Teardown

Daily digest · 2026-09-08

Scan #048: an Odyssey $74M DFI-plus-equity growth round into distributed-renewables project finance across 50+ emerging markets, Locus Robotics printing a $41.6M existing-investors-only Series G four years after its last priced round, NormanMax raising into a Lloyd's syndicate that just had its stamp capacity cut 72%, UniUni going public via TSX SPAC at ~US$1B EV on the eve of US de-minimis enforcement — and four US incumbents where the compounding has quietly reversed: Publix printing its first negative comp with the ESOP stock cut, Erie Insurance Exchange in year two of an underwriting loss while Indemnity collects a 25% management fee, Louisiana-Pacific's SmartSide volumes down 18% for the first time, and Con Ed cut to 13% of its rate ask by the NY PSC while AI data-center load routes around NYC

Eight companies where the underlying math shifted in the last twelve months — four emerging companies with a single falsifiable question each (Odyssey's project-finance take-rate against DFI concessional pricing; Locus's Array/Nexera manipulation stack against Amazon Robotics Proteus/Sequoia at a 32%-lower implied valuation; NormanMax's parametric loss ratio against a Lloyd's Performance Management Directorate that has already cut its stamp capacity from £108M to £30M; UniUni's post-de-minimis pivot from Shein/Temu to US-domiciled shipper volume on a 1.2-star Trustpilot rating) and four US incumbents whose 2026 numbers force a re-underwriting: Publix's first negative comp print plus internal ESOP stock cut, Erie Insurance Exchange's second year without an underwriting profit while the reciprocal structure keeps Indemnity's public shareholders paid, Louisiana-Pacific's SmartSide volume decline that says even the good half of the P&L has demand elasticity, and Con Ed's 87% rate-case denial plus AI-load routing around Manhattan.

Today’s eight sit on one thread: the operating environment has quietly re-priced, and the balance-sheet math each of these companies was underwriting three years ago no longer holds. Four emerging companies each face a single, falsifiable question with a printable answer inside 24 months — a DFI take-rate print, a Series-G-priced follow-on, a Lloyd’s PMD capacity decision, and a post-de-minimis shipper-mix migration. Four US incumbents just printed the first numbers that mark a regime change — Publix’s first negative comp in memory and an internal-stock cut, Erie Insurance Exchange’s second year of underwriting loss with an AM Best downgrade, Louisiana-Pacific’s SmartSide volume shock, and Con Ed getting cut to 13% of its rate ask while the AI-load wave routes around NYC.

Odyssey Energy SolutionsEnergy · Emerging. Boulder-based Emily McAteer’s ex-Bloomberg-NEF marketplace-plus-software for distributed renewable energy project finance in emerging markets closed a $74M round on September 1, 2026 ($27M equity from Broadscale, FMO and Al Mada Ventures; $47M debt from BII, BIO, FEI and EEGF), taking total capital to ~$94M. Odyssey connects 6,000+ solar installers and EPCs in 50+ countries with DFI and commercial financiers, and has cumulatively facilitated $3.6B of capital across 1.5 GW of projects, with India activity up 205% YoY. The open question is whether the platform monetizes that facilitated volume at a take rate — via procurement fees, financing spread and SaaS — that supports a venture return, or whether the model stays structurally captive to concessional DFI programs where fee compression, currency risk on the installer side and slow disbursement cycles cap gross margins at services-business levels.

Locus RoboticsLogistics / Supply Chain · Emerging. The Wilmington, MA warehouse AMR pioneer — spun out of Quiet Logistics in 2014 — printed a $41.6M Series G in September 2026 from Tiger Global, Goldman Sachs Asset Management, G2 Venture Partners and Scale Venture Partners at a ~$1.35B secondary-market-implied valuation, 32% below the ~$2B 2022 Series F post-money and with no new growth investor on the round. The company has ~4,000 pick-assist AMRs deployed across ~50 3PL and enterprise customers (DHL, GXO, UPS, Boots, Radial) generating ~$180M ARR on a RaaS subscription. The open question is whether Locus converts that install base into the manipulation layer of the warehouse via April 2026’s Locus Array launch and the May 2026 Nexera Robotics acquisition before Amazon Robotics’ Proteus/Sequoia stack eats the enterprise end, Symbotic swallows the full-facility ASRS deals, and Geek+ commoditizes the collaborative-AMR base — on the balance sheet of a flat-to-down insider round.

NormanMax Insurance HoldingsInsurance · Emerging. Bradley Meier’s (Universal Insurance Holdings founder) 2023-vintage parametric-insurance platform in Miami — a US E&S carrier plus Lloyd’s Syndicate 3939 plus two coverholder MGAs plus a servicing/analytics stack — kicked off a Stonybrook Capital-led equity raise in January 2026 into what Lloyd’s Performance Management Directorate has already flagged as distressed. Syndicate 3939’s stamp capacity was cut from £108M in 2025 to £30M in 2026, a 72% reduction, after successive underwriting losses on weather and CAT parametric covers. The falsifiable question is whether Syndicate 3939 can hit its ~$50M 2026 GWP plan and post a sub-100% combined ratio through a Category-4 US landfall — proving that no-deductible index-triggered covers can be priced correctly and rebuilding Lloyd’s PMD confidence — or whether Lloyd’s exits the trade and the platform becomes an E&S-only US MGA rump.

