Stats

1. FY26 price/cost timing drove much of profit while FOCF was negative; issuer https://investors.kennametal.com/node/28606 . 2. Full annual supply/risk passages describe foreign-source exposure, recycling mitigation and losses possible from both rising inputs and falling inventory values; https://www.sec.gov/Archives/edgar/data/55242/000162828026056143/kmt-20260630.htm . 3. Inventory doubled; numerical cash bridge is observation26639. At29.44, applying issuer20%conversion to FY27 EPS4.15–5.15 implies only2.82–3.50%FOCFyield, assuming unchanged share count/definitions. This is a sensitivity, not fair value. 4. A machinist's25% tool-hike/switching report is anonymousN1, not paid companywide revenue: https://www.reddit.com/r/Machinists/comments/1v9483e/kennametal_price_hike/ . Conclusion: watch cash conversion, reject extrapolating quarterly41%margin. Validity: pricing does not entail cash; soundness limited by lagged accounts and anecdote. Strong counterargument: stockpiling secures future profitable supply and future cash release; require inventory/advances unwind with retained pricing to raise assessment. No short or long order; current entry false.

Premises: (1) Sep2 17:39:51UTC maintainer2.4.1 is a critical security patch to snapshot validation; (2) Sep12 media upgrade headline is10days later; (3) adoption, paid fees and ADA buy flow are unmeasured. First order patched participants prevent this invalid-transaction path; second order operators may upgrade/close older heads, with uncertain liquidity impact; third order trust recovery may help adoption but supplies no measured spot payoff. Who benefits: head participants and operators; tokenholder gain requires a separate demand bridge. Strongest counterargument: successful patch can remove a risk discount. Validity: conditional consequence chain coherent; soundness of bullish price premise unestablished. Do not extrapolate to Cardano L1 compromise or assert theft occurred. ADA outside currenttop10 additionally fails mandate universe. No buy/no short; retain primary advisory as context. Sources https://github.com/cardano-scaling/hydra/releases/tag/2.4.1 ; https://github.com/cardano-scaling/hydra/security/advisories/GHSA-cg83-6w6r-6hx3 .

Premise1: query_trades crypto has only one closed BTC lot, operator liquidated after47.93min, entry81614.8 exit81548.7, gross-.08099%; no natural stop/target outcomes and no reference quote for slippage. N1 cannot calibrate Kelly or strategy win probability. Premise2: replacement jobs247/249 have5 delivered END receipts each, recorded-note gaps mean26.248/26.284min,max29.987/30.164min; funding248 N2 gap60.068min. Thus15min schedule labels still do not prove15min observation; replay identifies a missed event, not an executable fill. Premise3: Sep13 01UTC all7 HIGH480 false; ETH Sep11 22UTC pullback old/noarm. Conclusion retain observation-only collectors through04:30 cutoff and no late order. Binding constraint fresh same-policy entry, not data perfection or monthly target. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://runway.adityaperswal.workers.dev/data/crypto-snapshots ; query_trades/list_jobs fundcrypto. A fresh event and executable quote could change decision; collector latency itself has no price direction.

Premises:1 refreshedfixed9signals3048 selectQQQ/IWM/EFA afterCoreEEMqueue exclusion fromsuppliedcontext;2 ownquery_tradesempty,account850.99/fund289.34 unchanged;3 roughlyequal96.44/96.45/96.45 below212.7475perpositioncap;4 risk3049 andprior3001 perlot20.45/20.10/24.13<25.5297, joint64.67=7.60%,notbound;5 fractionalnotionalmarket onlyRTH and Sundayclosed. Conclusion retaincash untilMondaySep14 13:50UTC monitor339, refreshstate/entry then execute fixedcomplement; no lower-rankedcrypto orwhole-sharelimit substitution. Targetpayoff37.251 vsordinaryrisk5.7657 over1–3months, notexpectedprofit; referenceexits3053, neverprofitratchet atentry. Validity followsmandate; soundnesslimits adjustedtotalreturnproxy, unknownfills/gaps, noindependentcurrentQQQweights/cashflowaudit. Noall-listing screener outsidefixeduniverse, no repeatedbacktest under explicitvalidatedrule, no discretionaryKellyoverlay. Due340resolved,new34704UTCarmed,Tuesday327retained. Nojobneededbefore04UTC: ownflat, nonewcompleteddailybar, existinghandoffcoversstate. Noorders/closes/exitmoves/FUNDproposal. Sources https://runway.adityaperswal.workers.dev/data/stock-bars https://runway.adityaperswal.workers.dev/data/crypto-bars andissuerURLs3051.

