Berkshire Hathaway-A

Berkshire Hathaway reported a few more changes to its 13F equity investment portfolio for the second quarter, including selling stakes in more holdings we have historically ascribed to Todd Combs and Ted Weschler.

Why it matters: Berkshire's quarterly 13F filing has historically provided insight into the company's stock trading activity in the US markets. As of the end of June 2026, Berkshire had $299.3 billion in 13F reportable equity securities, including its US holdings but excluding foreign investments held abroad, such as the insurer's stakes in the Japanese Sogo Shosha trading houses. Including those holdings, the insurer had $344.3 billion in stock holdings at the end of the second quarter. Based on Berkshire's recent 10-Q filing, the company recorded net purchases of $19.8 billion (comprising $3.7 billion in reported sales and $23.5 billion in purchases) during the June quarter.

The bottom line: With Greg Abel taking over as CEO at the start of the year, we expected Berkshire to winnow the stock portfolio to fewer, more concentrated holdings. The firm moved faster than we expected, eliminating holdings we had ascribed to Todd Combs (who left at the end of last year) and to Ted Weschler during the first quarter, with a few holdovers sold in the June quarter. Among the major changes made during the second quarter were the sale of 30.2 million shares of Bank of America (for an estimated $1.6 billion), 9.3 million shares of Lennar Class A ($830 million), 4.2 million shares of Capital One ($795 million), and 11.0 million shares of Kroger ($700 million). Major purchases during the June quarter included another 24.5 million shares of Alphabet Class A (for an estimated $7.9 billion), 23.6 million additional shares of Alphabet Class C ($7.6 billion), 17.5 million additional shares of Delta Air Lines ($1.4 billion), and 12.4 million additional shares of Lennar Class B ($1.1 billion).

We already knew about Berkshire's additional commitment to Alphabet, which reached an agreement with the insurer to sell $5 billion of its Class A shares to Berkshire at $351.81 per share and another $5 billion of its Class C stock at $348.20 per share in early June. So, based on the fact that Berkshire bought even more tells us that the company was either already adding to its Alphabet stake before the announced deal or added more shares after the June transaction.

Berkshire also added to its stakes in Macy's (buying 4.3 million additional shares for an estimated $90 million) and New York Times Cl A (553,465 additional shares for $45 million), while making a new money purchase in D.R. Horton (buying 3,564 shares for an estimated $0.5 million). This was not, however, Berkshire's first purchase of D.R. Horton; it acquired 1.5 million shares during the first quarter of 2025, which it sold off by the end of the third quarter of last year.

We also question the move, given that Berkshire had acquired homebuilder Taylor Morrison Home Corporation for close to $6.8 billion in an all-cash transaction at the end of May 2026 (which closed near the end of July), which led us to expect the firm to sell off stakes in other homebuilders—like Lennar and NVR. But it looks like the company basically swapped nearly all of its Lennar Class A holdings for Class B shares during the second quarter, continued to hold NVR, and reintroduced D.R. Horton to the portfolio.

For the rest of the company's sales during the June quarter, Berkshire completely eliminated its stake in Constellation Brands (selling 632,890 shares for an estimated $90 million) and trimmed its stake in Ally Financial (2.0 million shares for $85 million). All in, Berkshire reported an estimated $18.1 billion of stock purchases during the second quarter in its 13F filing, which was $5.4 billion less than the insurer reported in its 10-Q filing. We already know that Berkshire bought shares in Tokio Marine for around $1.8 billion during the June quarter, with the remaining $3.6 billion likely invested in the Japanese Sogo Shosha trading houses.

The company's US stock investment portfolio remained fairly concentrated at the end of the second quarter, with its top 20 positions accounting for 99.8% of the 13F portfolio. Berkshire held 26 stock positions, with Class A and Class B shares counted as a single holding, at the end of the June quarter of 2026. This was down from 39 stocks at the end of last year.

As for portfolio dynamics, the top five stock positions in the US stock portfolio at the end of the second quarter of 2026—Apple (22.0% of the portfolio), American Express (17.1%), Alphabet (12.6%), Coca-Cola (10.9%), and Bank of America (9.2%)—accounted for 71.9% of the insurer's stock holdings (up from 67.1% at the end of March 2026 and 70.9% at the end of the fourth quarter of 2025).

