VEKTOR · 2026 INVESTMENT THESIS

Physical gates, or stalled AI investment.

The 2026 thesis is built on the reality that the era when software alone was enough has ended. Every innovation we care about now requires a physical gate to be cleared. If the gate is not cleared, the trillion dollar AI investment stalls.

We have identified 12 stocks that sit at those physical gates for the 2027 cycle. The full list and sizing logic are available to approved accounts only.

12

BOTTLENECK STOCKS · 2027 CYCLE (INVITATION ONLY)

210

SIMULATED DECISION EVENTS

26

EVENT CATEGORIES

25K

MINIMUM · INVITATION ONLY

2026 PUBLISHED SLEEVE · FIVE GATES · YTD

Equal-weight bottleneck book on the site today

Five stocks at equal weight for 2026. The numbers below are committed static marks (regenerated from public closes when we refresh the file). The 2027 twelve stock mandate is separate and not listed here.

2026 PUBLISHED SLEEVE · YTD

Five stocks at equal weight (20% each, rebalanced at inception). Figures are actual sleeve and position YTD returns from internal investor reporting through . S&P 500 (SPY) benchmark for the same window: -0.61% YTD. Values are published as committed static marks in the codebase (not fetched live on each page load).

+21.18%

Sleeve YTD

-0.61%

S&P 500 (SPY) YTD

+21.79%

Alpha vs SPY

Sleeve (YTD)S&P 500 (YTD)
TickerCompanyGateWeightYTDContribution
VRTVertivThe thermal gate20%+82.16%+16.43%
CEGConstellation EnergyThe carbon-free power gate20%-18.90%-3.78%
CATCaterpillarThe reshoring gate20%+38.02%+7.60%
GOOGLAlphabet (Google)The infrastructure platform gate20%+1.36%+0.27%
AMZNAmazonThe logistics and compute gate20%+3.28%+0.66%

Position YTD and contribution are actual marks as of 2026-04-09. Contribution = weight × position YTD. The chart shows a cumulative path with month-end style progression through the year, ending at these YTD totals (not intraday live).

Where performance is shown for a calendar year or year-to-date period, figures are actual returns for that sleeve and period from internal investor reporting materials—not hypothetical or illustrative scenarios. These materials are for informational discussion only and do not constitute an offering. An offering may be made only by delivery of a confidential offering memorandum to appropriate investors. PAST PERFORMANCE IS NO GUARANTEE OF FUTURE RESULTS.

2026 gates · why each position matters

VRT+82.2%

Vertiv

The thermal gate

AI compute density has exceeded the limits of air cooling for dense clusters. Vertiv sits at the transition to liquid cooling and rack level thermal and power architecture, one of the hard gates before hyperscale halls can run at target density.

CEG-18.9%

Constellation Energy

The carbon-free power gate

Hyperscalers are hitting the wall of the public grid. Constellation is among the few operators with existing nuclear scale to anchor firm, carbon-free baseload without a full greenfield thermal construction cycle.

CAT+38.0%

Caterpillar

The reshoring gate

Domestic fabs, battery plants, and heavy industrial reshoring move earth first. Caterpillar remains the industrial capacity gate for the physical buildout when policy points capital at onshore capacity.

GOOGL+1.4%

Alphabet (Google)

The infrastructure platform gate

The vertically integrated stack from research through cloud is a compute gate in its own right: model depth, TPU/GPU scale, and GCP footprint together set how fast frontier workloads can deploy at institutional size.

AMZN+3.3%

Amazon

The logistics and compute gate

AWS is still the foundational layer of much of the cloud economy; Amazon’s logistics engine remains a physical gate for how goods and compute capacity move at national scale.

2025 · VEKTOR CORE · ACTUAL SLEEVE RETURNS

Two sleeves. One framework.

Vektor Core Growth and Vektor Core Safe are actual full-year 2025 sleeve returns from internal investor reporting, with stated weights on public marks.

Full year 2025

+60.8%

S&P 500

+17.9%

Alpha

+42.9 pts

Win / loss names

13W / 3L

Best / worst

APP +320% · YELP -15%

Max drawdown −14% to −18% (April 2025 tariff shock). Recovery roughly 6 to 8 weeks. Sharpe 1.8 to 2.2; Calmar 3.5 to 4.3.

