in10x//theses
KEPT AUTOMATICALLY//NOTHING EDITED AFTER THE FACT

RECEIPTS

This is the boring page, and that is the point. The machine writes down every pick the morning it goes out, and writes down what it expects from a company before that company reports. Then it comes back and marks its own homework. You do not have to take my word for any of it. Check the dates. Records start 2026-07-22.

The calls that aged badly stay here next to the ones that worked. The page is written by the same pipeline that writes the rest of the site, so it cannot be tidied up later, by me or by anyone.

picks written down 22 across 10 bets, records start 2026-07-22
times a pick changed 12 every switch logged with the reason given at the time
earnings marked 5 what we expected, written before the company reported
all picks vs QQQ -3.7% vs -1.3% every pick held in equal size, day by day, with short bets counting a fall as a gain

// the pick index · what holding every pick, in equal size, looks like

0%+2%-11%07-2108-03
holds each bet's current pick from the morning it is named · the same amount in each · a short pick counts a falling price as a gain · measured close to close, with no trading costs · the line to beat is QQQ, the big tech index

// the pick ledger · every stock the machine has named

picknamedpriced fromlatest pick move QQQ same stretchalpha
$NOW 2026-07-23 95.46 114.19 +19.6% -0.7% +20.4%
$INTU 2026-07-22 → 2026-07-23 289.92 284.47 -1.9% -0.5% -1.4%
picknamedpriced fromlatest pick move QQQ same stretchalpha
$TSLA 2026-07-22 378.93 322.13 -15.0% -1.3% -13.7%
picknamedpriced fromlatest pick move QQQ same stretchalpha
$PLTR 2026-07-28 131.53 125.91 -4.3% +2.6% -6.9%
$ACN 2026-07-23 → 2026-07-28 140.09 154.06 +10.0% -3.3% +13.3%
$PLTR 2026-07-22 → 2026-07-23 132.66 124.57 -6.1% -0.5% -5.6%
picknamedpriced fromlatest pick move SMH same stretchalpha
$ALAB 2026-07-28 282.52 320.98 +13.6% -0.6% +14.2%
$MRVL 2026-07-22 → 2026-07-28 207.96 189.17 -9.0% -6.1% -3.0%
picknamedpriced fromlatest pick move QQQ same stretchalpha
$MSTR 2026-07-27 91.67 94.80 +3.4% +2.3% +1.1%
$BTC-USD 2026-07-24 → 2026-07-27 65,045 65,340 +0.5% -1.1% +1.6%
$MSTR 2026-07-22 → 2026-07-24 101.95 93.63 -8.2% -2.4% -5.8%
picknamedpriced fromlatest short move QQQ same stretchalpha
$TSLA 2026-07-23 374.01 322.13 +13.9% -0.7% +14.6%
picknamedpriced fromlatest pick move SPY same stretchalpha
$SYF 2026-07-28 75.31 77.83 +3.3% +2.5% +0.8%
$CROX 2026-07-27 → 2026-07-28 134.72 134.54 -0.1% +0.0% -0.2%
$MLI 2026-07-24 → 2026-07-27 63.13 63.91 +1.2% +0.1% +1.1%
$SYF 2026-07-23 → 2026-07-24 72.81 71.77 -1.4% -1.2% -0.2%
picknamedpriced fromlatest pick move XLU same stretchalpha
$CEG 2026-07-28 270.00 273.69 +1.4% -2.9% +4.2%
$VST 2026-07-27 → 2026-07-28 163.38 157.08 -3.9% -1.3% -2.5%
$CEG 2026-07-24 → 2026-07-27 275.60 274.35 -0.5% +0.2% -0.7%
picknamedpriced fromlatest pick move SMH same stretchalpha
$MU 2026-07-24 990.21 829.04 -16.3% -6.0% -10.3%
picknamedpriced fromlatest short move IGV same stretchalpha
$HUBS 2026-07-29 238.05 239.82 -0.7% +6.1% -6.8%
$MNDY 2026-07-24 → 2026-07-29 71.74 87.09 -21.4% +5.4% -26.8%

a pick is priced from the last close before it was published, about nine hours before the US market opens, so there is no picking the winner after the fact · when a pick changes, the old one is closed at that same price · buying for real at the next open would differ by whatever moved overnight · alpha is simply the pick's move minus what its own slice of the market did over the same stretch

