Inside Aqr Capital Management Llc’s 2026-Q1 Playbook: Doubling Down on AI Rails

The biggest adds table reads like a who’s who of AI infrastructure and the businesses that monetize it. AQR is paying up in size for the platforms that sell compute, bandwidth, and high-value software, even when near-term P&Ls are under pressure.

• Microsoft: AQR lifted the position by +60.2%, adding about $1.38B and taking it to 1.69% of the book. With the stake already up 38.0% versus their average cost, they are clearly leaning into Azure and AI co-pilot economics rather than de-risking a winner.
• Nvidia: Shares are up +22.4%, roughly a $1.00B incremental bet, with the position now their largest single name at 2.50%. Sitting on a 122.2% gain vs cost and still adding, AQR is effectively saying the GPU cycle is nowhere near done.
• Broadcom and Alphabet: Broadcom’s stake grew +37.8% (about $512.3M added) and Alphabet’s GOOGL line +28.1% (about $354.2M). That’s a paired bet on networking, custom silicon, and the cloud/ads flywheel that finances AI capex.
• Adobe and Snowflake: Adobe is up +81.6% in shares (about $459.7M added) despite being down 29.0% from AQR’s cost; Snowflake is up +49.9% with a modest unrealized loss as well. AQR is explicitly averaging down into large-cap software names reset by market skepticism on AI monetization pace.
• Costco and Booking: Costco’s shares jumped +211.9% (about $695.6M added) and Booking +79.0% (about $582.9M). These aren’t cheap hedges; they’re high-multiple, global, premium-demand franchises — a vote that the top of the consumer stack remains durable.

Taken together, the adds show AQR is not diversifying away from growth volatility after a negative quarter; it is concentrating into AI rails and asset-light platforms with long duration cash flows.

If the buys are about owning AI and premium demand, the trims reveal what AQR views as expendable ballast. The pattern is consistent: monetize defensives and mature growers, free capital for higher-beta compounders.

• Walmart and PepsiCo: AQR cut Walmart shares by -22.9% (about -$351.5M) and PepsiCo by -11.1% (about -$113.3M), even though both sit comfortably above cost. That’s a straightforward rotation out of low-vol retail and staples just as they’re doing what they’re supposed to do.
• Big pharma: Merck was reduced by -23.3% (roughly -$315.3M), with Johnson & Johnson trimmed -8.6% (about -$81.2M). Both positions remain nicely in the green, but health care’s share of the book slid from 13.70% to 12.74%, showing a gentle de-emphasis on traditional defensives.
• Insurance and Berkshire: Chubb was cut -9.5% (around -$130.4M), and Berkshire Hathaway is down -2.0%. AQR is not abandoning insurance ballast — Travelers and PNC were actually increased — but it is pruning where valuations look full.
• Mature tech and pandemic winners: Meta is down -10.4%, Verisign -6.7%, and Zoom cut aggressively by -31.4% (about -$303.5M). All three are solidly profitable versus cost, yet AQR is reallocating from cash-generative, lower-growth or de-rating stories into AI-levered names that have sold off.

These trims don’t read as a macro de-risk, but as a deliberate raise of the portfolio’s growth and duration profile: out of safe cash cows and into more volatile, longer-tail opportunities.

The sector chart confirms what the individual trades hint at: AQR is letting technology pull more of the load while modestly shrinking classic shock absorbers. Tech’s share in this slice of the book rose from 41.04% to 43.23%, driven not just by the megacaps but by semis (Nvidia, Broadcom, Micron, Western Digital) and software/data platforms.

Consumer Discretionary ticked up to 16.63% from 16.34%, but the mix tilted more premium. Adds in Costco, Booking, TJX, Amazon, FedEx, and Airbnb offset a sizable Walmart trim, shifting exposure toward global and higher-margin demand rather than mass-market defensiveness.

Health care slipped from 13.70% to 12.74% as AQR rotated within the sector: out of Merck and Johnson & Johnson, into managed care (Centene, Cigna) and high-growth biotech like Eli Lilly and Regeneron. Finance eased from 5.09% to 4.60%, with Chubb trimmed but Travelers and PNC scaled up, consolidating into names that can still grow book value at compelling rates.

Consumer Staples fell meaningfully, from 1.91% to 1.45%, on PepsiCo selling. Utilities, Industrials, and Basic Materials (via Newmont) remain steady ballast, with small net reductions but notable adds to industrial winners like Comfort Systems, EMCOR, and defense primes such as Lockheed and General Dynamics. The portfolio is still diversified, but the marginal dollar is clearly going to compute, cloud, and premium services rather than traditional safety trades.

This quarter’s shifts show AQR behaving less like a purely systematic value-and-momentum shop and more like a conviction allocator around structural profit pools. The message: they believe AI infrastructure, cloud software, and premium global consumer franchises will out-earn classic defensives over the next leg of the cycle.

They are willing to average down in high-quality software (Adobe, Snowflake) and to extend into already very profitable trades (Nvidia, Apple, Broadcom, Alphabet). That combination tells you they see recent volatility in AI and growth software as a price reset, not the end of a regime.

On the risk side, the book carries less downside cushion from staples, pharmacies, and low-beta tech than it did a quarter ago. The remaining ballast is more targeted — insurers with pricing power, utilities with regulated returns, gold exposure via Newmont, and durable industrial cash generators like Comfort Systems and EMCOR.

For observers, the key is to watch whether AQR keeps pressing this AI-and-premium-demand tilt if markets stay choppy. If they continue to trim mature defensives and pandemic-era tech winners to fund semis, cloud, and experiential consumer names, you’re looking at a sustained shift in how this franchise expresses its macro and factor views.

Conversely, any future rebuild in Consumer Staples or a swing back into big pharma would signal a more cautious stance on growth and on the durability of the AI capex cycle. For now, the 2026-Q1 13F says AQR is comfortable being paid — or punished — for being early and large in the economics of compute.