While an enormous share of market attention continues to focus onartificial intelligence, data centres and technology companies that could benefit from this transformation, some of the world's best-known investors are making decisions that deserve careful consideration. Berkshire Hathaway holds an impressive amount of liquidity, but has also resumed deploying capital in new transactions. Michael Burry continues to challenge widely held beliefs, while Bill Ackman selects a relatively small number of businesses on which to base his return expectations. Reading these moves as a single signal of distrust in technology would, however, oversimplify matters: their approaches differ, and some of their holdings remain heavily exposed to innovation.
The temptation is to ask immediately: ‘What are major investors buying? Should I buy it too?’. Individual transactions, however, can change quickly and often become public only several weeks later. Nor do we always know the average purchase price, hedges, objectives or all the other investments held by the same investor. What we can try to understand is how they think when deciding where to allocate capital: what value they assign to a business, which risks they accept and how much they are willing to pay for future results.
Portfolio information arrives with a delay
Before discussing names and figures, a clarification is needed. In the United States, many major investment managers disclose certain holdings through Form 13F, a periodic document filed with the SEC, the authority overseeing US financial markets. The usual deadline is within 45 days of the end of the quarter, with adjustments for holidays and weekends. The document provides a snapshot of positions at the reporting date: it does not show the portfolio in real time, does not cover every type of investment and does not report short sales. An accurate list of securities may therefore tell only part of the strategy. The SEC's guidance clearly explains these limitations.
As of 30 September 2026, therefore, we cannot treat a position reported for 30 June as if it had been verified this morning. The same applies to a statement attributed to an investor: it may describe a general belief or a decision made under conditions that have since changed. Following these documents is useful for studying a method; using them as a shopping list requires much greater caution.
Berkshire: almost 365 billion dollars, but also new investments
Let us start with Berkshire Hathaway, the company built over more than sixty years by Warren Buffett and now led by Greg Abel. Since 18 September 2026, Buffett has been chairman emeritus and remains on the board, while Howard Buffett has taken over as chairman. This distinction matters: discussing Buffett's historical approach makes sense, but does not mean automatically attributing every current company decision to him. The change is set out in Berkshire's official announcement.
As of 30 June 2026, adding cash, cash equivalents and short-term US Treasury securities and subtracting purchases of those securities that had yet to settle gives approximately 364.7 billion dollars. This is the consolidated figure that can be reconstructed from the quarterly financial statements, rather than money sitting entirely idle in a current account. Comparison with March also changes the significance of the news: the balance fell from around 380.2 billion. In the second quarter, Berkshire made net equity purchases of almost 20 billion and repurchased approximately 4.5 billionof its own shares, as reported by Reuters on 8 August.
These figures do not describe a company that has simply sold everything while waiting for a crash. My interpretation is that caution and willingness to invest can coexist. A large reserve allows a company to meet operating and insurance obligations and seize exceptional opportunities, but holding it also has a cost: liquid assets may earn less than productive investments over the long term. Size also matters. A transaction attractive to a small saver may be too small to affect Berkshire's results; a group of this size needs investments worth billions to make a real difference.
This is the concept often summed up by the expression ‘dry powder’, meaning capital available when an opportunity arises. In September 2008, during the financial crisis, Berkshire agreed to invest 5 billion dollars in Goldman Sachs preferred shares, carrying an annual dividend of 10% and additional rights to purchase common shares. The terms are described in the transaction announcement. Having resources during a difficult period can provide strong negotiating power, although investing during a crisis still carries risks.
US debt: distinguishing data from forecasts
The United States' fiscal position also forms part of the background. This is contextual information, not a proven explanation for Berkshire's decisions. Different figures are often placed side by side in this discussion: debt is an accumulated stock, whereas the deficit is the shortfall generated during a particular period. Moreover, debt held by the public excludes debt held by other federal government entities but includes securities held by the Federal Reserve; it therefore differs from gross federal debt and does not mean only debt owed to citizens and private investors.
For a recent reference point, in the report published on 9 September 2026 , the Congressional Budget Office, Congress's budget analysis office, estimates a deficit of around 2,000 billion dollars in the first eleven months of fiscal year 2026. This covers October 2025 to August 2026, rather than the entire year. The CBO's monthly review must be distinguished from the projections published in February, which indicated approximately 7,400 billion in spending, 5,600 billion in revenue and 1,900 billion in deficitfor the full fiscal year. These summary figures are rounded separately, so subtracting them does not reproduce the reported deficit exactly.
