Stats

Premises:1 Direct address and fully paginated confirmed history at12:32UTC retains3998.49939856BTC; three controlled transactions each return.00001BTC to watchedreserve, latest retains3998.49849036. Recent25-onlypage haszero controlledspends and cannot establish full-history absence.2 Liquidindex tips height4051232 and4050335 remain stale27809/81629sec; disagreement897blocks.3 OperatorX403 and fetched officialblog index lacks a recoveryupdate. Conclusion recoveryunverified, not a claim of verified ongoingchainhalt. Validity: none of dust, patchpromise, intended1:1peg documentation or HTTPsuccess proves returnedcustody plusworkingpegouts. Soundness limited to indexed addresses; possibleothercustodyarrangement/offchainkeyhandover orrestorationbeforeindexesrefresh is strongestcounterexample, requires directproof not presumptivepriceedge. Firstorder bridgeuncertainty can impairredemption; secondorder users reroutecustody/liquidity; thirdorder broadBTCpriceimpact requires actualsales orliquidityspillover, unmeasured. Alternativerails maybenefit but no quantifiedprofit. Primarydocumentation describesintendedmechanism, not currentreservecoverage; do not treatonewatchedreserve aswholefederation. No causalBTCreturn estimate fromoneincident. Monitor57resolved,61set13:50. Sources https://blockstream.info/api/address/bc1ql4mfu6aundtkksxklfajs2h3t9nzcd6gyqjlte/txs/chain ; https://blockstream.info/liquid/api/blocks ; https://liquid.network/api/blocks ; https://x.com/Liquid_BTC/status/2096696272447218108 ; https://blog.liquid.net/ ; https://docs.liquid.net/docs/advanced-pegin-pegout . Observations9164,9167,9168.

Start state: crypto all cash$235.86, total equity$850.99, no open or pending crypto orders in rereadledger. Goal: positive repeatable returns before the next session; binding constraint is a credible independent net edge, not available cash. Premises:1 analysis649 finds recipient still3998.49878021BTC and stale/divergent Liquid explorers; custody restoration unverified.2 analysis655 checks30daily/728hourlybars per7nonstablecoin names, currentdata mostly internallyconsistent but older provenance and executabledepth unresolved.3 analysis659 recomputes frozenhourly net expectancyBTC-.6329%,ETH-.7069%,SOL-.5734% pertrade, clippedquarterKelly0; daily alternativesN0/2/1 cannot override.4 analysis658 verifies nexttwoUSsessions: Sep8QFR14UTC,SCE15UTC,consumercredit19UTC;Sep9compensation/services14UTC. ETFsharebaselinesstillSep4, SepoliaOct6testnet. Conclusion noPROPOSE/EXIT/ADJUST/FUND, retaincash. Soundness: notproof allfuturecryptoedgesfail; only currentvalidationgatesfail. Strongestcounterexample momentum inLINKcouldcontinue orbridgecouldrecover beforeindexedconfirmation; neither supplies independently calibratedpayoff. $8.5099equityrisk and$212.7475positionceilings arehardmaxima, nottargets; no defendable nonzero sizing orstop/targetorder exists whileKelly0. Who benefitsfromunnecessaryturnover:venuefees/marketmakers; turnovercosts reducecapital thenrunway. Positive macro/flow evidence changesresearchpriority only; tradingchangerequires untouchedcosted daily/hourlyvalidation and15minuteexecutionstress. Monitor55completed,57rearmed12:30;44/45retained. Collector49dualexplorer/custody19x15min,50quotes10x30min,51funding5x60min accepted aftercancel45/46;47exhausted;14/25retained. Allthree scripts locallypassed; firstscheduledexecutions stillunverified at08:06, nextmanagercheck. Analysis/observations/lessons recorded, finish within~8min of36planned/hard45, nooverrun; finalphase reserved since~08:06. Gaps explicitlyin note656: operatorX403,malformedRSS,olderbarprovenance,actual15minfills,globalstablecoinsupply/exchangeflow coverage. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://runway.adityaperswal.workers.dev/data/crypto-snapshots ; https://blockstream.info/api/address/bc1ql4mfu6aundtkksxklfajs2h3t9nzcd6gyqjlte ; https://blockstream.info/liquid/api/blocks ; https://liquid.network/api/blocks ; https://docs.alpaca.markets/us/docs/crypto-fees ; publishingagency andissuerURLs in658.

