Harris Associates L P 13F Portfolio

The biggest adds tell one story: Harris wants long-duration, high-ROE compounders in the financial and payments ecosystem, plus a few mispriced software platforms.

On the financial infrastructure side, they made outsized dollar commitments to names that monetize volumes and volatility rather than pure credit spread:

• Marsh & McLennan (MRSH) was ramped by +68.1%, a roughly $586.4M add, turning it into a core fee-based risk and insurance platform holding.
• Charles Schwab (SCHW) and Intercontinental Exchange (ICE) both saw ~+11% and +10.5% share increases, respectively, each with >$210M in fresh capital, reinforcing a structural bet on brokerage, custody, and exchange economics.
• Bank of America (BAC) and Capital One (COF) were lifted by +22.1% and +11.0%, signaling Harris still sees upside in scaled deposit franchises and card/consumer credit despite big gains from their low-$20s and low-$100s average costs.

Outside pure finance, they’re rebuilding exposure to high-quality software at less-loved entry points. Salesforce (CRM) was increased +8.3% even though it sits about -20.0% below Harris’ average cost, and Adobe (ADBE) was lifted +23.3% while still slightly underwater. That looks more like averaging into core enterprise franchises than performance-chasing.

They also leaned heavily into Sysco (SYY), up +48.0% (about $441.6M more), a quiet but telling conviction that foodservice distribution remains a pricing-power story in a still-tight labor and supply environment. Rounding it out, Alphabet’s non-voting GOOG line was more than doubled (+118.4%, +$371.0M), underscoring that they still want exposure to the Alphabet engine, just in a different capital-allocation wrapper.

The funding sources for this push into financial plumbing and selective software are exactly where you’d expect from a disciplined value shop: mega-cap tech winners, energy, and fully rerated financials.

Alphabet’s GOOGL line is the poster child: Harris cut the position by -25.4%, freeing up an estimated $760.8M even as the name sits roughly +356.6% above their cost basis. They didn’t abandon the story — the add to GOOG shows that — but they are clearly rotating out of their fattest embedded gains.

In cyclicals and defensives, the scissors are sharp. ConocoPhillips (COP) was trimmed -12.9% (about -$233.9M) and Targa Resources (TRGP) -9.0% (about -$210.7M), reducing exposure to energy price risk after strong appreciation. State Street (STT) and Citigroup (C) were cut by -22.5% and -5.6%, respectively, shedding roughly $307.2M combined from banks where multiples have already normalized and returns on equity may be near peak.

Health care is also a clear source of cash. Centene (CNC) was slashed -21.1% (about -$190.6M), and smaller trims in Merck, Elevance peers, and Zimmer Biomet indicate Harris is less eager to pay today’s multiples for managed-care and pharma than for fee-based financial franchises. Add smaller nicks in travel-exposed Delta (DAL) and various banks, and the pattern is consistent: de-risk cyclicals and regulated reimbursements, keep the structural growers.

The sector chart understates the real story. Finance edges up to 36.71% of the book from 35.81%, but the composition is tilting decisively toward exchanges, brokers, insurers, and diversified fee platforms.

Within that financial bucket, Harris is building out an ecosystem: Schwab, ICE, Nasdaq (NDAQ), Marsh & McLennan, Willis Towers Watson (WTW), Equifax (EFX), and Carlyle (CG) create a mesh of brokerages, data, risk intermediaries, and alternative asset managers. Regional and money-center banks still matter — COF, BAC, Wells Fargo (WFC), First Citizens (FCNCA) — but the marginal dollar is going to fee and flow rather than pure balance-sheet duration.

Health care steps down from 15.09% to 14.51%, mainly via cuts to Centene and small trims in pharma and medtech, suggesting less comfort with reimbursement and pricing risk versus more transparent fee structures elsewhere. Energy-linked exposure (COP and TRGP) falls from 2.86% to 2.44%, despite strong absolute P&L, which looks like Harris acknowledging how much of the easy money in the oil and midstream trade is already captured.

Consumer Discretionary ticks up to 14.1% from 13.64%, but again via services and distribution rather than fashion or hardware: Sysco, Airbnb (ABNB), Amazon (AMZN), Netflix (NFLX), and Aercap (AER) dominate. Technology, at 12.53% versus 12.66%, is being quietly rebalanced — less concentrated in one Alphabet line, more spread across Salesforce, Adobe, Roper (ROP), CDW, and TE Connectivity (TEL).

Taken together, Harris Associates’ 2026 Q2 moves sketch a clear forward thesis: own the rails of the financial and digital economy, not the most volatile end-customers or commodity exposures.

On the financial side, they now have a deep bench of names that earn from volumes, volatility, and client assets more than from directional market calls — exchanges, brokers, global insurers, and scaled banks with strong funding. That mix should, in their view, compound through a wide range of rate and credit regimes, especially if capital-markets activity stays elevated.

In technology and consumer, the emphasis is on scalable platforms with recurring revenue and strong ecosystems that have seen some multiple compression: Salesforce, Adobe, Netflix, and the two-share-class Alphabet structure. The decision to average into underwater software tells you Harris is comfortable underwriting a multi-year earnings and margin story, not just next quarter’s guide.

The trims in energy, health insurers, and certain banks suggest they see those trades as largely played out at current prices, at least relative to the opportunity in financial infrastructure and high-quality software. For anyone reading their 13F tea leaves, the message is straightforward: when the cycle gets noisy, Harris wants to sit on the toll booths, not the traffic.