1) One of my favorite stocks for the past three decades, Berkshire Hathaway (BRK-B), reported second-quarter earnings on Saturday (press release here and 10-Q here). So let's take a look...
It was a solid quarter, as operating earnings (excluding foreign exchange gains) rose 5.2%:
Berkshire provides comments on each business segment in its 10-Q:
After-tax earnings from insurance underwriting declined 13.1% in the second quarter and increased 3.6% in the first six months of 2026 compared to 2025... GEICO produced lower underwriting earnings in the first six months of 2026 compared to 2025 [$2.4 billion versus $4 billion], which were partially offset by increased earnings from reinsurance and other primary insurance business.
After-tax earnings from insurance investment income declined $308 million (9.1%) in the second quarter and $522 million (8.3%) in the first six months of 2026 versus the same periods in 2025, attributable to lower interest income, reflecting lower interest rates.
After-tax earnings of BNSF increased 6.3% in the second quarter and 9.5% in the first six months of 2026 compared to 2025. Earnings in 2026 benefited from higher shipping volumes and improved operating efficiencies, partly offset by increases in fuel costs and the impact of higher effective income tax rates...
After-tax earnings of BHE increased 26.9% in the second quarter and 11.5% in the first six months of 2026 compared to 2025, which reflected higher earnings from the U.S. utilities and natural gas pipelines businesses, partially offset by lower earnings from other energy businesses.
Cash flow from operations jumped 11.2% to $11.2 billion, capital expenditures rose from $4.9 billion to $5.6 billion, and free cash flow grew 6.6%.
There are hopeful signs here that CEO Greg Abel is starting to more tightly manage Berkshire's vast operations and improve profitability. (Though he has work to do at Geico, where pre-tax earnings declined 45% in the second quarter due to "higher claims frequencies and average severities.")
Berkshire's followers are also closely watching Abel to see what he might do with the company's huge cash hoard. It currently sits at a staggering $366 billion – larger than a majority of companies in the S&P 500 Index.
There was promising news on the capital allocation front, as Berkshire was a net buyer of stocks during the quarter for the first time in 14 quarters. It acquired $23.5 billion worth of shares – most notably, a $10 billion investment in Alphabet (GOOGL) – while only selling $3.7 billion worth.
On July 24, Berkshire also completed its acquisition of Taylor Morrison Home for $6.8 billion plus $1.7 billion of debt.
In addition, Berkshire bought back $4.2 billion of stock in the second quarter and an additional estimated $3 billion in July. This represents a sharp turn from recent years, as you can see in this chart:
As a result, Berkshire's share count dropped 0.75% in four months – from 2.16 billion at the end of March to 2.14 billion at the end of July. That's 2.3% annualized, which becomes meaningful over time...
I'm glad to see a resumption of share repurchases, especially since Berkshire had been trading at a meaningful discount to its intrinsic value, as I calculated in my May 5 e-mail.
Tomorrow, I'll share my latest calculation of Berkshire's intrinsic value, so stay tuned!
2) Last Tuesday, I analyzed the historical financials of software company Veeva Systems (VEEV) and concluded that they're among the best I've ever seen.
Yet I noted that the stock only trades at:
... 22.8 times this year's consensus analysts' estimates of $9.06 per share.
That's a market multiple for a far above-average business – just the kind of setup I like. But this assumes that AI doesn't impair the business, as many investors fear.
Many of my readers asked me for additional insights on Veeva. So I wanted to share three in-depth pitches I came across...
The first is by Daan Rijnberk, an independent investor who published this on Substack on June 20: Veeva Systems – A Deep Dive into a Misunderstood SaaS Stock, Now on Sale.
It was particularly well timed, as the stock was then at $153, close to a six-year low – it has since rallied to around $230 today.
Rijnberk notes that the stock had been crushed due to the AI-driven "SaaSpocalypse":
However, Veeva Systems doesn't fit the AI disruption story it's currently being priced for. Every dollar in pharma [research and development] and commercial operations runs through software that has to survive [a Food and Drug Administration] audit, and that single fact changes everything about how vulnerable a company actually is to a clever new agent or a horizontal competitor's AI push. Veeva isn't a generic [customer relationship management ("CRM")] vendor or a horizontal productivity tool vulnerable to the next clever AI agent; it's the compliance backbone for an entire global industry, embedded so deeply into the regulated workflows of pharmaceutical and biotech giants that ripping it out isn't a software decision; it's a multi-month regulatory ordeal.
He argues that Veeva has strong growth prospects:
Today, 35 of the top 50 pharma companies globally use Veeva Vault across clinical, quality, regulatory, medical, and commercial processes, reflecting significant growth in product adoption in recent years. Most of these started with just CRM, but as Veeva rapidly expanded its product count over the last decade, its value to customers has grown significantly. And the runway remains large, so I expect this to remain its primary growth lever.
Positively, Veeva's cross-sell position remains unusually strong, as it has an existing relationship with virtually every large pharma and biotech company on earth, a product suite that spans nearly every regulated workflow those companies run (commercial, clinical, regulatory, quality, safety), and a platform architecture (Vault) where each additional module shares the same validated infrastructure as the last, meaning the marginal cost and friction of adding another product for the customer is far lower than evaluating and onboarding an entirely new vendor...
