Key Themes

Capital allocation engine outperformance. CSU has deployed capital at 17-18% returns across a $13.6B capital base, with management distributing acquisition authority across 1,100+ businesses to scale deal volume beyond the 10-deal annual ceiling that constrains peers. Crossroads claims this structure allows CSU to redeploy $2B+ annually at attractive spreads without sacrificing returns.

Sticky vertical software fundamentals prove resilient. CSU's portfolio of niche software businesses generates $5B in recurring maintenance revenue annually with 25-year average customer lifespans and 3-8% annual price increases. Crossroads claims AI threatens the narrative around software more than the cash flows themselves, as switching costs, regulatory requirements, and workflow complexity remain prohibitive.

AI creates acquisition opportunity, not terminal value risk. Crossroads alleges the SaaSpocalypse created a buyer's market where terminal-value uncertainty forces sellers to accept lower prices while CSU benefits from both lower purchase multiples on future acquisitions and improved post-acquisition economics as AI reduces the cost to improve and support those businesses. The firm warns that fear of AI's impact has widened the spread between what CSU pays and what those businesses are worth in its hands.

Founder transition managed by decentralized culture. Mark Leonard's departure as president in September 2025 triggered an 11.5% stock decline, but Crossroads claims CSU's culture and capital-allocation authority were never reliant on a single individual. The firm asserts successor Mark Miller inherits a system already operating without Leonard, eliminating succession risk that plagued other founder-led software businesses.

Valuation offers asymmetric upside across three scenarios. At C$2,526 average cost versus C$2,895 September close, Crossroads models 14-27x FCFA2S valuations depending on AI's severity, with base-case scenarios implying $200-280 per C$100 invested over five years from cash flow compounding alone, independent of multiple expansion. The firm claims even the "software Armageddon" scenario leaves CSU capable of generating mid-teens returns through disciplined deployment of free cash flow into depressed assets.

Organic growth deceleration warrants monitoring. Maintenance and recurring revenue growth slowed to 2% in Q2 2026 from historical 5-6% range, which Crossroads acknowledges could signal AI erosion rather than temporary headwinds. The firm identifies this as a critical warning indicator requiring multi-quarter persistence before altering its thesis, alongside return-on-invested-capital deterioration below 15% or evidence of post-Leonard centralization.

Read more research at Phalanx Research.

Key Themes

Earnings quality concerns. MaxLinear's improving results are driven by accounting adjustments rather than genuine operational strength, with adjusted EPS margins too thin to absorb adverse developments.

Weak earnings cushion. Quarterly adjusted EPS beats average only 4 cents, indicating MXL's reported outperformance lacks durability and leaves the company vulnerable to earnings misses.

Unsustainable recovery narrative. The company's sharp revenue and earnings improvement masks accounting-driven results that do not reflect genuine business momentum or durable profitability.

Read more research at Phalanx Research.

Key Themes

Flawed hyperscaler narrative. Pure claims hyperscaler traction via Meta software licensing, but Meta will assemble white box hardware using direct flash purchases and ODMs, not buy Pure products. The engagement has not expanded to other Meta data centers, and prospects at other major hyperscalers are dim given differing technology stacks.

Commodity hardware advantage in AI. Ultra-high performance data infrastructure competitors like DDN, WEKA, VAST Data, and Hammerspace will dominate AI-centric storage applications by leveraging commodity hardware, leaving Pure deriving little benefit from AI spending trends.

Flawed total cost of ownership claims. Pure's Pure//E TCO analyses ignore fundamental hyperscaler storage economics. HDDs retain a 5-6x TCO advantage versus flash-based solutions, a conclusion supported by hyperscaler storage architects.

Stagnant core market with mature fundamentals. Pure's enterprise storage products are undifferentiated with stagnant market share. All-flash array (AFA) revenues are projected to grow at single-digit rates as the market matures, confined to three primary competitors trading share in a saturated segment.

Legacy software business disguised as SaaS. Pure's "software" business is largely maintenance, support, and pre-purchased upgrade revenue that does not merit SaaS valuation multiples, contrary to bull narratives.

Read more research at Phalanx Research.

Key Themes

Undifferentiated, interchangeable technology. CoreWeave possesses no proprietary technology, no defensible IP, and no competitive edge in solving power constraints. Its bare metal GPU rental services are interchangeable commodities, not differentiated solutions.

