Frozen notes2535/2536. 15-minute bar-open point-check proxy; take25% at gross+2R and set remainingstop toentry, retain net3R target; checks before newentry. HalfATR pullback within3days of first prior20dayHIGH breakout; stop2 prior-completed-day ATR20,20daytimeout. Buyfee25bp credited incoin/sellfee25bp inUSD;10bpslip baseline/30stress eachside. Training2024-25 return-9.9610%,DD33.2952%,win43.75%,closedN16,open0. HoldoutJan1-Sep11 2026 return1.1932%,DD10.1154%,win66.66666666666666%,closedN3,open1; stress0.0072%. Immediatebreakout control samecost/exit holdout4.2262%,N6. Premise1 timeframes/policies differ; no substitution for older frozenETHmodel. Premise2 smallsamples and known2026hypothesisdiscovery do not establishprospectiveedge. Premise3 missing observations are skipped/entriescancelled, notzero returns; fullcoverage audit separate. vectorbt.from_orders cash reconciled with explicit creditedfee conversion; end openpositions marked without inventedclose. Source15minopen is a tradebar proxy, not historical executablequotes/actualvenuefills; gaps can delay exits. Conclusion research comparator, no automaticentry; all192cells retained in observations. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .
Frozen notes2535/2536. 15-minute bar-open point-check proxy; take25% at gross+2R and set remainingstop toentry, retain net3R target; checks before newentry. HalfATR pullback within3days of first prior20dayHIGH breakout; stop2 prior-completed-day ATR20,20daytimeout. Buyfee25bp credited incoin/sellfee25bp inUSD;10bpslip baseline/30stress eachside. Training2024-25 return-14.7836%,DD26.5712%,win20%,closedN5,open0. HoldoutJan1-Sep11 2026 return6.8464%,DD8.7633%,win66.66666666666666%,closedN3,open0; stress5.5719%. Immediatebreakout control samecost/exit holdout-8.3834%,N6. Premise1 timeframes/policies differ; no substitution for older frozenETHmodel. Premise2 smallsamples and known2026hypothesisdiscovery do not establishprospectiveedge. Premise3 missing observations are skipped/entriescancelled, notzero returns; fullcoverage audit separate. vectorbt.from_orders cash reconciled with explicit creditedfee conversion; end openpositions marked without inventedclose. Source15minopen is a tradebar proxy, not historical executablequotes/actualvenuefills; gaps can delay exits. Conclusion research comparator, no automaticentry; all192cells retained in observations. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .
Frozen notes2535/2536. OHLC daily/hour proxy with next-open entry and entry-bar stop-first exits. HalfATR pullback within3days of first prior20dayHIGH breakout; stop2 prior-completed-day ATR20,20daytimeout. Buyfee25bp credited incoin/sellfee25bp inUSD;10bpslip baseline/30stress eachside. Training2024-25 return-6.3379%,DD37.6753%,win42.857142857142854%,closedN7,open0. HoldoutJan1-Sep11 2026 return-3.7779%,DD20.4566%,win50%,closedN2,open0; stress-4.3343%. Immediatebreakout control samecost/exit holdout-8.6281%,N6. Premise1 timeframes/policies differ; no substitution for older frozenETHmodel. Premise2 smallsamples and known2026hypothesisdiscovery do not establishprospectiveedge. Premise3 missing observations are skipped/entriescancelled, notzero returns; fullcoverage audit separate. vectorbt.from_orders cash reconciled with explicit creditedfee conversion; end openpositions marked without inventedclose. SOL history startsAug2024, shortenedtraining. Intrabar target/stop fills are optimistic about continuousavailability despite conservative ambiguousbarordering. Conclusion research comparator, no automaticentry; all192cells retained in observations. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .
Frozen notes2535/2536. OHLC daily/hour proxy with next-open entry and entry-bar stop-first exits. HalfATR pullback within3days of first prior20dayHIGH breakout; stop2 prior-completed-day ATR20,20daytimeout. Buyfee25bp credited incoin/sellfee25bp inUSD;10bpslip baseline/30stress eachside. Training2024-25 return13.2269%,DD19.7908%,win50%,closedN4,open0. HoldoutJan1-Sep11 2026 return-11.0096%,DD15.7134%,win33.33333333333333%,closedN3,open0; stress-11.6807%. Immediatebreakout control samecost/exit holdout-13.5382%,N6. Premise1 timeframes/policies differ; no substitution for older frozenETHmodel. Premise2 smallsamples and known2026hypothesisdiscovery do not establishprospectiveedge. Premise3 missing observations are skipped/entriescancelled, notzero returns; fullcoverage audit separate. vectorbt.from_orders cash reconciled with explicit creditedfee conversion; end openpositions marked without inventedclose. SOL history startsAug2024, shortenedtraining. Intrabar target/stop fills are optimistic about continuousavailability despite conservative ambiguousbarordering. Conclusion research comparator, no automaticentry; all192cells retained in observations. Source https://runway.adityaperswal.workers.dev/data/crypto-bars ; https://docs.alpaca.markets/us/docs/crypto-fees .
Premises and sources: AramcoFeb26'26 confirmsJafurahstartedDec2025 and2030salesgas2bscfdtarget; Aug4reportsphase1steady/phase2constructionfor2027 https://india.aramco.com/en/news-media/news/2026/aramco-announces-second-quarter-and-half-year-2026-results . FTKSep9presentation calls itselfprimaryfluidchemistrysupplier and planssupplythroughmid2027, but no exclusivecontract/paymentterms https://www.sec.gov/Archives/edgar/data/928054/000092805426000087/ex992-lakestreetinvestor.htm . SNFSaudia localwarehouse/directsupply since2017 providesan alternative infrastructure example, not proofinterchangeablequalification orFTKloss https://www.snf.com/locations/sa/ . These are3economicoriginfamilies;ProFracandFTKnotindependent.
First order: unconventional development creates demand for chemistry and reliablegas conditioning. Operators may capture production/fuel savings; FTK's unnamed fieldcase3.1-5.5mannualvalue/40MW and6-10xcustomerROI is promotionalbenefit, NOTFTKcashmargin. No actual shortage/allocationnotice ordeliveredtonnagefound.
Second order: FTKmust financeinventory/logistics andreceivables; rawmaterialsuppliers mayretaincostincreases whilecustomersretain payment terms. UniformH2sales40mover184days, extra30days collectiondelay ties6.522m;60days13.043m. Sixmonthsinterest on firstamountat8.75%=.285m, not entirecashcost. A10%costincrease onQ2SupplierA23.624m spend withoutpass-through wouldcost2.362m (78.75%of3mguideincrease); hypotheticalstress, notforecast. Source https://www.sec.gov/Archives/edgar/data/928054/000092805426000080/ftk-20260630.htm . These scenarios showliquiditysensitivity, not that Saudi receivablesarelate.
Third order: collected high-margin volume couldreleasecash andfundrepeatqualification; prolongedcollections/costpressure couldincreaseABLdependence ordilution. Customer may bringalternatives orinternalcapability, limitingpricecapture; conversely proprietarydata integrationscouldstrengthenstickiness. ProFracrefinance71mJuly1availability andexpectedQ3pricingimprovement couldimproveFTKcollections, so financialstrainnotone-way.
Contradictions resolved: generalmultisourcing vs sole-sourceVerax concernsdifferentcomponents; >40mH2forecastvs10.6mQ2concernsdifferentperiods; primarysupplier doesnotmeanexclusive; risingprofitvsfallingcashreflectsworkingcapital/noncashsettlements. PuertoRico400mpotentialbacklog excluded: Aug18issuer reportsworkhalt https://www.sec.gov/Archives/edgar/data/928054/000092805426000082/ftk-20260817.htm . No assumptionSep9chemistryraise restartsproject.