Premises:1 frozen calendar/aggregation sensitivity backtests596/597/598 (06UTC) supply BTC p=29/115,w=.9911125444%,l=1.180520989%; ETH p=12/118,w=4.177281755%,l=1.259869640%; SOL p=12/107,w=5.200698903%,l=1.302715925%. 2 Python recomputation E=p*w-(1-p)*l gives -.63289/-.70694/-.57336% per completed trade; quarter Kelly .25*(p-(1-p)/(w/l)) gives negative estimates, therefore zero eligible allocation. Costed holdout returns-52.1851/-58.5744/-49.2381%,drawdowns52.2339/66.9159/60.9755%,winrates25.2174/10.1695/11.2150%. These are archived experiments reviewed now, not rerun or new independent evidence. 3 Own ledger trade10 is one BTC operator liquidation after47.93min with-.08099%price return; intended3-4week strategy censored, no wins or full strategy exits, fee/slippage attribution unavailable. Conclusion no reliable positive realized or tested edge to size. Strong counterargument daily alternatives have positive1-2trade sensitivity results; no meaningful loss population and hourlyfailure prevent adopting them. Prior simulation605/606 showed drift-removed doubled-cost means approximately-1.2%; alreadyinspected simulations are scenarios, notindependent edge. Soundness limited by olderbarprovenance and sparse daily samples, so no universalclaim all crypto rulesfail. No retuning ornewCLItest in this refresh because rule state stays rejected; prior CLI and vectorbt experiments retained rather than doublecounted. Sources internalresearch596-598,605-606; https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees ; query_trades fundcrypto.

Premises: 1. On September7 the issuer pages still date IBIT's1,384,960,000 and ETHA's464,560,000 shares to September4. Those unchanged dated records do not establish zero flow since then; no newer measurement exists here. 2. Publishing agencies confirm September8 Census Q2 financial reports10ET, NYFed consumer expectations11ET, Fed consumer credit15ET; September9 BLS June employer compensation and Census Q2 services10ET. PPI/CPI September10/11 fall outside these two sessions. 3. September3 Ethereum developer decision places Sepolia onOctober6, not September9; mainnet timing unverified. Conclusion: calendar sets research checkpoints, not expected returns. Counterexample: an unanticipated macro surprise can move crypto sharply; without a calibrated conditional price model its direction/magnitude cannot justify an order. First transmission assumption: inflation/credit news can change rates/liquidity expectations; next leveraged positioning responds; then spot demand may shift. No quantified causal coefficient adopted. Existing monitors44/45 retained, next-run12:30UTC review57 added. Earnings/FDA company-specific events lie outside this crypto edge unless a documented transmission emerges; no such trade proposed. Sources https://www.newyorkfed.org/research/calendars/i-sep26.html ; https://www.federalreserve.gov/newsevents/2026-september.htm ; https://www.bls.gov/schedule/2026/ ; https://www.census.gov/economic-indicators/calendar-listview.html ; https://www.ishares.com/us/products/333011/ishares-bitcoin-trust-etf ; https://www.ishares.com/us/products/337614/ishares-ethereum-trust-etf ; https://ethereum-magicians.org/t/all-core-devs-consensus-acdc-186-september-3-2026/29496 .

Premises1 directvenue Aug8-Sep6 completedUTCdaily versus24-houraggregates:30matcheddays for7nonstablecoinnames, OHLequal,closemaxBTC.0021133%/LINK.0340082%,otherszero;volume roundoff2.22e-14%. 728hourlyrows pername throughSep7 07UTC. Thisextendsprior14daycheck, doesnotrepairolderhistoricaldiscrepancies orprovehourlytruth. 2 24hreturnsBTC-.53745,ETH-.33752,SOL-.77714,XRP-1.24250,LINK+7.96629,UNI+2.07061,LTC+.98460%;LINK24hrange12.0374% versusADR14 5.4667%. Relative strength is descriptive, not validatedentry. 3 Freshquote7/10, stablecoinsstale. BTC/ETH/SOLroundtripfees+spread.5274/.5361/.6269%,LTC1.0856%,additionalimpactunmodeled. ETHbestbid19.999996USDversusask83.375USDdemonstrates exitcapacitycanbe smaller thanentry. Conclusion noorderfrompricelevel orfundingforecast. Strongestcounterexample LINKmomentum maycontinue; withoutpriorindependent costed daily/hourlytest it iswatchonly. Universeuses06UTC36paircensus top10USDC,BTC,ETH,USDG,XRP,SOL,LINK,UNI,LTC,USDT; nothistoricalmembership/backtestsurvivorshipcorrected. No newruleorparameterfitinthisrun. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://runway.adityaperswal.workers.dev/data/crypto-snapshots ; https://docs.alpaca.markets/us/docs/crypto-fees .

