Stats

Premises: 1. Signals must precede fills. 2. Costs and out-of-sample windows must be explicit. 3. Stock profits do not establish option profits. Rule: adjusted close SMA10>SMA50, shift state one bar, buy next close on entry and sell next close on exit. Reset $1000 cash at Jan2026, warm indicators with prior data, close at final bar. Fees and slippage each5bp per side; stress each15bp. Fixed fast5/10/20 and slow50 sweep recorded separately, no holdout selection. Training2022-2025: {"symbol":"IWM","fast":10,"slow":50,"cost":0.001,"window":"train","return_pct":7.763242958011107,"drawdown_pct":17.333153286915394,"n":13,"win_pct":38.46153846153847,"avgwin_pct":6.60573409618851,"avgloss_pct":2.9414965176713834}. Test2026 throughSep4: {"symbol":"IWM","fast":10,"slow":50,"cost":0.001,"window":"test","return_pct":11.401279414640214,"drawdown_pct":7.001675041876066,"n":2,"win_pct":100,"avgwin_pct":5.624661773141211,"avgloss_pct":0}. All1173 daily bars per symbol from https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=IWM&timeframe=1Day&start=2022-01-01&end=2026-09-05&adjustment=all&feed=iex . Implemented vectorbt. CLI also run --symbols XLF,XLE,GLD,EEM,IWM,TLT --strategy sma-cross --fast10 --slow50 --days1709 (flags had spaces); inspected source /app/agent/src/backtest/bars.ts: CLI omits adjustment, so its aggregate cannot validate this adjusted chronological holdout. Small test count, selection of ETFs, IEX-only coverage and changing regimes limit inference. No observed loss in a small sample does not imply zero loss probability. No option return sample or executable quotes, so no option allocation.

Premises: 1. Signals must precede fills. 2. Costs and out-of-sample windows must be explicit. 3. Stock profits do not establish option profits. Rule: adjusted close SMA10>SMA50, shift state one bar, buy next close on entry and sell next close on exit. Reset $1000 cash at Jan2026, warm indicators with prior data, close at final bar. Fees and slippage each5bp per side; stress each15bp. Fixed fast5/10/20 and slow50 sweep recorded separately, no holdout selection. Training2022-2025: {"symbol":"GLD","fast":10,"slow":50,"cost":0.001,"window":"train","return_pct":77.71279328532353,"drawdown_pct":12.12802180040049,"n":12,"win_pct":58.333333333333336,"avgwin_pct":10.580961872550269,"avgloss_pct":2.022247524487106}. Test2026 throughSep4: {"symbol":"GLD","fast":10,"slow":50,"cost":0.001,"window":"test","return_pct":1.5192051340824604,"drawdown_pct":20.314652789434117,"n":2,"win_pct":100,"avgwin_pct":0.7582119864399353,"avgloss_pct":0}. All1173 daily bars per symbol from https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=GLD&timeframe=1Day&start=2022-01-01&end=2026-09-05&adjustment=all&feed=iex . Implemented vectorbt. CLI also run --symbols XLF,XLE,GLD,EEM,IWM,TLT --strategy sma-cross --fast10 --slow50 --days1709 (flags had spaces); inspected source /app/agent/src/backtest/bars.ts: CLI omits adjustment, so its aggregate cannot validate this adjusted chronological holdout. Small test count, selection of ETFs, IEX-only coverage and changing regimes limit inference. No observed loss in a small sample does not imply zero loss probability. No option return sample or executable quotes, so no option allocation.

Required CLI executed Sep6 06UTC: node /app/agent/dist/backtest.js --symbols BTC/USD,ETH/USD,SOL/USD --strategy mean-reversion --lookback 20 --entry 2 --exit 0 --days 979. Whole-period gross returns BTC7.473%,ETH27.784%,SOL11.298%; N28/26/19; rawdaily980/980/742bars include partialSep6. Source code strategies.js defines rollingclose z<=-2 state, exit when false; --exit is not used by per-symbol mean-reversion. No explicit costs/holdout and known rawdaily defects; different from preregistered3daytrendpullback/ADR exits. Thus control is exploratory only and cannot support allocation or be substituted for costed daily/hourly holdout. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest/strategies.js ; /app/agent/dist/backtest/bars.js .

