Stats

Numpyseed2026090502; eachsymbolindependently drawthree complete24h logreturnblocks from2026history,10k72hpaths. XRP/LINK/UNI245usabledayblocks,PAXG193; previousclose-to-currentclose hourreturns; no unseenintrahourmovement or15minvenuechecks. ReferenceSep4close with1ADR20stop,2ADR20target widths respectively7.609/15.218%,5.912/11.823%,7.126/14.251%,3.685/7.370%. Observedhourlyclose targetbeforestop frequencies3.71/5.67/8.51/.60%; stopfirst9.53/22.25/24.30/13.93%; neither86.76/72.08/67.19/85.47%. Unconditionalresampling, notactual stop-touch probabilities,signal-conditioned forecast orweekend-specific sample.72hterminal5th/95th XRP-7.764/+10.303%,LINK-8.471/+10.095%,UNI-10.435/+13.670%,PAXG-4.236/+4.179%. A measured2ADRtarget can stillbeunlikelyinside3days; do notstretchholdingwindow merelytofit. Nominal1%equityrisk8.5108 yields notionalceilings111.86/143.97/119.44/212.77, notguaranteedlosslimits. Actualsize0 dueuncertainnetedge. Validitysupports caution aboutattainability; soundnesslimitedbyresamplingpastblocks,separateassets,stationarity andcloseonlypaths. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; artifacts/stat_risk.py/stats-risk.json.

Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; Jan1-Sep4 2026 net natural-exit samples XRP6,LINK14,UNI13,PAXG8. Numpy20,000 IID resamples seed2026090502: mean trade95% intervals XRP[-1.194,18.765]%,LINK[-1.726,2.184]%,UNI[-2.933,2.634]%,PAXG[-1.402,.419]%. IID diagnostic assumes independent trades; current-universe selection and regimes limit interpretation. Positive-resample frequencies94.375/51.585/38.65/13.8% are not posterior probabilities. Daily OLS onBTC with intercept: betas1.237/1.176/1.181/.299, correlations.871/.874/.677/.446; sample247 each exceptPAXG204. Similar XRP/LINKbeta weakens diversification narrative. Rawbinarymean-payoff quarterKelly notional formula .25*(p/loss-(1-p)/win) gives3.889/.281/-.468/-7.550 capitalfractions, clearly unstable and constrained beforecaps. Lowermeanboundsnegative => authorizedallocation0. PremisespositiveXRPreturn andrate66.67% wins depend on6trades; strongestbullcase largepayoff, but no precise populationedge established. Need prospective fixedrule evidence and testedvenueexits; do not pickbestlookback frominspectedholdout. UNI terminalopenmark excludedfrompayoffs rather than countedwinner.

Registered note125 before results. Signal completed daily close > previous20 closes, desired long nextopen while true, exit nextopen whenfalse; no stop/target overlay. Baseline25bpfee+5bpslippageeachside; vectorbt allcash10k longonly, independentcash/shareledger maximumerror<1e-6. TrainingSep2024-Dec2025, diagnostic chronologicalholdoutJan1-Sep4 2026; present-dayuniverse selection means not independentprospective evidence. PAXG beginsFeb12 2026: no pre2026training; holdout label not justified for thissymbol. Daily return -4.103506577337702%, DD 5.412997437937084%; identicaldaily signal onhourly returns match with DD 6.3723479589041006%. Separate20observed-hour rule return -78.88693847013198%, DD 78.97980465858033%, win 14.393939393939394%, N264; reject costlyturnover, gaps mean20observedbars notexact20hours. Doublecost daily -8.597865801961252%. No terminalopenposition. No current completedSep4breakout; mandatoryvenueexits/+1Rmanagement and15minfills untested. XRP positive6trades insufficient forsize; others unstable/negative. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; artifacts/test.py and backtests.json, all metrics observationssaved. Failed initialtruncatedhourly outputs discarded; current figures final.

Registered note125 before results. Signal completed daily close > previous20 closes, desired long nextopen while true, exit nextopen whenfalse; no stop/target overlay. Baseline25bpfee+5bpslippageeachside; vectorbt allcash10k longonly, independentcash/shareledger maximumerror<1e-6. TrainingSep2024-Dec2025, diagnostic chronologicalholdoutJan1-Sep4 2026; present-dayuniverse selection means not independentprospective evidence. Training return -35.39128324423077%; lookbacks15/20/30 tested only training, fixed20retained. Daily return 13.0683957825984%, DD 11.86325956746298%; identicaldaily signal onhourly returns match with DD 18.182406721058307%. Separate20observed-hour rule return -83.3970961790418%, DD 83.39709617904185%, win 18.373493975903614%, N332; reject costlyturnover, gaps mean20observedbars notexact20hours. Doublecost daily 4.271874489078182%. UNI open terminalposition contributes to markedreturn;13naturalexits average negative. Sep4false signal would exitSep5open outsidecutoff, not missingexitlogic. No current completedSep4breakout; mandatoryvenueexits/+1Rmanagement and15minfills untested. XRP positive6trades insufficient forsize; others unstable/negative. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; artifacts/test.py and backtests.json, all metrics observationssaved. Failed initialtruncatedhourly outputs discarded; current figures final.

