Part I  ·  The Investment Thesis

The origin of alpha.

Over 250 years of evidence: every time input costs collapse, systems reorganize — and the companies that become the new infrastructure produce outsized returns. The 2020s are a turning point. This is the long argument for why.

I. A Perez-Style Turning Point

Techno-revolutions arrive in half-century cycles. We are at one now.

Carlota Perez's framework teaches that techno-revolutions move through four phases: installation, a turning point (typically a financial crisis and bubble), deployment, and maturity. The deployment phase is where the dominant companies of the next era are built and capitalized.

The dot-com cycle delivered networking. The mobile cycle delivered phones-in-pockets and the cloud. The current cycle is delivering something larger: a re-foundation of how knowledge is produced, how energy is generated and delivered, how machines and humans collaborate, and how value moves between them. The infrastructure underneath is being rebuilt simultaneously across each of these dimensions.

That is the origin of alpha.

II. The Three Forces

Three forces, aligned for the first time in a generation.

01
The cost of knowledge is collapsing.

AI is democratizing access to expertise at machine speed — the same shift printing made to literacy. Foundation models compress decades of expertise into accessible APIs; agentic systems compress entire workflows. In the last year, 17 distinct "knowledge trades" endowed the cumulative knowledge of mankind to their generative AI equivalents. Cost-of-expertise is the input price falling fastest.

02
Industrial policy alignment.

The US, EU, and China collectively commit more than 60% of global R&D, increasingly directed into the substrates we target. The CHIPS and Science Act, the Inflation Reduction Act, the EU Chips Act, and parallel programs in Japan, Korea, India, and beyond represent the most coordinated state-level push into deep-tech industrial capability since the early Cold War. Public capital de-risks the early stages; venture follow-on captures the upside.

03
Energy + compute demand surge.

AI inference, electrification, and onshoring are driving a structural demand surge for power and advanced compute. Datacenter electricity demand is on track to roughly double from 2024 levels by 2030. Hyperscaler capex is at an all-time high. Grid modernization, transmission, advanced nuclear, and storage are all in the pipeline. Compute and energy are the two binding constraints of the deployment phase.

III. The Pluralist Industrial Era

Connectivity itself is becoming contested terrain.

The R&D figure above is one quantitative manifestation of a deeper structural shift. The post-1945 / post-1991 era of unipolar US-led globalization has given way to a pluralist industrial era in which multiple regional power centers — the US, the EU, China, India, and increasingly others — are simultaneously running their own industrial strategies, technology-export controls, and capability-denial regimes.

Industrial policy is not occasional or anomalous; it is the operating environment for the next cycle. The geo-strategist Thomas P.M. Barnett describes the United States as having transitioned from a market-making Leviathan (1945–2008) to a market-playing superpower competing within a structurally multipolar system. Two of his observations are especially load-bearing for our thesis:

Evernet Capital does not position itself as a national-security fund, and the strategy is not predicated on any specific geopolitical outcome. The point is structural: regardless of which administration sits in Washington, Brussels, Beijing, or Delhi, the operating environment for the next cycle is one in which industrial policy, export controls, friend-shoring, and rule-set competition are durable features rather than transient conditions.

IV. A Precedent We Lived Through

Electronic trading, the futures market, and the shape of what comes next.

If "input cost collapses → systems reorganize" sounds abstract, here is the recent precedent one of us had a front-row seat to during his time at CME Group — the cleanest illustration we know of how the framework plays out, and the same shape now repeating across the six families.

For 167 years, futures price discovery happened in open-outcry pits — raucous, hand-signaling, paper-ticket trading floors in Chicago and New York. Each trade carried high friction: a floor broker, a runner, a clerk, an exchange seat that traded for hundreds of thousands to millions, and a paper-based reconciliation chain. Spreads were wide because liquidity was rationed by the physical capacity of the pit.

Then the input cost collapsed. CME Globex launched in 1992 as the world's first global electronic futures platform. By 2004 it had recorded its one-billionth transaction. By the early 2010s, electronic execution was approaching 90% of CME volume. On July 6, 2015, after a 167-year run, CME closed open outcry on almost all of its futures pits; the remainder closed permanently in 2021. The marginal cost of a trade fell from dollars to fractions of a cent, and the system reorganized around the new economics.

Three Observations From That Experience

That bear directly on the current moment.

  • 01
    Volume did not just shift — it expanded by orders of magnitude. When per-trade cost collapsed, demand surged. New participants entered who could not previously justify the friction. The market is multiples larger than when human-only execution was the bottleneck.
  • 02
    The infrastructure layer captured most of the value. Floor traders adapted, were displaced, or found new niches. The structurally largest winners were the exchange and infrastructure operators — clearing, market data, and algorithmic execution firms built on top. The picks-and-shovels won.
  • 03
    The transition window was narrow and irreversible. Once execution costs fell below a threshold, the equilibrium flipped. Exchanges that did not move lost share permanently. The lesson: when an input cost is collapsing, the cost of being late is asymmetric. Too early is recoverable; too late is not.

The forces driving the present moment — collapsing cost of knowledge under AI, industrial-policy alignment, and surging energy and compute demand — are doing to the broader economy what electronic trading did to the futures market. We have seen this movie. We know how it ends.

V. The Financial Layer of the Evernet

The connective tissue across all six families.

Of the six families, Interoperable Industrial Infrastructure is the one we have the deepest operating history in. It is also the family where the regulatory, technological, and competitive landscape has shifted most dramatically in the last 36 months. Three forces unique to the financial layer have come into alignment:

Surface layer, orchestration layer, financial layer — all three are being built in parallel. The financial layer is the most regulated and the slowest to move, which makes its current acceleration the most meaningful signal in the system. When the financial layer catches up to the surface and orchestration layers, the entire stack flips from prototype to production. — Thesis, Part 3: Three Layers of the Evernet

This is why the financial layer is structurally weighted alongside Applied AI and Compute in our allocation. It is the connective tissue. Vertical agents in financial workflows, programmable money for agentic commerce, hardware-rooted identity, post-quantum cryptography — these are the building blocks of the next financial architecture, and they touch every other family in the portfolio.

VI. How We Win

Pattern recognition across multiple cycles.

The team has invested through digital infrastructure, AI, and blockchain waves. The discipline behind the Evernet thesis was not invented in 2026 — it has been compounding for over three decades. A 1994 industry framework segmented an emerging IT industry into three enabling substrates — Platform, Content, and Conduit — screened a 230-company universe down to a 38-name target list using six quantitative and eight qualitative criteria, and elevated to the top of their segments the names that became the infrastructure of the next era.

The 2026 framework in front of you is the same analytical shape: identify the enabling substrates, screen for survivors, underwrite the loops that compound. Six tech families instead of three sub-industries; AI, energy, and materials instead of computers, telecom, and semiconductors; agentic commerce instead of the original Information Appliance. The discipline travels.

Conclusion

The companies that will define the next industrial cycle are being formed now.

Evernet Capital is built to back them.