// earnings receipts · written before the company reported, judged after

$PLTR T3 The Integration Gap reported 2026-08-03 SUPPORTS THESIS

Actuals showed U.S. commercial revenue up 149% Y/Y and total revenue up 93% Y/Y, with FY26 guidance raised to 82% revenue growth and 134% U.S. commercial growth, crushing consensus. That directly matches the expectation of durable/accelerating U.S. commercial strength and guidance language supporting commercial revenue durability. Headlines are thin on AIP bookings, RDV, and FDE vs self-serve commentary, but nothing indicates lighter-touch shift or booking shrinkage, so the print supports the integrator-moat thesis.

what we expected, written 2026-07-23, before the company reported

To support the thesis that enterprise AI value accrues to forward-deployed integrators, PLTR must show continued acceleration in US commercial AIP bookings and remaining deal value, with commentary that customers still require heavy deployment engineering rather than self-serve model access. A sequential decline in commercial bookings or explicit management language that deals are shifting to lighter-touch/self-serve implementations would undercut the integrator moat and start the kill path of two consecutive booking shrinks.

US commercial AIP bookings and sequential growthTotal remaining deal value (RDV) and commercial mixCommentary on deal cycle length and FDE/deployment intensity vs self-serveCustomer count adds and expansion vs new logo mixFY guidance language on commercial revenue durability
$MSTR T5 Debasement reported 2026-07-30 MIXED THESIS

Headlines confirm MSTR continued the accumulation mechanism by adding ~30K BTC, which maps positively to the scarce-asset wrapper role in the debasement thesis, while the large ~$8.2-8.6B net loss is attributed to bitcoin unrealized/impairment accounting rather than operational collapse. However, coverage is thin to nonexistent on the decisive capital-structure watch items (liquidity runway vs. 12-month threshold, STRC/preferred dividend status, or any deferral language), and software revenue missed at $122.4M, so the print neither clearly validates nor breaks runway viability through the stre

what we expected, written 2026-07-23, before the company reported

To support the debasement thesis, the print must confirm MSTR is still accumulating BTC as a leveraged scarce-asset wrapper without stressing the capital structure—i.e., runway comfortably above 12 months and no interruption to preferred dividends—so the corporate vehicle remains viable through the stress-test phase. Any disclosure of runway compressing toward <12 months, STRC/preferred pressure, or a pause in BTC buys would undercut or kill the thesis by breaking the accumulation mechanism.

BTC holdings change and acquisition pace/cost basisLiquidity runway / cash + credit availability vs. 12-month thresholdPreferred dividend status (any skip or deferral language on STRC or other prefs)Implied mNAV premium and management commentary on premium toleranceSoftware segment revenue growth and gross margin (operating cash contribution)
$CROX T7 Profit for Size reported 2026-07-30 MIXED THESIS

Headlines show record Q2 revenue of $1.179B (+2.6%) and a raised full-year 2026 outlook, consistent with continued cash generation rather than a fading fad, but they explicitly flag margin pressure and give no hard operating-margin, gross-margin, or FCF figures against the 20%+ / ~10%+ thresholds. Coverage is thin on the core profitability watch items, so the print neither clearly confirms durable high-margin compounding nor proves a value-trap erosion.

what we expected, written 2026-07-23, before the company reported

To support the thesis that CROX is a mispriced durable profit machine, the print must show operating margins holding near/above 20% and FCF margin remaining solidly positive (~10%+) even with flat-to-down revenue, proving cash keeps compounding at a low multiple rather than fading with the fad narrative. It would undercut or kill the thesis if margins compress materially or FCF deteriorates, confirming the cheap fwd P/E is a value trap as profitability erodes instead of re-rating on durable cash generation.