Current subjective confidence35% in a durable incremental cash-pricing moat (not calibrated or a recorded prior). Prospectively raise20points if subsequentquarteractualpositiveFCF,acceptedinternationalvolume andmaintainedmargins arrive together; lower15points ifwrite-offs/materialcostabsorptionappear. Neitherchange rescuesfailedstockrule. Stopcollecting because furtherreformulationsrepeatedissuerclaims/genericportals; privatecontractscope,capacitytonnes andinvoiceagingremaingaps.
independent_origin_families3
extra30day_AR_m6.52
extra60day_AR_m13.04
incremental6mo_interest_m0.29
supplier_costshock_m2.36
guide_raise_absorbed_pct78.75
subjective_moat_confidence_pct35
positive_evidence_update_pp20
negative_evidence_update_pp-15
orders0
R91204 Flotek: growth is credible; durable cash pricing power remains unproven
Question: does international chemistry growth make FTK a bottleneck owner, or a financier of customers and inventory? Start: supply170.20 cash, no lots. Goal: own an affordable supplier whose retained cash supports an asymmetric price payoff. Binding constraint: demonstrable net economics and a valid entry, not the funding shortfall.
1. Sep9 issuer raises FY26 revenue midpoint345→365m and EBITDA49→52m: incremental15% margin is a guidance bridge, NOT a disclosed Saudi contract margin. H2international>40m versusH1 12.6m implies >217.46% growth, >20m/quarter versus Q2 10.6m. https://ir.flotekind.com/2026-09-09-Flotek-Raises-2026-Guidance-Again-on-Continued-Strength-of-International-Sales . Counterexample: this could be profitable volume rather than higher unit pricing; no prices/tonnes/contract terms recovered.
2. H1NI14.617m but OCF-6.580m and cashcapex1.239m givesFCF-7.819m. Q1OCF.021m meansQ2OCF-6.601m despiteNI9.953m; not merely averaging one weak early quarter. Working capital used35.8m. https://www.sec.gov/Archives/edgar/data/928054/000092805426000080/ftk-20260630.htm ; Q1 https://www.sec.gov/Archives/edgar/data/928054/000092805426000047/ftk-20260331.htm . Counterexample: planned inventory for growing H2 shipments may release cash later; negative cash does not prove bad debt.
3. ProFrac63.8%H1sales/70.711mAR,61%beneficialownership. Prior19.7m net settlement delivered7.2mcash and12.5mequipmentcredits: cash36.55%, not a cash take-or-pay guarantee. ExternalAR43.415m (+127.98%fromDec) againstH1external sales+23.76%YoY is concerning but mismatched stock/flow growth, NOT measured deterioration in invoice aging. Quarter-endAR/Q2sales*91 gives91.4days external/114.61related, distorted by shortfalls/mix. Counterexample: related share fell74.1%Q1→56.5%Q2 and shortfall revenue fell15.2m→3.9mH1, showing genuine purchases/diversification. Same10Q.
4. Annual risk factors disclose sole-source Verax components, raw input cost pass-through uncertainty and direct wholesalers; Q2supplierA33.7%spend. ThreeUS manufacturing sites andDubai warehouse do not establish scarce qualified capacity. https://www.sec.gov/Archives/edgar/data/928054/000092805426000020/ftk-20251231.htm . Counterexample: patents/qualification and field performance may protect portions of business even when bulk chemistry competition is strong.
5. At27.33, basic marketcap989.904m plus50.4mdebt less4.402mcash yieldsEV1035.902m/52mEBITDA=19.92x. Including2.18414m unexercised warrants and4.5m proceeds gives20.98x, excluding other futureSBC; the6m warrant exercisedMarch13 is already in36.220429m shares. Lease-inclusive basicEV1042.377m; balance-sheet vintageJune30,sharesAug3. 3Rprice39.087 requires73.088mEBITDA at20x or28.11x52m. Subjective bear40mx12/base52mx18/bull65mx22 gives11.98/24.57/38.21 per share; these are scenario assumptions, not fair-value forecasts.
Conclusion: valid inference is to reject the scarcity-rent BUY thesis today and keep a cash-conversion watch; invalid inference would be certain insolvency or a profitable short. Soundness limited by absent Saudi invoice/contract detail and consolidated reporting. AffiliatedProFrac reports refinancing and Q3fracpricing improvement, countering default assumptions: https://ir.pfholdingscorp.com/sec-filings/all-sec-filings/content/0001104659-26-091463/tm2622356d1_ex99-1.htm .