Berkshire's top 10 holdings—which included the top five holdings above, and Chevron (4.7%), Occidental Petroleum (4.3%), Chubb (3.9%), Moody's Corporation (3.7%), and Kraft-Heinz (2.6%)—accounted for 91.0% of the insurer's stock holdings at the end of the June quarter (up from 90.7% at the end of the first quarter of 2026 and 88.3% at the end of last year).

From a sector-allocation perspective, the financial services sector accounted for 30.9% of the 13F portfolio at the end of the second quarter of 2026 (down from 38.2% at the end of December 2025 and 36.6% at the end of June of last year—largely due to Berkshire's sale of large chunks of Bank of America shares, as well as other financial services holdings, this year). Technology stocks, meanwhile, accounted for 22.8% of the portfolio at the end of the June quarter (down from 23.4% at the end of last year and 23.8% at the end of the second quarter of 2025—as gains in the company's Apple stake offset share sales during the back half of last year).

This left financial services and technology as the insurer's two largest sector bets, accounting for 53.7% of the US stock portfolio at the end of the second quarter of 2026, down from 61.6% at the start of the year, as well as 60.4% at the end of the June quarter of 2025. Further down the list were communication services, which hit 14.8% of the portfolio at the end of the June quarter (up from 3.6% at the end of last year and 1.7% at the end of the second quarter of 2025), mostly due to the large investments in Alphabet in the first half of this year.

Consumer defensive names accounted for 14.2% of the portfolio at the end of the second quarter of 2026 (down from 14.9% at the end of last year and 16.5% at the end of the June quarter of 2025). Energy holdings also diminished, representing 9.0% of Berkshire's US stock holdings at the end of June 2026 (down from 11.2% at the end of 2025 and 11.1% at the end of the second quarter of 2025). The rest of the $299.3 billion 13F equity portfolio at the end of the June quarter consisted of holdings in the industrials (5.7%), healthcare (2.1%), consumer cyclical (0.5%), and basic materials (0.1%) sectors, as defined by Morningstar.

For more details on our current thinking and key concerns about Berkshire Hathaway, please see our recent special report, "Berkshire Hathaway Enters a New Era With Greg Abel at the Helm," published on May 1, 2026. We also published another special report, "Abel Puts His Imprint on Berkshire's Stock Portfolio," which took a deeper look at the insurer's stock trades in the first quarter, as well as its valuation using the look-through earnings approach.

Our fair value estimate for Berkshire Hathaway remains in place at $765,000 per Class A after updating our forecasts for the company's operating businesses and insurance investment portfolio to incorporate changes since our last revision. Our valuation is equivalent to 1.45 and 1.35 times our estimates for Berkshire's book value per share, respectively, at the end of 2026 and 2027. For some perspective, during the past five (10) years, the shares have traded at an average of 1.51 (1.45) times trailing calendar year-end book value per share. We use a 9.0% cost of equity in our valuation and assume that Berkshire, at the very least, pays the 15% corporate alternative minimum tax on adjusted financial statement income.

Our fair value estimate is derived using a sum-of-the-parts methodology, valuing each of Berkshire's four operating segments separately and adding them back together for our firmwide estimate. After reviewing our projections for 2026-30, including updating the value of the insurance investment portfolio, we've lowered our valuation for Berkshire's insurance operations slightly to $393,900 per Class A share (from $398,900 previously).

While most of this was due to adjustments to earned premium growth and underwriting profitability, some of it was tied to expected changes in the value of the investment portfolio, as well as yield, over the next five years. Our forecast assumes the firm generates overall earned premium growth of 8.0% on average annually during 2024-28 (down from 8.3% previously), compared with 7.0% (8.0%) average annual growth during 2021-25 (2016-25), aided by stable pricing in commercial property and casualty lines and improved results from Geico.