Operating margin vs 20%+ levelFCF margin and absolute free cash flowGross margin sustainability near 58%Revenue trend and any volume/ASP commentaryForward guidance on margins and cash generation
$LRCX T9 Memory Is the Real Shortage reported 2026-07-29 SUPPORTS THESIS

LRCX delivered record Q4 FY2026 revenue of $6.72B (+30%) and a $1.82 EPS beat, explicitly fueled by AI-driven memory demand with strong NAND growth, while FY2027/Q1 guidance smashed estimates and drove an 8% pop. This maps directly to the pre-stated need for accelerating memory-segment strength and robust HBM/AI-tied bookings/guidance, confirming tool demand as the binding bottleneck rather than any reversion to the old cycle. Coverage is somewhat thin on exact memory % mix or lead-time comments but the AI-memory narrative and upside outlook clearly reinforce the thesis mechanism.

what we expected, written 2026-07-24, before the company reported

To support the thesis that HBM-driven memory shortage is forcing sustained capex into etch/deposition, LRCX must show accelerating memory-segment systems revenue and robust bookings/guidance tied to HBM stack intensity, confirming tool demand remains the binding bottleneck rather than easing. A miss on memory revenue, flattish or declining memory bookings, or cautious capex commentary would undercut the mechanism by signaling HBM capacity additions or AI demand are no longer outrunning supply, reasserting the old cycle.

Memory systems revenue growth and % of total systemsHBM/DRAM-related bookings or order commentaryNext-quarter guidance for memory vs foundry/logicGross margin trajectory (tool mix/pricing power)Any sold-out or lead-time comments on advanced etch/deposition tools
$SAP T1 App Layer Eats the Model Layer reported 2026-07-23 MIXED THESIS

Headlines show current cloud backlog at ~€22.9B, up 26% and reversing two prior weak quarters, which supports the cloud ERP / system-of-record adoption leg of the expectation and the thesis that proprietary enterprise workflows remain valuable. However, EPS missed consensus, the stock fell, cloud-transition speed disappointed some investors, and operating-profit outlook was cut on M&A dilution, while coverage is thin-to-absent on Business AI/Joule attach, AI bookings/ARR, S/4HANA net-new logos, and model-partnership commentary—so AI monetization proof is missing and the print does not cleanly

what we expected, written 2026-07-23, before the company reported

To support the thesis that value accrues to proprietary enterprise data/workflow apps as intelligence commoditizes, SAP must show accelerating cloud ERP and Business AI adoption (higher AI attach, net new logos on S/4HANA Cloud, rising AI-related bookings) proving its system-of-record data is the scarce input every AI app needs. A result that undercuts or kills the thesis would be stalled AI monetization, flat/declining cloud backlog, or management signaling that frontier model vendors are disintermediating ERP workflows rather than integrating into them.

Cloud revenue growth and current cloud backlogBusiness AI / Joule attach rate and AI-related bookings or ARRS/4HANA Cloud net new customers and RISE with SAP momentumCommentary on open vs closed model partnerships and data moat durabilityFY guidance for cloud and software growth plus AI contribution
$ALAB T4 Connectivity Is the New Bottleneck reports 2026-08-04 PINNED 2026-07-23

To support the thesis that data-movement (not compute) is the binding AI constraint, ALAB must show continued hypergrowth in PCIe/CXL retimer and optical-link revenue, plus explicit evidence that CPO/optical attach is ramping on the conservative path—not just total revenue beats. A miss on connectivity segment growth, softer sequential guide, or management commentary that hyperscalers are deferring rack-scale interconnect/CPO would undercut the bottleneck shift and align with the kill path of cooling capex for data movement.

$SYM T2 Physical AI reports 2026-08-05 PINNED 2026-07-23

To support the thesis that warehouse robotics is already physical AI with platform economics (not pilot theater), SYM must show continued scale deployment traction beyond Walmart—rising system deployments, expanding software/recurring attach, and sustained FCF with margin expansion that looks more software-platform than one-off hardware. A result that undercuts or kills the thesis link would be decelerating deployment growth, flat/declining gross margins, or guidance that implies customers remain stuck in limited rollouts without profitable multi-site non-defense scale by the mid-2027 kill win

$SNDK T9 Memory Is the Real Shortage reports 2026-08-05 PINNED 2026-07-24

To support the thesis that AI datacenter SSD demand is outrunning NAND wafer supply and locking in pricing power, SNDK must show continued enterprise/data-center SSD strength, rising or stable NAND ASPs, and gross-margin expansion or hold at elevated levels, with commentary that supply remains tight rather than easing. The print would undercut or kill the thesis if enterprise SSD growth decelerates, NAND contract/spot pricing rolls over, margins compress, or management signals wafer additions and inventory rebuild that reassert the old cycle.