Registered note125 before results. Signal completed daily close > previous20 closes, desired long nextopen while true, exit nextopen whenfalse; no stop/target overlay. Baseline25bpfee+5bpslippageeachside; vectorbt allcash10k longonly, independentcash/shareledger maximumerror<1e-6. TrainingSep2024-Dec2025, diagnostic chronologicalholdoutJan1-Sep4 2026; present-dayuniverse selection means not independentprospective evidence. Training return -39.712282707450555%; lookbacks15/20/30 tested only training, fixed20retained. Daily return 0.2949758879143155%, DD 16.073828141163325%; identicaldaily signal onhourly returns match with DD 17.771428331495475%. Separate20observed-hour rule return -82.69256923103462%, DD 83.21785242055526%, win 17.159763313609467%, N338; reject costlyturnover, gaps mean20observedbars notexact20hours. Doublecost daily -7.78575922941096%. No terminalopenposition. No current completedSep4breakout; mandatoryvenueexits/+1Rmanagement and15minfills untested. XRP positive6trades insufficient forsize; others unstable/negative. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; artifacts/test.py and backtests.json, all metrics observationssaved. Failed initialtruncatedhourly outputs discarded; current figures final.

Registered note125 before results. Signal completed daily close > previous20 closes, desired long nextopen while true, exit nextopen whenfalse; no stop/target overlay. Baseline25bpfee+5bpslippageeachside; vectorbt allcash10k longonly, independentcash/shareledger maximumerror<1e-6. TrainingSep2024-Dec2025, diagnostic chronologicalholdoutJan1-Sep4 2026; present-dayuniverse selection means not independentprospective evidence. Training return 257.08276701988984%; lookbacks15/20/30 tested only training, fixed20retained. Daily return 49.30134045526893%, DD 10.42241544981568%; identicaldaily signal onhourly returns match with DD 13.486776844598625%. Separate20observed-hour rule return -71.61556289760512%, DD 75.69734866102567%, win 16.206896551724135%, N290; reject costlyturnover, gaps mean20observedbars notexact20hours. Doublecost daily 44.0220253594203%. No terminalopenposition. No current completedSep4breakout; mandatoryvenueexits/+1Rmanagement and15minfills untested. XRP positive6trades insufficient forsize; others unstable/negative. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; artifacts/test.py and backtests.json, all metrics observationssaved. Failed initialtruncatedhourly outputs discarded; current figures final.

Supersedes analysis67 USDG volume and continuity statements and observation273 windows. Initial tail(7) selected seven observations spanningAug26-Sep4, overstating seven-calendar-day amount4.103M. ExplicitAug29-Sep4 gives2.103309M across4availabledays; three missingdays unknown, Sep4reportedvolume0. USDG remainssecond by observedUSDnotional, but complete ranking cannot be certified without missing-day reconciliation. Last30calendarwindow has10rows and3zerovolumerows, not sixzero-volume days (old30observations). Its prior20close window also spans more than20days; no valid20day signal claimed. Other35names have7/7 recentdailyrows; testedXRP/LINK/UNI/PAXG complete latest21. Strongcounterargument no-trade days may simply be omitted, but do not assume that without data semantics. Corrected collector will explicitly label daycounts,missingcoverage andeligible0. Primary source https://runway.adityaperswal.workers.dev/data/crypto-bars . This audit changes data-quality assessment, not zero-order decision. Need calendar filtering before rollingwindow interpretation on sparseassets.

Premises and hypotheses kept separate. Facts: observations317-319,331,336-337 establish issuerfeechanges,sponsorship,USDCissuance,LINKcustomerintegration andhistoricalUNIburns. 1 USDC+.61B in7days implies0.827% from prior73.72B;30days+2.25B implies3.122% from72.08B. Then supply expands settlement capacity; users may deploy to payments/collateral rather than majors; withoutdestinationdata higherstock doesnot identify marginal tokenbuyer.2 XRPlogo raises awareness; adoptionmayfollow; tokenreturnsstillrequirebuyerflow,notadvertisingspend.3 CircleusesChainlinkreserves; serviceusage mayincreaseoraclefees; LINKpricebenefit depends paidtoken demand/retention absenthere.4 UNIJune5.1Mfee annualizes61.2M ifconstant,an assumptionnotforecast; burns canreduceunits; LPmigration/governancechanges/treasuryreleases mayoffsetmarginalbenefit. NoSeptemberburnflow or currentnetfloataudited,so nofairvalue multiple.5 PAXGprimaryredemptionfrictions change arbitrageeconomics; weekendsecondarypremium maypersist; cashbookcannot assume immediatephysicaldeliveryorfee-freeconversion. Strongbullcase confirmedinfrastructure andburnmechanisms are real; stillnot proof historical netpriceedge survives. Who benefits: sponsoreduniversity,oracleproviders,users,LPs/searchers/custodians accordingtomechanism; tokenholderbenefit conditional. Startflat,goalrepeatablenetgains,bindingconstraintindependent executablepayoffs. Sources https://www.circle.com/transparency ; https://www.circle.com/blog/cirbtc-a-clearer-reserve-standard-for-wrapped-btc ; https://gov.uniswap.org/t/temp-check-activate-v4-protocol-fees/26162 ; https://support.paxos.com/articles/2899561282-PAX-Gold-Fees ; FloridaAthleticsURLobservation318. Conclusionretaincashuntiltestsandcurrententryqualify.

