Stats

Premises1 primary date required: DRI Sep24BMO https://investor.darden.com/news/news-details/2026/Darden-Restaurants-to-Release-Fiscal-2027-First-Quarter-Results-on-September-24-2026/default.aspx ; PPIH Sep9BMO https://investors.permapipe.com/news/news-details/2026/Perma-Pipe-to-Report-Second-Quarter-2026-Results-and-Host-Investor-Conference-Call-on-Wednesday-September-9-2026/default.aspx ; WLTH Sep9AMC https://ir.wealthfront.com/node/7786/pdf ; MRK Sep21 goal https://www.merck.com/wp-content/uploads/sites/124/2026/08/MRK-06.30.2026-10Q-FINAL.pdf . 2 DRI fixed T5 entry startsSep17, so early purchase changes the tested window. Its close below20SMA/above50SMA is not a trained combined filter. 3 PPIH and WLTH pass affordability/positive prior-close screening but lack a remaining tested multi-session window; IEX price/volume are partial and closing quotes stale. 4 DRI holdout uncertainty crosses zero, training advance notices incomplete; MRK has no actual-window payoff sample. 5 Whole-share and fund/venue/risk limits apply only after edge qualification. Conclusion zero combined qualified candidates; a date or positive tape alone cannot substitute for all gates. Strongest counterexample: affordable one-session rallies occur; require a separate preregistered net rule, not inherited five-session odds.

Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=DRI&timeframe=1Day&feed=sip&adjustment=all . Completed Sep8 close212.92 is below SMA20 219.5185 but above SMA50 209.3456; five-session return-0.6254%. ATR14 5.4490 and ADR20 5.4510, recent low210.86/high229.7625, RV20 30.8874% annualized. Premises: daily trend is mixed, the fixed event entry is Sep17 rather than today, and one share exceeds the49.14 fund cap. A one-to-two-ADR upside reference is about5.45–10.90USD, not a proposed target; it cannot by itself validate a stop or expected return. Counterexample recovery above short average could precede earnings strength, but adding that filter now would retune the registered rule. No entry, exit or stop adjustment; book flat. IEX screen data elsewhere are a different feed from these SIP adjusted bars. Daily history cannot establish15minute execution.

1. Startaccount850.99/quant245.69cash; no positions. Goalrepeatablepositiveprofit;bindingconstraintnegativevalidatedpayoff. Fromfrozen301trademanagedhistory,samplejointportfolioreturns,not independentstocks. 2. Seed9090201,20000circular3-sessionblockpaths,16sessions approximatingremainingSep9–30tradingwindow,entire245.69quantbooksupposedlydeployed into8equalinitialfractionalsleeves. Base5/50/95profitUSD[-1.87048,0,1.11598],lossfrequency49.8%;stress[-2.30212,-.17552,.81874],loss58.445%. No samplemakes349.01USD;this isnotprooftruelikelihoodzero. Flatdays in historicalstrategyexplainmedian0;no missingreturns convertedtozero. 3. Longerdiagnosticseed9090200,20000joint3monthblockpaths60months,quant/accountconstantweight. Baseaccount5/50/95return[-3.15198,-1.89739,-.64159]%,loss99.325%;stress[-5.27339,-3.97671,-2.67820]%,loss100%infiniteMonteCarlosample only. Correction:actual horizon60months,not100months originallycontemplated. Excludesfixedoperatingcosts,newregimes,wholelots,15minfillgaps andfuturecashrate. Resamplingassumesstationarity;shortblocksdo notpreservemonthlyentrycalendar; hypotheticalfreshschedule isnotanapprovedlateentry. 4. QuarterKellybinary-average-loss approximationfloors0 inbaseandstress;notional0. Caps212.7475/position and8.5099nominalrisk limitlossplans,notguaranteedgaploss. Neither need349.01 nor parentpositivereturnsuppliesapprovededge. Sources archivedhorizon_sim.py/json anddiagnostics; https://runway.adityaperswal.workers.dev/data/stock-bars .

