Tekton All Caps: +49.2% YTD—Driven by a Sequence of Reallocations, Not One Winning Bet
Through July 31, 2026, the strategy was 39.1 percentage points above the S&P 500 and 38 percentage points above VT.
Tekton All Caps returned +49.2% in 2026 through July 31, net of an assumed annual management fee of 1.5%.
Over the same period, the S&P 500 returned approximately +10.1%, while the Vanguard Total World Stock ETF (VT) the strategy’s formal benchmark, returned approximately +11.2%.
The result is noteworthy. But the central message is not only the performance difference.
It is how the portfolio evolved throughout the year.
Investment portfolio performance of the Tekton AI-Enhanced All Caps Strategy compared to the S&P 500 Index and the Vanguard Total ETF during the January - July 2026 period.
First quarter: energy takes the lead
Early in 2026, Tekton’s artificial intelligence agents identified attractive characteristics in energy companies.
The portfolio developed meaningful exposure to companies including Petrobras, Equinor and Golar LNG, which displayed relative strength, cash generation and attractive valuations while broad global indices weakened.
Energy made an important contribution during the first quarter. However, it would be misleading to attribute the full-year result to one successful sector allocation.
What followed provides a broader picture.
April to June: the portfolio broadens into four themes
Following the June optimization, the portfolio had expanded across four complementary themes:
High-quality financial institutions;
Artificial intelligence and digital infrastructure;
Energy and natural resources;
Defensive infrastructure and utilities.
JPMorgan was then the largest holding. The portfolio also included exposure to Amazon, Palo Alto Networks, MercadoLibre, Petrobras, Chevron, Golar LNG, Duke Energy and NextEra Energy, among other companies. The allocation already demonstrated that the strategy was not tied exclusively to energy or to a single macroeconomic thesis.
July: another expansion
During July, the portfolio had evolved again. Amazon had become the largest holding in both implementations reviewed, with an exposure of approximately 10%. At the same time, JPMorgan—previously the largest position—had been significantly reduced.
Energy was not abandoned. The strategy retained diversified exposure across Petrobras, TotalEnergies, Occidental Petroleum, Ovintiv, EQT, Golar LNG and Chevron.
Alongside that allocation, the portfolio included companies connected to:
Digital commerce and cloud infrastructure;
Cybersecurity and enterprise software;
Semiconductor equipment;
Utilities and growing electricity demand;
Defense, infrastructure and logistics;
Healthcare and defensive businesses;
Financially strong institutions.
Based on the average of the two end-of-July implementations, approximately 31% of the portfolio was associated with AI, digital platforms and technology infrastructure, while approximately 25% remained invested in energy and LNG. The balance was distributed across utilities, industrials, defense, logistics, healthcare, defensive businesses and financial quality.
The strategy therefore did not replace an energy concentration with an indiscriminate technology concentration. It developed a combination of structural growth, cash generation, defensive resilience and economic diversification.
What adaptation means in practice
The strategy uses agents trained with Deep Reinforcement Learning, or DRL. These agents evaluate prices, company fundamentals, current portfolio positions and the results of prior decisions. They use this information to learn allocation policies designed to maximize long-term risk-adjusted returns.
The process operates at two levels:
Strategic optimization: approximately every 60 trading days, the system conducts a complete review of portfolio weights using updated prices and fundamentals.
Continuous monitoring: between full optimizations, the system may introduce a new position, reduce an existing holding or eliminate a security when a meaningfully better opportunity emerges.
Transaction costs, taxes and turnover are incorporated into the optimization.
A sequence of decisions, not a forecast
The process did not seek to predict a war, oil prices, interest rates or the next market-leading sector.
Instead, the agents repeatedly reassessed the relationship among prices, fundamentals and expected returns.
The portfolio’s evolution summarizes that process:
Energy led during the first quarter.
In June, the allocation broadened into four complementary themes.
By the end of July, Amazon had become the largest holding, financial exposure had been reduced and energy remained meaningful within a broader portfolio.
The +49.2% result does not guarantee that similar performance will continue. It does, however, illustrate a process designed to modify allocations as the opportunity set changes.
Tekton All Caps is not built around making one major forecast. It is built around producing a disciplined sequence of decisions.
To learn more about the strategy and its implementation through separately managed accounts, contact Tekton.
Important information
Performance information is provided solely for informational purposes and does not constitute a promise or guarantee of future returns. The +49.2% return covers the period ended July 31, 2026, is based on live implementation and assumes, for illustrative purposes, an annual management fee of 1.5%. The S&P 500 is presented as a widely recognized market reference; VT is the strategy’s formal benchmark. Past performance is not indicative of future results. Investing involves risk, including the possible loss of principal.
Comments