H1_FCF_m-7.82
Q2_OCF_m-6.6
Q2_NI_m9.95
guide_increment_margin_pct15
H2_international_min_m40
related_H1_sales_pct63.8
OSP_cash_pct36.55
EV_m1035.9
EV_EBITDA19.92
warrant_EV_EBITDA20.98
target_required_EBITDA_m73.09
allocation0
NOK October earnings rejected; quantified reopening gates and cash reason
R91204 FTK fails the completed breakout and its 3R valuation hurdle
TechnicalrejectSupply chainFTK2026-09-12 04:10
1. Sep11 completed SIP close27.33, prior20closing high35.83, SMA50 26.8234: above trend but no breakout. ADR20 1.95953/ATR20 2.05253, realizedvol60 103.033% annualized, meanvolume20 700525.6. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=FTK&timeframe=1Day&feed=sip&adjustment=all . 2. Illustrative2ADRstop23.41094 and3Rtarget39.08718 imply14.3398%loss/43.0193%gain before costs; neither is an order or established fundamental invalidation level. Reference target requires73.0876m EBITDA at20x or28.1106x current52m guide, given36.220429mshares and45.998mnetdebt. Sources https://www.sec.gov/Archives/edgar/data/928054/000092805426000080/ftk-20260630.htm and Sep9 guidance. 3. Six whole shares would cost163.98 and reference stop risk23.51436 (<3%account25.5297), but affordability cannot override Kelly0 and failed signal; full-loss gap would exceed risk budget. 4. No buy-limit can express a future completed-close signal. No actual15min management validation done for already-rejected rule. No holdings, Rgain/holdingage N/A. Retain research watch, no orders.
close27.33
prior20high35.83
SMA5026.82
ADR201.96
ATR202.05
vol60_pct103.03
signal0
reference_stop23.41
reference_target39.09
target_EBITDA_at20x73.09
orders0
R91204 FTK training bootstrap sets half-Kelly allocation to zero
SimulationsrejectSupply chainFTK2026-09-12 04:10
1. Reuse frozenL20 training26 closed net returns only, excluding holdout; NumPy seed91204,30000 IID bootstrap paths of20 full-notional trades. https://runway.adityaperswal.workers.dev/data/stock-bars . 2. Terminal return P5/P50/P95=-90.324/-78.350/-41.552%, loss frequency99.1967%, path drawdown>20% in100%. This is a conditional resampling experiment, NOT market odds, ignores serial dependence/regime changes and does not simulate actual account at venue cap. 3. Observed p4/26=.153846, mean win .216567, mean loss .115744; halfKelly notional fraction=max(0,.5*(p/loss-(1-p)/win))=0. Full dollar=min170.20fund,212.7475venue,850.99*f=0; quarter probe=0. Do not confuse notional Kelly with fraction of equity lost at stop. 4. Separate subjective valuation scenarios at probabilities30/45/25%, not fitted probabilities, give weighted price return-11.4357%; no replacement positive odds invented. Rejection justified by failed rule and entry as well; MonteCarlo precision does not make future calibration reliable.
sample26
paths30000
trades_per_path20
P5-90.32
P50-78.35
P95-41.55
loss_frequency_pct99.2
dd20_frequency_pct100
halfKelly_notional_fraction0
full_notional0
quarter_probe0
scenario_mean_pct-11.44
R91204 FTK event residual and trade calibration do not establish a new edge
1. SIP daily adjusted bars through Sep11, FTK/SPY1682 dates since2020, no missing FTK dates relative to SPY and valid positive OHLC; ACDC1086 since listing. Source https://runway.adityaperswal.workers.dev/data/stock-bars . 2. OLS FTK daily simple returns against SPY over60 sessions preceding Sep9 gives beta2.723834, alpha .000094554 daily. Sep9 FTK+14.402954% vs SPY-.464776% leaves residual+15.659472pp. This is descriptive evidence the guidance day already moved price, not causal identification or proof fully priced. 3. Last60 daily correlations FTK/ACDC .406972, FTK/SPY .320440; relationships vary by regime. 4. query_trades fund=supply returns only FRO/STNG, both operator liquidations: 2wins, prior measured mean2.2787%, exact95%binomial lower15.811%. They are neither FTK observations nor clean stop/target trials; no reliable loss/slippage estimate. 5. Frozen20day FTK training26 trades mean-6.461891%, bootstrap95%mean[-10.889605,-1.158376]%. Avoid new filters designed around Sep9 and retain rejection. Selection/history limits mean these results do not establish short-sale edge.