We expect the insurance operations to post an average annual combined ratio of 93.6% during 2026-30 (a slight adjustment from 93.2% previously), compared with 93.3% (96.2%) during 2021-25 (2016-25), primarily driven by easing pricing benefits in most business lines, as well as an easing of improvements at Geico. We expect the insurance investment portfolio (including market gains/losses and additional investments) to expand at a 4.4% CAGR during 2026-30.

Our valuation of Berkshire's railroad operations remains at $95,300 per Class A share, with our forecast assuming carload volume increases at a 0.1% CAGR during 2026-30, compared with 0.3% (negative 0.6%) during 2021-25 (2016-25). With average revenue per car/unit expected to increase 2.2% annually on average during the next five years, we have freight revenue expanding 2.3% on average annually during 2026-30, compared with 2.9% (negative 0.8%) during 2021-25 (2016-25). We expect BNSF's operating ratio to reach 58.0% at the end of 2030, putting it more on par with Union Pacific, with the assumption being that Berkshire's railroad finally adopts precision scheduled railroading.

Our fair value estimate for Berkshire's utilities/energy division remains at $59,600 per Class A share. This valuation represents a slight premium to forecast EV/EBITDA multiples for most other large regional regulated utilities. Our forecast assumes continued constructive rate case outcomes, with adjusted EBITDA for the firm's regulated US utilities expanding 4%-6% annually on average during 2026-30. We also include continued annual charges for West Coast wildfire litigation in our valuation.

As for Berkshire's manufacturing, service, and retail operations, we've increased our fair value estimate to $216,200 per Class A share (from $211,200 previously) after reviewing our projections for 2026-30. We assume average annual revenue growth of 3.7% during 2026-30, relative to 8.6% (7.6%) during 2021-25 (2016-25). We also expect to see pretax operating margins of 8.5% on average annually over the next five years, compared with 8.6% (8.5%) on average annually during 2021-25 (2016-25).

Our bull-case fair value estimate of $956,250 per Class A share assumes Berkshire's insurance segment premium growth and underwriting profits, as well as returns for its investment portfolio, come in higher than our expectations. We also assume stronger economic growth, with Berkshire's two noninsurance segments—manufacturing, service, and retailing and railroad, utilities, and energy—benefiting. We also expect Berkshire to put more capital to work in acquisitions and other investments than in our base case.

Our downside case fair value estimate of $612,000 per Class A share assumes Berkshire's insurance segment does not perform as well as we project in our base case, with Geico continuing its downward trajectory, and softer pricing in the property and casualty market affecting results for both BHRG and BHPG. We also assume growth at Berkshire's two main noninsurance segments stalls, with faltering economic growth torpedoing the firm's more economically sensitive operations.

We've historically believed that Berkshire's economic moat is more than just a sum of its parts, although the parts that make up the whole are moaty in their own regard. The insurance operations—Geico, Berkshire Hathaway Reinsurance Group, and Berkshire Hathaway Primary Group—remain important contributors to the overall business. Not only are they expected to account for 46% of Berkshire's pretax earnings on average during the next five years (and 51% of our firmwide valuation), but they are overcapitalized (maintaining a larger-than-normal equity investment portfolio for a property and casualty insurer).

They also generate low-cost float, which is the temporary cash holdings that arise from premiums being collected in advance of future claims. This allows Berkshire to generate returns on these funds with assets that are commensurate with the duration of the business being underwritten. And they have tended to come at little to no cost to Berkshire, given the company's proclivity for generating underwriting gains over the past several decades.

That said, we don't believe the insurance industry is particularly conducive to the development of maintainable competitive advantages. While there are some high-quality firms in the industry (with Berkshire having some of the best operators in the segments where it competes), the product that insurers sell is basically a commodity, with excess returns difficult to achieve on a consistent basis. Buyers of insurance are not inclined to pay a premium for brands, and the products themselves are easily replicable.

Competition among insurance firms is fierce, and participants have been known to slash prices or undercut competitors to gain market share. Insurance is also one of the few industries where the cost of goods sold (signified by claims) may not be known for years, providing an incentive for companies to sacrifice long-term profitability in favor of near-term growth. In reinsurance, this dynamic can be even more pronounced, as losses in this business tend to be large in nature and may not be realized for many years after a policy is written.