1. Frozen managed252/1.5R2021–25holdout301trades:win35.8804%,W1.154196%,L.913672%,mean-.171713%;6bp/side portfolio-6.455432%/DD7.243026%. 2. NumPyseed9090200,20000circular3-monthblockbootstrap samples wholemonths jointly acrossassets,usingpooledsum/count so correlated ETF outcomes are not independent observations. 95%meanCI[-.314774,-.030477]% pertrade;16bp/side mean-.371170%,CI[-.514338,-.228008]%. 3. Monthlyregression againstequal-initial-weight8ETFbuyhold:managedbeta.041989,alpha-.136128%/month,HAC3lag t=-2.62162,R2.04850;stressalpha-.257074%,t=-4.58224. Buyhold42.4684% gross over5years,not matched risk. 4. Parent19.21533% costed/3.57774%CAGR,meanCI[-.038505,.903911]%;stress10.05885%/1.93539%CAGR,CI[-.237810,.702467]%. Parentpositive total doesnotestablish alpha ormanagededge. All12managedsweepcellsnegative,not independent12tests. Validity conditionaloncollectedadjustedprices/laggedsignals; soundnesslimitedbyreusedhistoricalera,proxyassets,actualfillunknown andindependentsleevedrift. Own3operatorliquidationsp1/3,avgwin7.93994%/loss5.26143%,mean-.86097%,cannotcalibrateplanned4–6weekexits orslippage withnoquote benchmarks. Prior skepticism strengthened for this adaptation; would change only withnewpredeclaredindependentpositive netevidence andactualexecution. Sources archivedQTR9090200diagnostics.json,trend.py; https://runway.adityaperswal.workers.dev/data/stock-bars .

Ran node /app/agent/dist/backtest.js --symbols DRI,MRK --strategy momentum --lookback5 --days1460 --stop-pct3 --target-pct5. 1001 daily bars and 133 trades. Source https://runway.adityaperswal.workers.dev/data/stock-bars . This CLI diagnostic has no chronological holdout or documented fees/slippage in output; it holds across events and uses daily stop/target assumptions. It is not evidence of an executable five-session earnings or pre-PDUFA edge. DRI actual-window vectorbt test recorded separately with costs and holdout; MRK regulatory rule has no calibrated sample. Positive MRK headline return cannot be transported to Sep21 HYPERION. Reject adoption of both generic rules. Command/output stored cli_dri_mrk.json.

Ran node /app/agent/dist/backtest.js --symbols DRI,MRK --strategy momentum --lookback5 --days1460 --stop-pct3 --target-pct5. 1001 daily bars and 127 trades. Source https://runway.adityaperswal.workers.dev/data/stock-bars . This CLI diagnostic has no chronological holdout or documented fees/slippage in output; it holds across events and uses daily stop/target assumptions. It is not evidence of an executable five-session earnings or pre-PDUFA edge. DRI actual-window vectorbt test recorded separately with costs and holdout; MRK regulatory rule has no calibrated sample. Positive MRK headline return cannot be transported to Sep21 HYPERION. Reject adoption of both generic rules. Command/output stored cli_dri_mrk.json.

Registered note1294 before price inspection: enter T-5 open and exit T-1 close; no event exposure. Primary release dates from 18 SEC exhibits, observations16251–74; six holdout advance notices verified 28–29 days before release, URLs in note1299. Training2022–24 N12; untouched2025–June2026 holdout N6. Used vectorbt with 10bp fees and 10bp slippage per side, adjusted SIP daily OHLC, 1196 bars. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=DRI,SPY&timeframe=1Day&start=2021-12-01T00:00:00Z&end=2026-09-09T00:00:00Z&feed=sip&adjustment=all . Training net return1.4581%, DD6.6115%; holdout net5.6622%, DD4.9869%, p50%, cash-basis mean0.951163%, average win3.216773%, loss1.314447%. Stress25bp slippage plus10bp fee per side still +3.7773%, DD5.1290%, mean0.648762%. Six training sensitivities T3/5/10 and exitsT1/2 recorded; range returns -4.3463% to+9.3334%, not used to choose holdout parameters. Bootstrap50000 seed20260909 mean95%CI[-0.8285,2.9660]%, includes zero. Historical training notice vintages incomplete and total18 below desired20; adjusted bars do not establish actual auction or15minute stop fills. Positive evidence merits prospective watch, not adoption. Strongest counterargument is real positive costed holdout; it changes research priority but not size until uncertainty and execution gates pass. Current primary notice https://investor.darden.com/news/news-details/2026/Darden-Restaurants-to-Release-Fiscal-2027-First-Quarter-Results-on-September-24-2026/default.aspx ; paper entrySep17open/exitSep23close, not today. Local reproducible dri_test.py/dri_test.json. Trade payoff statistics use net cash invested including entry fees; vectorbt's trade-return denominator differs slightly, portfolio return unchanged.

