Skip to content
All hypotheses

Hypotheses

OKX Cross-Sectional Lottery (MAX) Reversal: Dollar-Neutral Short the Highest Max-Daily-Return Alts / Long the Lowest, Weekly Rebalance (13 OKX USDT Swaps, 1-DAY, 3-Parameter)

Dollar-neutral, multi-instrument, cross-sectional strategy on 13 OKX USDT perpetual swaps with daily history from no later than January 2021: ADA, LINK, DOT, ATOM, AVAX, UNI, XLM, TRX, BCH, ETC, AAVE, ALGO and SOL. I…

Hypothesis

Dollar-neutral, multi-instrument, cross-sectional strategy on 13 OKX USDT perpetual swaps with daily history from no later than January 2021: ADA, LINK, DOT, ATOM, AVAX, UNI, XLM, TRX, BCH, ETC, AAVE, ALGO and SOL. I checked the catalog for these 13; each has 2,080-2,470 daily bars through 2026-10-05. NEAR, DOGE, XRP, LTC and FIL were left out because their OKX daily history starts 2026-09-24. The signal is the MAX effect, or lottery demand (Bali, Cakici & Whitelaw 2011, JFE). Grobys & Junttila (2021) and others……Show moreShow less

Dollar-neutral, multi-instrument, cross-sectional strategy on 13 OKX USDT perpetual swaps with daily history from no later than January 2021: ADA, LINK, DOT, ATOM, AVAX, UNI, XLM, TRX, BCH, ETC, AAVE, ALGO and SOL. I checked the catalog for these 13; each has 2,080-2,470 daily bars through 2026-10-05. NEAR, DOGE, XRP, LTC and FIL were left out because their OKX daily history starts 2026-09-24. The signal is the MAX effect, or lottery demand (Bali, Cakici & Whitelaw 2011, JFE). Grobys & Junttila (2021) and others document it in crypto. Coins with the largest single-day return over the recent window attract lottery-seeking, leveraged retail demand and are overpriced, so they underperform next week. Coins with the smallest recent maximum are neglected and outperform. Every rebalance_days days, at the daily close, the strategy ranks all 13 coins by MAX = the highest single daily close-to-close return over the last max_lookback days. It shorts the n_per_side highest-MAX coins and buys the n_per_side lowest-MAX coins, with equal notional on each side: gross exposure 1.0x equity, 0.5x long and 0.5x short, account leverage 1. The book is market-neutral by construction. This mechanism differs from the promoted Hyperliquid cross-sectional momentum rotation. Momentum ranks by the trailing cumulative return. MAX ranks by the right-tail extreme, which is only weakly correlated with that return, and it trades in the reversal direction. Venue choice: OKX has 2.5% of the corpus, against a ≥5% quota target. The strategy is long+short (18.7% of the corpus against a ≤55% long-only target) and multi-instrument. It uses no BTC and no supplementary feed (OHLCV only, per L189). Execution substitutions: orders are market orders at the rebalance bar's close. On a 24/7 venue this equals the next bar's open, which is the use_equivalent for next_bar_open_fill. The per-leg catastrophic stop is an intrabar protective stop: a reduce-only _submit_stop_market matched inside the bar by the platform's simulated venue. Live trading mirrors those fills, so no exchange-resting order is required. Netting accounts are fine: each instrument holds at most one net position.

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