In that same CBO scenario from February 2026, debt held by the public rose from 101% of GDP in 2026 to 120% in 2036, while net interest increased from 3.3% to 4.6% of GDP. These are projections based on particular economic and legislative assumptions, not outcomes already realised. The message remains relevant: persistent deficits can increase debt and the interest burden, leaving less room for other budget decisions. They are not enough, however, to establish when the stock market will fall or how an individual share will move.
The idea that the Federal Reserve can simply create money and make the problem disappear also misrepresents how the system works. The Fed and the Treasury have distinct responsibilities, and the central bank explains that it buys Treasury securities in the market without directly subscribing to new issues. A Fed purchase does not make government debt disappear. Monetary policy can influence interest rates, credit and demand, but it does not eliminate the problem of a budget that keeps spending more than it receives.
Apple: even a great business can account for too much of a portfolio
In recent years, Berkshire has substantially reduced its stake in Apple. This does not, however, justify concluding that it sees no prospects for the company. At 30 June 2026, Apple remained its largest listed equity holding, valued at around 66 billion dollars. Over the same period, Alphabet had risen to third place, at around 37.8 billion: another reason to avoid reading these decisions as a general flight from technology. The figures are reported by Reuters based on portfolio disclosures.
An investor may sell an excellent company for several reasons: excessive concentration, its valuation, liquidity needs, tax considerations or alternative opportunities. These are possible motives, not a list of causes we can confidently attribute to every Berkshire sale. The general principle nevertheless matters: an excellent business is not automatically an excellent investment at any price. The P/E, the ratio of share price to earnings per share, connects what we pay with the company's results. It can be calculated using historical earnings or expected earnings, which remain estimates.
Apple's results help illustrate the distinction. In its third fiscal quarter of 2026, ended 27 June, the company recorded 109.4 billion dollars in revenue, up 16%, a gross margin of 50.1% and diluted earnings per share of 2.02 dollars, up 29%. The Apple announcement filed with the SEC states that tariff refunds contributed approximately 2 percentage points to gross margin and 0.11 dollars to earnings per share. Results are strong, but components that might not recur also need to be distinguished. Assessing the investment therefore requires both the quality of the company and the price demanded for a share in its future results.
BYD: even a successful investment can come to an end
The story of BYD, the Chinese electric vehicle manufacturer, offers another example. Berkshire began investing in the company in 2008, started reducing its stake in 2022 and completed its sale in 2025. The energy subsidiary's disclosure showed a zero value for the holding as early as the end of March 2025, as reported by Reuters in September of the same year. This marked the end of a highly successful investment, not necessarily a mistake needing correction.
Selling does not mean disowning the original decision. The balance between opportunity and risk can change: a sector becomes more competitive, margins come under pressure, valuations incorporate higher expectations or political uncertainty increases. In early 2026, Reuters described weak sales and competitive pressure in China's automotive market. This context came after Berkshire's exit and does not prove its motives. The broader lesson is that risk also depends on the environment in which a company operates: competition, rules, tariffs and international relations can affect results as much as management's abilities.
Michael Burry: challenging the consensus
Michael Burry became famous for betting against the US mortgage market before the financial crisis. His case requires an update, however: Scion Asset Management closed at the end of 2025. On 28 September 2026 , Reuters reported his role as senior adviser to Minerva Investment Management, a new fund seeking short-selling opportunities, including those arising from risks in private credit, meaning credit extended through institutions and instruments outside traditional public bond markets. This is news about his current role, not a complete picture of his personal investments.
The interesting feature of his historical approach is his willingness to defend an analysis that differs from the prevailing one. Investing is described as contrarian when it seeks value in situations the market views negatively or challenges expectations considered overly optimistic. This may lead to buying an overlooked business, but also to developing a bearish thesis. In both cases, the quality of the analysis matters more than simply enjoying going against the crowd: the consensus can be wrong, but it can also be right even when it seems excessive to us.
A share that has crashed is not automatically cheap
A stock that loses 50% is not necessarily a bargain. Imagine, in a purely illustrative example, a share that traded at 100 euros and now costs 50. If the estimated value of the business has meanwhile deteriorated enough to justify only 30 euros per share, the current price remains high relative to that estimate. The fall tells us the market is paying less than before; it does not tell us whether it is paying little. The task is to distinguish a business facing temporary difficulties from one steadily losing customers, margins or competitive strength.
Before buying after a crash, therefore, concrete questions are needed: how much could the company earn in five or ten years? Is its debt sustainable? Does its competitive advantage still exist? Are there financial maturities that might force it to raise capital on unfavourable terms? Buying when others are afraid works only if the price provides adequate compensation for the risk. Some companies recover, others remain weak for years and others do not survive. Going against the crowd is not, by itself, a valuation method.
