Low-Volatility Anomaly (Lottery-Demand Short) on 13 OKX USDT Perps: Long the 3 Lowest-Realized-Vol Alts, Short the 3 Highest, Inverse-Vol Sized, Weekly Rebalance (OKX 1-DAY, long-short, pure OHLCV)

Hypothesis

A cross-sectional low-volatility factor on OKX USDT perpetual swaps. Data comes from OkxCollector as 1-DAY OHLCV bars. In the catalog, every leg has data from 2020 or 2021 through 2026-10-01. The universe is 13 coins, each with at least 2077 daily bars: ADA, DOT, LINK, ATOM, ETC, BCH, XLM, TRX, AAVE, UNI, ALGO, SOL and AVAX. XRP, DOGE, LTC, NEAR and FIL are left out on purpose because OKX has only 8 days of history for them. Every rebalance_days bars, rank the coins by realized vol over vol_lookback days. Go long the n_per_side lowest-vol coins and short the n_per_side highest-vol coins. Each leg is sized by inverse volatility, and the two books are matched to equal ex-ante volatility, which makes the portfolio roughly beta-neutral. Total gross exposure is 1.0x equity. Rolling vol must be computed in O(1) per bar (deque with running sums), and ranks only on rebalance bars, to avoid the recent sandbox_timeout deaths.

Fills under-weight buckets: OKX is 1.8% against a ≥5% target, long-short is 15.9%, and it is multi-instrument. It does not use BTC. The factor is new to the corpus. It trades rarely and holds for weeks, so it stays clear of fee_edge. There is no calendar trigger, no supplementary-feed gate and no options leg.

verification_loop: Verification failed (Layer 1.5 — hypothesis/config consistency) [class=hypothesis_mismatch]: - optimization_plan fixes leverage at 1.0

Implementation

Ranks 13 OKX USDT perpetuals by rolling daily realized volatility, holds the three lowest-volatility coins long and three highest-volatility coins short, and rebalances on a calendar-anchored weekly schedule. Inverse-volatility sizing allocates 1.0x gross equity exposure across the two books.

The config now sets leverage to the hypothesis’s fixed 1.0. The sole code change makes the gross-notional calculation explicitly use that configured leverage; at 1.0 it preserves the intended exposure and prior signal logic.

Verification Results

Verification failed (Layer 1.5 — hypothesis/config consistency) [class=hypothesis_mismatch]: - optimization_plan fixes leverage at 1.0

Iteration History

code_identical: iteration 2 resubmitted byte-identical code to iteration 1 — the previous feedback was not acted on, so the result cannot change. Either change the code to address it, or state recommendation='abandon' with the reason if no code change can help. Previous feedback: Verification failed (Layer 1.5 — hypothesis/config consistency) [class=hypothesis_mismatch]: - optimization_plan fixes leverage at 1.0

Outcome Summary

Resolve the leverage mismatch between the optimization plan and hypothesis/configuration before verification.

After three iterations, the strategy was abandoned in a verification loop. Layer 1.5 found a hypothesis/configuration mismatch because the optimization plan fixed leverage at 1.0.

The strategy aimed to go long the three lowest-volatility coins and short the three highest-volatility coins among 13 OKX USDT perpetuals, with inverse-volatility sizing and weekly rebalancing.

No backtest results or trade metrics are available. Backtesting and optimization were not reached.