Invariantum

AWK holds a legally protected local water monopoly on level 1 demand, intact and compounding: EPS +8.1%, rate base growing 8-9%. But it is declined at the strength gate. The return is set by regulators, not the company: Pennsylvania granted $75m of $160m requested at a 9.55% allowed ROE. A capped monopoly is strong, not exceptional. The live watch is the direction of allowed returns.

WM's moat is an exceptional landfill permit monopoly: new sites cannot be permitted in populated areas, so existing ones are irreplaceable toll booths on an unavoidable flow. Vertical integration, route density and rising compliance costs deepen it. Q2 2026 confirms condition, with 5.7% core price, 30.9% margin and free cash flow up 35%. Demand is level 1: waste never stops. All three gates pass. The only realistic risk is self inflicted M&A.

Synopsys: exceptional switching-cost moat in a foundry-certified EDA duopoly, fully intact — retention ~100%, 85% recurring, ~$11bn backlog, 41.6% margin. No entrant in decades. Declined at gate three: demand is level 3, resting on the semiconductor trajectory rather than a permanent human need. A structural plateau would shrink revenue without ever breaching the moat, so the drawdown may not reverse. No Layer 2 entry.

Coinbase: the most trusted, most regulated US crypto exchange, gaining share (record 10.3%) and diversifying into services (48% of revenue), yet revenue fell 19% to $1.2B with a $359M loss. The moat is a trust-and-regulation lead, contestable rather than locked, with weak, eroding trading-fee pricing power. Demand is level 4: hype-cyclical crypto that may not return to its peak. Declined on both moat and demand gates; not a Layer 2 candidate.

NVIDIA holds an exceptional moat via CUDA software lock-in, full-stack integration, and annual chip cadence, unbreached by any outside competitor at 80% share and 75% margins. The singular risk is that its hyperscaler customers (Google, Amazon, Microsoft, Meta) are building custom silicon to escape it, funded by NVIDIA's own pricing power. The moat is intact today but under a slow, largely invisible customer-competition threat.

VeriSign holds an exclusive contractual monopoly over .com domains, generating 68% operating margins and record growth (179.1M domains). No competitor can enter, but pricing is capped by a US Department of Commerce agreement, not set freely. Condition is fully intact; the key watches are future contract repricing terms and whether domains stay central as AI changes internet navigation. Verdict: Exceptional/Intact.

Visa's network moat is exceptional and intact where profit concentrates: developed markets and cross-border payments (up 13%). Government rails (UPI, Pix) have displaced cards domestically in India and Brazil, cutting India's card share from 43% to 21%, but this stays contained to emerging markets. Pivotal watch: whether state rails reach developed economies, turning contained erosion structural.

ASML holds a near absolute EUV lithography monopoly (100% market share); no substitute exists below the leading node, and no competitor (Nikon, Canon, China's SMEE) is within a decade of closing the gap. Strong Q2 2026 results confirm demand. Real risks (customer concentration, export controls) hurt market size, not the monopoly itself. Verdict: Exceptional/Intact. Long term watch: potential paradigm shift away from node shrinking chip architecture.

Ferrari's moat is controlled scarcity: it caps volume below demand, never discounts, and its order book is full through 2027, giving luxury-grade 39% margins. But it's strong, not exceptional, for two reasons: real apex rivals exist (Lamborghini, McLaren etc.), and its desire is tied to the combustion engine it's now leaving. The Luce EV drew a hostile reception. Verdict: Strong / Intact, under a brand-transition question. Watch: do electric Ferraris sell.

LVMH is not one moat but a machine for owning them: it buys brands and compounds them with retail, real estate and capital no independent house can match, funded by a few deep brand moats (Louis Vuitton above all) and locked behind entrenched family control. Strong, not exceptional, because rivals run the same machine and it depends on one operator. The live question is Arnault's succession to five heirs. Verdict: Strong / Intact, under a governance question.