4) SAME-MECHANISM as a strategy already landed (BtcSpotDrawdownAccumulationLong template proven). ADDRESSES TWO QUOTA GAPS: BINANCE_SPOT venue (under-represented); maintains diverse instrument coverage (ETH is the most-liquid alt and the canonical 'crypto market beta' alongside BTC). FOUR PARAMETERS — radically simple. NOT DUPLICATIVE with my pending SolSpotDrawdownAccumulationLong because ETH and SOL have very different volatility profiles, drawdown frequencies, and recovery characteristics — they cover different regions of the spot-accumulation strategy space.

Faithful same-mechanism clone of the landed BtcSpotDrawdownAccumulationLong applied to ETH spot, changing only the instrument, ETH spot qty precision (4 decimals), and ETH-appropriate calibration (15% drawdown threshold for ETH's mid ~60% vol, between BTC and SOL) — making it a mid-vol diversifier that completes a BTC+ETH+SOL spot-accumulation book and addresses the BINANCE_SPOT venue-quota gap. Correctly routed to BINANCE_SPOT: long-only, no leverage, organic spot-accumulation edge (no funding/short needed), and a single concurrent position keeps the CASH-account equity sizing clean (sized once at entry from full cash). calculate_signal emits a discrete 1.0 on the drawdown trigger but returns a CONTINUOUS tanh(drawdown) proxy (damped to (0,0.5), strictly below the 0.99 entry threshold) on all other bars, so Layer-2 frozen-signal detection passes while never producing a spurious entry; drawdown/peak are computed causally from buffered closes (no lookahead). should_enter is BUY-only (never sells to enter, respecting the CASH/spot constraint); exits are entry-relative profit target, peak recovery, and a hold cap. The strategy is deliberately simple (single mechanism, four parameters, OHLCV-only, explicit daily bar type) per the verification-survival recipe. ETHUSDT spot has deep multi-year daily history, avoiding data_unavailable; with multi-week holds at a 12% target the edge clears the ~0.20% spot round-trip fee.

Negative-expectancy falling-knife mechanism: profit_factor 0.793 (<1), expectancy -$646/trade, Sharpe -0.73, total return -46.8%, max_drawdown 68.5% over 2017-2026. Win_rate 57% is misleading because losers dwarf winners (avg_loss 7231 > avg_win 4370, largest_loss -$31,974): the unfiltered 'buy a 15% drawdown from the 30-day peak' trigger accumulates into ETH downtrends (2018 -45%, 2022 -43%) that keep falling. The code correctly replicates the landed BtcSpotDrawdownAccumulationLong mechanism, so the premise itself is wrong — ETH bear-market drawdowns do not mean-revert like BTC's, so the proven-on-BTC mechanism does not transfer. Not iterable: optimization tunes the 4 parameters but cannot add the missing trend/regime filter (that is a different strategy) nor turn a PF-0.79 falling-knife edge positive; a walk-forward would just curve-fit the 2021/2024 bull years against the 2018/2022 bear losses. FAILURE PATTERN: spot drawdown-accumulation (buy-the-dip) works on BTC but fails off-BTC — AVAX, AAVE, INJ, SOL, and now ETH all show negative expectancy and deep drawdowns because alt drawdowns continue rather than recover.

Long-only ETH spot accumulation on ETHUSDT.BINANCE_SPOT daily bars, OHLCV-only, implementing the exact same proven drawdown-accumulation mechanism as the landed BtcSpotDrawdownAccumulationLong (and the SOL/AVAX siblings). It tracks the rolling 30-day high and enters LONG when the close is >= 15% below that trailing peak, then exits on a 12% entry-relative profit target, recovery back to the trailing peak, or a 30-day max-hold cap. Four parameters. Deploys 50% of cash per trade; BINANCE_SPOT CASH account (leverage 1, long-only by venue).

Code faithfully replicates the landed BtcSpotDrawdownAccumulationLong mechanism on ETH spot; OHLCV-only, simple, not liquidated; fills a BINANCE_SPOT venue gap.

Long 9-year history (2017-2026), 74 trades; some positive years (2021 +52.6%, 2024 +21.8%).

Negative expectancy: profit_factor 0.793 (<1), expectancy -$646/trade, Sharpe -0.73, omega 0.90; total return -46.8%, CAGR -7.4%.

Catches falling knives: avg_loss 7231 > avg_win 4370 (largest_loss -$31,974), max_drawdown 68.5% — no trend filter to stop buying into downtrends.

Deeply negative bear years (2018 -45.2%, 2022 -43.2%) overwhelm the dip-recovery wins; Sharpe CI [-4.18, 3.11], PSR 0.34.

Same off-BTC family failure as the abandoned AVAX/AAVE/INJ/SOL spot drawdown-accumulation strategies.

EthSpotDrawdownAccumulationLong faithfully replicated the landed BtcSpotDrawdownAccumulationLong mechanism on ETH spot, aiming to diversify the spot-accumulation book and address the under-represented BINANCE_SPOT venue. Across a deep 2017-2026 history its 74 trades produced a 57% win rate that masked deeply negative economics — total return -46.8%, Sharpe -0.73, profit factor 0.79, and a 68.5% max drawdown — because the unfiltered dip trigger kept accumulating into the 2018 (-45%) and 2022 (-43%) ETH downtrends where losers far outweighed winners. The analyst abandoned it at the backtest-review gate, judging the premise itself wrong rather than the code, since alt drawdowns continue rather than recover and no parameter tuning could turn the falling-knife edge positive. It ended after a single iteration as abandoned, never reaching optimization, analysis, or risk review, and confirming a documented off-BTC failure pattern shared with the abandoned AVAX, AAVE, INJ, and SOL versions.

The drawdown-accumulation buy-the-dip mechanism that works on BTC does not transfer off-BTC: ETH (like AVAX, AAVE, INJ, SOL) suffers bear-market drawdowns that keep falling rather than mean-reverting, and tuning 4 parameters cannot add the missing trend/regime filter needed to stop buying into downtrends.

It was abandoned at the pre-optimization backtest-review gate (verdict: abandon) for negative-expectancy falling-knife behavior — it failed max_drawdown, Sharpe, profit_factor, and avg_trade_return criteria — so optimization and all later stages were never reached.

A long-only ETH spot buy-the-dip strategy on ETHUSDT.BINANCE_SPOT daily bars that accumulated when price fell ≥15% below its 30-day trailing peak and exited on a 12% profit target, recovery to the prior peak, or a 30-day hold cap — porting the already-landed BtcSpotDrawdownAccumulationLong mechanism to ETH to fill the BINANCE_SPOT venue quota gap.

Over 9 years (2017-08-17 to 2026-05-18) it took 74 trades with a 57% win rate but negative expectancy: total return -46.8%, CAGR -7.4%, Sharpe -0.73, profit factor 0.79, expectancy -$646/trade, and a 68.5% max drawdown, with average losses ($7,231) dwarfing average wins ($4,370).