Unsustainable financial model. CoreWeave generates returns below its cost of capital, destroying shareholder value through weak unit economics and expensive debt financing. Equity holders receive nothing during contract terms and depend entirely on residual value from five-year-old chips to break even.

Extreme customer concentration risk. Microsoft represents 70% of CoreWeave's revenue and recently rejected an expansion option, instead signing a larger strategic contract with competitor Nebius. This concentration exposes CoreWeave to massive revenue risk from a single customer defection.

Debt-fueled valuation bubble. CoreWeave has inflated from a crypto mining operation to a $75 billion "AI hyperscaler" in three years through financial engineering and timing, not lasting innovation. Secured lenders capture healthy coupons while equity investors absorb obsolescence risk on aging hardware.

No defensible position in AI infrastructure. CoreWeave lacks a full-stack platform like leading clouds and serves only as a stopgap assembler for large customers. As hyperscalers scale proprietary infrastructure, CoreWeave's role becomes increasingly irrelevant.

Read more research at Phalanx Research.

Key Themes

Unsustainable premium valuation model. BMNR's strategy of selling overvalued stock to buy crypto and grow token-per-share worked when scarcity created enthusiasm, but conditions have vanished—MicroStrategy's own premium has collapsed from 2-2.5x to below 1.5x, proving the model is broken.

Massive dilutive capital raises masked as accretive. BMNR issued $10 billion in stock in three months ($170 million daily) while claiming accretion; the recent $365 million offering at $70 was actually discounted once warrant values are factored in, sacrificing credibility for cash.

Deteriorating operational disclosure. BMNR quietly removed updated share counts from weekly press releases, making ETH-per-share accretion harder to track as the NAV premium narrows and share dilution accelerates.

Destroyed competitive moat and market saturation. Over $100 billion in planned crypto-treasury capital raises and proliferating ETFs have eliminated scarcity, collapsed NAV premiums across the sector, and rendered BMNR's generic strategy commoditized and redundant.

Thesis rests on premium that must evaporate. BMNR's entire investment case depends on investors paying a market premium for Ethereum exposure when direct ETH purchase, staking, and ETFs offer superior cost, liquidity, and simplicity with no premium—ensuring BMNR trades at parity or below.

Read more research at Phalanx Research.

Key Themes

Data fabrication and selective disclosure. Fervo altered production temperature charts between 2024 and 2026 filings, omitting extreme fluctuations evident in Nevada state data. The company cherry-picked a single metric to hide a devastating March 2024 stress test where injecting 33% more water caused production temperatures to plummet 90°F and water losses to spike from 21% to 57%, results never disclosed to investors or in SEC filings.

Overstated thermal stability claims. Fervo claimed its pilot plant exhibited "consistent, stable temperature output" and "no premature thermal decline" in its IPO registration statement. Nevada state data reveals thermal decline accelerating to 24°F per year—16x above economically viable rates—with production temperatures inexplicably plummeting 90°F in a single month in June 2026, contradicting the company's narrative.

Uneconomical water loss and concealed Cape Station test results. The pilot plant exhibited 30% water loss rates, 30x higher than the sub-1% required for commercial viability. At Cape Station, Fervo ran two flow tests (Frisco and Bearskin pads) but disclosed only 24 hours of data from a 30-day test, omitting water loss metrics. Former employees allege water loss at Cape Station was "much worse than they wanted to see" and that test results were withheld because showing the full data "will put a lot of questions in the market in terms of viability."

Water rights and "non-consumptive" designation at risk. Fervo obtained Utah water rights based on a "non-consumptive" claim that water lost to subsurface leak-off will "eventually" migrate back through 6,000 feet of granite to recharge the source aquifer. Industry experts, former employees, and energy analytics firm Enverus characterize this theory as unsupported, with one expert calling it "the rosiest picture they could present to regulators." If reclassified as consumptive use, Cape Station's water access is jeopardized.

Induced seismicity causing equipment damage and water loss. Fervo CEO Latimer called seismicity a "perception issue" and claimed only "several" yellow-level events occurred. Company data reveals 58 yellow-level events with acceleration to 39 events in a single 30-day period. A magnitude 3.1 earthquake—Fervo's first red-level event—occurred 24 days ago, triggering mandatory 24-hour shutdown protocols that Fervo has not publicly disclosed. Former employees allege seismic events are causing "major problems" with well casing failures and opening pathways to deep faults, exacerbating water loss at "much higher loss ratios."