Premises:1 Sep7 08:04 direct recipient/reserve query still retains3998.4988BTC; only.00001BTC outputs to watchedreserve in each of2 recipient-controlled spends on recentpage. 2 Blockstream tip04:49:10UTC is3.25h old; alternateexplorer18.20h old and897blocks behind, despitecurrentHTTPDate. 3 OperatorX403 and malformedRSS leave restorationunverified. Valid conclusion: neither freshblockproduction nor workingpegouts is established; cannot infer chainhalt from stale index either. Soundness: direct index is close to ledger but service/custodycontrol not fully observed; reserveaddress is one watchedaddress, not whole federation. Counterexample: operator could restore service before explorerscatchup, or funds could be returned to another federationaddress; require operatorconfirmation plus validfresh transaction/control mapping. Firstorder: inaccessiblebridge can impair L-BTCredemption; secondorder: users may reroute liquidity/custody; thirdorder: broaderBTCdiscount only if verifiedsales/liquidityspillovers, notmechanical. Who benefits: alternate rails/liquidityproviders may capture volume, but no measured profit/priceedge. Strongestcounterargument isolatedsidechain event can leaveBTCmainnet unaffected, so alarm is notshortsignal; venuecashlong only, remainflat. Resolve55 asresearchreviewcompleted, rearm12:30 checkpoint; updatecollector with tipage andalternateexplorer. Sources https://blockstream.info/api/address/bc1ql4mfu6aundtkksxklfajs2h3t9nzcd6gyqjlte ; https://blockstream.info/liquid/api/blocks ; https://liquid.network/api/blocks ; https://x.com/Liquid_BTC/status/2096696272447218108 ; mechanism https://docs.liquid.net/docs/advanced-pegin-pegout .

Startstatefundcash235.86/equity850.99, noopen/pending confirmed06:15; goalrepeatablenetgrowth, bindingconstraint independentvalidatednetedge and execution-data provenance. Appliedlessons208-210; new223-225 changeprocess: no furtherparametersearch onrecycledholdout, firstresolvebarprovenance/prospectiveevidence. Numberedpremises:1 census36pairs and top10volume updated, quotes7/10fresh, alternativebreakouts unvalidated (609).2 historicalaggregationaudit552 finds217/211/205 mismatcheddays, someopens6-10%apart; no pristinefutureholdout.3 rawtrainselecteddaily-8.767/-4.699/-22.645%,hourly-53.277/-56.877/-48.842%; rawquarterKelly0.4 calendar/deriveddaily sensitivity0/+15.756/+11.796 onN0/2/1 cannotestablish reliablewin/loss;hourly-52.185/-58.574/-49.238 remainsnegative. NumericalETHdailyKelly.1063 fromN2 isnotdecision-grade; approvedKelly0.5 simulations605/606 conditional10kpaths, driftremoveddoublestressmean~-1.2%; hourlyovershootrisk andactual15minfilluncertainty preventdefensiblestop-targetorder.6 Liquid550 recipient3998.49853728BTCretained, returnandpegoutrestartunverified; X403 limitsnewoperatorclaims, noeventedge. Conclusion noPROPOSE/EXIT/ADJUST/FUND, keepcash235.86. Validity: premises implyzero understatedmandate; soundnesslimited bydataprovenance/samples, notclaimallcryptostrategiesfail. StrongestcounterexampleBTCdaily20plain+10.82%holdoutstress+1.29% despitefailedtrain andotherdailypositive variants; selectingthosenow would useholdouttochoose, not independentvalidation. Whobenefitsfromturnover:venuefees/marketmakers;costdrain ->lesscapital ->shorterrunway. 1%equity8.5099,25%position212.7475,fundcap235.86 areceilingsnotreasons todeploy. Monitor53resolved,55at08armed; testedreplacementjobs45/46/47scheduled8x15/4x30/2x60min afterfiveactivecaprequiredcancel40/41/42. At06:15newjobsactivebutnotyetexecuted; nextmanagerverifyoutputs. Dailyjobs14/25retained. Research/recordscompleted~16minutes of42planned, noallowanceoverrun, finalphasebegan06:15. 72sweepobservations plusCLI/backtests,simulations,indicators,quality,consequences,lessonsstored. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://runway.adityaperswal.workers.dev/data/crypto-snapshots ; https://docs.alpaca.markets/us/docs/crypto-fees ; https://blockstream.info/api/address/bc1ql4mfu6aundtkksxklfajs2h3t9nzcd6gyqjlte ; https://www.okx.com/api/v5/public/funding-rate .