1. CompletedSep8dailybars: all8nonstabletop10 haveSMA20>SMA100. BTC78448.681,ETH2484.527,SOL103.36;20dayreturns13.1839/10.3693/21.1092%; ATR14 simplemeantrue range2271.0718/88.76049/5.6218. One/twoATR measured ranges are evidence, not proposedtargets. 2. LatestcompletedhourSep9 01UTC: SMA20>SMA100 onlyLINK,LTC,UNI,XRP (4/8); BTC,ETH,SOL,HYPEfail. BTC20hour+.4040%,ETH+1.0431%,SOL+.8619%, so positive short returncan coexistwithslowertrendfailure. Noneof10includingstables crossesprior20dayclosinghigh in census. 3. Download738dailybarsmostassets,HYPE204; hourly17712BTC/ETH/LTC,17707SOL,17708LINK,17709UNI,17684XRP,4884HYPE. Missing2/2/2/7/6/5/30/1 respectively; zero-volume bars materiallylimitexecution. Initial8workers hit429; retriedonlymissingfiles with2workers andbackoff; completepagination85BTC/77ETH/84SOL/etc recorded. 4. FixedSMA diagnostics1356–1371 reject: dailyonly1closedloss plusopen; allhourly held-gapinvalid, sevennegativebaseline andHYPEpositiveunqualified. Combiningindicatorsdoesnotfixpayoffvalidation. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; observationssavedpersymbol,timeframe. Conclusion no signalapprovedsize; preserve2026diagnostic asreusedhistory and testfutureflowobservationsprospectively.

1. NumPy seed2026090902; sample10000 joint overlapping3day logreturn blocks from last60 completed dailybars throughSep8. Only58 distinct blocks, preserving cross-major dependence within each block. Fee haircut.9975²; no slippage/stops/targets/15minexecution. Unconditional risk, not an entry-conditioned forecast. 2. Simulated mean BTC+.72379%,ETH+1.36121%,SOL+1.28556%; lossprob49.84/46.94/47.79%; simulated fifthpercentile-3.36973/-3.90019/-5.85392%, dollar losses at212.7475 cap7.16901/8.29756/12.45408. SOL exceeds8.5099 equityriskbudget. 3. Exact linear-interpolated fifthpercentiles across58distinct blocks -3.21421/-3.66867/-5.10359%; worst3day -4.13597/-4.26211/-7.24518%. Quantile differsfrompreviousseed because tailhasfewdiscreteblocks, notnewinformation; no confidencefrom10000repeats. 4. No newdailybar sincepriorrun, so dependence/volatility unchanged. Positiveunconditionaldrift and lowerBTC historicaltail do notcalibrateentrywinrate. Kellygate alreadybindsat0; stopslippage couldexceedriskbudget. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; test.py and simulation.json, complete observation. Conclusion size0, no invented exit odds.