Failed pilot plant performance and scaling claims. Fervo claimed its pilot plant "achieved all technical objectives" and was "phenomenal from a technical demonstration standpoint." In reality, the pilot missed virtually every stated performance target, generating just 41% of forecasted capacity while operating at less than 50% of nameplate capacity. The company falsely claimed this failure proved its technology is "proven at scale."

Unsustainable PPA pricing and cost structure. Fervo sold 658MW of PPAs at $83-$99/MWh, below its own estimated cost of $107/MWh for new geothermal. The company claims it will reduce CapEx from $7,000/kW to $3,000/kW through a 57% reduction, but costs have risen from $5,000 to $7,000 over two years. Former employees describe the $3,000 target as a "try to get the stock price up" claim, with completions costs alone estimated at $3,000/kW—equal to Fervo's entire long-term target.

Hidden completion cost problem and drilling plateau. Fervo relentlessly highlights drilling cost reductions while concealing completion costs (hydraulic fracturing), which former employees describe as "the 300-pound gorilla in the room" and Fervo's "hidden problem." Drilling cost charts were altered between 2024 and 2026 SEC filings, omitting Well 15 data showing rising costs. An EGS expert states Fervo is "very close to the plateau" on drilling advancements with no credible path to offsetting completion cost reductions.

Transmission capacity collapse and California delivery failure. Fervo claimed in its May 2026 prospectus to have 290MW of transmission capacity with Cape Station Phase 2 delivering power to California by 2028. In April 2026, Fervo admitted to regulators there was "no viable pathway" for Utah geothermal to reach California before 2032. The company later disclosed its California transmission pathway is "being retired in June 2027," reassigning capacity to an undefined Google data center with vague "2030+" delivery timelines, effectively admitting it cannot fulfill California PPA obligations.

PPA default risk and penalty exposure. Fervo's $7.2 billion+ of PPAs include daily delay damages exceeding $500,000 weekly and default triggers for delays exceeding 180 days or capacity factors below 80% over two-year periods. The pilot plant operated at less than 50% capacity, making PPA defaults highly probable if pilot performance replicates at commercial scale.

Read more research at Phalanx Research.

Key Themes

Single-token balance sheet with no operating business. As of June 30, 2026, SDEV held $119.2 million in SKY tokens representing 94% of total assets against only $7.0 million in cash. The company operates no stablecoin network, payments platform, or software service—it is purely a token holder with staking as its only revenue line.

Undisclosed insider control and conflict of interest. CEO Michael Kazley simultaneously runs SDEV and manages R01 Fund LP, which supplied 74% of tokens in the January 2026 financing and holds 32% of warrants. Sky Frontier Foundation, which supplied tokens and stablecoins, placed its Director on SDEV's board. All four purchasers retain contractual consent rights over any change to the SKY-only investment strategy for 24 months.

Devastating capital destruction masquerading as profitability. SDEV reported $511.3 million of net income in H1 2026, entirely composed of $5.8 billion in non-cash warrant revaluation gains offsetting $5.3 billion in non-cash issuance losses. Operating activities consumed $7.8 million of cash while the company spent $84.5 million buying additional SKY tokens, leaving only five months of cash runway at current burn rates.

Massive dilution through warrant trap. The January 2026 private placement issued 167.5 million pre-funded warrants at an effective $0.85 per share (post-split) exercisable in three tranches through January 2027. Pre-funded warrant holders pay only $0.01 per share at exercise while holding embedded gains of 96%+ and face no price floor for selling. The first two tranches alone equal 166% of current common shares outstanding.

Coordinated exit structure with embedded dump mechanism. After each warrant tranche unlocks, purchasers may sell up to 10% of 30-day average daily volume daily—a limit that increases when Friday run-ups artificially inflate trading volume. SDEV's stock has repeatedly crashed the trading session after Friday rallies since January 2026. The company registered 212.9 million shares for resale as of September 2026, over four times the common stock outstanding.

Token price dependency with realized losses mounting. SDEV's SKY position is carried at $0.052 per token as of June 30, 2026—15% below the $0.0615 rate paid in the financing and 20% below its own $0.065 average purchase price. The company took a $28.0 million unrealized loss in H1 2026, including a $50.6 million loss in Q2 alone, while holding nearly all tokens in single custody through one infrastructure vendor.