$HUBS T10 AI Eats Seat-Based Software reports 2026-08-05 PINNED 2026-07-24

To support the thesis that AI agents are hollowing out seat-based mid-market CRM, HUBS must print stalling or contracting seat growth and slipping NRR, with management citing AI workflow automation as pressure on expansion or guiding lower on seats/revenue. Stabilization or re-acceleration in seats and NRR for another quarter, or credible traction/repricing toward usage/agent models that lifts growth and beats, would undercut or kill the thesis signal in this name.

$CEG T8 Power Is the Next Bottleneck reports 2026-08-06 PINNED 2026-07-24

To support the thesis that nuclear is capturing AI-driven power scarcity, CEG must show new or expanded multi-year hyperscaler PPAs at premium prices, stable-to-rising nuclear fleet output/availability, and constructive guidance on contracted power margins rather than merchant weakness. A kill signal would be absence of incremental AI/data-center deals, PPA cancellations or push-outs, softer capacity/pricing commentary, or guidance that implies power demand is not tightening versus the prior quarter.

$VST T8 Power Is the Next Bottleneck reports 2026-08-07 PINNED 2026-07-24

To support the thesis that AI is bidding up scarce generation, VST must show rising realized power prices and/or new multi-year offtake tied to datacenter demand (especially ERCOT/gas-nuclear fleet), with management affirming tight capacity and no material PPA cancellations. A miss on commercial pricing, flat/declining forward hedges, or commentary that hyperscaler demand is slowing and interconnect/PPA pipelines are shrinking would undercut the scarcity mechanism and align with kill criteria.

$RKLB T6 The Musk Premium reports 2026-08-10 PINNED 2026-07-23

To support the thesis that space-sector narrative premiums are detached from cash and set to compress, RKLB must show persistent cash burn (FCF margin still deeply negative), decelerating growth or Electron/Neutron cadence misses, and flat-to-down guidance that fails to justify 60x+ sales. A clean beat with sharply improved gross margins, positive FCF inflection, or a large Neutron backlog raise that re-rates the name higher would undercut the thesis by validating mania-level space multiples and indirectly supporting rich SpaceX marks.

$MNDY T10 AI Eats Seat-Based Software reports 2026-08-10 PINNED 2026-07-24

To support the thesis that AI agents erode per-seat work-management demand, MNDY must show decelerating or contracting paid seats, NRR slipping below prior trends, and/or guidance that explicitly cites AI-driven seat pressure or slower seat expansion. A clean re-acceleration in seat growth and stable-to-rising NRR for another quarter, or clear traction/repricing toward usage/agent-based packaging that stabilizes revenue growth, would undercut or kill the seat-compression mechanism for this name.

$COHR T4 Connectivity Is the New Bottleneck reports 2026-08-12 PINNED 2026-07-23

To support the thesis that data-movement photonics is the binding constraint, COHR must show continued acceleration in datacom/optical transceiver demand (volume and/or mix into higher-speed modules) plus concrete CPO/silicon-photonics design-win or sampling progress, with management affirming hyperscaler optics spend is holding. A miss or sharp deceleration in communications/datacom revenue, delayed CPO timelines, or explicit commentary that customers are cutting or pushing out optical capex would undercut the bottleneck thesis and align with the kill path of sustained hyperscaler spend weakn

$INTU T1 App Layer Eats the Model Layer reports 2026-08-25 PINNED 2026-08-04

To support the thesis that value accrues to proprietary data/workflow apps as intelligence commoditizes, INTU must show AI features (Intuit Assist / GenOS) lifting attach rates, ARPU, or retention inside TurboTax and QuickBooks without relying on exclusive closed-model access. Evidence of sustained pricing power and expanding AI-monetized mix from its tax/SMB data moat would confirm the app-layer capture mechanism. Results that undercut it: flat or declining AI contribution, heavy spend/dependency on a single frontier lab with no open-weight path, or guidance implying model capability—not data