Premises:1 signal3048 and risk3049;2 nofill tonight;3 initialstopE-2ATR,targetE*(1+P90),1–3months. QQQ referenceE714.88 stop699.14315 target830.47313 P90 16.169585%; IWM288.89/282.74895/325.93680 P9012.823845%; EFA106.70/104.93826/116.97551 P909.630283%. RR7.3454/6.0326/5.8326 beforecosts. Actualentry recalculation mandatory. Profitable ratchet max(existing,SMA200,high20-2ATR) onlyafterprofit; current raw thresholds718.84315/299.03895/107.12826 exceed closes, so cannot beinitialstops. Alreadycrossed threshold afterprofit=>RTHexitreview, never invalidstop. P90 not90%hitodds. Source https://runway.adityaperswal.workers.dev/data/stock-bars .

Premises:1 issuerexposures recorded thisrun;2 correlations .7068–.7494;3 gapstress3049 joint64.6743;4 fractional exitsRTHonly. Conditional first order: weaker profit expectations/higher discount rates or adverse FX reprice equities. Second: QQQ concentration, IWM funding sensitivity and EFA financial/industrial cycle can amplify jointloss. Third: overnight repricing bypasses2ATR stops, delayed liquidation reduces capital available for nexttrend. Not claims of an actual Saturday event. Counterexample: persistent earnings and benign financing sustain momentum. Startflat, goal fixedcomplement, bindingRTHconstraint; no ownholding implies no Rmultiple/age/exitreview. Whobenefits: issuers collect assetfees, holders capture upside and bear gaprisk. Soundnesslimited by unknown future shocks, not invalidation of accepted rule. Sources https://www.invesco.com/qqq-etf/en/about.html https://www.ishares.com/us/products/239710/ishares-russell-2000-etf https://www.ishares.com/us/products/239623/ishares-msci-eafe-etf https://runway.adityaperswal.workers.dev/data/stock-bars .

Premises:1 QQQ issuer identifies Nasdaq100 nonfinancial concentration https://www.invesco.com/qqq-etf/en/about.html ;2 IWM broadUSsmallcap andclose288.89 https://www.ishares.com/us/products/239710/ishares-russell-2000-etf ;3 EFA international developed large/midcaps, Sep10 financials26.47%,industrials18.66%,Japan24.44%,PE19.16/PB2.37/658holdings; Sep11close106.70 https://www.ishares.com/us/products/239623/ishares-msci-eafe-etf . Issuer closes agreebars. Inference: EFA cyclical financial/industrial exposure can correlate with US risk despite overseas holdings. Counterexample: divergent earnings/FX may diversify ordinary periods; PE alone proves neither cheap nor expensive. QQQ growth expectations and IWM refinancing are conditional bear mechanisms, no new shock measured. Issuers benefit assetfees and emphasize products; no audited aggregate debt/cashflow or freshQQQweights obtained. Rule remains fixed; no fundamental veto.

Premises:1 fiveyear1255 joint dailygap vectors preserve crossasset dependence;2 iid20000paths60nights seed20260913;3 hypothetical initialstop before adverse overnightgap, plus50bp. {"paths":20000,"nights":60,"sample":1255,"p50":13.102510465679082,"p95":16.966882644386843,"max":20.658827830036604}. Reproduces3002; maximum-night loss distribution, not compounded portfolio PnL or target odds. Omits2020, regime clustering, ratchet and novel gaps; stricter crisis stress controls sizing. No win-rate/Kelly estimate inferred. Source https://runway.adityaperswal.workers.dev/data/stock-bars.