Premises:1 CircleSep3rawHTML reportsUSDC74.33B,7dnet+.61B,30d+2.25B; issuer page is source, not proof of on-exchange buying.2 IBITSep4shares1,384,960,000 establish baseline; absent matched prior count no creation delta; page mixesSep3price/volume withSep4NAV/AUM, so AUMchange must not be labelledflows.3 Florida AthleticsSep4 confirmsmultiyearRipplefield-logo sponsorship; no mandatoryXRPpurchase or token-holdercashflow quantified.4 PAXGissuer newredemptionbands affect primaryconversion, not Alpaca spottradefees; creationwaiver endsSep30. Conclusion demand narratives remain uncalibrated watches. Validity: evidence establishes infrastructure/liquidity supply but not token return expectancy. Soundness limited to issuerreports,roundedUSDCvalues,missingdestinationflows andETFhistory. Strongbullcase more settlementcapital and visibility could precede adoption; need actual token demand and independentnetpayoffs. SourceURLs: https://www.circle.com/transparency ; https://www.ishares.com/us/products/333011/ishares-bitcoin-trust-etf ; https://d2b5htfb6s9xp9.cloudfront.net/news/2026/9/4/football-florida-athletics-debuts-xrp-cryptocurrency-logo-at-ben-hill-griffin-stadium ; https://support.paxos.com/articles/2899561282-PAX-Gold-Fees . Startcash235.86,goalrepeatablenetgain,bindingconstraintpositive executableedge; noorder.

Premises:1 Alpaca published support inventory lists36 USD pairs; May19 expansion release is primary corroboration of added assets, but same issuer not independent source.2 Fully paginated daily bars Sep2024-Sep4 give7day sum(v*vw) ranking USDC,USDG,BTC,ETH,SOL,XRP,PAXG,LINK,UNI,USDT. USDG was missing from prior33-name sample and displaces AAVE. 3 USDG Sep4 traded volume0 despite7day4.103M, sixzero-volume days last30; activity concentration is not continuously available liquidity.4 Only USDT closes above prior20 closes, with prior anomalous pricing; no nonstable signal. Conclusion no entry based on screen; current top10 is published-list/7day/USD-only, not verified account Assets inventory or cross-quote consolidated ranking. API gateway /data/assets404; no account credentials available; do not claim exhaustive live tradability. BTC/ETH/SOL core verdict66 unchanged. BONK/LDO current mover headlines do not satisfy seven-day top10 or validated signal. Current-universe selection introduces survivorship bias in historical tests. Sources https://alpaca.markets/support/what-cryptocurrencies-does-alpaca-currently-support and https://alpaca.markets/blog/alpaca-adds-new-crypto-coins-to-offering/ ; https://runway.adityaperswal.workers.dev/data/crypto-bars . Observations272-307 retain every screened row, including rejected names. Top candidates new audits XRP/PAXG/LINK/UNI registered note125.

Current explicit status supersedes broad WATCH#20/#24 and old ADOPT#5: BTC/ETH daily20 tested implementation REJECT; all BTC/ETH/SOL hourly20 implementations REJECT; SOL daily20 WATCH only, not adopted. 1 Registered next-open daily net diagnostic holdout Jan1-Sep4: BTC-7.44%,ETH-15.57%,SOL+22.23%; separate20-hour losses-83.70/-80.99/-87.42%. 2 Identical DAILY signal hourly replication matches returns/payoffs, hourly-mark DD16.52/25.33/15.01%; independent cash ledger confirms. SOLN12 win50 meanW6.123 meanL2.270, IIDmean95%CI[-1.213,6.606]%, positive selection-tainted diagnostic not independent proof. 3 CompletedSep4 signals falseBTC/ETH/SOL/UNI; all-cash ledger zero open/pending, no stop/target to reset. 4 No validated fee-inclusive measured-range/15-minute venue exit or independent untouched holdout. Conclusion no entry/exit/adjust proposal; allocate0 notional. Deduction valid under mandate; soundness limited by sparse volume/missing hours, selected historical sample and assumed slippage. Rejecting implementations is not proof all momentum false. RawSOLquarterKelly7.866% is not authorized: conditioning/selection uncertainty and missing exits bind before caps. Nominal1%equity=$8.5108,position cap212.77,fund235.86. Goal repeated net gains, binding constraint independent implementable edge; monthly$348.92 shortfall does not change criteria. Prospective frozen window startsSep5 under note92; earliest reviewOct5 and require>=30closed trades, no retuning/early adoption. Range/simulation results illustrate risk, not fill guarantees. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees ; primary catalyst analysis52. Source claims and test metrics saved observations; artifacts local /tmp/codex-run-fYGHo5. Watch11 remains research-only at01:30 with76runs at entry, full output verified; phrase validation pending means independent implementation validation, now historical daily/hourly audits are complete. Newtested funding/OI job13 every120min24runs; first scheduled sandbox output must be checked next run. ETHmonitor10 Sep28 review, no exact mainnet date; no invented SOLdate. No action in other funds. Gaps: complete tradable-volume universe,ETF/exchange/stablecoin flow changes, full unlock inventory, independent edge and15-minute fill stress. Research finished early; no thread exceeded allowance, final recording phase preserved.

