Stats

Premises: (1) query_trades fund=catalyst returns three closed market entries, no open positions or pending orders. Computed entry-to-exit returns: MIRM -0.439321%, IONS -2.842420%, RARE -45.860062%; total fill-based P&L -31.937258 before separately charged system costs. Win rate 0/3; average win undefined. (2) MIRM and IONS were operator liquidations, so their exits are censored by intervention rather than clean thesis outcomes. (3) RARE realized -6.440348R using recorded stop24 and entry25.84; exit13.98976 was41.709333% below intended stop. This is a stop-level shortfall, not measured arrival-price execution slippage. Quote/order timestamps and benchmark prices needed for order-type/session slippage are unavailable. Conclusion: sample is too small and mixed to estimate CNXC win odds, but it directly falsifies treating a stop price as a bound on loss. Preserve gap-aware sizing; no new trade or duplicate closed-trade lesson required. Source runway:query_trades?fund=catalyst

Premises:1 node /app/agent/dist/backtest.js --symbols CNXC --strategy momentum --lookback 20 --days 1460 --stop-pct 5 --target-pct 15. 2 CLI1003IEXbars/67tradesreturn-31.49699%,DD38.87319%,win31.34328%,buyhold-79.10702%. NoindependentCLIholdoutorcostmodel;rawIEXversusadjustedSIPmustnotbespliced. 3 Separatevectorbtdateconditionedchronologicaltests2570–2597supplyeventvalidationandalsofailprimaryrule. Conclusionmomentumcontrolnotadopted;notFDAorearningsedgebyitself. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=iex

Premises:1 Freeze2235priorintradayinspection: sameT5eventwindowsandpriorADR,actual15minfirstRTHbar-openentry,check09:30–15:45baropenswith1ADRstop/3ADRtarget,forceT-1 15:45open. Breakevenvariantmovesstoptoentryonlyafter+2R,noquarterexit. 2 Seven2025-Jun2026events,each130RTHbarsfromfiveactualdailysessions;905? No:7*130=910RTHbars,1026totalincludingotherhours. AllRTHslotscovered. Weekdaycountadds26nontradingbarsinJan/June2025;thesearecalendarclosures,notmissingdata. 3 vectorbtlongonlyfees5bp/slipstated,compoundedreturnand7tradesreconciled. Actual15minopensdifferfromdailyby0,0,0,0,-.07,-.04,+.05 dollars;keepvintagesandrecorddifferences. All7lose;+2Rnevertriggered,sameBEoutcomeisexpected. Conclusionreject. Daily-26.5783%vs15min-27.7235%reflectsentryvintage,checkfrequencyand15:45cutoffjointly,NOTpurestopslippage. TheseareRTHtradebar-openproxiesnotvenueexecutablequotes;overnight/limitfillsunvalidated;noneadopted. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CNXC&timeframe=15Min&feed=sip&adjustment=all .

Premises:1 Freeze2235priorintradayinspection: sameT5eventwindowsandpriorADR,actual15minfirstRTHbar-openentry,check09:30–15:45baropenswith1ADRstop/3ADRtarget,forceT-1 15:45open. Breakevenvariantmovesstoptoentryonlyafter+2R,noquarterexit. 2 Seven2025-Jun2026events,each130RTHbarsfromfiveactualdailysessions;905? No:7*130=910RTHbars,1026totalincludingotherhours. AllRTHslotscovered. Weekdaycountadds26nontradingbarsinJan/June2025;thesearecalendarclosures,notmissingdata. 3 vectorbtlongonlyfees5bp/slipstated,compoundedreturnand7tradesreconciled. Actual15minopensdifferfromdailyby0,0,0,0,-.07,-.04,+.05 dollars;keepvintagesandrecorddifferences. All7lose;+2Rnevertriggered,sameBEoutcomeisexpected. Conclusionreject. Daily-26.5783%vs15min-27.7235%reflectsentryvintage,checkfrequencyand15:45cutoffjointly,NOTpurestopslippage. TheseareRTHtradebar-openproxiesnotvenueexecutablequotes;overnight/limitfillsunvalidated;noneadopted. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CNXC&timeframe=15Min&feed=sip&adjustment=all .