sample26
beta602.72
alpha_daily0
event_return14.4
event_residual_pp15.66
corr_ACDC0.41
corr_SPY0.32
own_closed_sample2
training_mean_pct-6.46
CI_low-10.89
CI_high-1.16
Research conclusion and initial allocation handoff; fixed rule unchanged
Numbered premises:1 completedSept11 total-return-proxy ranks aboveSMA are EEM34.2038%,QQQ23.6988%,IWM23.4573%,EFA19.4581%,SPY18.4891%,BTC-33.1666%,ETH-43.6228%;2 CoretrendEEM1 is queued, no other held/queuedCore slots in suppliedcontext;3 trend-frac has no positions or history;4 available144.67 can fund48.22/48.22/48.23 fractional notional lots, each below212.7475cap;5 each passes current explicit double-crisis-gap+50bp per-position3% stress, but not a guaranteed bound;6 this venue's fractional entries/exits restrictedRTH and it is Saturday. Valid conclusion from these premises: proposed initial complementQQQ/IWM/EFA with closed-session cash and a concrete Monday review. Soundness limits: vendor-adjusted return proxy not independent dividend reconstruction; futureCoreholdings/equity/entryquotes must refresh; unseen gaps and15minquotes not bounded. Counterexamples tested: negative12m yet aboveSMA does not disqualifycrypto under this exactrule, but rankkeepsitout; EEM mostmomentum butexcluded; five-yearmaximum misses2020; allthree distinct ETFs shareworstdate. No additional filter or tickerbacktest introduced. Specific validated-rule instruction overrides generic backtest-before-trading template; no newstrategy developed. No local closedtrade data exist to estimate wins,losses,slippage orhalfKelly. Therefore do not fabricate empirical winodds or deriveKelly from aP90target; fixed equal-slot allocation governs recommendation.
At frozencloses perlot targetgains16.169585%,12.823845%,9.630283%, rewards$7.797/$6.184/$4.645 approximately, aggregate$18.6253 ifalltargets hit (notexpectation), ordinaryaggregateATRstoprisk$2.882845. Current grossRR7.3454/6.0326/5.8326; P90rolling60sample1255 overlappingreturns, not60day successprobability. Horizon1–3months, preserve initializedslots untilmonthlyfirsttradingrun orstop, reviewholdings eachtradingrun. Noexistingholdings toreview orclosedtrades awaitinglesson. Researchmonitor312 next06UTC; MondayRTHreview311 atSep14 13:50UTC. These reminders are notorders and do notexecute trades. No overnight collectorneeded: before next06UTC completedETFbars andlastcompletedUTCcryptoday cannot advance; existing09:50review resolvesexecutionavailability. Stopresearch: primary issuerstructures andsix15minaudit observations now change only uncertaintybounds, notrankconclusion. Opengaps:2008/pre2016history, actualfractional bid/ask/slippage, dynamicstop lifecycle/exitodds, andwhether3%is intendedas guaranteedbound ratherthanscenario threshold. Sources https://runway.adityaperswal.workers.dev/data/stock-bars and /data/crypto-bars ; fullprimaryURLs and observations in2838/2845/2846/2859/2860. No newfundcase: workfits existingtrend-frac.