RequiredCLIrun SMA20/50days979 returnedBTC+52.1948%,DD38.3724%,win33.333%,N12; ETH+75.4890%,DD46.9620%,win50%,N8; SOL-37.4426%,DD67.0018%,win37.5%,N8. BTC/ETH980bars,SOL742 includepartialSep6. Localbars.js hasonepagerequest/noendfilter; CLIwholeperiodnoOOS/explicitcostmodel. Notnumericallyequivalentvectorbtholdout306-308. Conclusion exploratorycontrolonly; do notselectgrosspositiveBTC/ETHasvalidatedstrategy. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest/bars.js .

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee0bp plus slippage2bp each way. Gap stops at open; ambiguous daily bars stop first. 19 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.009944%, average win0.814326%, average loss0.609413%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage45bp each way. Gap stops at open; ambiguous daily bars stop first. 19 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.965262%, average win0.115454%, average loss1.092405%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 19 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.269581%, average win0.552549%, average loss0.867493%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max40 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 23 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean0.114560%, average win1.332183%, average loss0.822073%; quarter Kelly0.021499. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 23 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean0.060697%, average win1.208300%, average loss0.822073%; quarter Kelly0.012558. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max10 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 23 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.031775%, average win0.995613%, average loss0.822073%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee0bp plus slippage2bp each way. Gap stops at open; ambiguous daily bars stop first. 19 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.068713%, average win0.400042%, average loss0.490593%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage45bp each way. Gap stops at open; ambiguous daily bars stop first. 19 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-1.023470%, average win0.000000%, average loss1.023470%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 19 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.328197%, average win0.162090%, average loss0.684770%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max40 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 23 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.249348%, average win0.703346%, average loss0.513986%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max20 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 23 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.249348%, average win0.703346%, average loss0.513986%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Frozen protocol note294. Actual training data Jan2016–Dec2021 (requested2015 unavailable); holdout Jan2022–Aug2026. Quarterly first-session open after prior-quarter close; prior ATR20 sets stop1ATR and target1.5ATR, max10 sessions then next open; no reentry that quarter. Long-only, cash-only vectorbt order ledger and independent compounded trade returns agree within1e-8. Fee5bp plus slippage10bp each way. Gap stops at open; ambiguous daily bars stop first. 23 completed cohorts, not independent merger deals; no unfinished endpoints counted. Training10/40 variants are sensitivity checks; frozen20 used later. Net mean-0.235721%, average win0.683850%, average loss0.491158%; quarter Kelly0.000000. Drawdown includes initial cash and session-ending marks. Daily adjusted bars approximate distributions and fills; neither sparse trading nor 15-minute stop/partial-profit execution validated. This is an exit adaptation, not exact original-paper trading. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; local study.py, trades.json, metrics.json.

Preregistered note292 measurement experiment, nottrade-ready exitrule. 1 Vectorbt SMA20>50 on priorcompletedUTCdailycloses; enterfollowing00UTCopen, exitfollowingopenwhenfalse. 25bpfee+10bpslippage eachside, doublecoststress. TrainJan2024-Dec2025 (SOLstartAug2024); frozenholdoutJan1-Sep5 2026; flatatstart, forcedliquidationlastday00UTC. No rangeexits/stops/targets, so no tradeproposal evenifpositive. 2 Fourvariants rawdaily,rebuiltdaily, hourlyusingrebuiltsignals, hourlyusingrawsignals. Missinghourday makesreconstructeddailyNaN andsuppressesrolling50dayfilter; nofill. Identicalsignals+executiontimesyieldidenticaldaily/hourreturns; onlymarkingdrawdown differs. Rawdailyholdout -42.07581462700807%,N5; rebuilt/hour -34.622129406621696%,N4; doublecost -36.427338917791616%. Hourlytrain 6.677033703761973%,N4. 3 Training-only parameter sensitivity10/40 and30/60 recorded, noholdoutselection. Every60row fullcashledger matchescompoundednetcapitalreturns<=5.7e-14pp; fee-inclusivecapitalreturn replacesvectorbtentryvalue-onlyreturndenominator. Terminaldaily/hourtimealignedbeforefinalresults. 4 SmallN,forcedterminalwinner,priorinspected2026data,imputedopenfills andunvalidated15minrangeexitmean noindependentedge. BTCtinypositiveplug-inKellyisnotadopted:negativecompoundedreturnanduncertainty. Conclusion rejectnewallocation; datarepaircannotrevivepreviousETHratio/LTC/fundingrejections. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; directvalidation https://data.alpaca.markets/v1beta3/crypto/us/bars .