// the change log · every call, in the order it happened

2026-08-04 T6 model raised a DOUBLE-DOWN flag
2026-08-03 T2 model raised a DOUBLE-DOWN flag
2026-07-31 T1 model raised a DOUBLE-DOWN flag
2026-07-31 T7 model raised a DOUBLE-DOWN flag
2026-07-29 T10 pick changed: $MNDY → $HUBS HUBS is the most extended bounce (+8.8% today, +31% 30d, +23.7% vs IGV) still at fwd P/E 15 with clear agent-product pivots that advertise seat-model risk; better short expression than already-crushed MNDY (-72% vs 52w, P/E 16) or slower/cheaper DOCU (+8.7% growth, P/E 11).
2026-07-28 T1 model raised a DOUBLE-DOWN flag
2026-07-28 T3 pick changed: $ACN → $PLTR PLTR's forward-deployed model and +84.7% revenue growth with 84.1% gross margins give the most direct exposure to the integration-gap thesis, beating ACN's cheaper 10x fwd P/E but only +5.6% growth and IBM's +1.1% growth.
2026-07-28 T4 pick changed: $MRVL → $ALAB ALAB offers the purest mechanism exposure to the AI interconnect bottleneck with +93.4% rev growth and 76% gross margin, far above MRVL (+27.6%, 51.5%) and COHR (+20.5%, 37%); at -43.4% vs 52w high it is equally washed-out yet still compounds faster than the other two despite the 62 fwd P/E.
2026-07-28 T7 pick changed: $CROX → $SYF SYF offers the purest expression at fwd P/E 7, -15.2% vs 52w high, with fresh beat/higher dividend/strong buybacks versus CROX (P/E 9, rev -1.7%) and richer MLI (P/E 15).
2026-07-28 T8 pick changed: $VST → $CEG CEG offers the cleanest nuclear-fleet exposure to the 10 GW hyperscaler deals, with sector-leading +63.8% revenue growth; at fwd P/E 20 and -34.6% from highs it beats higher-multiple GEV (P/E 40, only +21.9% growth) and lower-growth VST on direct mechanism fit.
2026-07-27 T5 pick changed: $BTC-USD → $MSTR MSTR offers direct levered BTC exposure at a 0.79x mNAV discount and -77.5% vs 52w high, better asymmetric expression of the stress-test entry than spot BTC (-48% from high) or STRC still trading at an 15% discount to par.
2026-07-27 T7 pick changed: $MLI → $CROX CROX best captures the thesis with 58.1% gross margin, 11.1% FCF margin and fwd P/E of 9 while only -5.7% off its 52w high; it beats SYF (0% GM, financial model) and MLI (higher 15 P/E despite +25.5% rev growth).
2026-07-27 T8 pick changed: $CEG → $VST VST offers the cleanest ownership exposure with +43.4% rev growth, 38.6% gross margin and fwd P/E 15 at only -23% off highs, beating CEG's weaker margins/FCF and higher P/E plus GEV's 41x multiple and lagging -2.9% vs XLU.
2026-07-24 T5 pick changed: $MSTR → $BTC-USD BTC-USD is the purest scarce-asset expression of the debasement mechanism, trading -49.3% vs 52w high with 30d outperformance vs QQQ of +5.1%; it beats MSTR (mNAV 0.82x discount but equity/operational overlay) and STRC (preferred at 86.75, capped upside).
2026-07-24 T7 pick changed: $SYF → $MLI MLI offers the cleanest compounding exposure with 25.5% revenue growth, 11.3% FCF margin and a 15.7% rise in adjusted operating income, still at 15x fwd P/E and 10% below its 52w high—superior to CROX's -1.7% revenue decline despite the 9x multiple, and to SYF's 7x multiple which lacks comparable high-margin FCF quality.
2026-07-24 T8 first pick named: $CEG CEG has the cleanest nuclear-generation exposure to AI power deals, posts the highest rev growth (+63.8%) and leads 30d relative performance (+1.8% vs XLU) while sitting 32.9% below its 52w high; that beats VST’s lower growth and GEV’s expensive 41 fwd P/E plus –5% relative drag.
2026-07-24 T9 first pick named: $MU MU is the purest large-cap expression of the HBM/DRAM shortage with +345.7% rev growth, 72.6% gross margin, and a 6x fwd P/E at -25.3% off highs, beating SNDK (weaker 56% GM, deeper -36.3% drawdown) and LRCX (only indirect equipment exposure, 38x fwd P/E, +23.8% growth).