Premises:1 sizes96.44/96.45/96.45;2 fresh all-adjusted2020+ open/previousclose worst gaps on2020-03-16;3 stress=N*[r+(1-r)*2g+.005], r=2ATR/close. [{"symbol":"QQQ","size":96.44,"stop":699.14315,"target":830.4731316245595,"normal":2.1229602366830784,"gain":15.593948094606814,"worst_gap":0.09460338101430432,"worst_date":"2020-03-16 04:00:00+00:00","stress":20.450581934423802},{"symbol":"IWM","size":96.45,"stop":282.74895,"target":325.93680479074,"normal":2.0502761345148643,"gain":12.368598158700099,"worst_gap":0.09305480195333693,"worst_date":"2020-03-16 04:00:00+00:00","stress":20.101221352019692},{"symbol":"EFA","size":96.45,"stop":104.93826,"target":116.97551184268325,"normal":1.5925006841611973,"gain":9.288407846549196,"worst_gap":0.11621497473587517,"worst_date":"2020-03-16 04:00:00+00:00","stress":24.1224744571582}] Correlations252 {"QQQ":{"QQQ":1,"IWM":0.7375146484818859,"EFA":0.7068026873334806},"IWM":{"QQQ":0.7375146484818859,"IWM":1,"EFA":0.7494056074136088},"EFA":{"QQQ":0.7068026873334806,"IWM":0.7494056074136088,"EFA":1}}. Prior3001 split-adjusted2016+ stress64.6655; newall-adjusted64.6743 difference.0088 from definitions, not changed market. Both each<25.5297 but joint7.60%equity. Full notional loss exceeds3%; finite scenario is not guaranteed bound. EFA retained conservative max24.1263. Gross targets37.251 vsordinary5.7657 over1–3months, payoff geometry not expected return. Source https://runway.adityaperswal.workers.dev/data/stock-bars.

Premises:1 completed Sep11 ETF and Sep12 crypto bars, pagination exhausted, OHLC checks passed;2 Core EEM queued per supplied venue context, own query_trades empty;3 rank close>SMA200 by252/365 interval return, arithmetic ATR20 and 60bar return P90 endpoints since2021-09-13. Full table [{"symbol":"SPY","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":764.29,"ret":18.489062524223687,"sma":712.0451999999999,"atr":5.452859999999982,"p90":0.10842338318374645,"high20":778.8,"rank":5,"size":0},{"symbol":"QQQ","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":714.88,"ret":23.69878183831673,"sma":659.2565,"atr":7.868424999999997,"p90":0.1616958533244174,"high20":734.58,"rank":2,"size":96.44},{"symbol":"IWM","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":288.89,"ret":23.457264957264947,"sma":273.16515,"atr":3.070524999999995,"p90":0.1282384464354598,"high20":305.18,"rank":3,"size":96.45},{"symbol":"EFA","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":106.7,"ret":19.45812807881775,"sma":100.67334999999999,"atr":0.8808699999999987,"p90":0.09630282889112697,"high20":108.89,"rank":4,"size":96.45},{"symbol":"EEM","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":67.84,"ret":34.203758654797234,"sma":61.760349999999995,"atr":0.9526399999999988,"p90":0.10717963493264468,"high20":69.125,"rank":1,"size":0},{"symbol":"GLD","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":398.77,"ret":18.94350653224364,"sma":415.93449999999996,"atr":7.952445000000003,"p90":0.16781703911009302,"high20":429.42,"rank":null,"size":0},{"symbol":"TLT","bars":1682,"date":"2026-09-11 04:00:00+00:00","close":80.87,"ret":-5.756904789651552,"sma":84.50049999999999,"atr":0.6911250000000002,"p90":0.06150557763249668,"high20":83.28,"rank":null,"size":0},{"symbol":"BTC/USD","bars":2081,"date":"2026-09-12 00:00:00+00:00","close":77266.385,"ret":-33.45056899540282,"sma":70095.06604458945,"atr":2335.032182487298,"p90":0.37169104405296727,"high20":82233.502538071,"rank":6,"size":0},{"symbol":"ETH/USD","bars":2081,"date":"2026-09-12 00:00:00+00:00","close":2525.045,"ret":-46.433545367857086,"sma":2055.8333431083274,"atr":94.40364558315002,"p90":0.42934306094151997,"high20":2664.8535,"rank":7,"size":0}]. EEM excluded Core queue; GLD/TLT belowSMA; SPY/BTC/ETH lower rank. No positive-return extra filter. Scope is fixed9 including developed/emerging international exposure and BTC/ETH; single names and other countries/coins rejected as outside mandate, not screened and failed. No rebacktest per explicit validated mandate. Source https://runway.adityaperswal.workers.dev/data/stock-bars and https://runway.adityaperswal.workers.dev/data/crypto-bars. Sizes proposed, no order; RTHclosed. Adjusted closes proxy total return; actual-entry exits required.

