Bristol Gate U.S. Equity Q3 2026 Portfolio Review

Microsoft was the portfolio’s largest contributor, rising 37.8% for the quarter, while KLA and Applied Materials were the largest detractors, falling 35.3% and 29.2%. NVIDIA, Apple, and Meta—none of which the portfolio owned—were relative detractors despite gains of 14.3%, 15.2%, and 28.8%. The portfolio purchased Garmin, sold Verisk, and rebalanced outliers toward approximately equal weights during the quarter.

Analysis

The key signal is attribution, not a sector call: gains in unowned NVIDIA, Apple, and Meta created relative headwind, while the owned semiconductor-equipment positions detracted. That does not establish that the portfolio lost money or that those unowned stocks are attractive now; benchmark, starting weights, portfolio return, and valuation context are missing. The split between NVIDIA’s gain and declines in KLA and Applied Materials is consistent with market leadership favoring AI compute over equipment, but one quarter cannot establish a durable change in capex economics.

The equal-weight rebalancing could have trimmed the strongest contributor and added to laggards. That is disciplined only if position sizing reflects forward risk and fundamentals; otherwise it may mechanically reduce exposure to momentum and increase exposure to a weakening equipment cycle. Microsoft’s contribution also warrants a concentration check: contribution alone does not show whether the portfolio’s return depended on one name.

Near term, the main risk is chasing recent relative winners or treating detractors as bargains. Over 1–3 months, compare KLA and Applied Materials’ order/backlog and guidance trends with compute demand and customer capex plans; over 6–18 months, a broadening or pause in AI infrastructure spending could reverse the compute-versus-equipment divergence. The Garmin purchase and Verisk sale cannot be evaluated without rationale, sizing, and valuation. No directional trade is justified from this attribution snapshot alone.

Key Decisions for Investors

• Request portfolio and benchmark returns, beginning/ending weights, and security-level contribution before changing exposure; distinguish absolute loss from relative opportunity cost.
• Review rebalancing records to determine whether the equal-weight reset trimmed Microsoft or added to KLA and Applied Materials, and test resulting concentration and downside under an AI-capex slowdown.
• Put KLA and Applied Materials on an earnings watch: reassess only against order, backlog, and guidance evidence. A sustained deterioration would falsify a mean-reversion case; improving trends alongside stable compute demand would weaken the underweight case.
• Do not chase NVIDIA, Apple, or Meta solely because they were relative detractors through non-ownership. Revisit only after benchmark-relative valuation, portfolio exposure, and forward earnings expectations are available.