Preregistered note292 measurement experiment, nottrade-ready exitrule. 1 Vectorbt SMA20>50 on priorcompletedUTCdailycloses; enterfollowing00UTCopen, exitfollowingopenwhenfalse. 25bpfee+10bpslippage eachside, doublecoststress. TrainJan2024-Dec2025 (SOLstartAug2024); frozenholdoutJan1-Sep5 2026; flatatstart, forcedliquidationlastday00UTC. No rangeexits/stops/targets, so no tradeproposal evenifpositive. 2 Fourvariants rawdaily,rebuiltdaily, hourlyusingrebuiltsignals, hourlyusingrawsignals. Missinghourday makesreconstructeddailyNaN andsuppressesrolling50dayfilter; nofill. Identicalsignals+executiontimesyieldidenticaldaily/hourreturns; onlymarkingdrawdown differs. Rawdailyholdout -7.101324055413278%,N3; rebuilt/hour -13.138652032340486%,N3; doublecost -14.94374057915982%. Hourlytrain 120.2419847750352%,N5. 3 Training-only parameter sensitivity10/40 and30/60 recorded, noholdoutselection. Every60row fullcashledger matchescompoundednetcapitalreturns<=5.7e-14pp; fee-inclusivecapitalreturn replacesvectorbtentryvalue-onlyreturndenominator. Terminaldaily/hourtimealignedbeforefinalresults. 4 SmallN,forcedterminalwinner,priorinspected2026data,imputedopenfills andunvalidated15minrangeexitmean noindependentedge. BTCtinypositiveplug-inKellyisnotadopted:negativecompoundedreturnanduncertainty. Conclusion rejectnewallocation; datarepaircannotrevivepreviousETHratio/LTC/fundingrejections. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; directvalidation https://data.alpaca.markets/v1beta3/crypto/us/bars .

Preregistered note292 measurement experiment, nottrade-ready exitrule. 1 Vectorbt SMA20>50 on priorcompletedUTCdailycloses; enterfollowing00UTCopen, exitfollowingopenwhenfalse. 25bpfee+10bpslippage eachside, doublecoststress. TrainJan2024-Dec2025 (SOLstartAug2024); frozenholdoutJan1-Sep5 2026; flatatstart, forcedliquidationlastday00UTC. No rangeexits/stops/targets, so no tradeproposal evenifpositive. 2 Fourvariants rawdaily,rebuiltdaily, hourlyusingrebuiltsignals, hourlyusingrawsignals. Missinghourday makesreconstructeddailyNaN andsuppressesrolling50dayfilter; nofill. Identicalsignals+executiontimesyieldidenticaldaily/hourreturns; onlymarkingdrawdown differs. Rawdailyholdout -4.423418958113871%,N5; rebuilt/hour -2.419298196890395%,N3; doublecost -4.447148458677293%. Hourlytrain 57.32145831754195%,N8. 3 Training-only parameter sensitivity10/40 and30/60 recorded, noholdoutselection. Every60row fullcashledger matchescompoundednetcapitalreturns<=5.7e-14pp; fee-inclusivecapitalreturn replacesvectorbtentryvalue-onlyreturndenominator. Terminaldaily/hourtimealignedbeforefinalresults. 4 SmallN,forcedterminalwinner,priorinspected2026data,imputedopenfills andunvalidated15minrangeexitmean noindependentedge. BTCtinypositiveplug-inKellyisnotadopted:negativecompoundedreturnanduncertainty. Conclusion rejectnewallocation; datarepaircannotrevivepreviousETHratio/LTC/fundingrejections. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; directvalidation https://data.alpaca.markets/v1beta3/crypto/us/bars .