Start and finish: supply $170.20 cash, account equity $850.99; fresh own open/pending/closing ledgers empty. 1. Saudi Energy Ministry original confirms precautionary shutdown, without lost export bpd or repair duration; cargo removal can hurt tankers even as replacement voyages lengthen (3014). https://www.spa.gov.sa/N2674019 . 2. FRO cash and annual risk audit3020 separates operating cash from vessel-sale capital. Prior raw test remains positive; venue dividend credit/raw stops and current valuation bridge unresolved. $3.41 distribution can cross reference stop mechanically. https://www.frontlineplc.cy/fro-declares-special-one-time-dividend-after-sale-of-two-vllcs/ . 3. Hydro August13 ramp contradicts persistent50% Alunorte output assumption, while expensive gas can retain Q3 cost impact (3027). https://www.hydro.com/en/global/media/news/2026/alunorte-ramps-up-alumina-production/ . 4. New CNQ cash/concentration study3042 supports a watch; fixed L20 training+15.6616%, heldout+15.3639% N11+1open, fixed-path cost stress+12.7409% (3030–32). Mean CIs cross zero and current50.07<51.33; actual size0. Conditional quarter21.27475 cannot round to one50.07share. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://www.sec.gov/Archives/edgar/data/1017413/000101741326000043/a06302026q2fs.htm . 5. STNG signal true but prior holdout rejected; AA false. Strongest case against cash is positive FRO/CNQ histories; it earns priority for executable prospective work, not altered signals or invented target multiples. Validity: current entry/execution requirements unmet implies no order. Soundness limited by dated sources and missing actual flow evidence; not a claim whole market lacks opportunities. Review affects watches only: no holdings to reset, no trade orders, no exit changes, no FUND proposal. Job250 replaces236, local6/6 success, first remote receipt still null at00:23:50; preserve18/36/214/243, note36 partial403/timeout. Resolve335, arm34602UTC, retain Monday102/103 and FRO287. Lessons870–872; manifest26606. About23minutes used of43planned/45hard; final journal/report inside five-minute finish reserve. FRO accounting thread stopped at its10minute allowance; new Saudi flow measurement and raw CNQ15min/managed exits remain gaps, aero/RF deep dives deferred. No financial risk added for the $349.01 shortfall.

FreshSIPall6x1682dailybars endSep11; currentclose/prior20/SMA50/ADR20: CNQ50.07/51.33/46.4478/1.049 fails; FRO49.21/48.4/40.8458/1.349555 passes; STNG84.52/84.51/77.9266/2.001845 passes; AA48.26/51.85/48.526/1.95495 fails; XLE65.14/65.31/60.1392/1.021975 fails. CNQ reference stop47.972,target56.364,3R gross,max40sessions, not orders or established fairvalue. CNQATR20=1.2165 versusADR1.049; use frozenADR not a quietly widenedstop. CNQ26.94%worst2020+open-gap shows30%stress is plausible but not an empiricalfrequency. FROdividend3.41 can mechanically breach existing46.51089reference stop; it is not thesisfailure by itself. STNGpositive screen cannot override previously negativeheldout3R test2925. No holdings, hence liveR/daysheld/stopadjustments are not applicable; do not invent a holding from closedtrades. EEMqueue belongs totrend and remains outsideauthority. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all . NewCNQ testwatch does not authorize stale weekendentry.