Direct exchange API retrieved Sep5 00:04UTC, BTC/ETH/SOL USDT swaps. https://www.okx.com/api/v5/public/funding-rate?instId=BTC-USDT-SWAP ; funding-rate-history same instId&limit=100 ; open-interest same instId&instType=SWAP (replaceBTC withETH/SOL). Current funding forecasts0.01034/0.58253/-0.84163bp for08UTC, versus100 realized8h settlements Aug3-Sep5; empirical percentiles4/46/2. Last settled0.00702/0.57593/-0.85247bp. Forecast-versus-realized comparison is descriptive and can change before settlement. OI$2.192bn/$1.478bn/$287.585m is single-venue baseline, not change or extreme. If forecast stayed constant, simple annualized carry+.113/+6.379/-9.216%,72h total+.000931/+.052428/-.075747%; assumptions, not predictions, and spot receives none. Premise1: low funding weakens generic crowded-long assertion on this venue. Premise2: it does not prove a short squeeze, since net directional exposure, liquidation levels and OI history absent. Premise3: daily price breakout false in all3. Conclusion: no conjunction producing a tested buy. Strongest bear interpretation negativeSOL funding may reflect persistent adverse expectations; strongest bull interpretation shorts may later cover. Neither identifies expected spot win/loss. Scheduled tested job13 every120min24runs (60s timeout) stores timestamped forecast/settled funding andOI to build changes; informational eligible=0, local test all3 fresh/exit0. Binance public endpoint451 prevented measurement there; OKX is independent source, not corroboration. No derivatives orders within cash-only mandate.

Start state0holdings/0queued, goal repeatable net returns; binding constraint validated payoff after costs, not crypto trading hours. No existing stops/targets to change. BTC strongest bear case: failed closing breakout plus unverified fresh ETF creations. And then1: chasing yesterday's signal incurs two-sided costs;2: reversal consumes an uncalibrated stop budget;3: correlated ETH/SOL replacement risk compounds losses. Counterargument: ETF demand and momentum could revive, but neither is calibrated current edge. ETH bear case: upgrade repricing changes may disrupt a subset of contracts. Then1: builders must alter gas assumptions;2: implementation delays or higher state-operation cost can slow some activity;3: token returns still depend on adoption, issuance/burn and valuation, not throughput alone. Counterargument: most contracts unaffected, fixes and capacity expansion may improve adoption. SOL bear case: Q3 versusOctober date conflict and current close below109.15 trigger. Then1: anticipated launch demand may precede actual activation;2: delay/rollout risk can reverse positioning;3: limitedN12 price-rule calibration and BTC beta1.20 amplify sizing error. Counterargument: successful150ms-finality target could improve applications; cannot translate that engineering target into investment payoff. Who benefits: users/builders from successful capacity; liquidity providers collect spread/fees on premature turnover; exchange perpetual funding paid between leveraged parties is not income for this spot book. Attack cash: missing a rally is opportunity cost, but a funding deadline changes incentives, not conditional expectancy. Current evidence supports0risk, not assertion majors must fall. Facts end at recorded prices, tests, issuer schedules; economic chains are hypotheses. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://www.ishares.com/us/products/333011/ishares-bitcoin-trust-etf ; https://blog.ethereum.org/2026/08/24/glamsterdam-repricing-testing ; https://solana.com/upgrades ; https://solana.com/upgrades/alpenglow .

Source https://runway.adityaperswal.workers.dev/data/crypto-bars Jan1-Sep4 2026 hourly data. NumPy seeded20260905 draws3 complete24-hour log-return blocks from246days,10000paths per symbol,72h horizon. Unconditional day-block bootstrap, NOT forecast or weekend-conditioned strategy test; preserves intraday clustering but breaks cross-day dependence, samples each symbol separately, ignores intrahour stop ordering/15-minute venue delay. Illustrative stop1ATR/target1.5ATR: BTC3.392/5.088%,ETH4.337/6.505%,SOL6.986/10.479%. Stop-first/target-first/neither probabilities: BTC32.06/15.56/52.38%;ETH32.51/16.08/51.41%;SOL17.20/6.38/76.42%. Terminal5th/95th pct BTC-6.184/+7.006,ETH-8.939/+10.032,SOL-9.011/+9.617. 95th adverse path losses7.508/10.722/10.492%. Separately observed35 Sat00UTC-Mon23UTC windows using hourly lows:5th-percentile adverse excursion BTC-7.051%,ETH-6.750%,SOL-9.198%; worst -11.414/-20.240/-18.188%. Observed windows include Monday, not pure weekend. Median favorable excursions2.644/3.741/3.510%, materially below illustrative targets. Hard cap nominal1%equity=$8.5108; at1ATR stop notional ceilings min25%equity/stop risk are212.77/196.25/121.83. These are ceilings only, actual authorized size0. Premise: simulated and observed paths can cross nominal stops; conclusion: no claim stops guarantee1%maximum realized loss or targets achievable. Counterexample: a validated signal-conditioned distribution could improve odds; absent here. No mean-reversion or weekend entry inferred. Artifacts simulation.json,weekend.json,risk_analysis.py,weekend.py.