Historical precedent of pharmaceutical shell manipulation. SDEV was formerly NovaBay Pharmaceuticals under CEO David Lazar, who controlled 94% of supply and orchestrated a rug pull after a short squeeze. Current insiders acquired control in October 2025 and immediately converted the shell into a token treasury, following an identical concentration and exit playbook.

Read more research at Phalanx Research.

Key Themes

AI expansion capital demands. Oracle's AI buildout creates substantial financial pressures that revenue growth alone cannot sustain, raising questions about whether capital investments will translate into actual earnings and cash flow generation.

Earnings quality concerns. Oracle's Q1'27 earnings beat masks deteriorating cash flow dynamics and earnings durability, with accounting footnotes revealing pressures hidden in headline results.

Cash conversion risk. The company's AI expansion threatens cash flow conversion, with investors facing uncertainty about whether growth will materialize as durable cash generation or remain trapped in balance sheet investments.

Read more research at Phalanx Research.

Key Themes

Misleading depreciation claims. Nvidia's assertion that GPUs retain sufficient value to justify slower depreciation schedules is propaganda designed to counter legitimate criticism about accelerating chip obsolescence and value destruction.

Historical precedent for collapse. The 1968 computer leasing bubble—where leveraged speculators claimed infinite demand and exponential earnings growth while dismissing depreciation concerns as "middle-aged"—directly parallels Nvidia's current investor messaging and presages similar market failure.

Speculative leverage masking weakness. Then as now, optimistic rhetoric about technological transformation and unlimited upside obscures the fundamental economic reality: rapidly depreciating assets purchased with thin capital cannot sustain claimed returns, and the market eventually reprices this weakness violently.

Read more research at Phalanx Research.

Key Themes

Accounting-driven profit illusion. Insperity's HRScale platform generates extra profit by reclassifying operating costs from expense to capitalization under US GAAP, representing a six-of-one, half-a-dozen-of-another accounting shift rather than genuine improvement.

Unsustainable guidance. The company's 56% midpoint adjusted EBITDA growth projection is subject to considerable uncertainty and fails to account for competitive pressures and economic headwinds.

Cost reduction offset by premium inflation. While Insperity offloads health claims volatility, this cost cutting is met by higher insurance premiums to pay for reduced volatility, negating savings.

Structural cash flow deterioration. NSP has not generated sufficient operating cash flow to fund both capital expenditures and dividend payments on a cumulative basis since end-2023, forcing reliance on additional borrowing to maintain cash position.

Vulnerable market position. Insperity is a low-profit, low-growth business ill-equipped to fight tough competition and faces significant economic headwinds.

Read more research at Phalanx Research.

Key Themes

Zero geological proof for headline narrative. GLND's central claim of 13.04 billion barrels is a 3U prospective resource estimate for undiscovered accumulations assuming every horizon across 58 prospects is productive. The company has drilled no exploration wells, made no commercial discovery, and holds zero proved reserves against $17.51 million in capitalized unevaluated property costs.

Collapsed drilling timeline with capital deferral. GLND pushed the contractual deadline for its first required well from December 31, 2026 to December 31, 2028, then paid £500,000 to formalize the extension while simultaneously pursuing acquisition of its drilling partner at a 43% premium—evidence the company cannot meet its original commitments.

Cash burn and imminent insolvency. GLND burned $32.6 million in six months from a $70 million April offering, consuming $28 million on exploration activities alone. At this pace, the remaining $37.42 million cash balance would be exhausted in under seven months before drilling anything, while another 20.8 million dilutive securities loom against 43.7 million shares outstanding.

Compromised auditor and internal controls. GLND dismissed its prior auditor (which issued going-concern language) and appointed MaloneBailey, censured by the PCAOB in May 2024 for pervasive quality control violations in testing revenue and auditing accounting estimates—the exact categories critical to validating $17.51 million in capitalized exploration costs with no proved reserves to support them.