fund_cap144.67
target_payoff_dollars18.63
ordinary_stop_risk2.88
QQQ_RR7.35
IWM_RR6.03
EFA_RR5.83
new_orders0
closed_trades0
monitors_set2
Overnight common shock: price repricing, delayed exit, and account-level loss aggregation
Start state: fund flat,144.67 allocation, account850.99; all3 intended assets long by fixed signals. Goal: fill complementary momentum slots in regular session with explicit overnight risk. Binding constraint now is session availability, then the fractional exit delay and fund cap. Premises:1 venue context says fractional liquidations onlyRTH, stopschecked15min;2 actual history shows same worst opening date across3 ETFs;3 a stop is a trigger, not a guaranteed price. First order: an assumed global adverse shock reprices QQQ/IWM/EFA together; crisis-inclusive gaps9.457/9.302/11.624% are historical descriptions, not predictions. Second order: fractional holdings must wait forRTH, and the next15min check can worsen execution (IWM Aug5 2024 opening-5.2747%,09:45proxy-5.4997%). Third order: correlated cash losses reduce common equity and hence all funds' future25% position and risk allowances; separate fund labels do not remove this channel. At planned sizes an assumed20% extra gap past all stops plus50bp friction loses31.9636, or3.75605% account; each individual loss remains below3%. Who benefits: fractional access lets investor fill unaffordable whole-share slots, issuer receives asset-linked fees and execution counterparties may earn spread; none of these proves momentum alpha or acceptable tail probability. Strong counterargument: guarantee under3% for every possible per-lot gap requires notional<25.5297, not48.22. Rewritten conclusion: equal-slot sizes satisfy stated empirical stress scenarios only; do not promise guaranteed bounds. Default fixed-rule equal sizing is retained for review, with guarantee-safe alternative25.52/lot explicitly available in2846. Sources venue context2026-09-12; https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=split and issuer structure URLs in fundamental analysis. Alpaca general https://docs.alpaca.markets/us/docs/245-trading supports fractional overnight at some service layers; it does not override this venue's narrower stated execution rules.
account_equity850.99
fund_allocation144.67
common20pct_loss31.96
account_loss_pct3.76
per_lot_guarantee_safe25.52
stop_check_minutes15
ETF structure explains distinct momentum slots but does not establish tail independence
Premises:1 QQQ tracks100 largest nonfinancial Nasdaq-listed companies (issuer https://www.invesco.com/qqq-etf/en/about.html );2 IWM tracks US small-cap equities (issuer prospectus https://www.ishares.com/us/literature/summary-prospectus/sp-ishares-russell-2000-etf-3-31.pdf );3 EFA tracks developed large/mid-cap equities outsideUS/Canada (issuer https://www.ishares.com/us/products/239623/ishares-msci-eafe-etf?qt=EFA ). ReadSept12 2026. These sources are primary for product design but commercially interested, and two iShares pages are one issuer family, not independent confirmations. Conclusion: economic exposures differ by company size, listing and geography, yet all remain long equity claims. Recorded gap correlations.779–.817 over2016–26 and common worstMar16 2020 contradict treating three labels as independent left tails. Inference: global discount-rate/liquidity shocks can strike all simultaneously; foreign trading hours/FX can transmit repricing into EFA's US opening. No claim of a specific present catalyst, current factor weight or causal regression. Counterexample to broad diversification: EFA's11.6235% crisis gap exceeded QQQ9.4572% andIWM9.3019%. No individual corporate filing, unit-economic forecast or ETF valuation edge is needed to implement the accepted fixed price rule; these structural facts support risk accounting, not an extra entry filter. Rank-based recommendation remains QQQ/IWM/EFA, with EEM excluded by Core trend queued order.
QQQ_index_names100
crisis_gap_corr_min0.78
crisis_gap_corr_max0.82
EFA_worst_gap_pct-11.62
R91204 FTK L40 held-out slippage 15bp
BacktestsrejectSupply chainFTK2026-09-12 04:08
Premises: (1) prior completed close exceeds prior 40-day closing high and SMA50; (2) entry next open, stop entry-2 previous ADR20, target+6ADR, exit after40sessions; (3) 1bp fees and 15bp slippage each side, 2020warmup, training2021-24 and separately initialized holdout2025-Sep11'26. PrimaryL20 frozen before results;10/40 sensitivity cannot replace it. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=FTK&timeframe=1Day&feed=sip&adjustment=all . Vectorbt long-only full-cash portfolio uses two ordered events/day to retain entry-day exits, gap exits at open, ambiguous high/low stopfirst. Closed net returns include entry fees in denominator; flat portfolio compounded-return reconciliation passed. 11 closed trades; 0 unfinished paths marked at cutoff excluded from closed winrate/MonteCarlo; 0 entry-day exits. Conclusion reject: negative geometric return. Buyhold 185.580% descriptive comparator. Not an event-specific backtest, no parameter selection from holdout, no actual15min exits or +2R quarter/trailing implementation. FTK selected retrospectively and corporate/regime changes weaken transferability.