arXiv subjects

Arati Uday Kamat

Publications and source records attributed to Arati Uday Kamat.

7 recordsLinked to original sources

arXiv2026

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

We report a 15-day paper-traded autonomous memecoin trading deployment 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 Solana venue lacks published measurement: (i) time-of-day return patterns on a 24/7 permissionless venue; (ii) whether decision-time filter stacks are net-positive against counterfactual returns of rejected tokens; (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 (n=17, mean -11.60 percent) versus all others (n=173, mean +1.82 percent) 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 (the full-cohort 17.9 percent is a censored lower bound). 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), we deposit audit.py (MIT), an assertion-based reproduction script that exits zero iff every headline number reproduces from the deposited CSVs. Companion dataset: Zenodo concept DOI 10.5281/zenodo.20043301. The paper's principal contribution is measurement infrastructure and three transferable design lessons.

q-fin.TR↗

arXiv2026

Outcome-Classified Precision Auditing of Filter Rules in Algorithmic DEX Trading: Evidence from 2,400 Rejection Events

This paper reports a precision audit of a production filter stack against a 13-day window of post-rejection forward-market observations on Solana DEX trading (2026-04-10 to 2026-04-23, UTC). The audit yielded 99,510 follow-up samples across 2,402 unique rejection events spanning eight active filter rules. We classify each event under a five-tier outcome rule and report per-filter distributions. The descriptive result is a save-to-miss ratio of 3.7 : 1 (418 windowed measured-drawdown saves against 112 misses). This ratio is a descriptive count, not by itself evidence of filter skill: a fair-game price path also reaches a 50 percent drawdown before a 100 percent rise more often than the reverse, and a rebuilt analysis supersedes it. Correction (October 2026): I withdraw the early-death tier, the wider 14.8 : 1 ratio that rested on it, and the matched lifecycle comparison (48.9 versus 57.6 percent of mints reaching the gone state). In 1,235 of the 1,236 single-sample "early-death" events the same token was rejected again about 10 minutes later and the tracker restarted its record, so these events are tracker artefacts, not token deaths, and gone is a tracker state, not confirmation of a rug pull. I also withdraw the claim that every adequately sampled filter is individually net-positive (filter_7 has 3 saves and 3 misses). Per-filter estimates may be affected because follow-up records are keyed by token rather than by rejection decision. Details are in the correction notice on page 1.

q-fin.TR↗

arXiv2026

Coordinated Sniper Cohorts on Pump.fun: Detection of 1,012 Persistent Wallet Rings and a Contamination-Adjusted Estimate of Coordination-Specific First-Hour Buyer-Flow Lift

CORRECTION (Oct 2026): buyer records include sells and miss many buys; all numerical results below are withdrawn pending a corrected analysis. Motivated by Kyle (1985) informed order flow and the Meiklejohn et al. (2013) wallet-clustering tradition, we ask whether persistent coordinated wallets causally raise first-hour buyer flow on the Solana pump.fun bonding-curve marketplace. Using 1,578,333 buyer observations from 166,098 launches over 13.4 days (2026-06-11 to 2026-06-25), a two-stage detection pipeline (intra-launch first-buyer-window extraction plus cross-launch persistent-cohort surfacing via union-find on co-occurrence graphs) identifies 1,012 persistent wallet cohorts (2 to 12 wallets, 2,965 addresses) that systematically co-fire as early buyers. Under a contamination-adjusted estimator excluding cohort wallets' own buyer events from the outcome, 1:1 nearest-neighbour propensity-score matching with a 0.2-SD caliper on ten launch-quality covariates (Rosenbaum-Rubin, 1983) yields a first-30-minute buyer-count lift of +16.1% (95% CI [+13.0%, +19.4%]) on 5,419 matched pairs; the corresponding SOL-inflow lift is +6.3% ([-0.5%, +15.1%]), not distinguishable from zero. The naive contaminated same-universe pooled contrast is +130.9%; approximately half is arithmetic contamination from cohort wallets' own buys in the outcome (dropping to +63.9% after cohort exclusion), and most of the remainder is absorbed by PSM on launch-quality covariates. A parallel activity-matched placebo across 100 seeds produces lifts with median +189.6%, above zero and above the real lift in 100/100 seeds, indicating the placebo estimator is biased; we retain it as a bias diagnostic. Of 5,419 treated launches, 382 (7.0%) had zero non-cohort buyers in the first 30 minutes. We release the full cohort catalogue, detection code, PSM script, and robustness artefacts as RED-COHORT-2026-v1 under CC-BY-4.0.

q-fin.TR↗

arXiv2026

Do Two On-Chain Observation Pipelines See the Same Tokens? Cross-Pipeline Coverage on the Solana pump.fun Launchpad