Governance failure and value destruction. GLND appointed Carol Craig to its Audit Committee, the board member responsible for financial reporting oversight. Craig simultaneously sits on Twin Vee PowerCats with GLND Chairman Larry Swets Jr.; Twin Vee executed a defective Nevada reincorporation and invalid reverse splits while compounding share dilution 370-to-1 in one year, and Sidus Space, the company Craig founds and chairs, is down 99.8% from its IPO on a split-adjusted basis.

Predatory capital structure and insider exits. Citadel Multi-Strategy Equities converted 1.25 million pre-funded warrants (priced at nominal exercise cost) for 1.25 million free shares representing 2.9% of the company, then exited 96% of that position within five months while retaining only warrant optionality—evidence of a position constructed and executed solely for immediate profit extraction, not shareholder conviction.

Stock promotion and undisclosed paid advocacy. GLND signed Rubenstein Public Relations in September 2025 for $10,000-15,000 monthly plus $100,000 in stock compensation and discretionary bonuses tied to "campaign success," then saw paid social-media posts from influencer Mario Nawfal (previously investigated for bot-manipulation and investor fraud allegations) promoting the stock with undisclosed compensation on the morning of September 21, 2026.

Ineffective internal controls and vendor equity compensation. Management certified that disclosure and internal controls over financial reporting were ineffective as of June 30, 2026 due to lack of formalized documentation. GLND paid vendors $317,500 in equity (85,000 shares) for services rather than cash, printing dilutive shares while capital erodes on an undrilled exploration program.

Read more research at Phalanx Research.

Key Themes

Revenue concentration and operational risk. Ethos relies on three carriers for 88% of revenue, creating acute vulnerability to any single carrier's operational failure or relationship deterioration.

Distributor defection risk. The company's largest third-party distributor by head count publicly ranked Ethos last among competing platforms, indicating potential loss of key revenue channels and market credibility collapse.

Systemic operational fragility. A TruStage system outage in July halted premium draft processing for three weeks, triggering widespread policy lapses and exposing the company's inability to manage customer-critical functions independently.

Read more research at Phalanx Research.

Key Themes

Revenue concentration risk. Ethos relies on three carriers for 88% of revenue, creating acute dependency on partners who can experience operational failures or exit the relationship.

Distributor confidence collapse. Family First Life, Ethos' biggest third-party agency by headcount, ranked Ethos last among its insurance partners publicly, indicating a breakdown in distributor confidence and competitive positioning.

Operational vulnerability. TruStage's system outage stopped premium drafts for three weeks, triggering policy lapses—demonstrating how dependent Ethos is on carrier infrastructure it does not control.

Read more research at Phalanx Research.

Key Themes

Dangerous revenue concentration. Ethos relies on just three carriers for 88% of revenue, creating acute vulnerability to any single carrier's operational failures or strategic decisions.

Carrier system failures trigger mass policy lapses. TruStage's July outage prevented premium payment drafts for at least three weeks, causing a wave of customer policy lapses that exposed Ethos' operational fragility.

Key distributor publicly demotes Ethos. Family First Life's founder ranked Ethos last among insurance partners at a conference and required its representatives to apologize, signaling deteriorating trust and potential loss of this major distribution channel.

Read more research at Phalanx Research.

Key Themes

Massively overstated economics. FISN's pre-IPO LCOE guidance of $50-70/MWh disappeared after the June 2026 IPO. Iceberg's calculations using FISN's own disclosed assumptions show FOAK LCOE of $289/MWh and NOAK LCOE of $148/MWh, both uncompetitive versus natural gas at $103-110/MWh and even fuel cells at $140/MWh.

Understated capital costs hide deteriorating fundamentals. FISN's claimed NOAK capital cost of $84mn is overstated by 34%. Applying industry standard scaling factors and third-party cost estimates yields a revised NOAK cost of $112mn and LCOE of $170/MWh, making the reactor more expensive than most firm energy sources except conventional nuclear.

Unprecedented and unsupported regulatory timeline. FISN claims NRC approval and first commercial deployment by end-2027, a timeline shorter than any SMR precedent. The company already missed its DOE pilot program targets while peers advanced; it took nearly a year to complete just the first stage (NSDA) versus three months for competitor Antares, contradicting claims that PWR technology accelerates licensing.

Six-month deployment claim is false. FISN's own prospectus discloses a two-year development timeline despite marketing a six-month deployment advantage. The company's simplified design strategy is non-unique, and peers pursue identical approaches. Most critically, FISN adds an unconventional major step—drilling a one-mile hard-rock borehole up to 60 inches in diameter—which executives admitted "pushes the limits" and introduces unproven technical and regulatory challenges.