Meanwhile, Veeva reports it still has only penetrated 16% of its [total addressable market ("TAM")], suggesting there is plenty of room to grow with its existing product catalog through cross-selling... [The] company also continues to add products to its platform, further growing its TAM and addressable spend per customer.
As for the threat from AI, he believes that it "provides a significant opportunity for Veeva":
According to management, AI could boost efficiency in the life sciences industry by 15% by 2030 by addressing pain points in compliance-heavy workflows such as regulatory approvals, safety reporting, and clinical trials, meaning AI may expand the value of what Veeva already offers rather than replace it.
Veeva was also pitched twice on stock-idea website Value Investors Club. Both write-ups are only available to subscribers, so I'll include extended excerpts here...
The first is from user "Manchu" on November 22, 2025, who believes that Veeva "offers the opportunity to buy a market-leading software company with a long runway for growth at a long-term low relative valuation multiple."
At the time, investors were worried about the company "transitioning its Veeva CRM from the Salesforce platform to an internal platform." But Manchu argues that this risk may be overrated for multiple reasons:
From a market share perspective, Veeva believes it will have an opportunity to win back lost clients, as the Salesforce transitions may be "challenging." Veeva also has opportunity to gain share outside of the top 20, which contributes around half of the business.
Manchu shares Rijnberk's view that AI looks to be more of an opportunity than a risk:
The combination of an extremely tight regulatory/risk environment around customer data and security in the biopharma space, and the need for highly specialized data should support Veeva's incumbency benefits. Veeva plans to launch some AI products in the 2026-2027 timeframe and this could represent a possible revenue expansion opportunity in time.
Almost eight months later, with the stock down 23%, user "Lyncroft" pitched Veeva on July 9, arguing that it was cheap at around 15 times non-GAAP operating income for fiscal 2027:
That multiple is too low for a founder-led vertical software company growing revenue mid-teens with 40%+ non-GAAP operating margins. It is much too low if Veeva can turn its workflow position into a regulated labor automation platform.
Like Manchu, Lyncroft dismisses the "risk associated with Veeva moving customers from Salesforce-based Veeva CRM to Vault CRM":
Salesforce has won some large biopharma decisions and that transition will remain noisy. But Veeva is no longer a CRM story. CRM Suite was roughly 75% of revenue ten years ago, is about 20% today, and Veeva expects it to be about 10% by 2030.
Lyncroft is also bullish on Veeva's AI prospects:
In May 2026, Veeva announced Falcon, an agentic platform for drug development. The initial use cases are narrow: trial master file document intake and quality control, health-authority correspondence, and safety case triage and intake. Early-adopter availability is planned for November 2026...
Veeva is focusing Falcon on repetitive, document-heavy, compliance-heavy workflows. They are expensive to run manually and hard for generic AI tools to automate safely because the tool needs workflow context, user permissions, current documents, customer-specific rules, and an audit trail.
Lyncroft also highlights the track record of Veeva's co-founder and CEO, Peter Gassner:
He co-founded Veeva in 2007 and still runs the company. He beneficially owns 15.6 million shares, or 9.4% of the company. His compensation is unusually aligned. The board has emphasized long-term stock options over cash, and his 2024 CEO option grant has a $236.90 exercise price and is intended to be his only equity compensation until at least 2030. He does not participate in an annual incentive plan and, apart from a below-market salary, has not been paid cash compensation since the IPO.
Gassner has a strong history of meeting his long term goals. Veeva laid out a $1 billion revenue run-rate goal in 2015 and passed it in 2019, ahead of schedule. In 2019 he laid out a $3 billion 2025 revenue run-rate goal which again they beat. At the end of 2024 they laid out a 2030 goal of $6 billion in revenue.
According to Lyncroft, the main risk is if Veeva fails to own the agentic AI layer:
AI-native startups could own that layer while Veeva remains the system of record underneath it. That would still leave Veeva with a good software business, but it would cap the AI upside and likely erode pricing power over time.
The defense is that regulated agents need more than a model. They need the current document, the right workflow state, the user's permissions, the customer's rules, and evidence for auditors. Veeva has those assets because customers already run critical workflows inside Veeva.
Regarding valuation, Lyncroft thinks the stock could soar past $400:
The base business alone supports a solid return. If Veeva reaches its $6 billion 2030 revenue run-rate goal and earns a high-30s non-GAAP operating margin, it would produce roughly $2.3 billion of non-GAAP operating income. After tax, that is about $1.8 billion of earnings power. A low-20s multiple plus current cash supports an equity value of roughly $45-50 billion, or about $275-300 per share. That is 45-60% upside without giving Falcon much credit.
Falcon creates the chance for a much better outcome. If Falcon becomes a $500 million high-margin business, it can likely add $5-7 billion of equity value. If it becomes a $1 billion regulated labor platform, it could add $10 billion or more of value. That moves the stock from a base-case $275-300 outcome toward $330 in a good case and $400+ in a bull case.
I think Rijnberk, Manchu, and Lyncroft all make strong cases for Veeva...
My team and I at Stansberry's Investment Advisory will take a closer look at the company. If we decide to recommend it, as always, our subscribers will be the first to know. You can become one by clicking here.
Best regards,
Whitney
P.S. I welcome your feedback – send me an e-mail by clicking here.