PRIMECAP's $132B Contrarian Book: LLY 7.5% Anchors Q1

PRIMECAP Management, the contrarian Pasadena-based active manager that runs the original Vanguard Capital Opportunity, Vanguard PRIMECAP, and Vanguard PRIMECAP Core sleeves, filed its 2025Q4 13F with $132.11 billion in reported value across 319 positions. That position count is the lowest in the visible nine-quarter window — down from 336 in Q3 2024, 333 in Q2 2025, and 327 in Q3 2025. PRIMECAP has been actively pruning the book heading into 2026, and the names that survived the cull are the cleanest read of where the firm's bottom-up research process actually sits.

The single most important data point in the file is the LLY position at 7.54%, sized at $9.96 billion. For context, that is more than triple the NVDA position (2.43%) and roughly equal to the combined weight of MSFT (2.67%) plus GOOGL (3.58%) plus NVDA. PRIMECAP has historically built positions over many years and held them through extended drawdowns; sizing Eli Lilly to 7.54% of a 319-position book is not a momentum trade — it is a bottom-up thesis that has been compounding for multiple cycles. With the Q1 2026 13F due May 15, the question is whether PRIMECAP trimmed into Lilly's strength or held through the GLP-1 and metabolic-health volatility.

The healthcare conviction stack nobody is reading correctly

Add up the named healthcare positions in the top-10 and the picture gets more interesting. LLY 7.54% + AZN 2.57% (AstraZeneca, the rare large foreign issuer in PRIMECAP's book) + AMGN 2.21% + BSX 2.24% (Boston Scientific medtech) = 14.56% of the entire $132 billion portfolio in pharma, biotech, and medical devices. That is structurally larger than the megacap-tech exposure (MSFT 2.67% + GOOGL 3.58% + NVDA 2.43% = 8.68%) and roughly double the semis cluster (MU 5.36% + KLAC 2.65% + NVDA 2.43% = 10.44%).

For a manager whose investor letters consistently emphasize sector-bottom-up rather than top-down allocation, that healthcare concentration is the cleanest external evidence of where the research process has aggregated conviction. The Q1 2026 print will reveal whether PRIMECAP held through the obesity-drug news flow or trimmed Lilly into the late-2025 strength. Either reading carries information.

The semis pair trade hidden inside the top-10

The second-largest position is MU at 5.36% ($7.08 billion), and the fifth-largest is KLAC at 2.65% (KLA Corp, the semis-inspection equipment maker). Together with the smaller NVDA at 2.43%, that is 10.44% of the book in semis exposure spread across memory (MU), wafer-inspection equipment (KLAC), and AI accelerator demand (NVDA). PRIMECAP's traditional semis playbook leans into long-cycle equipment names like KLAC and Lam Research rather than the headline AI accelerators, so seeing MU sized to 5.36% — a memory cycle bet — is consistent with the contrarian discipline. The book is positioned for a memory-and-equipment recovery that runs longer than the spot-AI demand cycle.

The pie chart reveals the broader structure: 33.4% in the top-10 names, 66.6% spread across the remaining 309 positions. PRIMECAP has historically built mid-tier positions slowly — many of the names sitting in the 0.5-1.5% band today have been held since the early 2010s and are still being added in small increments. Q1 2026 is the first quarter where the 319-position book at year-end can be tested against whether further pruning continues. If position count drops below 315, PRIMECAP is concentrating more sharply; if it expands, the contrarian book is reabsorbing names that previously rolled off.

The AUM trajectory tells the contrarian story

PRIMECAP's reported 13F value moved from $118.47 billion in Q3 2023 to $132.11 billion in Q4 2025 — an 11.5% gain across nine quarters. That is dramatically slower than the AI-megacap-driven AUM expansion at peer active managers (Citadel +35%, D.E. Shaw +90%, Capital Research Global +49% over comparable windows). The chart shows the full nine-quarter sequence: a peak at $138.13 billion in Q2 2024, a $17 billion drawdown to $120.96 billion by Q1 2025, then a four-quarter recovery to $132.11 billion.

That drawdown-and-recovery shape matters. Most large active managers expanded steadily through the 2024-2025 AI cycle; PRIMECAP gave back ground through 2024H2 and 2025H1. The most likely explanation: a contrarian book underweighted in the megacap-AI complex (NVDA at just 2.43%) and overweighted in healthcare lagged the headline beta during the rally and recovered as the trade broadened. The Q1 2026 print will reveal whether the contrarian discipline held through Q1 chop or whether PRIMECAP rotated into the consensus AI names.

What to watch on May 15

Three signals worth marking when PRIMECAP's Q1 2026 13F-HR drops. First, the LLY weight — if it climbs above 8% the book has reinforced the healthcare thesis; if it compresses below 6.5% PRIMECAP has trimmed into strength. Second, whether MU stays at 5.36% — the memory cycle bet has been the cleanest counter-cyclical position in the portfolio, and any shrinkage signals the firm has called the cycle peak. Third, the position count itself — if it drops below 315, the book is concentrating; if it rebuilds above 325, names that rolled off are returning.

For ongoing position changes between filings, the active 13D/G feed surfaces threshold-crossing trades across the names PRIMECAP holds, and the institutional signal feed aggregates rotation patterns across active managers. Readers comparing PRIMECAP's contrarian book to the megacap-AI-heavy active sleeves can pull consensus-overlap views to see which positions are unique to PRIMECAP's research process and which appear across multiple peer books.