Customer pipeline is largely phantom. Of FISN's claimed 18.5GW pipeline, only ~4GW ties to identifiable counterparties, all with non-binding agreements. Great Plains Industrial Park, listed as the pilot site, generated only $2.5mn annualized revenue and remains "mostly overgrown" with no committed large electricity consumers. Blue Owl and Endeavour Energy agreements contain explicit carve-outs for non-binding status and zero commitments.

Imminent funding crisis and massive dilution. FISN has less than twelve months of cash runway with ~$350mn in unfunded capital needs through commercial construction and project equity contribution. This funding gap equals 76% of current market capitalization and will be funded entirely through equity, implying ~43% dilution to existing shareholders. The company faces deteriorating investor appetite as nine SMR stocks now compete for capital, with Westinghouse expected to list as a superior alternative.

Founders' track record: decade of failed commercialization. Deep Fission recycles Deep Isolation, the founders' prior venture founded in 2016 that failed to commercialize deep-borehole nuclear waste disposal after ten years. Deep Isolation's 2025 revenue of $6.1mn was 59% from unrelated consulting subsidiary Freestone and 39% from government feasibility studies—zero revenue from actual waste disposal commercialization.

Natural gas is the proven winner for AI data centers. Roughly 75GW of booked on-site data center generation is natural gas-based, with lead times as short as twelve months. FISN's reactor remains unproven, years from commercialization, and uncompetitive on cost. Even NuScale, with a more conventional design and multi-year regulatory head start, has sold zero reactors three years after NRC certification.

Read more research at Phalanx Research.

Key Themes

Hidden off-balance sheet liabilities create existential risk. EquipmentShare allegedly promises 10% first-loss guarantees and equipment repurchases at 55-56% of original cost to OWN Program investors—undisclosed commitments that create approximately $500 million in liabilities that would wipe out all tangible common equity, with total buyback exposure potentially approaching a material portion of the ~$5.5 billion equipment fleet enrolled in the program.

Undisclosed related-party dominance undermines growth narrative. The Premiere Group (formerly Schlacks Rentals), controlled by co-founders Jabbok and William Schlacks, owns hundreds of millions in equipment enrolled in the OWN Program—far exceeding the <$1 million disclosed on earnings calls—suggesting that purported independent third-party demand driving the asset-lite growth story is artificially inflated through founder-controlled entities.

Inflated profitability metrics mislead investors and creditors. EquipmentShare inappropriately excludes OWN Program distributions to equipment financiers from its Equipment Rental Segment Adjusted EBITDA calculation, reporting 40%+ margins when true margins are approximately 14% once investor payouts are properly accounted for, with margins deteriorating sharply from 34% (FY '23) to 14% (1H '26).

Margin loan and stock pledge create forced liquidation risk. Co-founders pledged ~27.2 million shares (~40% of their holdings) worth ~$886 million at IPO close as collateral for margin loans containing undisclosed loan-to-value covenants, creating risk that a stock price decline could trigger forced selling that devastates share price, particularly as major shareholder Romulus Capital has already begun liquidating its 26% stake following its 2025 lawsuit alleging misconduct.

Opaque insider enrichment through capital funnels. EZ Equipment Zone, Armada Fleet Management, and Bevel Financial—entities closely connected to the Schlacks family—funnel capital into and finance the OWN Program while collecting substantial undisclosed fees, with these related-party intermediaries controlling 50%+ of OWN Program capital flows despite minimal SEC disclosure.

Read more research at Phalanx Research.

Key Themes

Capital cycle peak positioning. The Big 5 hyperscalers are trapped in the middle of a capital cycle peak matching prior bubbles in 2000 and 2007, with net investment-to-GDP ratios signaling imminent market reversal and asset write-downs.

Hidden off-balance sheet leverage. The companies carry $3 trillion in aggregate purchase commitments, future leases, guarantees backing third-party debt, and construction-in-progress that are obscured from standard balance sheet analysis.

Deteriorating free cash flow dynamics. Free cash flow at the most profitable companies in the world (excluding Apple) is turning negative while borrowing accelerates, exposing unsustainable capital spending models.