Ran actual CLI node /app/agent/dist/backtest.js --symbols SRRK --strategy momentum --lookback20 --days2000, live venuefetch succeeded. CLI1376bars,72trades,return-50.2208%,DD71.2093%,win26.3889%; no costed independent holdout inCLI itself. Separate vectorbt test: daily prior-close sign of5/10/20session return executed nextIEXopen; fixed through2024 training,2025-Jan–2026-Sep4holdout; lastopen liquidation,no optimizedstops. Baseline20 training-67.6483%,N52; holdout-36.3390%,DD42.5445%,N25,win36%,avgW5.3818%,avgL5.4618%. Fee5bp+slip10bp per side. Stress5+25bp:holdout-40.9389%. Heldout5day-45.3578%,10day-9.47835%; allcostedreturnsnegative. Sweep is robustness diagnostic, notselection. Entry/exitpricesIEXpartialvenue;dailyopenfillsnotactualquotes,no15minutevenueexecutionreplication. Positive arithmeticmean for10day atbasecostdoesnotrepairnegativecompoundedreturnortrainingfailure. BaselinequarterKelly=max(0,(.36-.64/(5.3818/5.4618))/4)=0. This rule is not a PDUFA run-up and cannot validate one. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=SRRK&timeframe=1Day&feed=iex&adjustment=all ; exactCLIjson,12sweeptradeCSVs andscriptretained.

Preregistered note293 fixed T-20 close to T-5 close before issuer PDUFA. Aug6 2025 issuer notice establishesSep22 deadline before entry. IEX all-adjusted bars2021-Jan4–2026-Sep4,1425sessions; source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=SRRK&timeframe=1Day&feed=iex&adjustment=all . Vectorbt fee5bp/slippage10bp each side, cash100: Aug22close35.495 ->Sep15close33.66,return-5.453807%, marked-to-closeDD12.119469%,N1,win0%; sole loss5.453807% of invested cash, average win undefined. Trade-record denominator differs slightly:-5.456534%. XBI same-window gross+2.706883%. Stress30bp total each side approximately-5.73702%. OriginalCRL signedSep22,publicSep23; this window actually exited beforedecision, yetlost. One completed original submission is not a matched resubmission corpus; chronological event holdoutN0 because2026pending. Current rule would enterSep1 andexitSep23 (weekdaycalendarexcludingLaborDay); entryalreadygone, no catch-up variation authorized. Validity: reject deployment because evidence lacks independent positive expectancy, not becauseN1provesallPDUFArunupsfail. No event early-hazard frequency inferred. Sources https://investors.scholarrock.com/news-releases/news-release-details/scholar-rock-reports-second-quarter-2025-financial-results-and ; https://download.open.fda.gov/crl/CRL_BLA761463_20250922.pdf . Price/source rows and scripts archived in this run.

Premises: exactrulemustmatchimplementation; useadjustedprices,nextcloseexecution,costs,andheldoutwindow; stocksuccesscannotvalidateoptionreturns. RequiredCLIcommand node /app/agent/dist/backtest.js --symbols NFLX --strategy mean-reversion --lookback20 --entry2 --exit0 --days1344 (actualshellusedspacesbetweenflagandvalue). CLIreturns+25.683%,DD11.605%,win61.538%,N26,rawbuyhold-73.465%. Sourceaudit /app/agent/src/backtest/bars.ts omitsadjustment; strategies.ts meanReversion line130–137 longsiffz<=-entry andotherwiseexits, ignoringexitflag. Thus notregisteredholduntilz>=0. CLIresultinvalidasvalidation, notfixedbysimplycomparinggrossnumbers. Independentvectorbt on922adjustedIEXbars: z20=(close-rollingmean)/samplestd; entryz<-2,exitz>=0,signalslagonebar,nextclosefills; fee/slippageeach5bp(base10bponeway),15bpeach(stress30bp); resetcash1000 atJan2026,indicatorwarmupfrompriorhistory,noinheritedholdings; forcedwindowendclose. Train752bars2023–2025 +50.793%,DD15.237%,N13,win69.231%; test170bars2026throughSep4 -6.395%,DD19.124%,N3,win33.333%,avgwin8.045%,avgloss6.923%; stress-7.511%. Preregisteredsweepthreshold1.5/2/2.5 givesall6heldoutcost/thresholdvariantsnegative (-6.395 to-7.511%), only1–3trades. Nochoosingbestholdout. Underlyinglongmeanreversionisadiagnosticnotputstrategy, zeroinferenceforoptionpwin; noactualoptionbidaskhistorygivenpriorOPRA403. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=NFLX&timeframe=1Day&start=2023-01-01&end=2026-09-05&adjustment=all&feed=iex . All12train/testparameterresultssavedobservations.