Primary venue bars retrieved Sep5, completed Sep4, https://runway.adityaperswal.workers.dev/data/crypto-bars ; indicative snapshots /data/crypto-snapshots. BTC79676.7475 vs81269.4915 prior20close max; ETH2456.376 vs2515.933; SOL101.95 vs109.15; UNI6.18484 vs6.32576. All false; Sep3 BTC/UNI alerts expired when Sep4 close failed. Daily returns BTC-1.960%,ETH-2.067%,SOL-1.941%,UNI-2.228%. Core remains above20-day SMA but that alone is not registered entry. ATR14 simple mean true range BTC2702.40/ETH106.53/SOL7.122; ADR20 BTC2921.85/ETH118.75/SOL6.530. Illustrative1ATR stop/1.5ATR target percent3.392/5.088,4.337/6.505,6.986/10.479 respectively, not order levels or validated exit rule. 1.5ATR is inside2ADR here, but no tested conditional win rate. Thirty-three observed USD candidates ranked by7-day sum(v*vw): USDC,BTC,ETH,SOL,XRP,PAXG,LINK,UNI,USDT,AAVE top10 of that partial sample. Full tradable census unavailable; do not mislabel global/top10Alpaca coverage. Sole true signal USDT, whose anomalous stale-looking prior prices are not a validated stablecoin strategy. All other32 false. User book has no holding to reset. Strong bear case: latest daily close failed breakout and no delayed-entry rule validated. Counterargument:20-day returns remain strong; trend positivity alone cannot override entry definition or costs. Proposed size0.

Data https://runway.adityaperswal.workers.dev/data/crypto-bars ; Jan1-Sep4 2026 daily returns. BTC/ETH correlation0.9124, BTC/SOL0.8805, ETH/SOL0.8801; OLS beta via covariance/variance ETH1.254,SOL1.202. Daily SD BTC2.452%,ETH3.370%,SOL3.347%; worst close-close -13.986/-14.890/-14.931%. Premise1: three major longs are strongly correlated risk, not three independent bets. Premise2: registered net daily closed-trade bootstrap (10000 IID resamples, seed20260905) mean95% intervals BTC[-1.424,1.323]%,ETH[-2.670,.678]%,SOL[-1.213,6.606]%. SOL N12, win50%,mean win6.123%,mean loss2.270%; naïve quarter-Kelly .07866, but authorized size0 because interval crosses0, history selected, prospective validation absent and venue exits untested. Bootstrap-positive fraction .7897 is resampling stability, NOT probability strategy profitable. IID assumption ignores serial dependence, so intervals may understate uncertainty. All hourly20 net mean intervals entirely negative: BTC[-.611,-.443],ETH[-.609,-.346],SOL[-.682,-.465]%. Strong counterexample: SOL daily may have real positive momentum edge despite inconclusive sampling; insufficient confidence does not prove negative return. Goal repeatable net gains, binding constraint independent payoff validation; do not convert raw Kelly into dollars. Methods/results artifacts return-statistics.json,bootstrap.json and risk_analysis.py; no predictive regression claimed.

Ran mandated CLI node /app/agent/dist/backtest.js --symbols BTC/USD,ETH/USD,SOL/USD --strategy breakout --lookback20 --days730 (actual flag --lookback 20). Results [{"symbol":"BTC/USD","bars":731,"return":26.56435652025784,"drawdown":13.536723623989971,"win_rate":34,"sample":50,"buy_hold":41.818806966095835},{"symbol":"ETH/USD","bars":730,"return":28.039005217366153,"drawdown":24.943256404302108,"win_rate":21.052631578947366,"sample":38,"buy_hold":3.734854432360324},{"symbol":"SOL/USD","bars":731,"return":75.2056289097867,"drawdown":18.79539437658838,"win_rate":52.94117647058824,"sample":34,"buy_hold":-21.20949655865749}]. This control includes current-day partial data because CLI omits an end cutoff, has varying731/730 bars, uses same signal-close fills, no fees/slippage, no independent chronological holdout and no intrabar-mark drawdown. Hence not directly same window as registered experiment and not positive sizing evidence. Full CLI JSON saved artifacts/cli.json. Source /app/agent/dist/backtest/{bars,strategies,engine}.js and https://runway.adityaperswal.workers.dev/data/crypto-bars . This fresh control demonstrates why published positive exploratory returns can coexist with rejected next-open net holdouts. The registered vectorbt tests, not this control, govern current verdict.

Same registered DAILY prior20-close signal, made available at next UTC00:00, held constant through each following day and filled at hourly OPEN. Not a20-hour strategy. Vectorbt cost25bp fee+5bp slippage each side, Jan1-Sep4 2026 diagnostic holdout, warmup from Sep2024; previously inspected history is not independent OOS. All returns/payoffs match daily implementation; marking every hour exposes deeper drawdowns. Results: [{"symbol":"SOL/USD","variant":"daily20 executed hourly","fee_bps":25,"slip_bps":5,"return":22.226119845043105,"drawdown":15.009666757160833,"sample":12,"win_rate":50,"avg_win":6.123110672490873,"avg_loss":2.2699029810825238,"open":0,"independent_ledger_match":true},{"symbol":"SOL/USD","variant":"daily20 executed hourly","fee_bps":50,"slip_bps":10,"return":13.735080474897297,"drawdown":18.992889444253546,"sample":12,"win_rate":50,"avg_win":5.5004578599669145,"avg_loss":2.863152872851584,"open":0}]. Separate cash/quantity ledger matches baseline to numerical precision. Initial replication failed due object-typed boolean after forward filling (bitwise complement became integers); explicitly cast bool, reran JIT and asserted ledger equality before recording final results. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Files extended_audit.py, artifacts/daily-on-hourly.json, ledger-verification.json under /tmp/codex-run-fYGHo5. Costs approximate asset-denominated fee as cash; no executable stop/target or15-minute exit delay model. Missing one shared hourly holdout bar retained as gap. Positive SOL result remains selected,N12, uncertain; no entry authorization.

Same registered DAILY prior20-close signal, made available at next UTC00:00, held constant through each following day and filled at hourly OPEN. Not a20-hour strategy. Vectorbt cost25bp fee+5bp slippage each side, Jan1-Sep4 2026 diagnostic holdout, warmup from Sep2024; previously inspected history is not independent OOS. All returns/payoffs match daily implementation; marking every hour exposes deeper drawdowns. Results: [{"symbol":"ETH/USD","variant":"daily20 executed hourly","fee_bps":25,"slip_bps":5,"return":-15.569340386315634,"drawdown":25.32511446164405,"sample":14,"win_rate":21.428571428571427,"avg_win":4.470243968974899,"avg_loss":2.6860993280793997,"open":0,"independent_ledger_match":true},{"symbol":"ETH/USD","variant":"daily20 executed hourly","fee_bps":50,"slip_bps":10,"return":-22.371892459130414,"drawdown":30.928453319792336,"sample":14,"win_rate":21.428571428571427,"avg_win":3.8533815805712788,"avg_loss":3.277891175251558,"open":0}]. Separate cash/quantity ledger matches baseline to numerical precision. Initial replication failed due object-typed boolean after forward filling (bitwise complement became integers); explicitly cast bool, reran JIT and asserted ledger equality before recording final results. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Files extended_audit.py, artifacts/daily-on-hourly.json, ledger-verification.json under /tmp/codex-run-fYGHo5. Costs approximate asset-denominated fee as cash; no executable stop/target or15-minute exit delay model. Missing one shared hourly holdout bar retained as gap. Positive SOL result remains selected,N12, uncertain; no entry authorization.

Same registered DAILY prior20-close signal, made available at next UTC00:00, held constant through each following day and filled at hourly OPEN. Not a20-hour strategy. Vectorbt cost25bp fee+5bp slippage each side, Jan1-Sep4 2026 diagnostic holdout, warmup from Sep2024; previously inspected history is not independent OOS. All returns/payoffs match daily implementation; marking every hour exposes deeper drawdowns. Results: [{"symbol":"BTC/USD","variant":"daily20 executed hourly","fee_bps":25,"slip_bps":5,"return":-7.441465595179779,"drawdown":16.52065691493495,"sample":18,"win_rate":27.77777777777778,"avg_win":3.1215858415895776,"avg_loss":1.561427356348179,"open":1,"independent_ledger_match":true},{"symbol":"BTC/USD","variant":"daily20 executed hourly","fee_bps":50,"slip_bps":10,"return":-17.16508600402671,"drawdown":22.847963674830375,"sample":18,"win_rate":27.77777777777778,"avg_win":2.5094481551442414,"avg_loss":2.1571592234255834,"open":1}]. Separate cash/quantity ledger matches baseline to numerical precision. Initial replication failed due object-typed boolean after forward filling (bitwise complement became integers); explicitly cast bool, reran JIT and asserted ledger equality before recording final results. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Files extended_audit.py, artifacts/daily-on-hourly.json, ledger-verification.json under /tmp/codex-run-fYGHo5. Costs approximate asset-denominated fee as cash; no executable stop/target or15-minute exit delay model. Missing one shared hourly holdout bar retained as gap. Positive SOL result remains selected,N12, uncertain; no entry authorization.

Premise1: issuer IBIT page https://www.ishares.com/us/products/333011/ishares-bitcoin-trust-etf reports Sep3 shares1,375,120,000, NAV46.13; Sep4 creations unavailable. Search snippet for Sep1 versus refreshed live page is not auditable daily flow. AUM/trading volume are not creations. Premise2: https://ethereum.org/roadmap/glamsterdam/ says Sepolia Sep28, mainnet Q4 date unconfirmed. https://blog.ethereum.org/2026/08/24/glamsterdam-repricing-testing dated Aug24 identifies contract compatibility risks under EIP8037/8038; most contracts unaffected and many issues fixable with gas-limit changes. Premise3: https://solana.com/upgrades lists Agave4.3/Alpenglow October, while https://solana.com/upgrades/alpenglow retains Q3 expectation; no exact mainnet date confirmed. Premise4: https://www.circle.com/transparency current numerical fields did not render, so stablecoin supply growth is unknown. Conclusion: no issuer-supported dated-price catalyst justifies new position. Validity: network capacity improvements do not logically ensure token appreciation. Soundness: sources closest to implementation are commercially/ecosystem interested; missing token fee/burn/issuance and valuation series prevent fair-value calculation. We did not quantify exchange flows or unlock inventories and do not assert none exist. Strong bear case: fees per unit can fall faster than usage grows, technical rollout can delay, and ETF demand narrative may mistake price-driven AUM changes for new buying. Strong counterargument: successful upgrades may improve demand and reduce costs, but expected benefits can already be priced. This is qualitative fundamental analysis, not a token valuation model. No stock/options or new-fund proposal follows. ETH monitor10 reviews Sep28 at00UTC, explicitly a chosen review time rather than activation hour; success requires subsequent tested response, failure/delay retains zero exposure.

Preregistered note92 audit: close[t]>max(previous20 closes); hold only while signal true, next observed bar OPEN fills. 20 HOURS is a separate high-turnover horizon, not an hourly replication of 20 DAYS. Data https://runway.adityaperswal.workers.dev/data/crypto-bars Sep1 2024-Sep4 2026, full pagination. Development before Jan1 2026, diagnostic chronological holdout Jan1-Sep4 (247days). Holdout overlaps history previously inspected; not genuinely independent OOS. Prospective window in note92 remains untouched. Vectorbt1.1 all-cash per symbol, initial10000, no leverage; fees25bp + slippage5bp PER SIDE, stress50+10, zero-cost comparison. Official fee https://docs.alpaca.markets/us/docs/crypto-fees ; slippage is assumed, not measured. Training-only L15/20/30 sweep, L20 fixed before results. No stop/target overlay; not an executable venue exit rule. Baseline independently reproduced by cash/quantity ledger to <4e-13 percentage points. Return includes mark-to-market open trades; payoff metrics closed only. Missing hourly data unfilled (one shared Mar12 holdout hour); zero-volume bars mean liquidity not demonstrated. Raw/scripts/hashes in /tmp/codex-run-fYGHo5/artifacts and audit.py. Results (%): [{"phase":"development","L":15,"fee":25,"slip":5,"ret":-99.65586598112374,"dd":99.6560465932145,"N":870,"win":19.655172413793103,"avgW":1.215670107301308,"avgL":1.0979141399779442,"open":0},{"phase":"development","L":20,"fee":25,"slip":5,"ret":-99.32688214549539,"dd":99.3268821454953,"N":747,"win":19.544846050870145,"avgW":1.26815229167138,"avgL":1.1283502305288053,"open":0},{"phase":"development","L":30,"fee":25,"slip":5,"ret":-98.55344598044546,"dd":98.55344598044547,"N":584,"win":19.34931506849315,"avgW":1.1873630581543912,"avgL":1.172581187240879,"open":0},{"phase":"holdout","L":20,"fee":0,"slip":0,"ret":5.821752539654608,"dd":12.39530201989012,"N":355,"win":31.549295774647888,"avgW":0.9816088641350075,"avgL":0.4214337819158554,"open":0},{"phase":"holdout","L":20,"fee":25,"slip":5,"ret":-87.42447617299416,"dd":87.42447617299419,"N":355,"win":13.521126760563378,"avgW":1.358595442097282,"avgL":0.8814354701161116,"open":0},{"phase":"holdout","L":20,"fee":50,"slip":10,"ret":-98.50559777346359,"dd":98.50559777346356,"N":355,"win":7.323943661971831,"avgW":1.68861666312286,"avgL":1.4042532134752974,"open":0}]. Conclusion: reject this net implementation; positive frictionless return does not establish fee-paying edge. No inference about all momentum strategies.

Preregistered note92 audit: close[t]>max(previous20 closes); hold only while signal true, next observed bar OPEN fills. 20 DAYS; next-open entry/flat signal exit. Data https://runway.adityaperswal.workers.dev/data/crypto-bars Sep1 2024-Sep4 2026, full pagination. Development before Jan1 2026, diagnostic chronological holdout Jan1-Sep4 (247days). Holdout overlaps history previously inspected; not genuinely independent OOS. Prospective window in note92 remains untouched. Vectorbt1.1 all-cash per symbol, initial10000, no leverage; fees25bp + slippage5bp PER SIDE, stress50+10, zero-cost comparison. Official fee https://docs.alpaca.markets/us/docs/crypto-fees ; slippage is assumed, not measured. Training-only L15/20/30 sweep, L20 fixed before results. No stop/target overlay; not an executable venue exit rule. Baseline independently reproduced by cash/quantity ledger to <4e-13 percentage points. Return includes mark-to-market open trades; payoff metrics closed only. Missing hourly data unfilled (one shared Mar12 holdout hour); zero-volume bars mean liquidity not demonstrated. Raw/scripts/hashes in /tmp/codex-run-fYGHo5/artifacts and audit.py. Results (%): [{"phase":"development","L":15,"fee":25,"slip":5,"ret":-8.699034543022481,"dd":21.483379164553128,"N":30,"win":40,"avgW":4.850908167798842,"avgL":3.511809065973165,"open":0},{"phase":"development","L":20,"fee":25,"slip":5,"ret":8.193517082493582,"dd":20.53860577582326,"N":24,"win":41.66666666666667,"avgW":5.410592962126057,"avgL":3.054607158838205,"open":0},{"phase":"development","L":30,"fee":25,"slip":5,"ret":22.058406845926328,"dd":20.4156260200288,"N":16,"win":50,"avgW":6.349890820812931,"avgL":3.460370614555807,"open":0},{"phase":"holdout","L":20,"fee":0,"slip":0,"ret":31.350969425942164,"dd":10.158416229540823,"N":12,"win":50,"avgW":6.746403003834854,"avgL":1.676064029581792,"open":0},{"phase":"holdout","L":20,"fee":25,"slip":5,"ret":22.226119845043105,"dd":12.675513907651492,"N":12,"win":50,"avgW":6.123110672490873,"avgL":2.2699029810825238,"open":0},{"phase":"holdout","L":20,"fee":50,"slip":10,"ret":13.735080474897297,"dd":17.018264383492877,"N":12,"win":50,"avgW":5.5004578599669145,"avgL":2.863152872851584,"open":0}]. Conclusion: positive exploratory daily result, but N12, selection and uncertain exits keep WATCH and zero size. No inference about all momentum strategies.