Understated depreciation and future write-downs. Depreciation expense is being systematically understated, mirroring the pattern from the 2000 TMT bust, with write-downs expected in 2028-2029 when the magnitude of capital commitments becomes undeniable.

Read more research at Phalanx Research.

Key Themes

Accounting misstatements and compliance failure. Hub Group disclosed significant accounting misstatements in February 2026 and has failed to produce corrected historic filings or 2026 quarterly filings required for Nasdaq compliance, risking delisting.

Deteriorating operational performance. The company's revenue, profits, and operating cash flows have declined since 2023, with management's recent guidance for Q1-Q2 2026 losses contradicting analyst consensus forecasts for profitability.

Failed strategic transformation. Asset-light business model acquisitions executed over recent years have not delivered expected results, forcing the company to reverse leadership decisions and bringing back former chairman David Yeager.

Market mispricing and hidden liabilities. Hub Group's stock trades at November 2025 levels despite material accounting restatements, and analyst coverage remains predominantly bullish despite deteriorating fundamentals, indicating the market underprices additional undisclosed problems.

Structural control preventing remediation. The Yeager family's 62% voting control and bloc voting agreement locks in incumbent management and prevents takeover solutions, leaving the company unable to access capital or strategic alternatives needed to address mounting problems.

Read more research at Phalanx Research.

Key Themes

Earnings quality deterioration. MKSI's reported Q2 EPS beat of 34 cents is largely attributable to favorable expense and reserve items rather than genuine operational performance, raising questions about earnings sustainability.

Hidden accounting benefits. Non-obvious accounting adjustments and reserve trends provided meaningful benefits to reported Q2 earnings, obscuring the true strength of underlying business performance.

Unsustainable working capital benefits. MKSI benefited from working capital trends in Q2 that inflated operating cash flow, representing a temporary tailwind unlikely to repeat.

Margin pressure and structural headwinds. The company faces meaningful margin pressure and increased capital investment requirements that conflict with the growth expectations embedded in its prior valuation.

Read more research at Phalanx Research.

Key Themes

Collapsed valuation thesis and extreme shareholder destruction. SWVL went public at $1.5 billion valuation on a $1 billion revenue projection. FY2025 revenue landed at $24.2 million; the stock has destroyed 90.4% of its public value to $143.7 million. Trust redemptions signaled investor disbelief at closing, with 84.6% of SPAC proceeds redeemed before the deal closed.

Artificial profitability through below-the-line gains. SWVL announced its "first profitable year" twice: FY2023 profit of $3.06 million was 84% driven by $18.83 million in "other income" from creditor settlements at 15 cents on the dollar. FY2025 profit similarly relied on non-operating gains, masking continued underlying operating losses.

Real-time operational collapse documented by customers. An independent review of 135 Google Play app reviews from January-September 2026 found 97.8% negative sentiment, with complaints accelerating from 0.4 daily to 0.79 daily by June. Riders describe broken technology, missing drivers, no customer support, and a company simultaneously discontinuing routes while management denies the discontinuation.

Dilutive capital raise with asymmetric governance terms. August 2026 Coefficient/HITE/Sofico placement issued 10 million shares at $1.45, more than doubling the 9.96 million pre-raise share count. Private investors negotiated board seats, pro-rata participation rights, and consent powers not available to public shareholders; the stock surged 50%+ on the announcement despite zero disclosed US contracts, revenue, or operating metrics.

Weak competitive position masking impossible US expansion narrative. SWVL claims to have "recently begun" US operations with $13 million backing it against Uber (75% US rideshare share), Zeelo, RidePal, and Enterprise mobility operations already entrenched in the corporate-shuttle segment. The US expansion narrative is a financing-timed press release, not a credible market-entry strategy.

Structural control weaknesses and auditor concerns. SWVL disclosed unremediated material weaknesses in internal financial controls across consecutive reporting periods and changed auditors after two consecutive going-concern opinions, signaling serious doubt about the reliability of reported figures management now relies on to market expansion.

Capital structure engineered to benefit insiders over public holders. A $14.5 million August 2026 raise involved HITE Hedge Asset Management, which settled SEC enforcement action in February 2023 for Rule 105 short-selling violations. The Coefficient lead investor holds a board designation contingent on 5%+ ownership and full contractual protections unavailable to public shareholders holding identical securities.

Read more research at Phalanx Research.