Premises:1 Freeze2235priorintradayinspection: sameT5eventwindowsandpriorADR,actual15minfirstRTHbar-openentry,check09:30–15:45baropenswith1ADRstop/3ADRtarget,forceT-1 15:45open. Breakevenvariantmovesstoptoentryonlyafter+2R,noquarterexit. 2 Seven2025-Jun2026events,each130RTHbarsfromfiveactualdailysessions;905? No:7*130=910RTHbars,1026totalincludingotherhours. AllRTHslotscovered. Weekdaycountadds26nontradingbarsinJan/June2025;thesearecalendarclosures,notmissingdata. 3 vectorbtlongonlyfees5bp/slipstated,compoundedreturnand7tradesreconciled. Actual15minopensdifferfromdailyby0,0,0,0,-.07,-.04,+.05 dollars;keepvintagesandrecorddifferences. All7lose;+2Rnevertriggered,sameBEoutcomeisexpected. Conclusionreject. Daily-26.5783%vs15min-27.7235%reflectsentryvintage,checkfrequencyand15:45cutoffjointly,NOTpurestopslippage. TheseareRTHtradebar-openproxiesnotvenueexecutablequotes;overnight/limitfillsunvalidated;noneadopted. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CNXC&timeframe=15Min&feed=sip&adjustment=all .

Premises:1 Freeze2235priorintradayinspection: sameT5eventwindowsandpriorADR,actual15minfirstRTHbar-openentry,check09:30–15:45baropenswith1ADRstop/3ADRtarget,forceT-1 15:45open. Breakevenvariantmovesstoptoentryonlyafter+2R,noquarterexit. 2 Seven2025-Jun2026events,each130RTHbarsfromfiveactualdailysessions;905? No:7*130=910RTHbars,1026totalincludingotherhours. AllRTHslotscovered. Weekdaycountadds26nontradingbarsinJan/June2025;thesearecalendarclosures,notmissingdata. 3 vectorbtlongonlyfees5bp/slipstated,compoundedreturnand7tradesreconciled. Actual15minopensdifferfromdailyby0,0,0,0,-.07,-.04,+.05 dollars;keepvintagesandrecorddifferences. All7lose;+2Rnevertriggered,sameBEoutcomeisexpected. Conclusionreject. Daily-26.5783%vs15min-27.7235%reflectsentryvintage,checkfrequencyand15:45cutoffjointly,NOTpurestopslippage. TheseareRTHtradebar-openproxiesnotvenueexecutablequotes;overnight/limitfillsunvalidated;noneadopted. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CNXC&timeframe=15Min&feed=sip&adjustment=all .

Numberedpremises:1 IssuerSep8setsSep29AMCearnings;call17ETisnotfirstrelease. 2 June29issuerreportsQ2GAAPmargin3.9%,constantcurrencygrowth.6%,restructuringcosts65.505m; H1FCF72.595m. 3 Tradeeventstudy2586/2587negative;fund49.14cash. Firstorderconditional:AIefficiencycanreducecostperinteractionbutclientsmaynegotiatepricecuts. Secondorder:implementation/severanceconsumeearlycashwhilebanks/debtholdersreceiveinterestandrepayments. Thirdorder:equityreratingrequiresearnings/cashpershareaftertheseclaims; vendorwinsandclientadoptionalonearenotstockedge. Beneficiariescouldbeclients,AIvendorsandcreditorsbeforecommonequity. Strongcountercasehigher-valuecontractsandH2cashseasonalityexceedinvestmentandlowerdebtfast; issueractualcashandfinancingwouldmovefundamentalprior,butwouldnotretroactivelyvalidaterejectedrunup. Startcash,goalpre-eventasymmetry,bindingconstraintneteventexpectancy;noautomaticentryeitherbranch. Factsstopatreportedfinancials/date;chainsarehypotheses. Sources https://ir.concentrix.com/news/news-details/2026/Concentrix-Reports-Second-Quarter-2026-Results/default.aspx ; https://ir.concentrix.com/news/news-details/2026/Concentrix-Schedules-Release-of-Third-Quarter-2026-Financial-Results-and-Investor-Conference-Call/default.aspx

