Strategies

ETH Relative-Volatility Acceleration: Weekly Defined-Debit Reverse Iron Condor

Outcome: Abandoned

EthRelativeVolatilityReverseIronCondor

Outcome Summary

This strategy tested whether ETH volatility accelerating faster than BTC's made ETH options cheap, buying a weekly reverse iron condor when the relative-acceleration signal cleared its threshold. After 2 iterations the backtest still produced zero option trades, so it had no return, Sharpe or per-trade figures to judge. The pipeline abandoned it for options_no_option_trades: 0 of 395 periods were entered, mostly blocked with no recorded reason and otherwise by the readiness gates. No backtest review, optimization, analyst or risk-officer review took place.

Strategy report

Hypothesis

Derived by the Research Lead from Imran Lakha's Deribit Insights article: test whether ETH volatility accelerating relative to BTC creates underpriced ETH convexity. The article discusses ETH/BTC option-volatility spreads in relation to realized spreads; it supplies no executable rules or profitability evidence. Source: [Options Dynamics: Reading the Signals in BTC and ETH](https://insights.deribit.com/industry/options-dynamics-reading-the-signals-in-btc-and-eth/). Buy an ETH weekly reverse iron condor: long the nearer-ATM call and put, short further-OTM call and put, all at one expiry. Use ETH-PERPETUAL.DERIBIT's own completed 4-HOUR bars as the untraded primary clock and BTCUSD.HYPERLIQUID synchronized 4-HOUR bars as an untraded reference. Extend RollingOptionsStrategy directly and use its historical ETH chain expansion; distinguish same-kind legs by strike, rather than a single C/P side map. Persist every resolved concrete option instrument ID in extra_instruments and the hypothesis execution manifest. The listed instruments are fixed clock/reference instruments; the options_universe specification below identifies the required dynamically resolved traded legs. No fixed expired option is a trading anchor. Ordinary continuous-market orders execute at signal-bar close, the registry's equivalent of next-bar-open execution. No atomic combo, independent opposing positions in one instrument, selectable same-bar policy, exchange-hosted protection, or inverse-perp sizing is required. Request up to 730 days and report actual synchronized chain coverage; historical underlying prices do not extend the option evidence span. Predict 15-25 qualifying weekly structures per covered year, not a daily scalping strategy.

Long-only experiments are 76.9% versus a 55% ceiling, while combined options are 4.7% versus a 15% target and cross-venue experiments are 9% versus 15%. This adds bought and sold option legs and cross-venue signal inputs, although it is long volatility and does not add traded Hyperliquid exposure or dollar-neutrality. The concentration object is empty, so no recent instrument/timeframe shares can be inferred. BTC is a reference because the source concerns ETH volatility relative to BTC; it is never traded. Snapshot and library searches found cheap-IV ETH straddles, skew convergence, and terminal-density butterflies, but no matching relative-realized-volatility gate on a bounded two-tail debit structure. Unlike the failed delta-hedged short strangles, this buys bounded convexity without repeated hedge turnover. Unlike the pending bounded-skew convergence strategy, it does not trade put/call asymmetry. These are substantive differences, not evidence of profitability. The source is historical commentary, not a claim about today's market regime. No performance pre-study was run.

options_no_option_trades: options strategy produced 0 option-leg trades (0 underlying trades) — degenerated to the underlying / smoke path; the option chain never traded (0/395 periods entered; held by no_reason_recorded x252, readiness_relative_acceleration_iv_or_debit_gate x143)

Implementation

Weekly equal-quantity ETH reverse iron condors gated by ETH/BTC relative realized-volatility acceleration and inner-option implied volatility. Both clocks remain untraded. Continuous-market entries execute at signal-bar close, the registry equivalent of next-bar-open execution.

Preserves the previous signal, sizing, research plan and configuration. Confirmed position-opened and position-changed events advance staged wing entry immediately after both inner longs fill, using the existing equal-quantity and in-flight guards. Actual filled legs are registered for scheduled roll exits. The fixed four-hour incomplete-entry timeout and unwind remain, including late fills racing timeout cancellation.

Iteration History

Verification failed (Layer 4 — QA review) [class=code_defect]: - [critical] The staged entry cannot complete under the configured four-hour bar interval and four-hour timeout. On the initial primary-bar callback, market-entry commands are queued by the simulated venue, so the protection-position checks cannot yet observe both inner-long fills and the short wings are skipped. Only _enter_leg submits wings, and RollingOptionsStrategy calls it only on primary bars. On the next primary bar, _enter_leg calls _check_partial before attempting any wing; now_ns - attempt_ns is already four hours, so _check_partial marks the expiry failed and unwinds the longs. Extra-bar callbacks check completeness but never advance wing entry. Consequently even a qualifying structure with successful ordinary market fills is aborted rather than becoming the required four-leg reverse iron condor. (line 227) — fix: Advance the staged entry after confirmed inner-long fills, for example through a fill-driven state machine that submits each short wing once both protecting longs are held. Keep the fixed four-hour timeout for genuinely incomplete structures and preserve equal quantities and duplicate-order guards. Verify queued ordinary market fills complete all four legs before the deadline, and separately verify a genuinely missing leg triggers the required unwind.

Advance the staged entry after confirmed inner-long fills, for example through a fill-driven state machine that submits each short wing once both protecting longs are held. Keep the fixed four-hour timeout for genuinely incomplete structures and preserve equal quantities and duplicate-order guards. Verify queued ordinary market fills complete all four legs before the deadline, and separately verify a genuinely missing leg triggers the required unwind.

The staged entry cannot complete under the configured four-hour bar interval and four-hour timeout. On the initial primary-bar callback, market-entry commands are queued by the simulated venue, so the protection-position checks cannot yet observe both inner-long fills and the short wings are skipped. Only _enter_leg submits wings, and RollingOptionsStrategy calls it only on primary bars. On the next primary bar, _enter_leg calls _check_partial before attempting any wing; now_ns - attempt_ns is already four hours, so _check_partial marks the expiry failed and unwinds the longs. Extra-bar callbacks check completeness but never advance wing entry. Consequently even a qualifying structure with successful ordinary market fills is aborted rather than becoming the required four-leg reverse iron condor.

Outcome Summary

Before backtesting, check that the resolved option chain actually overlaps the window set by the reference instrument's history, and that the entry gates can open on that chain; a strategy that never enters cannot be evaluated.

It was abandoned for options_no_option_trades: none of the 395 weekly periods opened an option position. 252 periods were blocked with no recorded reason and 143 by the relative-acceleration, IV or debit readiness gate. The option chain schedule in the config also listed 2019-era contracts that fall outside the backtest window.

Buy a weekly defined-debit ETH reverse iron condor on Deribit (long near-ATM call and put at +/-2%, short further-OTM call and put at +/-8%) whenever ETH realized volatility accelerated relative to BTC's, betting that ETH convexity was underpriced at those times.

No trades at all: the backtest recorded 0 trades, 0.0% total return and a 0.0 Sharpe, flagged unreliable because no day moved the equity curve. The usable window was 2024-04-18 to 2026-10-17, clamped by when the BTCUSD.HYPERLIQUID reference data starts (2024-01-22), and the signal diagnostic rejected the forecast as carrying no information (19 observations, fewer than the 30 needed).

Backtest Review

Max drawdown: 0.00%
Win rate: 0.0%

Backtest and paper results are hypothetical. Trading involves risk of loss.