Hour-Aware Adaptive Risk Management for Autonomous Memecoin Trading on Solana DEXs: Evidence, Theory, and Design Lessons from a 15-Day Deployment

Abstract

Hour-Aware Adaptive Risk Management for Autonomous Memecoin Trading on Solana DEXs: Evidence, Theory, and Design Lessons from a 15-Day Deployment Version v4 (2026-08-03). Kamat, Arati Uday. Independent Researcher. ORCID 0009-0000-4781-312X. This record deposits the v4 manuscript PDF. Substantive revisions from v3: 1. Added Section III "Theoretical Framework" (~700 words) — states three predictions from Kyle (1985), Precup-Sutton-Singh (2000), and Bailey-Lopez de Prado (2014).2. Added Section VII "Generalisable Design Lessons" (~600 words) and a Significance for the Field paragraph in Section I + expanded Section VIII.B.3. Expanded Section II Related Work from 1 subsection to 4, with 14 additional canonical citations.4. Added JEL classification codes G11/G12/G14/G17. Two pre-submission corrections were also made: entry-vs-exit hour recomputation in Section VI.A; observation-window censoring disclosure in Section VI.B. All headline empirical numbers unchanged from v2/v3. Companion dataset bundle (concept DOI, always resolves to latest version): 10.5281/zenodo.20043301. Companion preprints: SSRN abstract 6564803; arXiv 2606.08232. Abstract. We report a 15-day paper-traded deployment of an autonomous memecoin trading system on Solana decentralised exchanges (DEXs), designed as a controlled measurement of three microstructure questions on which classical equity theory offers well-defined predictions but on which the AMM-based Solana venue lacks published empirical measurement: (i) time-of-day return patterns on a 24/7 permissionless venue; (ii) whether decision-time filter stacks are net-positive against the counterfactual returns of the tokens they reject; (iii) whether small-sample cumulative-return statistics on a heavy-tailed venue are structurally robust or fragile. The 190-trade sample (March 29 to April 12, 2026) shows a 40.5 percent win rate, mean per-trade return +0.62 percent, cumulative +117.7 percent, skewness -1.21, excess kurtosis 6.61. Mann-Whitney U on three exploratorily identified worst entry hours yields p = 0.5634 (directional and non-confirmatory). A parallel counterfactual rejection-tracker collected 4,874 forward-sample observations across 184 rejection events; of 48 events observed for at least six hours, 27 (56.25 percent) reached a 50 percent drawdown from reference. Removing the top three trades (1.6 percent of sample) flips cumulative return unprofitable. The three findings connect to Kyle (1985) informed-flow, Precup-Sutton-Singh (2000) off-policy evaluation, and Bailey-Lopez de Prado (2014) deflated-Sharpe predictions. Alongside the trade log and rejection-sample corpus (CC-BY-4.0), the companion bundle deposits audit.py (MIT), an assertion-based reproduction script that exits zero iff every headline number in this manuscript reproduces from the deposited CSVs. The paper's principal contribution is measurement infrastructure and three transferable design lessons that generalise beyond this specific system. Competing interests. The trading system is the subject of pending U.S. Provisional Patent Applications #64/022,461 (filed 2026-03-30) and #64/099,108 (filed 2026-06-25), both Micro Entity. Manuscript release does not restrict re-use under this deposit's CC-BY-4.0 licence. Keywords: autonomous trading, memecoin, decentralised exchange, market microstructure, time-of-day effects, counterfactual evaluation, reject inference, fragility, Solana. JEL: G11, G12, G14, G17.