Insurers can develop durable cost advantages by either focusing on less commodified areas of the market or developing efficient and/or scalable distribution platforms. What they can't do is develop a competitive advantage through investing, even when gains are the result of the investing prowess of someone like Buffett. We believe insurers that consistently achieve positive underwriting profitability are better bets in the long run, as insurance profitability tends to be more durable than investment gains.

Given the growth of its auto insurance operations over the years, Geico has become one of the largest generators of earned premiums for Berkshire. The strength of the auto insurer's direct-selling operations has made it the third-largest US private passenger auto insurance underwriter, responsible for 11.6% of written premiums last year, compared with industry leader State Farm at 18.6%.

Much like its closest competitor, Progressive (which accounted for 18.6% of written premiums in the US during 2025), Geico has set itself apart by its scale in the direct-response channel. While scaling is typically difficult for insurance companies, personal line insurers like Geico and Progressive have done a better job of spreading fixed costs over a wider base, as their business models do not require as much human capital and specialized underwriters as other insurance lines.

Both firms are at the forefront of the ongoing shift into direct business from agent-derived business, and have had similar levels of underwriting profitability during the past decade. Geico is expected to produce an average annual combined ratio, including the impact of hurricanes and other natural disasters, of 90.6% during 2026-30 (ending 2030 at 93.1%) compared with Progressive at 92.9% with its direct operations (and 92.9% for all of its personal auto lines). We believe Geico, much like Progressive, has a narrow economic moat.

As for Berkshire's reinsurance arm, we believe BHRG has at best a narrow economic moat around its business. For a premium, reinsurers assume all or part of an insurance or reinsurance policy written by another insurer. While any insurance company can provide reinsurance, a handful of larger companies—Munich Re, Swiss Re, Berkshire Hathaway, Hannover Re, and Scor—hold sway over the lion's share of global reinsurance premiums underwritten. These policies often contain large, long-tail risks that, when priced appropriately, can generate favorable long-term returns.

That said, reinsurers compete almost exclusively on price and capital strength, making it almost impossible to build structural cost advantages. Losses in the reinsurance market are also lumpy and may not be realized for years after a policy is underwritten, magnifying the importance of disciplined and accurate underwriting. While Berkshire believes its catastrophe and supercatastrophe underwriting can generate solid long-term results, the volatility of these business lines, which have the potential to subject the firm to especially large losses, tends to be high.

Although we don't normally view reinsurers as benefiting from favorable competitive positions, there are some specialty lines where a long history of underwriting incidence and/or unique relationships has allowed a firm to build a maintainable competitive advantage. We believe Berkshire's reinsurance operations are unique. The company's overall balance sheet strength makes it capable of taking on large amounts of supercatastrophe underwriting (covering events like terrorism and natural catastrophes) that few companies have the capacity to endure, allowing them to name their price.

Berkshire has also historically had the luxury of walking away from business when appropriate premiums cannot be obtained, something its publicly traded peers cannot always do. While underwriting profitability has been less consistent because of the nature of the risks BHRG is underwriting, the company sticks with reinsurance, even if it proves to be unprofitable from time to time, because it generates float that can be invested for longer periods of time than short-tail business lines like auto insurance.

BHPG has been Berkshire's most consistently profitable insurance business over the past two decades, and we believe the segment has developed a narrow economic moat. What is even more remarkable about this is that BHPG is a conglomeration of more than a handful of different insurance operations, including Berkshire Hathaway Specialty Insurance, Berkshire Hathaway Homestate Companies, MedPro Group, Berkshire Hathaway Guard Insurance Companies, National Indemnity's primary group, and US Liability Insurance. These entities offer commercial insurance coverage as varied as healthcare malpractice, workers' compensation, automobile, general liability, property, and various specialty coverages for small, medium, and large clients.

By focusing more on specialty lines that require extensive experience or unique relationships to underwrite effectively, BHPG has been able to put together a continuous record of solid earned premium growth and underwriting profitability, which is a rarity in the insurance business; most P&C insurers are willing to take underwriting losses from time to time in order to generate earned premium growth, believing that they can make up the difference with investment gains.