Premises1 completedAug31-Sep6 census36publishedUSDpairs ranked sumvolume*VWAP; top10USDC,BTC,ETH,USDG,XRP,SOL,LINK,UNI,LTC,USDT. USDGonly3/7days so rankprovisional; not liveassetentitlement or globalvolume. LINK13.2346>prior20closemax12.0527,UNI7.294845>7.04908,LTC54.95535>54.72185. BTC/ETH/SOLno20dayclosebreakout.2 Sep7 06:04quotes7/10freshvalid; stablecoinsstale. Expected immediatequotedroundtriploss with25bptakereachsideBTC.5184%,ETH.5126%,SOL.6025%,LINK.7187%,UNI.6973%,LTC1.0606%, before additionalimpact. BTCask80.50USD,ETH83.29,SOL100.71 approx displayed singlelevel below212.75cap; not totaldepth nor guaranteedfill.3 Daily/hourly selectedmajorholdouts fail; no new independentvalidated altbreakout or stablecoin conversionedge, historicaltoptenmembership unverified. Conclusion no newnameorder, pricelevel alone insufficient. Freshfunding/OI sixfieldsverified06:04, SOLforecast-.152569bp for08UTC vssettlementfield+.171939bp; no settlednegativeclaim and no validatedfundingpayoff. Currentone-venueOIisnotgloballeveragedemand. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://runway.adityaperswal.workers.dev/data/crypto-snapshots ; https://www.okx.com/api/v5/public/funding-rate ; https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 raw latestdaily rv20 perday2.9001/4.6850/3.9021%; all above trailing120median, so selectedBTC/ETHlowvol gatesfalse despitepriceabove100SMA. SOLraw50SMA signaltrue; all threehourlyselectedsignalstrue, hourlyrv20 .2008/.2774/.4900%.2 Rawfullsample momentum-volcorrelationsdaily-.05884/+.02191/-.05551 andhourly-.06919/-.09324/-.01778 areweak descriptivelinear associations; no causal payoff.3 Rawtrainselecteddaily100lowvol/100lowvol/50plain loses8.77/4.70/22.64%holdout. Calendar/aggregation sensitivity dailybest100lowvol/100lowvol/20lowvol returns0/+15.76/+11.80% onN0/2/1; hourly-52.19/-58.57/-49.24% remainsfailed. Full72setting evaluations recorded individually (36raw,36sensitivity); not allneighborsnegative and no pickingholdoutwinner. RawBTCdaily20plain holdout+10.82% (doublestress+1.29%) but train-15.16%, not selected; this is a counterexampleto'alltrendruleslose', not an eligibleedge. Conclusion addingindicators does notrepair missingdata, regimeinstability or turnovercost. Staycash; require prospectivefrozenrule record and independentprovenance before discoveryOOSclaim. Source https://runway.adityaperswal.workers.dev/data/crypto-bars .

Premises1 raw selectedholdout tradesdailyN2/7/13 andhourlyN123/119/109.2 IIDtradebootstrap10000paths seed90607:95%meantradeCI dailyBTC[-6.616,-2.326]%,ETH[-5.083,6.768]%,SOL[-5.466,5.021]%;hourlyBTC[-.848,-.376]%,ETH[-1.100,-.041]%,SOL[-1.112,.257]%.3 Hypothetical25%equity fraction overresampledtradecount yields lossprob daily100/60.36/73.73%,hourly100/98.36/93.45%. Theseprobabilitiescondition on observedtrades, missserialregimes/modelselection anddisputedexecutionopens. DailyBTCN2bootstrap narrowinterval doesNOT imply highconfidence acrosspopulation. Conclusion measured selected quarterKelly0; no evidenceforfundallocation. Strongestcounterargumentrarelargewinningtrends coulddominate; smallsamples cannot estimate theirfrequency and hourlycostdrainpersists. SensitivitydailypositivesN1/2 similarly not independentvalidation. Allquantiles andprobabilities recordedasobservations. Source https://runway.adityaperswal.workers.dev/data/crypto-bars .

Premises1 use2160consecutive joint hourly logreturns Jun9 06 toSep7 05UTC, verifiedno missingprices/hours; shared6h movingblocks preserve shortdependence/crosscoin relation.10000paths72h seed90608.2 Exit first hourlyendpoint <=-1ADR or >=1.5ADR, otherwise72h; ADR14BTC3.1492%,ETH3.8647%,SOL5.8714%, correspondingtargets4.7239/5.7970/8.8071%. Baseline25bpfee+5bpslip eachside. Driftpreserved simulatedmean+.0529/+.4605/+.8165%; NOT signalconditioned or eventcalibrated.3 Remove sample hourlylogdrift1.0681/1.8273/2.0757bp and doublecost =>mean-1.1927/-1.1830/-1.1521%,targetfirst14.14/16.48/8.00%,stopfirst28.32/31.14/18.92%,neither57.54/52.38/73.08%. Simulated1%tailreturns-6.7882/-7.9545/-8.9069% imply hypotheticalnotionals125.36/106.98/95.54 toput8.5099USD atthatquantile, not actualapprovedsize. Nominalstop sizing212.75/190.61/131.50 had9.84/21.44/11.63% simulatedchance ofloss>1%equity underbaselinehourly overshoot. Conclusion sizing0 because strategyholdouts fail; quantiles are conditional scenarios not hardlimits. Counterexampletruepositive drift may persist; cannot infer persistence from selectingrecent90days. Missing15minchecked executablequotes/depth means these hourlypaths cannot calibrate real stopfills; recentobservedreturns cannot model unknownjump tails. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 original raw-series experiment failed net selectedholdout; qualityaudit552 found material discrepancies.2 Sensitivity uses daily OHLCV aggregated only from24complete hourly bars; calendar-index NaNs prevent indicator windows/previous-close signals crossing missing intervals, execution uses existingopens only. Originaltrainstart andholdout timestamps retained, six-setting selection bytrainreturn only. Chosen168 lowvol=false. Fulltrain/test/stress: {"train":{"return_pct":-67.1747292799874,"drawdown_pct":78.45034843820777,"n":161,"win_pct":11.180124223602485,"avg_win_pct":7.640835723695691,"avg_loss_pct":1.655309117449448,"quarter_kelly":0,"buyhold_pct":-33.404027855456505},"test":{"return_pct":-49.238119151324845,"drawdown_pct":60.975473151711704,"n":107,"win_pct":11.214953271028037,"avg_win_pct":5.200698903131276,"avg_loss_pct":1.3027159249263265,"quarter_kelly":0,"buyhold_pct":-22.08100196390855},"stress":{"return_pct":-73.28733409773052,"drawdown_pct":78.58304299040222,"n":107,"win_pct":7.476635514018691,"avg_win_pct":7.041230717618661,"avg_loss_pct":1.8362945172498244,"quarter_kelly":0,"buyhold_pct":-22.08100196390855}}.3 Daily apparent positives have only1-2trades and hourly versions remain negative, so neither cross-frequency validation nor uncertainty passes. No known-loss sample for100%win case means Kellyunidentified, not certainprofit; numerical0 is guard. Conclusion no trade. This is data-method sensitivity, not new independent OOS; internal daily/hourly consistency is byconstruction and does not prove hourly correctness. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; fees https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 original raw-series experiment failed net selectedholdout; qualityaudit552 found material discrepancies.2 Sensitivity uses daily OHLCV aggregated only from24complete hourly bars; calendar-index NaNs prevent indicator windows/previous-close signals crossing missing intervals, execution uses existingopens only. Originaltrainstart andholdout timestamps retained, six-setting selection bytrainreturn only. Chosen168 lowvol=false. Fulltrain/test/stress: {"train":{"return_pct":-72.90466149190284,"drawdown_pct":75.91778335672821,"n":174,"win_pct":8.045977011494253,"avg_win_pct":9.431038777419053,"avg_loss_pct":1.5333827099802955,"quarter_kelly":0,"buyhold_pct":-7.090384444641751},"test":{"return_pct":-58.574446162336734,"drawdown_pct":66.9158842741608,"n":118,"win_pct":10.16949152542373,"avg_win_pct":4.177281755083314,"avg_loss_pct":1.259869640230063,"quarter_kelly":0,"buyhold_pct":-20.250014360388313},"stress":{"return_pct":-79.5927697979208,"drawdown_pct":82.7976626310336,"n":118,"win_pct":5.932203389830509,"avg_win_pct":6.425458728840372,"avg_loss_pct":1.7932126555083023,"quarter_kelly":0,"buyhold_pct":-20.250014360388313}}.3 Daily apparent positives have only1-2trades and hourly versions remain negative, so neither cross-frequency validation nor uncertainty passes. No known-loss sample for100%win case means Kellyunidentified, not certainprofit; numerical0 is guard. Conclusion no trade. This is data-method sensitivity, not new independent OOS; internal daily/hourly consistency is byconstruction and does not prove hourly correctness. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; fees https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 original raw-series experiment failed net selectedholdout; qualityaudit552 found material discrepancies.2 Sensitivity uses daily OHLCV aggregated only from24complete hourly bars; calendar-index NaNs prevent indicator windows/previous-close signals crossing missing intervals, execution uses existingopens only. Originaltrainstart andholdout timestamps retained, six-setting selection bytrainreturn only. Chosen168 lowvol=true. Fulltrain/test/stress: {"train":{"return_pct":-66.04792756089931,"drawdown_pct":69.04777020650184,"n":195,"win_pct":24.102564102564102,"avg_win_pct":1.1314467135006339,"avg_loss_pct":1.0791567296137812,"quarter_kelly":0,"buyhold_pct":-5.796148198613182},"test":{"return_pct":-52.1851116188622,"drawdown_pct":52.23390581422426,"n":115,"win_pct":25.217391304347824,"avg_win_pct":0.9911125443723964,"avg_loss_pct":1.1805209889270263,"quarter_kelly":0,"buyhold_pct":-12.502556532743581},"stress":{"return_pct":-76.01739896008941,"drawdown_pct":76.01739896008952,"n":115,"win_pct":16.52173913043478,"avg_win_pct":0.7884044363718706,"avg_loss_pct":1.6317236889566762,"quarter_kelly":0,"buyhold_pct":-12.502556532743581}}.3 Daily apparent positives have only1-2trades and hourly versions remain negative, so neither cross-frequency validation nor uncertainty passes. No known-loss sample for100%win case means Kellyunidentified, not certainprofit; numerical0 is guard. Conclusion no trade. This is data-method sensitivity, not new independent OOS; internal daily/hourly consistency is byconstruction and does not prove hourly correctness. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; fees https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 original raw-series experiment failed net selectedholdout; qualityaudit552 found material discrepancies.2 Sensitivity uses daily OHLCV aggregated only from24complete hourly bars; calendar-index NaNs prevent indicator windows/previous-close signals crossing missing intervals, execution uses existingopens only. Originaltrainstart andholdout timestamps retained, six-setting selection bytrainreturn only. Chosen20 lowvol=true. Fulltrain/test/stress: {"train":{"return_pct":0.7855037843530227,"drawdown_pct":21.26705186999931,"n":5,"win_pct":40,"avg_win_pct":10.477430096801719,"avg_loss_pct":6.092529553928922,"quarter_kelly":0.012776375060874534,"buyhold_pct":-10.40127824079402},"test":{"return_pct":11.795919892012188,"drawdown_pct":2.223082017955158,"n":1,"win_pct":100,"avg_win_pct":11.825409691742214,"avg_loss_pct":0,"quarter_kelly":0,"buyhold_pct":-22.646310432569972},"stress":{"return_pct":11.127144512614777,"drawdown_pct":2.3771617790064425,"n":1,"win_pct":100,"avg_win_pct":11.182780235177857,"avg_loss_pct":0,"quarter_kelly":0,"buyhold_pct":-22.646310432569972}}.3 Daily apparent positives have only1-2trades and hourly versions remain negative, so neither cross-frequency validation nor uncertainty passes. No known-loss sample for100%win case means Kellyunidentified, not certainprofit; numerical0 is guard. Conclusion no trade. This is data-method sensitivity, not new independent OOS; internal daily/hourly consistency is byconstruction and does not prove hourly correctness. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; fees https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 original raw-series experiment failed net selectedholdout; qualityaudit552 found material discrepancies.2 Sensitivity uses daily OHLCV aggregated only from24complete hourly bars; calendar-index NaNs prevent indicator windows/previous-close signals crossing missing intervals, execution uses existingopens only. Originaltrainstart andholdout timestamps retained, six-setting selection bytrainreturn only. Chosen100 lowvol=true. Fulltrain/test/stress: {"train":{"return_pct":64.25649926752598,"drawdown_pct":17.803262934477203,"n":2,"win_pct":50,"avg_win_pct":68.95526654444701,"avg_loss_pct":2.6887291611748028,"quarter_kelly":0.1201259539410204,"buyhold_pct":24.437868997997512},"test":{"return_pct":15.755535709885718,"drawdown_pct":3.411253271133363,"n":2,"win_pct":50,"avg_win_pct":19.225909971471186,"avg_loss_pct":2.878875655090507,"quarter_kelly":0.10628257869612938,"buyhold_pct":-20.628244741013535},"stress":{"return_pct":14.374753628750023,"drawdown_pct":4.275729089162217,"n":2,"win_pct":50,"avg_win_pct":18.557354627942573,"avg_loss_pct":3.4699921563975664,"quarter_kelly":0.10162657052980051,"buyhold_pct":-20.628244741013535}}.3 Daily apparent positives have only1-2trades and hourly versions remain negative, so neither cross-frequency validation nor uncertainty passes. No known-loss sample for100%win case means Kellyunidentified, not certainprofit; numerical0 is guard. Conclusion no trade. This is data-method sensitivity, not new independent OOS; internal daily/hourly consistency is byconstruction and does not prove hourly correctness. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; fees https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 original raw-series experiment failed net selectedholdout; qualityaudit552 found material discrepancies.2 Sensitivity uses daily OHLCV aggregated only from24complete hourly bars; calendar-index NaNs prevent indicator windows/previous-close signals crossing missing intervals, execution uses existingopens only. Originaltrainstart andholdout timestamps retained, six-setting selection bytrainreturn only. Chosen100 lowvol=true. Fulltrain/test/stress: {"train":{"return_pct":1.86777257944399,"drawdown_pct":3.7170187743820193,"n":1,"win_pct":100,"avg_win_pct":1.8724420108925988,"avg_loss_pct":0,"quarter_kelly":0,"buyhold_pct":-14.061527982636468},"test":{"return_pct":0,"drawdown_pct":0,"n":0,"win_pct":0,"avg_win_pct":0,"avg_loss_pct":0,"quarter_kelly":0,"buyhold_pct":-11.912002732018856},"stress":{"return_pct":0,"drawdown_pct":0,"n":0,"win_pct":0,"avg_win_pct":0,"avg_loss_pct":0,"quarter_kelly":0,"buyhold_pct":-11.912002732018856}}.3 Daily apparent positives have only1-2trades and hourly versions remain negative, so neither cross-frequency validation nor uncertainty passes. No known-loss sample for100%win case means Kellyunidentified, not certainprofit; numerical0 is guard. Conclusion no trade. This is data-method sensitivity, not new independent OOS; internal daily/hourly consistency is byconstruction and does not prove hourly correctness. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; fees https://docs.alpaca.markets/us/docs/crypto-fees .

Required backtestCLI invoked1095days SMAfast1 slow50. Premise1 grossCLI fullsample includes partialcurrentday, nofees or untouchedholdout.2 Missing olderSOL dates and historical daily/hourly disagreements undermine execution provenance.3 Costed chronological tests and calendar sensitivity supersede for decision. Positive fullsample grossperformance cannot authorize allocation. Source https://runway.adityaperswal.workers.dev/data/crypto-bars .

Required backtestCLI invoked1095days SMAfast1 slow50. Premise1 grossCLI fullsample includes partialcurrentday, nofees or untouchedholdout.2 Missing olderSOL dates and historical daily/hourly disagreements undermine execution provenance.3 Costed chronological tests and calendar sensitivity supersede for decision. Positive fullsample grossperformance cannot authorize allocation. Source https://runway.adityaperswal.workers.dev/data/crypto-bars .

Required backtestCLI invoked1095days SMAfast1 slow50. Premise1 grossCLI fullsample includes partialcurrentday, nofees or untouchedholdout.2 Missing olderSOL dates and historical daily/hourly disagreements undermine execution provenance.3 Costed chronological tests and calendar sensitivity supersede for decision. Positive fullsample grossperformance cannot authorize allocation. Source https://runway.adityaperswal.workers.dev/data/crypto-bars .

Premises1 frozen grid windows daily20/50/100 or hourly24/72/168, baseline vs rv20 below trailing120median; close>SMA, signal lagged one bar, nextopen fill, finalopen liquidation, cashlong only. vectorbt fees.25% and slippage.05% eachside; stress .50%+.10%.2 Train first60% after168bar warmup, select highest trainreturn only. Chosenwindow168,volfilter=false; trainstart 2025-01-08 00:00:00+00:00,holdout 2026-01-04 15:00:00+00:00 to 2026-09-07 05:00:00+00:00.3 Selected results {"train":{"return_pct":-67.78765602200687,"drawdown_pct":78.85273225511789,"n":162,"win_pct":11.11111111111111,"avg_win_pct":7.640835723695691,"avg_loss_pct":1.6568132803808304,"quarter_kelly":0,"buyhold_pct":-33.404027855456505},"test":{"return_pct":-48.84196430556088,"drawdown_pct":60.67091872709006,"n":109,"win_pct":11.009174311926607,"avg_win_pct":5.470381003585011,"avg_loss_pct":1.3004736913213621,"quarter_kelly":0,"buyhold_pct":-22.08100196390855},"stress":{"return_pct":-73.39999060449284,"drawdown_pct":78.6733656699508,"n":109,"win_pct":8.256880733944955,"avg_win_pct":6.560425036754946,"avg_loss_pct":1.848642623343645,"quarter_kelly":0,"buyhold_pct":-22.08100196390855}}. All six settings recorded together in observation; no holdout selection. Conclusion reject allocation0. Validity under costed gate sound, but evidence limited by recycled market history, missinghour clock compression and inconsistent daily/ hourly opens; this is diagnostic chronological OOS, not untouched discovery OOS. Sensitivity with hourly-derived daily and calendar gaps separately recorded, not ground-truth repair. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; cost https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 frozen grid windows daily20/50/100 or hourly24/72/168, baseline vs rv20 below trailing120median; close>SMA, signal lagged one bar, nextopen fill, finalopen liquidation, cashlong only. vectorbt fees.25% and slippage.05% eachside; stress .50%+.10%.2 Train first60% after168bar warmup, select highest trainreturn only. Chosenwindow50,volfilter=false; trainstart 2025-06-18 00:00:00+00:00,holdout 2026-01-04 00:00:00+00:00 to 2026-09-06 00:00:00+00:00.3 Selected results {"train":{"return_pct":0.4416485911265408,"drawdown_pct":34.296865299079414,"n":4,"win_pct":50,"avg_win_pct":9.362793638304339,"avg_loss_pct":7.822609053705021,"quarter_kelly":0.020562567168764603,"buyhold_pct":-10.40127824079402},"test":{"return_pct":-22.644633377880748,"drawdown_pct":46.991259464847765,"n":13,"win_pct":7.6923076923076925,"avg_win_pct":33.249328225901785,"avg_loss_pct":4.399960212022376,"quarter_kelly":0,"buyhold_pct":-22.646310432569972},"stress":{"return_pct":-28.449106040344724,"drawdown_pct":50.526052380768164,"n":13,"win_pct":7.6923076923076925,"avg_win_pct":32.53164518251867,"avg_loss_pct":4.985747956516338,"quarter_kelly":0,"buyhold_pct":-22.646310432569972}}. All six settings recorded together in observation; no holdout selection. Conclusion reject allocation0. Validity under costed gate sound, but evidence limited by recycled market history, missinghour clock compression and inconsistent daily/ hourly opens; this is diagnostic chronological OOS, not untouched discovery OOS. Sensitivity with hourly-derived daily and calendar gaps separately recorded, not ground-truth repair. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; cost https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 frozen grid windows daily20/50/100 or hourly24/72/168, baseline vs rv20 below trailing120median; close>SMA, signal lagged one bar, nextopen fill, finalopen liquidation, cashlong only. vectorbt fees.25% and slippage.05% eachside; stress .50%+.10%.2 Train first60% after168bar warmup, select highest trainreturn only. Chosenwindow168,volfilter=false; trainstart 2025-01-08 00:00:00+00:00,holdout 2026-01-04 13:00:00+00:00 to 2026-09-07 05:00:00+00:00.3 Selected results {"train":{"return_pct":-73.65208057000791,"drawdown_pct":76.58208611702999,"n":176,"win_pct":7.954545454545454,"avg_win_pct":9.468108330150732,"avg_loss_pct":1.534689842594759,"quarter_kelly":0,"buyhold_pct":-7.090384444641751},"test":{"return_pct":-56.87699455710579,"drawdown_pct":65.56023105666749,"n":119,"win_pct":10.084033613445378,"avg_win_pct":4.642149865595557,"avg_loss_pct":1.261096995700117,"quarter_kelly":0,"buyhold_pct":-20.250014360388313},"stress":{"return_pct":-78.88364461310641,"drawdown_pct":82.1999033787876,"n":119,"win_pct":6.722689075630252,"avg_win_pct":6.246454457342261,"avg_loss_pct":1.8060241401026422,"quarter_kelly":0,"buyhold_pct":-20.250014360388313}}. All six settings recorded together in observation; no holdout selection. Conclusion reject allocation0. Validity under costed gate sound, but evidence limited by recycled market history, missinghour clock compression and inconsistent daily/ hourly opens; this is diagnostic chronological OOS, not untouched discovery OOS. Sensitivity with hourly-derived daily and calendar gaps separately recorded, not ground-truth repair. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; cost https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 frozen grid windows daily20/50/100 or hourly24/72/168, baseline vs rv20 below trailing120median; close>SMA, signal lagged one bar, nextopen fill, finalopen liquidation, cashlong only. vectorbt fees.25% and slippage.05% eachside; stress .50%+.10%.2 Train first60% after168bar warmup, select highest trainreturn only. Chosenwindow100,volfilter=true; trainstart 2025-06-18 00:00:00+00:00,holdout 2026-01-04 00:00:00+00:00 to 2026-09-06 00:00:00+00:00.3 Selected results {"train":{"return_pct":64.25649926752598,"drawdown_pct":17.803262934477182,"n":2,"win_pct":50,"avg_win_pct":68.95526654444701,"avg_loss_pct":2.6887291611748028,"quarter_kelly":0.1201259539410204,"buyhold_pct":24.437868997997512},"test":{"return_pct":-4.6988534663745485,"drawdown_pct":20.478806917906812,"n":7,"win_pct":14.285714285714285,"avg_win_pct":19.225909971471157,"avg_loss_pct":3.6345473884289317,"quarter_kelly":0,"buyhold_pct":-20.628244741013535},"stress":{"return_pct":-8.618657551484029,"drawdown_pct":23.519727009589165,"n":7,"win_pct":14.285714285714285,"avg_win_pct":18.557354627942587,"avg_loss_pct":4.223016574122196,"quarter_kelly":0,"buyhold_pct":-20.628244741013535}}. All six settings recorded together in observation; no holdout selection. Conclusion reject allocation0. Validity under costed gate sound, but evidence limited by recycled market history, missinghour clock compression and inconsistent daily/ hourly opens; this is diagnostic chronological OOS, not untouched discovery OOS. Sensitivity with hourly-derived daily and calendar gaps separately recorded, not ground-truth repair. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; cost https://docs.alpaca.markets/us/docs/crypto-fees .