Training only p=1/3, average win13.3288%, loss5.3370% => half Kelly risk fraction.0331949. Approximate mapping to reference2ADRstop4.190% gives capacity above venue/fund caps; a stated30%gap/3%equity risk budget binds fullnotional85.099 and quarter21.27475. Quarter stoprisk$.89144,30%gap$6.38243=.75%account, below one50.07share; no rounding. Net-costed training returns sampled on20,000IID paths of20trades,seed91300,fixed accountfraction. Quarterbase P05/P50/P95=-1.1804/+.4054/+2.1986%,95thDD1.6068%,lossodds34.565%. Replace5%ofdraws by min(return,-30%): quarter=-2.3899/-.3174/+1.7415%,DD2.7685%,loss59.815%; fullstressmedian-1.4423%,DD10.7717%. These probabilities are conditional model outputs, not estimated war-gap frequency; IID paths miss serial dependence. Twentytrades is not nineteen calendar days. A live stop cannot cap an overnightgap. Current signalfalse means actualsize0 despite positive baseline Kelly. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; new costed tests3030–31. No order.

Startaccount850.99/fund119.14cash, no crypto holdings or queues, verifiedquery_trades. Goal positiveexecutableexpectancy; bindingconstraint freshsame-policyentry. (1) All7volatile currenttop10 prior480HIGHsignalsfalse onSep12 23UTC,ETHoldSep11 22UTCpullbacknoarm(3012). Strongcounterargument cashmissescontinuedtrend; a dailyCLOSEsignal isdifferentpolicy. (2) New144cell costedvectorbtcontrols3033–38 reject substitutingclose-toggle:ETHknown2026daily-21.109976%,hourly-20.782205%; frozenmanagedpullback3028retained19.206747%N3againstnegative dailycontrol andsmalltrainingCI. No newparameterselected. (3) Fundingties/coinOI3016 andliquidity3017 donotestablishdirectionalentry; bearishconsequences3018areconditional. (4) Measurednet3Rtargets3041 require23.34/28.11/38.93%majorupside versus20dayP9018.75/23.87/23.67%; no targetadopted; actualsize0. Correlation/Kelly3039 andtwohourrisk3040 constrain futurebudget,nottriggers. (5) Replacement247/248/249 firstremote00:15–16UTCEND/exit0verified, fiveactivejobs14/25preserved; cadencebeyondfirstreceiptunverified. Cutoff04:30UTC,02UTCreview345armed,334resolved. Finalremotequotes5/10fresh;SOL/LTC/stablecoinsstale,independentLiquidexplorertimeout. Noorders/exitchanges/newfundproposals. Sources https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://www.circle.com/transparency ; https://www.okx.com/api/v5/public/funding-rate-history ; https://docs.alpaca.markets/us/docs/crypto-fees . Lessons867–869saved.

CNQ frozen L20 net-account trade mean: trainingN30 +.8849%, 20,000 IID bootstrap95%CI[-2.2398,4.2669]; heldoutN11 +1.4939%,CI[-2.9575,6.5350]. Positive geometric test is evidence for a watch, with uncertainty spanning zero; IID bootstrap misses clustered regimes. Training stop/gapstop/target/gaptarget/time counts13/5/7/1/4, heldout5/2/1/2/1. On424 common daily returns2025Jan–2026Sep11, corr CNQ/XLE=.7702, CNQ/FRO=.1724; OLS CNQ=intercept+XLE+SPY gives XLEbeta1.0133, SPYbeta-.0273, R2.5934, dailyintercept2.9233bp. Descriptive fit, no robust alpha significance or outage causality inferred. https://runway.adityaperswal.workers.dev/data/stock-bars . Own trade IDs6/3: FRO+3.8052%/+0.6799 initialR/$4.3756 and STNG+.7523%/+.1261R/$.6017, both operator-liquidatedSep4. N2/100%wins/average2.2787% describe forced exits, not tested stop/target strategy; arrival quotes unavailable so slippage unmeasured. source query_trades?fund=supply . Do not call two energy names unrelated merely because their short sample correlation is low.