What is the money the company will generate in the future worth?
This is where Discounted Cash Flow, usually abbreviated to DCF, comes in: a model that estimates the present value of future cash flows. The principle is simple: receiving money tomorrow is not equivalent to receiving it today, because in the meantime we forgo other opportunities and face the risk that expectations will not materialise. Discounting means bringing those amounts back to the present using a rate consistent with time and risk. In a simplified example, 100 euros expected in one year, discounted at 10%, are worth approximately 90.91 euros today, because dividing 100 by 1.10 gives that result.
The model applies this reasoning to cash flows over many years and to a value assigned to the business beyond the period analysed. Consistent measures are essential, however: valuing cash flows available to all providers of finance gives a business value from which shareholders' value can be derived, accounting for debt, cash and other adjustments. Directly valuing cash flows intended for shareholders follows a different route. Taking a generic item called ‘cash’, multiplying it and dividing it by the number of shares is not enough.
DCF does not deliver a truth accurate to the last cent. Changing growth, margins, required investment or the discount rate can substantially alter the result. Claims such as ‘undervalued by 300%’ are also misleading: if estimated value were four times the price, the potential increase from the price to that value would be 300%, but the price's discount to value would be 75%. The denominators differ and, above all, value remains an estimate. The model is useful when it makes assumptions explicit and allows cautious and favourable scenarios to be compared, not when it gives a mathematical appearance to a conviction already formed.
The ability to generate cash creates more possibilities
The term free cash flow generally means cash remaining after the investment required by the business; the precise definition depends on the model used. It does not automatically equal accounting profit. A company can increase sales and report profits while absorbing substantial liquidity in trade receivables, inventory or plant. Revenue growth, although important, is therefore insufficient to demonstrate value creation: we must understand how much capital it requires and how much money actually remains available.
A company capable of generating cash can finance projects, reduce debt, acquire other businesses, distribute dividends or carry out buybacks, meaning repurchases of its own shares. If repurchases genuinely reduce the number of outstanding shares, each remaining share represents a larger stake in the business. The economic benefit nevertheless depends on the price paid, the financial strength remaining and the alternatives: repurchasing overpriced shares can destroy value, while new issues may offset part of the effect. Cash creates possibilities; how it is deployed determines the outcome.
Bill Ackman: a concentrated portfolio expanded in 2026
Another approach is that of Bill Ackman, founder of Pershing Square. In the half-year report published in August 2026, the manager disclosed six new investments: Visa, Mastercard, Netflix, S&P Global, Intercontinental Exchange and Alcon. All also appear in the Pershing Square Holdings fact sheet as of 31 August 2026, alongside Microsoft, Amazon, Meta, Uber, Fannie Mae and Freddie Mac, among others. Dates matter: the half-year report also includes updates after 30 June, and not all new positions were already held at the half-year end.
In the letter to investors, Ackman and chief investment officer Ryan Israel connect the decisions to business quality, cash generation and expected earnings growth. These are the manager's assessments, not guaranteed outcomes. The strategy remains concentrated compared with a global equity index fund: if assumptions prove correct, the best ideas can have a substantial impact on returns; if they are wrong, mistakes also carry more weight. Strong conviction alone does not reduce risk, and owning companies in different sectors does not rule out some depending on the same economic factors.
The lesson that really matters to me
Berkshire maintains a large reserve but resumes buying, Burry challenges widespread expectations and Ackman expands a portfolio built on selective convictions. These moves do not tell a story of a shared flight from artificial intelligence or provide a certain forecast of the stock market's next move. Instead, they show different ways of tackling the same problem: deciding what return to expect from capital, taking account of the price and what could go wrong. Even an experienced investor can misjudge valuation, timing or position size.
For me, this is the value of observing these investors. I may find a company they buy interesting, but before investing I must be able to explain to myself why the price makes sense, what assumptions I am making and how heavily an error would weigh on my portfolio. Their resources, constraints and operating terms differ from mine. The buyer's fame is no substitute for analysis. Studying their reasoning can help me ask better questions; the final decision must still be consistent with my objectives, time horizon and the risk I can bear.
Dates and reporting periods. Sources checked as of 30 September 2026. Berkshire data as of 30 June; Apple results for the fiscal quarter ended 27 June; Pershing Square Holdings portfolio as of 31 August. The first-eleven-month deficit is a CBO estimate published on 9 September; the fiscal scenario through 2036 is the one from February 2026. Disclosed positions do not represent portfolios in real time.