Caisse De Depot Et Placement Du Quebec 13F Portfolio

Rising conviction is clearest in the shift from AI beneficiaries to the tools that make AI possible. Lam Research (up 89.3% in shares, +$295.8M) and KLA (up 164.9%, +$225.9M) are now meaningful positions — classic semicap names that monetize wafer volumes rather than the exact trajectory of AI unit demand. Texas Instruments (shares up 29.2%, +$109.7M) and Intel (up 26.6%, +$66.5M) round out a bet that analog, embedded, and foundry capacity are the safer way to own silicon.

Microsoft is the only mega-cap AI platform they added to in size, with a 21.5% share increase worth about $398.1M. That says they’re discriminating among platforms: still happy to own Azure/OpenAI exposure, but less eager to keep riding Nvidia and Alphabet at current multiples.

In health care, they added aggressively to Eli Lilly (+17.4%, +$143.0M) and Merck (+17.1%, +$83.4M) and opened a new $273.9M position in AstraZeneca. That’s a clear GLP‑1 / oncology / immunology pipeline bet, paired with modest increases in UnitedHealth and AbbVie.

Factor sleeves also stand out. RSP, the S&P 500 equal‑weight ETF, was the single largest add (+294.8% shares, +$632.0M), and VTV, a value ETF, saw a 35.8% share lift (+$201.6M). They’re explicitly paying for idiosyncratic tech and pharma names with a broad tilt toward cheaper and more diversified U.S. equities.

The sell tape is less about fear and more about harvesting wins to rotate into what they see as better risk/reward. Nvidia was the single biggest funding source (shares down 20.8%, about -$657.0M), alongside Apple (-15.8%, -$506.1M) and Alphabet’s GOOGL line (-20.9%, -$496.1M). All three are sitting on large gains vs. cost, so this looks like disciplined risk management, not a macro call that AI is over.

They also lightened the highest‑beta semis: Micron (-31.6%, -$283.0M) and AMD (-25.3%, -$115.8M), while trimming Broadcom (-16.8%, -$139.2M). The proceeds have clearly been redirected into more diversified or upstream chip plays like Lam, KLA, TI, and Intel, where earnings are less hostage to a single GPU cycle.

Outside tech, Johnson & Johnson took a sizable cut (-24.3%, -$184.5M), even as other pharmas were being added, suggesting a shift from diversified pharma to more targeted growth pipelines. They also nipped at financials — JPMorgan (-10.0%), Citigroup (-5.2%), and Royal Bank of Canada (-5.8%) — likely as marginal funding for Progressive, Canadian industrials, and GM, all of which saw meaningful share increases.

Finally, they trimmed XLV, the health care sector ETF (-21.0%, -$71.5M), while buying single‑name pharma. That’s a classic move from “defensive sector beta” to “earnings we actually underwrite.”

On the surface, technology’s weight nudged down from 44.93% to 41.98%, but the internals changed far more than the headline. The cuts hit the most crowded names — Nvidia, Apple, Alphabet, AMD, Micron, Broadcom — while the adds went to semicap tools (Lam, KLA, Applied Materials marginally trimmed but still large), diversified chip suppliers (TI, Intel), and Microsoft. They are migrating from AI story stocks to AI infrastructure and capacity.

Health care crept up from 8.08% to 8.93%, but again via a quality upgrade: Lilly, Merck, UnitedHealth, AbbVie, and new AstraZeneca offset the J&J and XLV trims. That’s a shift from blanket sector exposure to specific cash-flow and pipeline risk.

The “Unclassified” bucket moved from 6.88% to 9.2%, but it’s really about structure: Berkshire, RSP, VTV, XLV, USHY, and HYG. Within that, RSP and VTV are doing the heavy lifting, turning what had been a fairly concentrated mega-cap book into something more factor‑balanced.

Elsewhere, sector weights are relatively stable: consumer discretionary at 11.96% (Costco slightly trimmed, TJX and WMT up), industrials flat around 10.5% (more TSLA and GM, less TFI), and finance essentially unchanged at 8.69%. The small bump in consumer staples, via Coca-Cola (+34.6% shares), and in utilities via TC Energy (+4.0%), rounds out a mild tilt toward dependable cash flows.

Put together, this is a late‑cycle, earnings‑driven stance: CDPQ is betting that AI capex and new drug platforms will keep compounding, but wants to own them through the picks-and-shovels and cash‑rich incumbents rather than the frothiest narrative stocks. The semicap build‑out, Microsoft add, and pharma upgrades all push in that direction.

At the same time, the surge into RSP and VTV, plus incremental size in Berkshire, looks like an explicit hedge against narrow leadership and potential multiple compression at the top of the S&P 500. If the market broadens out, they benefit; if it doesn’t, they still own the key AI and pharma engines.

The higher exposure to credit and income ETFs like USHY and HYG, alongside defensives like Coca-Cola and TC Energy, suggests growing comfort with credit risk but an awareness that rates may stay restrictive. That mix positions them reasonably for a “soft-ish” landing with pressure on valuations but continued earnings growth.

Investors reading this book should see a clear message: CDPQ is still pro‑equity and pro‑innovation, but it is no longer willing to pay any price for headline AI and mega-cap tech. The incremental dollar now prefers semicap throughput, GLP‑1 and oncology pipelines, and diversified value to the pure momentum trade of the last three years.