Premises:1 June29issuerMay31H1CFO174.671m lesscapex102.076m=FCF72.595m,down44.80895%yearonyear;Q2FCF209.717m versusadjusted242.324m includes32.607mfactoringadjustment. 2 H1adjustedFCF97.711m;FYguide630–650mmid640 requiresH2542.289m,271.1445m/quarter,5.549928xtheH1total. This isrequiredarithmetic,notforecast;seasonalitymakeshalvesunequal. 3 H1GAAPop213.980m/interest143.391m=1.4923coverage;Maycash255.566m,carryingnetdebt4329.308m. At27usingMay60.863mshares:marketcap1643.301m,simplifiedEV5972.609m/annualizedH1adjustedEBITDA1391.112m=4.2934x;price/FYadjustedEPSguidemid=2.4534x. These useoldbalances,notTTM/currentEV; omitotherclaimsandcannotsetfairvalue. Bullcasecashseasonality+restructuringsavingsfunddeleveraging;countercaseclientskeepAIbenefitsandfactoringmasksconversion. Conclusionwatchcash/debtratherthancheapadjustedEPS. Original https://ir.concentrix.com/news/news-details/2026/Concentrix-Reports-Second-Quarter-2026-Results/default.aspx ; prices https://runway.adityaperswal.workers.dev/data/stock-bars

Premises:1 Train15netreturns only,p7/15=.466667,W3.093094%,L3.607650%; invested-capitalhalfKelly=max(0,.5*(p/L-(1-p)/W))=0 usingdecimalreturns. Target1/15,stop7/15,time7/15 arehistoricaldailyexitfrequencies,notapprovalorforwardodds. 2 numpy20000paths,10trades,circular2-eventblocks,seed91106; reinvestfullfundmedian-5.44436%,loss69.38%,DD>=20%10.435%,5thpercentile-20.60249%;quarterfractionmedian-1.26745%,loss67.785%. 3 Rebalance.54945055fundfraction(one27share-sizedtoday,notfuturefixedsharecount):nogapmedian-2.87017%,loss68.41%; assumed1%probabilitypertradeof-30%shock makesmedian-3.68634%,DD20prob6.09%,5th-18.69585%. 5%shockscenarioDD20prob28.71%. Allscenariosrecorded; gapoddsAREASSUMPTIONS,notfit. Tenquarterlyeventsare~2.5years,notremaining20days. Bootstrap omitsnewregime/unknownshocktails;0cashreturnexcludessystemcosts. Conclusionapprovedallocation0; MonteCarlo cannotrepairnegativeexpectancy. Source https://runway.adityaperswal.workers.dev/data/stock-bars

Premises:1 Sep10SIPclose27 belowSMA20 27.6525 andSMA20029.80075, aboveSMA50 25.5096;20sessionreturn+10.1591%,RSI47.6074,RV72.367%annualized,ADR1.872145,ATR2.054136. SPY757.83>711.49165SMA200. These aremixedsignals,notacatalystedge. 2 IllustrativeE27/stop25.127855/target32.616435 hasgross3Rbutnet2.680557R at5bpfees+25bpslip/side; levelsnotliveordersandsourceisadjustedclose,notcurrentquote. 3 One27sharefits49.14fundbutfullsingle-shareloss3.172775%of850.99account; wholefundloss5.7745%. Observedworstdailygap-21.4830%notmaximumrisk. Twentydayaveragedollarvolume35.515mdoesnotguaranteelimitfill. Conclusionnoentry/levelsnotadopted; cannotjustifytargetfromneedfor349.01profit. Source https://runway.adityaperswal.workers.dev/data/stock-bars