UniUniEcommerce / Retail · Emerging. The Richmond, BC-based gig-driver last-mile parcel network is going public via TSX SPAC (MAK Acquisition, announced May 15, 2026) at ~US$1B enterprise value with ~US$285M total capital raised — including the March 2026 $30M equity + $55M RBC credit facility on top of the 2025 $70M Series D from Bessemer and Sinovation. UniUni moves roughly one million parcels a day at a $1.50-2.00/stop driver-pay rate with no reimbursement, carrying a 1.2-star Trustpilot rating across 2,200+ reviews and 1,527 BBB complaints against an A rating. The mechanism that got them here — cross-border Shein/Temu/AliExpress volume flowing under the US$800 de-minimis threshold — was killed globally on August 29, 2025 when the Trump administration eliminated the exemption. The open question is whether UniUni transitions to US-domiciled shipper volume (3PLs, mid-market DTC, TikTok Shop US inventory) fast enough to offset the tariff-driven contraction in Shein/Temu package counts, before the consumer-reputation and no-reimbursement gig-driver economics prevent premium-brand contracts at UPS/FedEx-adjacent prices.

Publix Super MarketsRetail / Grocery · Incumbent, at risk. The ESOP-owned Lakeland grocery chain founded September 6, 1930 by George W. Jenkins posted a rare earnings decline in 2026 — Q1 net income $794M, down 21.5% YoY (down 3% ex-mark); Q2 sales $15.7B, +1% but comparable-store sales -0.5%. Publix cut its internal stock price from $20.45 to $19.60/share on August 1, 2026 — the first cut in years — and has resumed store closures for the first time in a decade even as Aldi absorbs 200 converted Winn-Dixie boxes across the Southeast in 2025-27. Kevin Murphy took over as CEO from the retiring Todd Jones on May 31, 2026. Publix still runs 1,498 stores at ~4.1% national grocery share (Numerator, August 2026) — but Walmart (23.6%) and Costco (9.2%) are compounding at its expense, Medicare Maximum Fair Price pressure hit ten drugs in the Publix pharmacy book in both quarters, and the delivery model is still Instacart-only with no scan-and-go.

Erie IndemnityInsurance · Incumbent, at risk. The 100-year-old Pennsylvania reciprocal insurance exchange, managed by the publicly listed Erie Indemnity (NASDAQ: ERIE) under Hirt-Hagen Class B control, is in its second year of underwriting loss — Q2 2026 Exchange combined ratio 103.9%, still no profit — while Erie Indemnity itself collects a 25%-of-premium management fee no matter how the Exchange underwrites. Direct written premium growth halved from 9.2% in Q2 2025 to 3.3% in Q2 2026, AM Best downgraded the Exchange in September 2025 citing surplus adequacy, and the Stephenson & Beste litigation over the management-fee cap is grinding through Pennsylvania courts. Progressive, GEICO and the insurtech cohort (Root, Openly, Kin, Corgi) are taking share in Erie’s core PA/OH/VA/MD footprint while the agent-only distribution model can’t compete on DTC pricing. The public-shareholder-vs-policyholder structural mismatch is now the market’s problem.

Louisiana-Pacific CorporationConstruction / Building Products · Incumbent, at risk. The Nashville-based building-products company (NYSE: LPX, ~$5-6B market cap) — a 1972 Georgia-Pacific spinout with two segments (SmartSide engineered wood siding at ~55%+ of revenue and ~26% margins; commodity OSB at ~35%) — printed Q1 2026 revenue of $574M vs $724M prior-year and Q2 2026 revenue of $664M with EBITDA down $63M to $79M. The number that matters is Siding: for the first time, SmartSide volumes fell 18% in Q1, partially offset by 9% higher prices, with Q2 following at -11% volume / +7% price. The +price/-volume split says builders are pushing back on the margin-heavy segment that was supposed to be LP’s structural offset to commodity OSB. James Hardie is spending materially more on R&D, West Fraser+Norbord has consolidated OSB oversupply against LP, and mass-timber CLT is starting to displace OSB in mid-rise commercial. LP now guides -$120M full-year 2026 OSB EBITDA.

Consolidated EdisonEnergy / Utilities · Incumbent, at risk. Con Ed (NYSE: ED) — the NYC/Westchester T&D monopoly with a 52-year consecutive dividend-increase streak — is the utility the AI-driven load-growth wave is bypassing. Duke and Southern are guiding 7-8% annual EPS growth; Con Ed’s is 5-7%. The January 2026 NY PSC decision approved only ~13% of Con Ed’s proposed rate hikes, cutting the ask by 87%. Con Ed cannot own generation, Manhattan is not hyperscaler geography, and the CLCPA’s net-zero-2050 mandate plus the 2019-21 gas hookup moratorium precedent are constraints, not opportunities. Meanwhile, behind-the-meter storage economics in NYISO Zone J (the highest capacity prices in the ISO) now let VPP aggregators — Base Power, Lunar Energy, Sunrun tuned for high-rise NYC — bid exactly the load Con Ed most needs to defend. The 52-year dividend streak, the $38B five-year capital plan, and the monopoly franchise are all real. So is the structural bypass.

Full digest: teardown.ai/digest/2026-09-08.


Full deep dives