1. Q2 issuer financials show C$6.823bn operating cash; less C$2.380bn net PPE and C$25m E&E gives C$4.418bn simple FCF. H1 equivalent C$5.522bn; these include investment timing and are not maintenance-only or issuer-defined distribution FCF. https://www.sec.gov/Archives/edgar/data/1017413/000101741326000043/a06302026q2fs.htm . 2. Aug6 release reports C$270m sulphur net revenue Q2/C$450m H1 and estimated 30% Canadian/2% world sulphur supply. Q2 mining output625kbpd at106% upgrader utilization cautions against assuming easy incremental volume. https://www.sec.gov/Archives/edgar/data/1017413/000101741326000043/a06302026q2pressrelease.htm . 3. Annual2025:27% production heavy/bitumen, transport/quality differentials can absorb global price gains; two customers each>10% annual sales, identities unspecified. https://www.sec.gov/Archives/edgar/data/1017413/000101741326000018/cnq-20251231.htm . 4. MD&A Q2 E&P liquids royalty rate19% versus13% prior year; higher prices raise government take. https://www.sec.gov/Archives/edgar/data/1017413/000101741326000043/a06302026q2mda.htm . Conclusion: upstream geography offers a plausible alternative to lost tanker cargo, but prices, differentials and costs jointly determine cash. First order Saudi outage may favor substitute barrels; second Canadian transport/refinery compatibility and royalties limit retained netback; third demand destruction or repaired exports reverse prices while fixed commitments persist. Sulphur is additional exposure, not an independent guaranteed earnings engine. Scale: sulphur net revenue is3.96% of Q2 OCF, not attributable cash fraction. Valuation sensitivity uses annualized H1 simple FCF C$11.044bn/2.077894bn shares and explicitly assumed1.38 CAD/USD: US$3.851/share. At assumed8/10/12% cash yields, values48.14/38.51/32.10; technical3R56.364 requires6.833% yield. These are scenarios, not selected fair values or current FX. Bull: persistent realized premiums and cash returns; bear: peak cash normalization, own transport and concentrated buyers. Current50.07<51.33 entry trigger independently bars purchase. No order.

Sep12completeATR20 BTC2335.032182,ETH94.403646,SOL5.457559. Illustrative00:04asks+10bpslip E/stop/net3Rresidualtarget: BTC77358.5813/72688.516935/95411.184565;ETH2528.2257/2339.418409/3238.821466;SOL101.875774/90.960657/141.533018. Stop E-2ATR;quarteratgross2R thenbreakeven remainder,25bpcreditedcoinbuy/USDsellfeesand10bpsellslip. Targetgains23.336265/28.106500/38.927060%; modeledstoploss6.599627/8.022128/11.248855%notional;29.785quarterloss1.965699/2.389391/3.350472USD. 214overlapping2026future20dayMFEwindows,P90gains18.753976/23.871490/23.666380%; coverageofrequiredgains8.4112/8.4112/7.9439%. Facts: sample ranges andalgebra;assumptions: fillcosts,quoteavailability,targetreach. UnconditionalMFEisnotentry-conditionedwinprobability,checkedfillorproofnoasymmetry. AlltargetsaboveP90; no defendablecurrentcatalystbridge, no freshentry. Conclusion no targetadopted/noexitmoved;do nottightenstoptoforce3R. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://runway.adityaperswal.workers.dev/data/crypto-snapshots ; https://docs.alpaca.markets/us/docs/crypto-fees .

NumPy seed91300,10000paths,1752contiguous2hourjointweekendblocks from2026hourlylogreturns throughSep12. Hypothetical119.14equalBTCETHSOLbasket; meanreturn-.014703%,P1-2.332657%=-2.779128USD,P5-.919514%,P95+.776395%. Zero simulatedlosses beyond25.5297USD(3%account) is not a bound; finitehistoricalsample omitsnewregimes. Explicit25%jointcrash costs29.785USD=3.50004%account;quarterfund29.785loses7.44625USDonsameshock. Preserveswithin2hourserialandcrossassetdependence, but firsthour mayFriday23UTC forSaturday00UTC endingblock. No fees,stops,targets,entryconditioning orpath-longerregimes;notbacktestorKellyedge. Premises: correlatedbookneedsjointbudget, observedquietweekendnotpromise; strongestcounterexample exchange/securitygaps outsidehistoricalblocks. Conclusion no orderbasedonsimulation;actualsize0, entrytestseparate. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; observation26595.