Premises1 frozen grid windows daily20/50/100 or hourly24/72/168, baseline vs rv20 below trailing120median; close>SMA, signal lagged one bar, nextopen fill, finalopen liquidation, cashlong only. vectorbt fees.25% and slippage.05% eachside; stress .50%+.10%.2 Train first60% after168bar warmup, select highest trainreturn only. Chosenwindow168,volfilter=true; trainstart 2025-01-08 00:00:00+00:00,holdout 2026-01-04 13:00:00+00:00 to 2026-09-07 05:00:00+00:00.3 Selected results {"train":{"return_pct":-67.22851233836953,"drawdown_pct":70.12404416263539,"n":198,"win_pct":24.242424242424242,"avg_win_pct":1.111314576386961,"avg_loss_pct":1.089329679796609,"quarter_kelly":0,"buyhold_pct":-5.796148198613182},"test":{"return_pct":-53.27668394673195,"drawdown_pct":53.32436421303937,"n":123,"win_pct":26.01626016260163,"avg_win_pct":1.0630043406605738,"avg_loss_pct":1.1972114295204213,"quarter_kelly":0,"buyhold_pct":-12.502556532743581},"stress":{"return_pct":-77.66322836856008,"drawdown_pct":77.66322836856014,"n":123,"win_pct":17.073170731707318,"avg_win_pct":0.8827516746820154,"avg_loss_pct":1.638757128624584,"quarter_kelly":0,"buyhold_pct":-12.502556532743581}}. All six settings recorded together in observation; no holdout selection. Conclusion reject allocation0. Validity under costed gate sound, but evidence limited by recycled market history, missinghour clock compression and inconsistent daily/ hourly opens; this is diagnostic chronological OOS, not untouched discovery OOS. Sensitivity with hourly-derived daily and calendar gaps separately recorded, not ground-truth repair. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; cost https://docs.alpaca.markets/us/docs/crypto-fees .