DRI https://www.sec.gov/Archives/edgar/data/940944/000094094426000016/exhibit991-q4fy26.htm :1 traffic/mix andextraweekaffectheadlinegrowth;2 food/laborand875mplannedcapexcompeteforcash;3 cashafterinvestmentversusdividend/buybacksdeterminesfinancingneeds. Beneficiariescustomers,workers,suppliers,repurchasingshareholders;countercasepositivecompsandcashgrowthfundsustainableexpansion. MRK https://www.merck.com/wp-content/uploads/sites/124/2026/08/MRK-06.30.2026-10Q-FINAL.pdf :1 broaderHYPERIONlabelmayexpandeligiblepatient/treatmenttiming;2 physicians/payers/adverse-eventmonitoringdetermineactualuptakewhileBMSreceives22%net-salesroyalty;3 retainedcashnetofcommercial/researchcostsdeterminesMRKvalue. Countercase label adds little afterpriorZENITHapproval orpayerslimituse. No probabilityestimatedforapproval/stockorBMYbenefit;royaltyisarithmeticnottradeablearbitrage. KGF https://www.kingfisher.com/en/investors :1 housing/DIYdemandchangesvolumes;2 promotionsandfinancecostoffsetsales;3 FCF/workingcapitaldecidesbuybackcapacity;countercaseimprovedmixsustainsearnings. ASKUL https://www.askul.co.jp/news/INFO/ :1 resolvedloginissuesmayrestoreorders;2 repeatpurchasingandfulfillmentcostdecideactualrecovery;3 retainedcashandtrustdeterminevalue;countercaseresolvedbugsimmaterialtothequarter. Allconditionalhypotheses,notmeasuredpeerprofits. No separateFUND justified withoutcostedpayoffevidence.

