Balyasny Asset Management L P 13F Portfolio

The biggest dollar add is SPY, up +618.6% in shares with a ~$554.0M capital swing, and IVV enters as a new ~$532.1M line. That is Balyasny saying: stop overfitting to single-name factor noise, own the S&P, and express edge at the margins.

Under the hood, they are rebuilding their AI exposure away from the headline trade. Texas Instruments jumps by +7954.6% in shares to ~$448.9M, and Amphenol explodes +7406.9% to ~$256.7M — clean bets on the analog, connectivity and power-management hardware that actually moves bits and electrons.

They also lean into utilities and industrial defendants of the grid. AES is up +209.7% to ~$507.7M, and Digital Realty is effectively built from scratch, rising +23311.0% in shares to ~$202.6M — an obvious tie-in to data-center demand from AI workloads.

The other interesting build is in financials. US Bancorp is a new ~$267.9M position, HBAN is up +167.5%, and BLK grows +59.3% — a composite view that higher‑for‑longer rates plus ETF proliferation favor well‑run regional and money‑center banks and the asset‑gatherers who intermediate all this beta.

Finally, they add growth optionality in high-operating-leverage names where expectations have reset. ROKU’s stake is multiplied by +1277.1% to ~$434.8M, and EA is up +404.8%, a paired expression that streaming and interactive media monetization still have legs once the macro panic and ad‑cycle fear wash out.

The sell tape is blunt: reduce concentration in crowded, fully priced winners and recycle into cheaper, more cyclical risk. NVIDIA is cut -23.1% even though it still sits on a +91.5% gain vs cost; AAPL is down -42.3%, GOOGL -24.6%, and META -45.7% despite those being the canonical AI and internet bellwethers.

That’s not a “we hate AI” call; it’s “this part of AI already priced in.” They are keeping meaningful exposure but clearly think the second derivative — semicap, analog, bandwidth, data centers — offers a better risk/reward than the front-page GPUs and ad platforms after a multi‑year tear.

Trims in Datadog (-38.2%) and Snowflake’s relative underweight versus its growth in other software names show the same logic: lighten high‑multiple cloud monitoring and data platforms after a doubling-type move (DDOG is up +101.7% vs cost) and replace with more balanced software like Microsoft and ServiceNow, both aggressively increased.

Outside tech, they are similarly pragmatic. RTX gets halved (-51.6%), Norfolk Southern is reduced, and ETSY is cut -41.0%; these are funding sources where they still have gains or where idiosyncratic overhangs (rail accidents, aerospace complexity, small‑seller consumer sensitivity) cap near‑term upside.

The sector chart shows a deliberate re‑balancing. Technology still dominates at 30.84% of the book, but that’s down from 38.34% as they cash in some of the AI and software trade and redeploy into what looks like a more macro‑aware structure.

Unclassified — effectively ETFs and Berkshire — jumps from 2.51% to 9.81%, encapsulating the new SPY, IVV and XLF muscle. Financials move from 2.43% to 5.25% on HBAN, USB and BLK, while Health Care rises from 2.44% to 5.79% with a massive build in 3M and added Penumbra and Apogee.

Utilities tick up from 4.62% to 5.62% via AES and TXNM, signaling a preference for regulated, rate‑linked cashflows that still participate in electrification and data‑center demand. Industrials, conversely, drop from 16.24% to 10.9% as they trim RTX, Parker‑Hannifin and Thermo Fisher, migrating from capex‑heavy cyclicals toward more balance‑sheet‑resilient plays.

Consumer Discretionary dips from 28.74% to 24.49%, but that’s nuance, not abandonment. They’re exiting advertising‑dependent or balance‑sheet‑stressed names like WBD while reinforcing Costco, O’Reilly, UAL, Airbnb, EA and Spotify — a barbell of high‑income, services‑led and digital‑subscription demand rather than pure mall traffic.

Real Estate nudges slightly down to 2.5% even as DLR balloons, because ETSY is mis‑shelved there in the data and gets cut. The implicit message: own the infrastructure (data centers) not just the small‑seller platforms renting space atop it.

Pulling the moves together, Balyasny is setting up for a regime where macro volatility stays high, AI winners broaden along the stack, and cash-generative incumbents outperform moonshots. The new ETF spine gives them flexibility to dial exposure up or down quickly without chewing through single-name liquidity.

The AI stance is evolving from “own the GPU and the ad platform” to “own the power grid, interconnects, memory and data centers that make AI possible.” That shows up in the shifts from NVDA, META and Datadog into TXN, ASML, Amphenol, Seagate and Digital Realty.

On the defensiveness side, the growth in utilities, banks and health care tools implies a base case of higher‑for‑longer rates and persistent nominal growth, but not a deep recession. These are sectors that historically handle inflationary churn better than richly valued SaaS or story‑driven consumer names.

Yet this is not a bunker portfolio. Roku, EA, Spotify, Airbnb, Take-Two and UAL give them upside convexity if the US consumer holds together and digital entertainment and travel continue to compound; Amazon and Costco remain core as the most efficient distribution platforms in the system.

Expect Balyasny to keep toggling between index beta and targeted themes as the macro narrative shifts. The 2026‑Q2 13F suggests they are less interested in calling the exact AI champion and more interested in owning the infrastructure, cashflows and liquidity that will matter across several possible paths for the next cycle.