Cumulative return path (monthly marks)

Positions · Vektor Core Growth

TickerNameWeight2025Contribution
METAMeta Platforms Inc.8%+13.1%+1.05%
APPAppLovin Corporation7%+320%+22.40%
LRCXLam Research Corporation7%+139.2%+9.74%
AXTAAxalta Coating Systems Ltd.6%+28%+1.68%
FIXComfort Systems USA Inc.6%+120.9%+7.25%
DYDycom Industries Inc.6%+45%+2.70%
ENSEnerSys6%+18%+1.08%
EXELExelixis Inc.6%+55%+3.30%
FSLRFirst Solar Inc.6%+48.2%+2.89%
HRMYHarmony Biosciences Holdings Inc.6%+65%+3.90%
HWMHowmet Aerospace Inc.6%+42%+2.52%
PODDInsulet Corporation6%-10%-0.60%
NBIXNeurocrine Biosciences Inc.6%-5%-0.30%
RMDResMed Inc.6%+25%+1.50%
VRTVertiv Holdings LLC Class A6%+42.8%+2.57%
YELPYelp Inc.6%-15%-0.90%

Where performance is shown for a calendar year or year-to-date period, figures are actual returns for that sleeve and period from internal investor reporting materials—not hypothetical or illustrative scenarios. These materials are for informational discussion only and do not constitute an offering. An offering may be made only by delivery of a confidential offering memorandum to appropriate investors. PAST PERFORMANCE IS NO GUARANTEE OF FUTURE RESULTS.

DECISION MAKERS · 210 EVENT ENGINE · 2026

Simulation is how we time when each physical gate binds.

The same engine that maps documented decision makers (rate setters, regulators, utility and industrial policy) tracks when those actions land through the year. Portfolio construction turns those forward paths into the narrow set of companies that sit at maximum physical constraint, not the broad theme.

The five stock sleeve above is what we publish for 2026. The 2027 twelve stock book, with tighter mandate weights for approved accounts only, is the next map on the same principle.

Apply for access →

HOW THE RECORD IS BUILT

From documented actors to physical gates.

Vektor Securities does not start with screens or consensus estimates. It starts with the documented behavior of the people and institutions that move policy, rates, and large-scale capital allocation. The engine covers 210 discrete events across 26 categories; each is modeled before resolution with a specific call, confidence, and invalidation condition.

When those simulations point in a coherent direction, the next step is mapping what must exist in the physical economy for that direction to materialize. The portfolio holds the companies at the point of maximum constraint: the steps where scarcity bites before the rest of the chain can scale.

The five stock 2026 sleeve on this page is an actual equal-weight sleeve: YTD and per-name returns come from internal investor reporting for that period. We publish them as committed static marks in the codebase when we cut a reporting snapshot; they are not fetched live on each page load. SPY is the benchmark over the same calendar window.

The 2025 Vektor Core Growth and Vektor Core Safe sections are actual full-calendar-year sleeve returns from the same materials. Nothing on this page is an offering or a substitute for an offering memorandum.

Why weights can differ from your mandate?

Approved accounts apply Layer 5 sizing, liquidity, and risk constraints on top of the same bottleneck logic. The five-name sleeve here shows actual returns for that published basket at equal weight; it is not a full description of every client account or sleeve.

What would invalidate a gate?

A substitute technology or policy path that removes the constraint would force a rebuild of the bottleneck map. For example, a materially different cooling path, or baseload alternatives at scale without the same lead times, could do it. The simulation layer is what surfaces that shift early.

Event-level simulations

Committee and policy event models are part of how we pressure-test the macro path. The portfolio track record is the economic expression of where those paths point when they converge on physical scarcity.

Where performance is shown for a calendar year or year-to-date period, figures are actual returns for that sleeve and period from internal investor reporting materials—not hypothetical or illustrative scenarios. These materials are for informational discussion only and do not constitute an offering. An offering may be made only by delivery of a confidential offering memorandum to appropriate investors. PAST PERFORMANCE IS NO GUARANTEE OF FUTURE RESULTS.