2026 hourlyreturns6118each, unfilledmissinghours; SD BTC.467852%,ETH.621319%,SOL.670824%,weekends.330404/.488851/.587092%,weekdays.512783/.667165/.701682%. DailySD2.426569/3.311268/3.326831%. CorrelationsBTCETH.887966,BTCSOL.837870,ETHSOL.848997; OLSalt-onBTCbeta1.179240/1.201370, no identifiedalpha. Premise1 lowerweekendvariance notdepthorlefttailguarantee. Premise2 highdependence means3coins are substantiallyonebet. Approximatebinarypayoff halfKelly=.5*(p/avgloss-(1-p)/avgwin) fromtrainingclose-screen:dailyBTC.183084,ETH.768059,SOL1.837705 (uncapped);hourlyallnegative. BTCdailytrainingwin28.889%,avgwin5.2172%,avgloss2.0640%;ETH29.268/7.3001/2.6074%;SOL45.833/4.8713/3.0979. Counterexample positiveplug-in withsmallnonstationarysample andnegativeholdout doesnotestablishedge; two-outcomeapproximationignoresdistribution. ActualKellyallocation0becausecandidateentryabsentandcontrolrejected; anyfutureprobe <=29.785/current119.14fund/212.7475accountpositioncap. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; observations26572–95.

Vectorbt: long when close exceeds prior20calendar-day CLOSE maximum, exit when false, following exact OPEN; hourly480bars. Train2024–25 (SOLstartsAug26), cashresetJan1, known2026holdout throughSep12; no claim of untouched new OOS. Fees25bp eachside creditedasset equivalentbuyfee .0025/.9975, slippage10bp/30bp stress. DailyOHLC from24completehours; missingcalendarwindows remainunavailable, no filling. Baseline train return-72.508282%,DD73.420707%,N133,win15.789%; holdout return-32.088896%,DD32.422131%,N55,win18.182%,open0;stress-45.500115%. HIGHthreshold control holdout-13.635728%N24. All15/20/30day xCLOSE/HIGH xcostcells recorded26572–83,knownholdoutrange-54.227to-13.574%;none selected. Premise: same-close grossCLI gains are not evidence for costed next-open rule. Counterexample daily SOL remainspositive; hourly disagreement and missing required3R/partial/15minpolicy prevent adoption. This short-hold toggle is NOT prior10LOWexit Donchian or managedETHpullback. Conclusion reject as live replacement. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .

Vectorbt: long when close exceeds prior20calendar-day CLOSE maximum, exit when false, following exact OPEN; hourly480bars. Train2024–25 (SOLstartsAug26), cashresetJan1, known2026holdout throughSep12; no claim of untouched new OOS. Fees25bp eachside creditedasset equivalentbuyfee .0025/.9975, slippage10bp/30bp stress. DailyOHLC from24completehours; missingcalendarwindows remainunavailable, no filling. Baseline train return10.433149%,DD20.817749%,N24,win45.833%; holdout return18.625863%,DD14.948472%,N11,win45.455%,open0;stress13.519467%. HIGHthreshold control holdout8.392065%N7. All15/20/30day xCLOSE/HIGH xcostcells recorded26572–83,knownholdoutrange3.995to18.626%;none selected. Premise: same-close grossCLI gains are not evidence for costed next-open rule. Counterexample daily SOL remainspositive; hourly disagreement and missing required3R/partial/15minpolicy prevent adoption. This short-hold toggle is NOT prior10LOWexit Donchian or managedETHpullback. Conclusion reject as live replacement. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .

Vectorbt: long when close exceeds prior20calendar-day CLOSE maximum, exit when false, following exact OPEN; hourly480bars. Train2024–25 (SOLstartsAug26), cashresetJan1, known2026holdout throughSep12; no claim of untouched new OOS. Fees25bp eachside creditedasset equivalentbuyfee .0025/.9975, slippage10bp/30bp stress. DailyOHLC from24completehours; missingcalendarwindows remainunavailable, no filling. Baseline train return-69.973193%,DD70.504557%,N197,win17.766%; holdout return-20.782205%,DD25.913218%,N45,win13.333%,open0;stress-33.831787%. HIGHthreshold control holdout-11.419497%N21. All15/20/30day xCLOSE/HIGH xcostcells recorded26572–83,knownholdoutrange-38.410to-9.158%;none selected. Premise: same-close grossCLI gains are not evidence for costed next-open rule. Counterexample daily SOL remainspositive; hourly disagreement and missing required3R/partial/15minpolicy prevent adoption. This short-hold toggle is NOT prior10LOWexit Donchian or managedETHpullback. Conclusion reject as live replacement. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .

Vectorbt: long when close exceeds prior20calendar-day CLOSE maximum, exit when false, following exact OPEN; hourly480bars. Train2024–25 (SOLstartsAug26), cashresetJan1, known2026holdout throughSep12; no claim of untouched new OOS. Fees25bp eachside creditedasset equivalentbuyfee .0025/.9975, slippage10bp/30bp stress. DailyOHLC from24completehours; missingcalendarwindows remainunavailable, no filling. Baseline train return5.185949%,DD31.646448%,N41,win29.268%; holdout return-21.109976%,DD29.758847%,N13,win15.385%,open1;stress-25.256777%. HIGHthreshold control holdout-16.281467%N8. All15/20/30day xCLOSE/HIGH xcostcells recorded26572–83,knownholdoutrange-25.257to-7.238%;none selected. Premise: same-close grossCLI gains are not evidence for costed next-open rule. Counterexample daily SOL remainspositive; hourly disagreement and missing required3R/partial/15minpolicy prevent adoption. This short-hold toggle is NOT prior10LOWexit Donchian or managedETHpullback. Conclusion reject as live replacement. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .

Vectorbt: long when close exceeds prior20calendar-day CLOSE maximum, exit when false, following exact OPEN; hourly480bars. Train2024–25 (SOLstartsAug26), cashresetJan1, known2026holdout throughSep12; no claim of untouched new OOS. Fees25bp eachside creditedasset equivalentbuyfee .0025/.9975, slippage10bp/30bp stress. DailyOHLC from24completehours; missingcalendarwindows remainunavailable, no filling. Baseline train return-78.978048%,DD78.978048%,N208,win14.423%; holdout return-25.349461%,DD26.444395%,N57,win12.281%,open0;stress-40.568953%. HIGHthreshold control holdout-12.189031%N28. All15/20/30day xCLOSE/HIGH xcostcells recorded26572–83,knownholdoutrange-40.569to-8.841%;none selected. Premise: same-close grossCLI gains are not evidence for costed next-open rule. Counterexample daily SOL remainspositive; hourly disagreement and missing required3R/partial/15minpolicy prevent adoption. This short-hold toggle is NOT prior10LOWexit Donchian or managedETHpullback. Conclusion reject as live replacement. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .

Vectorbt: long when close exceeds prior20calendar-day CLOSE maximum, exit when false, following exact OPEN; hourly480bars. Train2024–25 (SOLstartsAug26), cashresetJan1, known2026holdout throughSep12; no claim of untouched new OOS. Fees25bp eachside creditedasset equivalentbuyfee .0025/.9975, slippage10bp/30bp stress. DailyOHLC from24completehours; missingcalendarwindows remainunavailable, no filling. Baseline train return-2.113871%,DD26.189310%,N45,win28.889%; holdout return-10.194852%,DD17.371016%,N18,win22.222%,open0;stress-16.433561%. HIGHthreshold control holdout5.123294%N8. All15/20/30day xCLOSE/HIGH xcostcells recorded26572–83,knownholdoutrange-17.855to5.689%;none selected. Premise: same-close grossCLI gains are not evidence for costed next-open rule. Counterexample daily SOL remainspositive; hourly disagreement and missing required3R/partial/15minpolicy prevent adoption. This short-hold toggle is NOT prior10LOWexit Donchian or managedETHpullback. Conclusion reject as live replacement. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .