Newsletter commentary Mar 2026

Very Little Can Be Predicted, Only Choices Can Be Made

By the end of February, amid ongoing negotiations, the United States and Israel initiated military action against Iran, triggering the closure of the Strait of Hormuz, disruption of key Middle East energy supply routes, and severe repercussions for the global economy. The economic impact intensified nonlinearly as the conflict persisted. This unjustified war, launched by the U.S. and Israel, inflicted widespread damage on the world economy.

Iran responded with resilient and asymmetric countermeasures, which proved both swift and effective. Despite its conventional military disadvantage, Iran succeeded in drawing the conflict into a protracted stalemate. Having entered the war with inadequate preparation and strategic foresight, Trump’s plan for a swift resolution collapsed. The conflict evolved very differently from previous engagements such as in Venezuela. The emerging deadlock, combined with the escalating global energy crisis, placed Trump in an increasingly difficult position. Whether to further escalate militarily or seek a face-saving exit became a pressing and fraught dilemma.

This conflict stands apart from all prior engagements, unfolding within a deeply interconnected global economy. Every nation and economic actor would be impacted by the disruption of Middle Eastern energy supplies. Much like the Renesas chip plant shutdown in Malaysia reverberated through the global auto industry during the pandemic, the severance of these energy routes would trigger countless similar and far larger shocks across sectors worldwide. Unlike the Russia-Ukraine war, the economic fallout from this scenario would be orders of magnitude greater. Beyond the three direct participants, the U.S., Israel, and Iran, every other economic entity, from corporations and individuals to other nations, would begin exerting pressure to shape the course of the conflict. An immediate cessation would swiftly become a unifying and urgent global demand.

We did not take the correct action at the outset. As markets adjusted, the portfolio experienced a large drawdown, mainly driven by materials. We wavered between expectations and reality. When facing irrational actors who have a significant impact on us, our seemingly rational analysis may in fact be irrational. A significant reduction in exposure would have been the best course of action, and clearly our response was not strong enough.

Looking ahead, the duration of the conflict would bring nonlinear accelerations in economic damage. What appears advantageous today could turn to loss tomorrow due to collapsed demand. Investment judgments even slightly longer-term have become exceedingly difficult. Certain structural opportunities, such as distributed and diversified energy systems, may require enduring severe volatility along the way. Whether the war ends soon or drags on would lead to dramatically different paths for medium-term energy prices. The manner of its conclusion would also significantly reshape future trade flows. Productive assets may be re-understood and revalued in its aftermath. We believe the U.S.'s crude unilateralism cannot resolve complex international issues and has cost it moral authority and credibility. Since the trade war began, the reliability and capability of the U.S. in global politics have been steadily questioned—a skepticism that will reverberate through financial markets in the medium to long term.

Meanwhile, the AI narrative has shown tentative but noteworthy shifts in March, or at least early signs of possible evolution. Questions are emerging: Could “agent + skills” serve as a viable paradigm for an extended period in the absence of AGI? If mutual distillation among global large models becomes unavoidable, how do we balance input and output? Could agents become the primary interface and starting point of our work and daily life? The token surge brought by agentic systems would impose stricter cost constraints. Porsche is exceptional, but sometimes a tractor can also carry people. Will NVIDIA’s comprehensive leadership in software ecosystems, single-chip performance, supply chain, and scaling face renewed cost-performance scrutiny? Will the high margins seen across the supply chain in recent years, driven by NVIDIA’s aggressive strategy and soaring demand, come under pressure? AI is now moving decisively out of labs and models into every industry. As inference gradually accounts for a larger share of activity, the value chain of the entire sector will be reconfigured.

We hold a simple intuition: if economic activity is divided into physical production, akin to moving bricks, and services that facilitate the circulation of those bricks, then in the AI era, the former may have limited room for compression, while the latter could face profound and sweeping efficiency gains.

The current investment landscape stands at the intersection of two transformative forces: near-term uncertainty arising from shifting geopolitical dynamics, which is part of a once-in-a-century transition, and the accelerating advance of artificial intelligence, which represents a once-in-a-millennium breakthrough. The distribution of profits, wages, taxes, and depreciation within the economy may undergo significant restructuring in a relatively short period, while the adaptation of human institutions and governance tends to follow a more gradual pace. This dynamic suggests that portfolio volatility could increase, and we are deliberately choosing the range of outcomes we are prepared and positioned to navigate.

We maintain the view that global economic and political pressures will continue to exert a restraining influence, encouraging reasoned behavior, including that of the United States and its leadership, and reducing the likelihood of escalation toward broader confrontation.