Mandated CLI executed SMA10/50 --days1344 on sixUS names before anyproposal;922bars each. Full-period native outputs: [{"symbol":"AEHR","returnPct":40.0021602846025,"drawdownPct":74.36484400892152,"winRate":25,"sample":12,"buyHold":368.3410263372251},{"symbol":"COHU","returnPct":44.44058175393799,"drawdownPct":40.8719973086827,"winRate":35.714285714285715,"sample":14,"buyHold":59.754369390647135},{"symbol":"AXTI","returnPct":992.4277589978756,"drawdownPct":74.57548265618689,"winRate":18.181818181818183,"sample":11,"buyHold":1308.4571428571428},{"symbol":"MP","returnPct":58.778171621743255,"drawdownPct":52.36997291189489,"winRate":40,"sample":10,"buyHold":125.53807947019865},{"symbol":"MU","returnPct":334.5504196438671,"drawdownPct":49.117752620514274,"winRate":42.857142857142854,"sample":14,"buyHold":1915.0883462378401},{"symbol":"NX","returnPct":-5.087252629102723,"drawdownPct":52.15464592547335,"winRate":50,"sample":10,"buyHold":-4.556856187290981}]. These are control runs, not holdouts and not the same model as preregistered note246 next-open costed vectorbt windows. Native control's signal timing/corporate-action handling differ; do not blend native wins with holdout returns to size. Governingtests use2023–24train/2025–Sep4holdout,5bpfee+10bpslip eachside,10/20stress,training-only3x3sweep; no parameter chosen from holdout. Canonical CLI JSON and independent audit in227partNXarchive,decodedSHAfb6f0085bae7adadf109f9f22ae7aae7da5c635b87657ecc0aa3b743304ea3bd. Zero authorization from full-period performance. Source https://runway.adityaperswal.workers.dev/data/stock-bars .

Ran actual /app/agent/dist/backtest.js --symbols HITI,HAIN --strategy momentum --lookback20 --days1000. DirectNodefetchhung; boundedfallbackPythonHTTPservedpreviouslyretrievedIEXadjustment=allbarsusingrequeststart/symbolfilters,childCLI SITE_URLpointedonlytolocalreplay. Exactinput/sourceandadapter archived. HITI686bars,N27return33.469686%,DD44.737554%,win33.333333%;HAINN30return-76.068273%,DD80.115761%,win20%. CLI rule is genericmomentum,notearningscalendar,notindependentholdoutandnotcostvalidatedhere;noeligibilityfrompositiveHITIheadline. Requiredcostedchronologicalholdoutisin244 (vectorbtT3N11). Premises:screeningmomentumcanlookpositive;drawdownlarge;doesnotidentifydatedeventeffect. ConclusionrejectCLIastradingevidenceandretain0size. Source https://runway.adityaperswal.workers.dev/data/stock-bars . Noreuseoflaterpriceorclaimsaboutconsolidatedfills.

Same note246candidate protocol, added aftercompletedUSrepair revealedNXtopliquidfreshmover, beforeNXhistory.2023-24training +10.5272%,N6;2025-Sep4holdout-36.1835%,DD53.2616%,N4naturalclosedalllosses plus1openmark. Fee5bpslip10bp eachside,nextopenSMA10/50/flatwindowstart. Stress10/20bp-37.0392%.Trainingonly5/50winnerholdout-9.8913%,N6,win33.33%,DD39.299%; alsofails, cannotretune. Closedpayoffmean-15.5929%,10000IIDbootstrapCI[-19.9132,-11.2727]%.10trade10%allocationempiricalsimulationmedian-14.5736%,conditionalallloss sample, nottrue100%forecastlossprobability. OLS420dailySPYbeta1.55985,corr.46099,annualintercept-15.5606%,descriptive. NativeCLIcontrolalsoarchived; vectorbtindependentcashreconciliationmax1.82e-12. Currentfreshcrossfalse; top1daypricejump+21.178% isdiscovery, notentryalpha. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; issuer https://investors.quanex.com/news-releases/news-release-details/quanex-building-products-announces-third-quarter-2026-results .