Skip to content
Toffel Capital

Archived thesis

Historical analysis, first published and last revised . It records the reasoning at the time, is not a current recommendation, and is no longer updated. All archived theses
Advanced Micro Devices logoAMD · Equity · SemiconductorsExited

Advanced Micro Devices

Historical view

This position has been exited. The thesis is kept as a record of the reasoning at the time and is not maintained as current research. See its decisions.
On this page
  1. Investment case
  2. Related decisions
  3. Sources and revisions

Investment case

Exited on valuation and relative preference

I exited the position because the narrative had moved ahead of the underlying story, and valuation had become less attractive, particularly relative to NVIDIA.

After a very large move the odds of a consolidation period looked high, and the same capital seemed better used elsewhere. The company did not become worse; the expected return at that price did.

What the trims got right and wrong

The trims did what they were supposed to do. AMD hit 11.45% in May 2026, above my band, and I cut it back over several sessions on size rather than on view. That is the policy working, and AMD still ended up the largest realized contributor in the account.

The part I would question is the gap between May and July. In May I said explicitly that the trim was not a thesis reversal. By late July I exited the rest on valuation. If the thesis was intact in May, something should have changed in the two months between, and I cannot point to what did. The honest read is that the July exit was partly a reaction to a position that had already been shrinking.

Related decisions

  1. ExitWeight 6.9% → 0%

    Closed the remaining AMD position after the May trims. The freed capital was reallocated the same day to ServiceNow and VanEck Semiconductor ETF. AMD was the single largest contributor to realized results across the account's history.

    In the decision log

  2. TrimWeight 9.2% → 6.6%

    Continued trimming AMD through May as the position appreciated above the target sizing band. Multiple sessions of small trims recycled gains while preserving core exposure to the AI compute thesis.

    In the decision log

  3. TrimWeight 11.5% → 10.6%

    Trimmed AMD in the account on May 1, 2026 after a significant run to bring the position back toward my 10% max position-size discipline. This was not a thesis reversal. The core AMD thesis remains intact, but the trim reflected concentration control, risk management, and a preference to preserve gains after the position had outgrown its intended role.

    In the decision log

  4. InitiateWeight 0% → 1.5%

    Rationale not recorded.

    In the decision log

Sources and revisions

  • Holdings snapshot: October 16, 2026
  • The analysis and views are Isaac Toffel's own.

Text revisions: Aug 27, 2026 · Aug 20, 2026.

PortfolioDecision log