Premises1 DRI Aug27 notice confirms Sep24BMO, not estimatedSep17: https://investor.darden.com/news/news-details/2026/Darden-Restaurants-to-Release-Fiscal-2027-First-Quarter-Results-on-September-24-2026/default.aspx . June25 SEC https://www.sec.gov/Archives/edgar/data/940944/000094094426000016/exhibit991-q4fy26.htm reports53-weekFY;Q4growth13.7% includesextraweek7.6%,OGcomps2.4vsLongHorn9.5. SimpleOCF-propertycapex-software/other1096.2m grows5.414%,but1364.7m shareholderreturns exceed268.5m;netshorttermdebt194m. FY27EPS11.10–11.35,capex875m. Counterexample cashgrowthandcompsarepositive;return/cashgapdoesnotproveinsolvency, andnext52-weekcomparisonmustbealigned. 2 MRK June30filing https://www.merck.com/wp-content/uploads/sites/124/2026/08/MRK-06.30.2026-10Q-FINAL.pdf confirmsSep21HYPERIONlabelreview. Q2Winrevair588m,3.5407%company16607m;22%BMSroyaltyimplies129.36m arithmetic,notcashreceived,remaining458.64mnotprofit. ExistingZENITHapprovaldistinct;WeliregOct4competingpipeline. Counterexample label expansioncouldaccelerateearliertreatment, butsalesgrowth75%isnotapprovalorstockwinprobability. 3 KGF Sep22explicitlyprovisional https://www.kingfisher.com/en/investors ,priorFYguidePBT565–625/FCF450–510mGBP notverifiedunchangedtoday. ASKUL Sep15dateonly https://www.askul.co.jp/corp/investor/calendar/ ;customerSep8resolvedloginnotice https://www.askul.co.jp/news/INFO/ doesnotquantifychurn. Validity: cash/royalty/datefactsinformwatchlists,nottradingedge. Soundness:primaryissuerincentivizedframing,unreadupdatedKGFQ1andforeignvenuecostgaps. Start49.14cash;bindingconstraintcostedactualwindowuncertainty. These improve candidate diversity but do not authorize allocation.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"XRP","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":-45.31520030422348,"drawdown_pct":68.05550120348127,"n":44,"win_rate_pct":20.454545454545457,"avg_win_pct":7.829663110407015,"avg_loss_pct":3.438492774858877,"expectancy_pct":-1.1336427074181266,"quarter_kelly":0,"open":0,"missing":30,"held_gap_invalid":true,"zero_volume":69}. Stress {"symbol":"XRP","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":-58.00375364562005,"drawdown_pct":75.17143779639743,"n":44,"win_rate_pct":13.636363636363635,"avg_win_pct":10.932408647804843,"avg_loss_pct":3.725359267777965,"expectancy_pct":-1.7265727338348549,"quarter_kelly":0,"open":0,"missing":30,"held_gap_invalid":true,"zero_volume":69}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"XRP","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":-12.576158684719962,"drawdown_pct":16.847593670961537,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":11.831399735049802,"expectancy_pct":-11.831399735049802,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Stress {"symbol":"XRP","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":-13.358289883070599,"drawdown_pct":17.429120215555372,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":12.360334054498919,"expectancy_pct":-12.360334054498919,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"UNI","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":-24.065349670624332,"drawdown_pct":59.702134711355704,"n":40,"win_rate_pct":20,"avg_win_pct":9.67953296923908,"avg_loss_pct":4.137901802815852,"expectancy_pct":-1.3744148484048655,"quarter_kelly":0,"open":1,"missing":5,"held_gap_invalid":true,"zero_volume":1681}. Stress {"symbol":"UNI","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":-40.44611675578691,"drawdown_pct":67.91804543656906,"n":40,"win_rate_pct":17.5,"avg_win_pct":10.377326346962166,"avg_loss_pct":4.58416563367424,"expectancy_pct":-1.9659045370628694,"quarter_kelly":0,"open":1,"missing":5,"held_gap_invalid":true,"zero_volume":1681}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"UNI","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":54.42818341582731,"drawdown_pct":27.840691563172683,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":15.77447261387305,"expectancy_pct":-15.77447261387305,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Stress {"symbol":"UNI","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":53.04660262119546,"drawdown_pct":28.272360177984247,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":16.279818835756174,"expectancy_pct":-16.279818835756174,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"SOL","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":-43.23618148390702,"drawdown_pct":60.72312265038189,"n":46,"win_rate_pct":21.73913043478261,"avg_win_pct":7.624812376997625,"avg_loss_pct":3.446120911624754,"expectancy_pct":-1.039396283663367,"quarter_kelly":0,"open":0,"missing":7,"held_gap_invalid":true,"zero_volume":124}. Stress {"symbol":"SOL","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":-56.92713016765196,"drawdown_pct":69.65503624102718,"n":46,"win_rate_pct":19.565217391304348,"avg_win_pct":7.761186383521048,"avg_loss_pct":3.9179357403007318,"expectancy_pct":-1.6328901073790798,"quarter_kelly":0,"open":0,"missing":7,"held_gap_invalid":true,"zero_volume":124}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"SOL","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":-7.2673895414414575,"drawdown_pct":16.278259939392502,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":15.473666528142124,"expectancy_pct":-15.473666528142124,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Stress {"symbol":"SOL","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":-8.097015266563403,"drawdown_pct":16.779096883987464,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":15.980812220518612,"expectancy_pct":-15.980812220518612,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"LTC","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":-52.24143230536836,"drawdown_pct":63.6873548585674,"n":40,"win_rate_pct":17.5,"avg_win_pct":3.201181880318882,"avg_loss_pct":3.075091125317925,"expectancy_pct":-1.9767433493314832,"quarter_kelly":0,"open":1,"missing":2,"held_gap_invalid":true,"zero_volume":2670}. Stress {"symbol":"LTC","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":-62.54400129506295,"drawdown_pct":71.2637004148044,"n":40,"win_rate_pct":17.5,"avg_win_pct":2.5823202344648957,"avg_loss_pct":3.6564070943463824,"expectancy_pct":-2.5646298118044086,"quarter_kelly":0,"open":1,"missing":2,"held_gap_invalid":true,"zero_volume":2670}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"LTC","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":-6.0422966924771915,"drawdown_pct":22.16149838847531,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":8.96627008670667,"expectancy_pct":-8.96627008670667,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Stress {"symbol":"LTC","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":-6.882882623918523,"drawdown_pct":22.627140852317286,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":9.512344073557632,"expectancy_pct":-9.512344073557632,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"LINK","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":-47.10564126651948,"drawdown_pct":64.1388375500173,"n":42,"win_rate_pct":16.666666666666664,"avg_win_pct":7.936627014382744,"avg_loss_pct":3.3530975309104454,"expectancy_pct":-1.471476773361581,"quarter_kelly":0,"open":1,"missing":6,"held_gap_invalid":true,"zero_volume":1183}. Stress {"symbol":"LINK","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":-59.010954906740324,"drawdown_pct":71.45019528772997,"n":42,"win_rate_pct":11.904761904761903,"avg_win_pct":10.244414035510268,"avg_loss_pct":3.7254668837559795,"expectancy_pct":-2.0623858219385687,"quarter_kelly":0,"open":1,"missing":6,"held_gap_invalid":true,"zero_volume":1183}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"LINK","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":13.57282319845375,"drawdown_pct":26.136839010113277,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":13.071973780277105,"expectancy_pct":-13.071973780277105,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Stress {"symbol":"LINK","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":12.556751987536268,"drawdown_pct":26.57870034555031,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":13.593486785812969,"expectancy_pct":-13.593486785812969,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"HYPE","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":41.49934526162804,"drawdown_pct":40.78782856459645,"n":30,"win_rate_pct":30,"avg_win_pct":17.108214488776355,"avg_loss_pct":4.607178679103651,"expectancy_pct":1.9074392712603507,"quarter_kelly":0.02787314936505654,"open":0,"missing":1,"held_gap_invalid":true,"zero_volume":3805}. Stress {"symbol":"HYPE","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":18.189741262485676,"drawdown_pct":44.901328382534146,"n":30,"win_rate_pct":30,"avg_win_pct":16.40615873211006,"avg_loss_pct":5.179329453420031,"expectancy_pct":1.2963170022389978,"quarter_kelly":0.019753511827571346,"open":0,"missing":1,"held_gap_invalid":true,"zero_volume":3805}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Absent development independently forces allocation0. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"HYPE","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":-0.7108071354784329,"drawdown_pct":29.95287976045705,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":7.519826451968189,"expectancy_pct":-7.519826451968189,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":16}. Stress {"symbol":"HYPE","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":-1.5990908602616742,"drawdown_pct":29.952879760457062,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":8.074553297802124,"expectancy_pct":-8.074553297802124,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":16}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Absent development independently forces allocation0. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"ETH","tf":"1Hour","fast":20,"window":"holdout","slip":0.003,"return_pct":-42.75159576364943,"drawdown_pct":58.68152338618442,"n":39,"win_rate_pct":17.94871794871795,"avg_win_pct":8.001791213141848,"avg_loss_pct":3.3104738828481173,"expectancy_pct":-1.280067327157611,"quarter_kelly":0,"open":0,"missing":2,"held_gap_invalid":true,"zero_volume":28}. Stress {"symbol":"ETH","tf":"1Hour","fast":20,"window":"holdout","slip":0.006,"return_pct":-54.69602149601879,"drawdown_pct":66.50804149274869,"n":39,"win_rate_pct":17.94871794871795,"avg_win_pct":7.3542115484231845,"avg_loss_pct":3.890381754278436,"expectancy_pct":-1.8721214178960934,"quarter_kelly":0,"open":0,"missing":2,"held_gap_invalid":true,"zero_volume":28}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Every hourly candidate has a missing interval while invested: diagnostic returns are not accepted execution outcomes; do not fill missing bars or count these as qualified tests. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.