Of the more than 75 noninsurance businesses that make up Berkshire's remaining businesses, Burlington Northern Santa Fe and Berkshire Hathaway Energy are lumped together under the railroad, utilities, and energy segment in Berkshire's financial statements. While their contribution to pretax earnings and our own fair value estimate for the firm is now overshadowed by the manufacturing, service, and retailing segment, they are far more transparent than the company's other operating segments. On a combined basis, BNSF and BHE are expected to generate 23% of Berkshire's pretax earnings on average during the next five years and contribute 20% to our firmwide fair value estimate.

The most interesting thing about these two businesses is that neither was a major contributor to Berkshire's pretax earnings over a decade ago. Buffett's shift into such debt-heavy capital-intensive businesses as railroads and utilities represented a marked departure from many of Berkshire's other acquisitions over the years, which tended to require less ongoing capital investment, had little to no debt, and produced higher returns on average. Still, had Buffett focused more on buying asset-light companies with fewer capital investment needs, this would have left his successors with even greater amounts of cash on the balance sheet to deal with.

During 2016-25, Berkshire generated an average of $19.6 billion annually in free cash flow. The amount of excess cash Buffett would have needed to find a home for would have been more than 50% higher had Berkshire purchased similar-size companies to BNSF and BHE with similar cash flow profiles that were not investing $12.1 billion annually on average in their combined property and equipment the past decade.

With BNSF, which was acquired in full in February 2010, Berkshire picked up a Class I railroad operator—an industry designation for a large operator with an extensive system of interconnected rails, yards, terminals, and expansive fleets of motive power and rolling stock. We believe that all the major North American Class I railroads benefit from colossal barriers to entry due to their established, practically impossible-to-replicate networks of rights of way and continuously welded steel rail. While barges, ships, aircraft, and trucks also haul freight, railroads are by far the lowest-cost option when no waterway connects the origin and destination, especially for freight with low value per unit weight.

Customers also have few choices and thus wield limited buyer power, with most Class I railroads operating as duopolies (and some being a monopoly supplier) to end clients in many markets. This provides the major North American Class I railroads with efficient scale. Believing that operators like BNSF will continue to leverage their competitive advantages of low cost and efficient scale to generate returns on invested capital in excess of the firm's cost of capital, we have awarded them wide moat ratings.

We think Berkshire Hathaway Energy is endowed with a narrow economic moat. Buffett built up this business through investments in MidAmerican Energy (supplanting a 76% equity stake taken in 2000 with additional purchases that finally lifted Berkshire's interest to 100% in 2024), PacifiCorp (acquired in full in 2005), NV Energy (acquired in full in 2013), and AltaLink (acquired in full in 2014). While BHE has picked up pipeline assets, which have wide-moat characteristics, most of its revenue, profitability, and ongoing capital investment is driven by its three main US-regulated utilities: MidAmerican Energy, PacifiCorp, and NV Energy.

We do not believe regulated utilities can establish more than a narrow economic moat, even with their difficult-to-replicate networks of power generation, transmission, and distribution. This is because their rates and returns are set by state and federal regulators. That said, we feel BHE has benefited greatly from being part of Berkshire's larger consolidated tax return, as well as from not paying a dividend to the parent company (with most of its publicly traded peers paying out as much as 60% of earnings as dividends annually). This has allowed BHE to invest far more capital (more than $20 billion during the past decade) in renewables than it could have on a stand-alone basis.

Berkshire's manufacturing, service, and retailing operations are now one of the largest contributors to pretax earnings, expected to account for 31% of pretax earnings on average annually for the next five years (and 29% of our estimate of the company's fair value). Given the lack of transparency into these operations, getting a handle on the profitability and economic moats of the wide array of businesses in the segment is difficult at best. Unlike BNSF and BHE, both of which file quarterly and annual reports with the Securities and Exchange Commission, there is little financial information available on the firms in the manufacturing, service, and retailing segment.