On-chain studies of memecoin launchpads usually rely on one data-collection pipeline, yet whether differently configured pipelines observe the same tokens is rarely measured. This paper compares the output mint sets of two separately configured pipelines from one research programme on the Solana pump.fun launchpad: a cohort-detection pipeline that flags coordinated early-buyer wallets, and a rejection-filtering pipeline that logs a trader-side observer's pre-trade filter decisions. Two consecutive, non-overlapping windows are analysed (v1: June 2026; v2: June-July 2026); in each, the rejection data are restricted to the interval spanned by the cohort detections under four timestamp rules. Only raw counts, observed proportions and Jaccard/Dice indices are reported. In v2, where the rejection stream covers the whole 17.5-day window, the cohort set contains 623 mints and the rejection set 1,742, with 5 overlapping mints (0.803% of cohort; 0.287% of rejection). In v1 a collection gap means the rejection stream covers only the final 14.25 hours (4.4%) of the 13.4-day cohort window; the 20,162 cohort and 53 rejection mints share 1 mint, and within the covered interval none, a count too small to be informative. Both windows show nearly disjoint outputs, agreeing in direction but not in magnitude. Collector configuration, not only market behaviour, therefore shapes which tokens an on-chain study observes, and the paper proposes reporting cross-collector overlap and collector coverage as a routine check. Data and a script that re-derives every number are openly available.

q-fin.TR↗

arXiv2026

Post-Rejection Follow-up Sampling: Measuring Outcomes of Rejected Decisions in Algorithmic DEX Trading

Filter-gated algorithmic trading systems on decentralised exchanges reject most candidate tokens they evaluate, yet the observed forward market trajectory of those rejected candidates is rarely measured on the same live venue that produced the rejection. This paper introduces Post-Rejection Follow-up Sampling (PRFS), an observational measurement methodology in which a separate tracking subsystem samples each rejected token's price and liquidity from the same live oracle path used by the rejecting scanner, at a scheduled cadence, from the moment of rejection out to a fixed analytic horizon. The methodology is defined through a formal specification of the observation window and the reason-attribution rule, a reference implementation (prfs v2.0.0) with executable coverage and reason-ledger tests, and a secondary cross-check implementation that re-derives every headline number through a distinct code path. The companion dataset contains 67,000 forward-observation rows across 2,997 rejection events collected on a single Solana automated-market-maker sub-venue during a launch-dynamics window of 8.63 calendar days across 457 unique mints. Under the primary reason-aware three-field event key, 1,455 events (48.55 percent) receive a matched forward observation; under the reason-agnostic two-field diagnostic key, 1,641 events (54.75 percent) match. Mint-level coverage is 100 percent. The 253-row reason-conflict ledger and 8,593 unaligned outcome rows are disclosed, and coverage is shown to be filter-associated (chi-square(6) = 401.29) and source-associated (z = 14.78). No causal or counterfactual claim of a hypothetical acceptance outcome is made; the reported estimand is the observed post-rejection forward return.

q-fin.TR↗

arXiv2026

Auditing Collector-Generated Graduation Labels on Pump.fun: Measurement Error and Temporal Non-Generalization

To examine whether collector-generated terminal labels on the pump.fun platform can be interpreted as platform-side graduation outcomes, and whether a pre-registered logistic association model generalises across a subsequent temporal holdout of the same cohort. A frozen primary cohort of 749,816 unique mints recorded by a V3 off-chain collector between 12 May 2026 and 10 June 2026 is analysed. The design combines a source-code and reconstructed-state measurement audit of the collector's terminal-classification mechanism with a pre-registered locked logistic association model fitted on a 15-day development cohort and evaluated on a subsequent 14-day validation cohort under the pre-registered stability-gate framework. Calendar-day cluster-robust covariance and day-block bootstrap procedures assess uncertainty. Development discrimination is high (AUROC 0.8594), but the model does not generalise to the held-out validation cohort (AUROC 0.4642; 500 of 500 successful bootstrap replicates; 95% percentile interval [0.4112, 0.5196] containing 0.5000). Validation calibration deteriorates substantially (slope 0.013, intercept +2.816); the market-cap functional form is not stable across pre-registered specifications; and of nine automated evaluations, two pass, six fail, and one is not evaluable. The collector's TIMEOUT label does not establish platform-side non-graduation. The paper treats outcome ascertainment as a first-order measurement problem rather than assuming that an off-chain collector's terminal classification measures the platform-side event. It pairs this audit with a pre-registered held-out temporal-generalisation test and reports the resulting negative finding as registered, alongside a fully reproducible correction history.

q-fin.TR↗

arXiv2026

A Multi-Venue Solana/DeFi Microstructure Data Corpus: The RED-2400 Family v2

Empirical research on decentralized and Solana-native market microstructure is constrained less by method than by data: the strongest results in the field rest on a small number of Ethereum-centric datasets, and public, record-level tapes of Solana venues, cross-chain flows, and on-chain oracle behavior remain scarce. We release and document the RED-2400 family v2, a corpus of five public benchmark datasets assembled by a single public-blockchain observer over a concurrent 57-day 2026 window (2026-05-08 to 2026-07-05) under one methodology (free-tier and public feeds only, no paid data, no private accounts, no trading), each with fixed-seed reproducibility scripts, SHA-256 manifests, and a CC-BY-4.0 licence. The five datasets characterize, respectively, Pyth oracle staleness against centralized-exchange prices (164,002 observations), an Aave utilization and Ethereum liquidation panel (18,750 liquidations plus a four-chain utilization series), spot-perpetual basis and funding coupling (244,719 observations), a Solana CEX-versus-DEX spread tape with realized impact (328,186 observations), and a Wormhole cross-chain source-flow tape (360,714 messages). Across the corpus we report three descriptive findings, one honest null, and one field-completeness characterization; none is presented as a causal estimate. The contribution is a coherent, reproducible benchmark family that complements the Ethereum canon and lowers the data barrier for Solana and cross-chain microstructure research.

q-fin.TR↗