Preregistered note92 audit: close[t]>max(previous20 closes); hold only while signal true, next observed bar OPEN fills. 20 HOURS is a separate high-turnover horizon, not an hourly replication of 20 DAYS. Data https://runway.adityaperswal.workers.dev/data/crypto-bars Sep1 2024-Sep4 2026, full pagination. Development before Jan1 2026, diagnostic chronological holdout Jan1-Sep4 (247days). Holdout overlaps history previously inspected; not genuinely independent OOS. Prospective window in note92 remains untouched. Vectorbt1.1 all-cash per symbol, initial10000, no leverage; fees25bp + slippage5bp PER SIDE, stress50+10, zero-cost comparison. Official fee https://docs.alpaca.markets/us/docs/crypto-fees ; slippage is assumed, not measured. Training-only L15/20/30 sweep, L20 fixed before results. No stop/target overlay; not an executable venue exit rule. Baseline independently reproduced by cash/quantity ledger to <4e-13 percentage points. Return includes mark-to-market open trades; payoff metrics closed only. Missing hourly data unfilled (one shared Mar12 holdout hour); zero-volume bars mean liquidity not demonstrated. Raw/scripts/hashes in /tmp/codex-run-fYGHo5/artifacts and audit.py. Results (%): [{"phase":"development","L":15,"fee":25,"slip":5,"ret":-98.98027822695771,"dd":98.98114637672167,"N":836,"win":17.105263157894736,"avgW":1.1588355581865888,"avgL":0.8942090604423029,"open":0},{"phase":"development","L":20,"fee":25,"slip":5,"ret":-97.33182984994359,"dd":97.34102716207289,"N":696,"win":18.103448275862068,"avgW":1.227928478444824,"avgL":0.8999617539377323,"open":0},{"phase":"development","L":30,"fee":25,"slip":5,"ret":-92.92567247386602,"dd":93.02844250804796,"N":535,"win":19.439252336448597,"avgW":1.1711264834517694,"avgL":0.8902230274841938,"open":0},{"phase":"holdout","L":20,"fee":0,"slip":0,"ret":43.619932920648694,"dd":8.982065444455966,"N":337,"win":33.82789317507419,"avgW":1.0375806726602939,"avgL":0.3568601989699852,"open":0},{"phase":"holdout","L":20,"fee":25,"slip":5,"ret":-80.98616414390666,"dd":81.06051996423044,"N":337,"win":13.94658753709199,"avgW":1.5583937954084368,"avgL":0.816463571875751,"open":0},{"phase":"holdout","L":20,"fee":50,"slip":10,"ret":-97.48281265008445,"dd":97.48281265008445,"N":337,"win":7.71513353115727,"avgW":1.9998189830661508,"avgL":1.3426282696477871,"open":0}]. Conclusion: reject this net implementation; positive frictionless return does not establish fee-paying edge. No inference about all momentum strategies.

Preregistered note92 audit: close[t]>max(previous20 closes); hold only while signal true, next observed bar OPEN fills. 20 DAYS; next-open entry/flat signal exit. Data https://runway.adityaperswal.workers.dev/data/crypto-bars Sep1 2024-Sep4 2026, full pagination. Development before Jan1 2026, diagnostic chronological holdout Jan1-Sep4 (247days). Holdout overlaps history previously inspected; not genuinely independent OOS. Prospective window in note92 remains untouched. Vectorbt1.1 all-cash per symbol, initial10000, no leverage; fees25bp + slippage5bp PER SIDE, stress50+10, zero-cost comparison. Official fee https://docs.alpaca.markets/us/docs/crypto-fees ; slippage is assumed, not measured. Training-only L15/20/30 sweep, L20 fixed before results. No stop/target overlay; not an executable venue exit rule. Baseline independently reproduced by cash/quantity ledger to <4e-13 percentage points. Return includes mark-to-market open trades; payoff metrics closed only. Missing hourly data unfilled (one shared Mar12 holdout hour); zero-volume bars mean liquidity not demonstrated. Raw/scripts/hashes in /tmp/codex-run-fYGHo5/artifacts and audit.py. Results (%): [{"phase":"development","L":15,"fee":25,"slip":5,"ret":5.074546192169164,"dd":33.36730432737971,"N":33,"win":27.27272727272727,"avgW":9.09382998262413,"avgL":2.878731642004194,"open":0},{"phase":"development","L":20,"fee":25,"slip":5,"ret":13.413365985730344,"dd":18.18321691858682,"N":26,"win":19.230769230769234,"avgW":14.041640552613908,"avgL":2.459762665071717,"open":0},{"phase":"development","L":30,"fee":25,"slip":5,"ret":24.292399949656083,"dd":16.5987981083772,"N":19,"win":21.052631578947366,"avgW":14.569092226518974,"avgL":2.105653587188495,"open":0},{"phase":"holdout","L":20,"fee":0,"slip":0,"ret":-8.17076158187183,"dd":19.76422684548176,"N":14,"win":21.428571428571427,"avgW":5.087735939106771,"avgL":2.0937209233623886,"open":0},{"phase":"holdout","L":20,"fee":25,"slip":5,"ret":-15.569340386315634,"dd":24.896705464423608,"N":14,"win":21.428571428571427,"avgW":4.470243968974899,"avgL":2.6860993280793997,"open":0},{"phase":"holdout","L":20,"fee":50,"slip":10,"ret":-22.371892459130414,"dd":30.324151062260917,"N":14,"win":21.428571428571427,"avgW":3.8533815805712788,"avgL":3.277891175251558,"open":0}]. Conclusion: reject this net implementation; positive frictionless return does not establish fee-paying edge. No inference about all momentum strategies.