In his latest memo, Shall We Repeal the Laws of Economics – Part III, dated September 22, 2026, Howard Marks addresses the interplay between rising demand for capital, higher interest rates and inflation. These dynamics are being driven by a strong economy, massive investments in AI, and substantial government debt and deficits. Some thoughts: Howard Marks highlights the sharp rise in the yield on the 30-year U.S. Treasury bond, which reached 5.3% on August 17, 2026. Since then, as of October 12, 2026, the 30-year U.S. Treasury yield has climbed further to 5.6%, its highest level in 24 years. With U.S. federal debt now exceeding $40T and substantial government deficits (~$2T for FY2026, ~6% of GDP, with ~$1.3T in interest costs), inflation in the U.S. is further being fuelled. This is happening at a time when the economy is running at near capacity (e.g., with an unemployment rate of only ~4%). The incremental liquidity created by this situation (i.e., government spending > tax revenue) stimulates economic growth and, in turn, inflation. At the same time, enormous investments are being made in AI. In recent years, these investments have largely been financed through companies' existing cash balances and OCF, including those of hyperscalers. These companies are now increasingly raising capital themselves through the issuance of debt and equity. This further increases the demand for capital, and with it, the price of capital. Given the high levels of government debt and deficits, alongside interest expenses that continue to rise as a proportion, it is reasonable to question the sustainability of this trajectory. According to Howard Marks, a change in behaviour is the only long-term solution to prevent the situation from deteriorating further. The hope, therefore, rests on the U.S. government adopting a more fiscally responsible approach, including steering toward a lower growth rate in government spending than the growth rate of GDP. Hereby, AI could become part of the solution. AI could help accelerate U.S. GDP growth in two ways: first, by stimulating economic activity through AI-related investment and, second, by driving further productivity gains. Should we sell our stocks? This is a question Howard Marks was recently asked by a fellow entrepreneur. In Marks' view, this is not where the answer lies. The problems do not concern the U.S. stock market or U.S. companies, but rather U.S. fiscal management and, by extension, the U.S. dollar. According to Marks, a potential risk lies in the latter: investors could become less comfortable with the USD and, therefore, could begin seeking assets denominated in other currencies. However, doing so would generally entail a different set of risks. Personally, I believe it is useful to reflect on these issues from time to time. This in order to identify the potential second-order effects that may emerge. These could affect certain companies in our portfolios or on our watchlists. That said, investing ultimately comes down to allocating capital in businesses that can generate long-term TSR in excess of the rate at which the purchasing power of our capital erodes. After all, this is the foundation for increasing the purchasing power of our capital. History teaches us that economic growth, inflation and currencies can fluctuate considerably (with currencies even being replaced). In my view, the greatest risks do not lie in these factors themselves, but in the quality of the value-creating mechanisms—the businesses—in which we choose to invest. The best companies in the world will succeed in continuing to create value even during periods of severe monetary, fiscal or government-policy turbulence, drawing on the physical, digital and intangible assets they have built and continue to develop across the many countries in which they operate. Perhaps, these assets will become even more valuable during times in which the value of money appears to be under increasing pressure.

Great paper by Michael J. Mauboussin and Dan Callahan (Morgan Stanley, 2026) on FCF, ROIC, and ROIIC, the relationships between them, and the market’s current expectations for these metrics at hyperscalers. Some thoughts: - Periods of large negative FCFs resulting from capital expenditures (CapEx) can have an extremely positive effect on total shareholder return (TSR) if the return on investment (ROI) from these efforts exceeds the cost of capital. - The spread between ROI and the cost of capital determines whether it creates (+) or destroys (−) incremental intrinsic value for companies. - Ultimately, TSR depends on changes in extrinsic value, but over longer periods, it is strongly correlated with the trajectory of intrinsic value. Although negative FCF resulting from CapEx can create financing challenges, intrinsic value can soar if the cash is deployed in investments that generate returns above the company’s cost of capital (e.g., Walmart generated substantial negative FCF from 1973 to 1986 while delivering a TSR of 33% CAGR during that period). - However, if (parts of) large CapEx outflows do not immediately translate into incremental revenue, a company’s return on invested capital (ROIC) can temporarily decline as a result of lower return on incremental invested capital (ROIIC) driven by a time lag between investment and returns. This could happen to a hyperscaler such as Amazon in the coming quarters, given its substantially expanding invested capital (IC) base and the lead time before the resulting investments translate into net operating profit after taxes (NOPAT). - It is therefore important to estimate the total incremental NOPAT generated over the lifetime of an investment, rather than judging CapEx solely by its immediate impact on KPIs such as FCF or annual ROIIC. In other words, it is important to look beyond the horizon to assess companies’ long-term value-creation mechanisms. - Meanwhile, we should remain vigilant about automatically endorsing large investments simply because they have a positive net present value (NPV). Changes in a company’s value-creation mechanisms and their potential impact on its overall valuation are even more important. - As investors, our job is to determine whether the market overvalues or undervalues companies’ FCF prospects, the underlying capital expenditures required to achieve them, and the returns those investments are likely to generate. After all, alpha is created by identifying mispricing between the intrinsic value and the market’s extrinsic valuation of a company.