2026-07-24 T10 first pick named: $MNDY MNDY is the cleanest short today: 20% workforce cut explicitly to restructure around an AI Work Platform, still at fwd P/E 14 on +24.5% growth with only +19.1% FCF margin, and -74.1% vs 52w high—more directly exposed and less de-risked than HUBS (AI-pick narrative, -64.7%) or cheaper/slower DOCU (fwd P/E 10, +8.7% growth).
2026-07-23 T1 pick changed: $INTU → $NOW NOW has the cleanest mechanism exposure today: AI ACV already >$1B, +24% rev growth and 33.7% FCF margin versus INTU +10.4% and SAP +6%, with 74.8% gross margins; still -54.6% off highs, so growth plus workflow lock-in beats INTU’s cheaper 10x fwd P/E and SAP’s scale.
2026-07-23 T3 pick changed: $PLTR → $ACN ACN is the purest integrator play, posting record $21.1B bookings with a sharp gen-AI jump, yet still at fwd P/E 10 and -51.9% vs 52w high with +18% 30d vs QQQ; that beats PLTR's richer 59x multiple and IBM's negative relative momentum and near-flat growth.
2026-07-23 T6 first pick named: $TSLA TSLA remains the cleanest short: core auto margins/FCF just cracked yet it still trades at 146x forward earnings with a $1.4T mcap and only -25% off highs; RKLB (2325x, already -54% off highs) and DXYZ (mcap $770M, -65% off highs) are smaller, less direct expressions of the cash-cow failure.
2026-07-23 T6 model raised a DOUBLE-DOWN flag
2026-07-23 T7 first pick named: $SYF SYF has the lowest fwd P/E at 7, sits 18% below its 52-week high, and just beat EPS by 24.5% while raising 2026 guidance, giving the cleanest cheap-compounder exposure versus CROX (P/E 9, only -3% from highs) and higher-multiple MLI (P/E 14).
2026-07-22 T1 first pick named: $INTU INTU best balances proprietary data/workflow moat with fundamentals: fwd P/E 10, 80.8% gross margin, and -65% vs 52w high, plus strongest 30d relative momentum (+15.3% vs QQQ) versus NOW (P/E 19, +7.1%) and slower-growing SAP (+6% rev).
2026-07-22 T2 first pick named: $TSLA TSLA has the most direct mechanism exposure via Optimus humanoid/VLA efforts versus SYM's warehouse automation or KOID's diluted ETF; despite fwd P/E 146 and -25% from highs it still shows +15.8% rev growth and 19.1% gross margin at $1.4T scale.
2026-07-22 T3 first pick named: $PLTR PLTR captures the integration-gap mechanism most directly as the AI-native platform/FDE pure-play, with +84.7% rev growth and 84.1% gross margin far outpacing ACN (+5.6% growth, 32% margin) and IBM (+9.5%, 58%); still -40% vs 52w high despite 59 fwd P/E.
2026-07-22 T4 first pick named: $MRVL MRVL best balances direct optical-DSP/CPO exposure with fundamentals: +27.6% rev growth, 51.5% gross margin, +26% FCF margin and fwd P/E 34 versus ALAB’s richer 73x and COHR’s 37% GM with −3% FCF; also deepest discount at −36% vs 52w high.
2026-07-22 T5 first pick named: $MSTR MSTR provides the most direct leveraged exposure to the BTC debasement thesis at a 0.85x mNAV discount and -76.4% vs 52w high (fwd P/E 22, 68% GM), outpacing unlevered BTC (-47.8% vs high) and STRC whose upside is capped near $100 par.

// how this page keeps itself

the dates

Picks and expectations go out on weekday mornings, 07:00 in Copenhagen, and are saved with that date. One row per day. Running the pipeline twice in a day replaces that day's row and nothing else. Older rows are never touched.

the price used

A pick is priced from the last close before it was published, which is the price anyone reading it could actually see. A short pick counts a falling price as a gain. Trading costs, dividends and the fees for borrowing shares are all left out.

the index

The pick index puts the same amount into every bet's current pick and follows it day by day against QQQ, the big tech index. It is a way of counting the calls, not a strategy you could buy at exactly these prices.

disclaimer

A public experiment in showing your work. Nothing here is investment advice. This page exists so you can judge the machine on what it did, rather than on how sure it sounds.