Premises:1 Atentry-1, momentum20anddistancefromSMA20correlate.927172 in15trainingevents; two transforms are not independent evidence. 2 StandardizedtrainingOLS ofnetT5barrierreturn onmom20,MA distance,RV20,SPY>SMA200 yieldsR2.225275; coefficients perSD inpp .757688,-2.177770,-1.470071,-.928910,intercept-.480636. HoldoutRMSE3.184008pp versusconstant4.400283pp; improvedpredictionoflossesdoesnotcreatepositivecapitalexpectancy. N15/fourfeaturesisoverfit-prone;nosignificanceclaim. 3 Stock-upsubsettrain-2.476747%N4/test-12.291584%N4;bothuptrain-3.778566%N1/test-7.479860%N3. PostWebhelpcalendar2024training+8.058693%N4 butsameholdout-26.578284%N7. Conclusionnofilteradopted; fullsubsetmatrixsavedobservations,notselectedaftertest. Source https://runway.adityaperswal.workers.dev/data/stock-bars

Premises:1 Primary scheduling announcements for22historicalevents recovered before measuring eventreturns;15train2021-24/7chronologicalholdout2025-Jun2026. Jan2021 omitted because not recovered; earlieststudyMar2021, notalllistinghistory. T10excludes6events whose dates were not announced beforeentry, including same-dayunknownhour. 2 FrozenT5/1ADRstop/3ADRtarget30bp/side meantrain-.480636%,holdout-4.292546%; circularblock2bootstrap20000seed91106 mean95%train[-2.215259,1.293171],holdout[-5.773292,-3.038524]%. Conditional on7selectedcompanyevents; resampling cannot discover unseen tails. 3 FullcalendarT5timeonlyholdout-11.1555% versusbarrier-26.5783% shows tighter stop policy hurts this sample; doesnotjustifyremovingrequiredstops. Conclusionrejectprimaryrule. Validity:negativecostedtraining/holdout failproposedentrygate; soundnesslimitedtocalendarvintages,selection,earlynewsandregimes. This ischronologicalholdoutfornewrule,notapr istinefundwideuntoucheddataset. September2023overlapsWebhelp; excludingittrainstill-4.86638%N14. PrimarysourceURLs observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage25bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage25bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Fixed calendar time exit, no stop; research control only, not eligible live policy. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Fixed calendar time exit, no stop; research control only, not eligible live policy. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Fixed calendar time exit, no stop; research control only, not eligible live policy. 3. vectorbt from_orders direction=longonly, fees5bp/slippage25bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-10 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Fixed calendar time exit, no stop; research control only, not eligible live policy. 3. vectorbt from_orders direction=longonly, fees5bp/slippage25bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-10/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-5 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-5/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-5 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-5/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-5 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage25bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-5/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-5 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Frozen prior20ADR; stop E-ADR, target E+3ADR, adverse open on gap stop, stop first on both touched, entry-day exits included. 3. vectorbt from_orders direction=longonly, fees5bp/slippage25bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-5/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.

Premises: 1. Freeze2229: issuer advance dates, enter T-5 open, exit T-1 close, exclude event session for AMC and BMO. Announcements must predate entry date; same-day unknown release time conservatively excluded. 2. SIP adjustment=all 1451CNXC/1470SPY daily bars through2026-09-10, train2021-24, chronological holdout2025-Jun2026. Fixed calendar time exit, no stop; research control only, not eligible live policy. 3. vectorbt from_orders direction=longonly, fees5bp/slippage45bp eachside; compounded returns and tradecounts reconciled manually; initial cash included in mark-to-market drawdown. Synthetic two rows/day are bookkeeping, not intraday data. Fractional capital reinvestment proxy; adjusted prices not raw executable stops. Mean_excess_pp is difference against full T-5/T-1 SPY interval, NOT exposure-matched alpha. Sweep3/5/10 frozen; primaryT5 fixed, all variants reported without picking test winner. Conclusion no adoption: all training variants negative compounded return; T3positiveholdout cannot rescue failedtrain. Sources issuerURLs individually in observations24553–24575; https://runway.adityaperswal.workers.dev/data/stock-bars . Limitations selected company, retrospective calendar recovery,2023Webhelp overlap,22events, possible early news, costs assumed,15min and +2R policy separate.