But given Buffett's penchant for acquiring companies with consistent earnings power, generating above-average returns on capital, holding little debt, and run by solid management teams, we believe many of the businesses in the segment are endowed with narrow economic moats. During 2025, the five largest companies (on a pretax earnings basis) in the MSR segment—Precision Castparts, Lubrizol, Clayton Homes, Marmon, and IMC/ISCAR—accounted for around half of pretax earnings. Each of these subsidiaries, by our estimates, has a narrow economic moat. When combined with the next five largest subsidiaries—Shaw Industries, Forest River, Johns Manville, TTI, and MiTek—this collection of businesses accounted for around 70% of the MSR segment's pretax earnings last year, with a moat rating overall that skews to the narrow end of the spectrum.

With Berkshire being run on a decentralized basis, the managers of the company's operating subsidiaries have been empowered to make their own business decisions. In most cases, the managers running these subsidiaries are the same individuals who sold their businesses to Berkshire in the first place, leaving them with a vested interest in the businesses they run. Barring a truly disruptive event in their industries, we expect these firms to continue to have the same advantages that attracted Buffett to them in the first place.

That does not mean that there won't be subsidiaries whose competitive advantages diminish over time (exemplified by the demise of the textile manufacturer that Berkshire Hathaway derives its own name from). It's just that the large collection of moaty firms that reside in Berkshire's MSR operations is more likely to maintain a narrow economic moat in aggregate, even as a few firms along the way succumb to changing competitive dynamics in their industries.

The decentralization in Berkshire's operations (on top of a less-than-adequate level of transparency for many of its operating companies) leaves the firm a bit exposed to ESG-related risks. Overall management is generally weak at diversified conglomerates like Berkshire because firms that have been constructed in this way tend to face challenges in terms of applying suitable systems for managing their ESG risks across the entire company, especially given (in Berkshire's case, more than any other conglomerate) the diverse nature of the products and services that they offer.

While Berkshire's operating businesses have generally provided the firm with a narrow moat on a combined basis, it was management's ability to produce additional excess returns from the cash flows thrown off by its disparate operations that historically pushed our moat rating into wide territory. But it became increasingly harder to justify that moat rating, not just because Buffett's departure will likely dampen future investment returns, but because we have continued to see slippage in some of the moat sources that support the economic moats in a few of its main operating businesses.

Berkshire's record of finding ways to invest the excess cash provided by its operating subsidiaries in projects that have, on average, earned more than its cost of capital has gotten thinner over the years. The firm has not only been fighting with the sheer size and scale of its operations, which have required larger and larger deals (or stock investments) to be meaningful (especially with an increasingly constricted opportunity set) but has had to contend with a growing cache of private capital chasing deals that might have been attractive to Berkshire (and with less acquisition discipline than management has generally brought to the table).

That's not to say Berkshire won't be able to put money to work in value-creating projects. Rather, the huge and growing sums of capital the firm must deal with will ultimately limit its ability to generate outsize returns. While we expect the company to continue to have sufficient competitive advantages to maintain a narrow economic moat, we think that many of the contributors to its economic moat from its operating companies have been diminishing. Although we see the potential for Berkshire to get back on track through more directed operational efficiencies, as well as some financial engineering, none of which were happening on Buffett's watch, this will take some time to come to fruition.

Book value per share, which is a good proxy for measuring changes in Berkshire's intrinsic value, increased at an estimated 18.1% CAGR during 1965-2025, compared with a 10.5% annualized return for the S&P 500 TR index.

Berkshire's stock performance has generally been solid, increasing at a 16.8% (14.3%) CAGR during 2021-25 (2016-25), compared with a 14.4% (14.8%) average annual return for the S&P 500 TR index.

At the end of 2025, Berkshire had $176 billion in insurance float. The cost of the firm's float has generally been negative during much of the past two decades.

Given its size, Berkshire's biggest hurdle continues to be its ability to consistently find deals that not only add value but are large enough to be meaningful.

Another big issue that has faced the firm has been the longevity of Buffett, especially following the death of longtime managing partner Munger in November 2023.

Berkshire's insurance operations face competitive and highly cyclical markets that occasionally produce large losses, and several of its noninsurance operations are economically sensitive and focused on US markets.