Frozen before download in note1292. Prior completed SMA20>SMA100 enters at next adjacent open; inverse exits, long-only. Development2025 fast10/20/30 and slips30/60bp; holdout2026Jan1 throughSep8 fast20 only, each period starts flat. Warmup begins2024Sep1; HYPE begins2026Feb17 so no2025training. Vectorbt from_signals fees25bp cash each side (approximation to credited-asset fees), baseline slip30bp/stress60bp. No stop/target/profit-taking or15minute execution simulated; not venue-ready. Baseline {"symbol":"ETH","tf":"1Day","fast":20,"window":"holdout","slip":0.003,"return_pct":12.13625209357799,"drawdown_pct":16.300196389017486,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":14.786455826177548,"expectancy_pct":-14.786455826177548,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Stress {"symbol":"ETH","tf":"1Day","fast":20,"window":"holdout","slip":0.006,"return_pct":11.133033064203527,"drawdown_pct":16.80090210623466,"n":1,"win_rate_pct":0,"avg_win_pct":null,"avg_loss_pct":15.29771252374835,"expectancy_pct":-15.29771252374835,"quarter_kelly":0,"open":1,"missing":0,"held_gap_invalid":false,"zero_volume":0}. Premises: current universe introduces survivor selection; 2026 already researched so not pristine unseen validation. Only one closed losing trade plus one open position: positive marked returns do not estimate average win or a tradeable probability. Conclusion reject for trading; even positive raw Kelly is overridden to0. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees . Reproducible scripts in this run's saved note.