Our 2026 Market Outlook

(A forecast. Which is a terrible thing to write.)

No one knows the future.

If anyone tells you they do, they are either selling something or auditioning to look foolish later.

That said, history leaves clues. Patterns repeat. Incentives rarely change. Human behavior is remarkably consistent, even when the technology is not. We may not know exactly what will happen next year, but we can study what has happened before and identify what is most likely to move markets.

Forecasting is awkward.

Get it right and people call you a genius.

Get it wrong and the internet never forgets.

Still, I think it is important to share how I see the world heading into 2026, not to predict headlines, but to help clients and readers understand the forces shaping the global economy. This is not about trading the next six months. It is about context.

Here are three things I am watching most closely.

1. The Silicon Fault Line

Why Taiwan, China, Japan, and South Korea matter more than most people realize

Every era has a hidden dependency that only becomes obvious during stress.

In the 1970s, it was oil.

In 2008, it was housing leverage.

Today, it is advanced semiconductors.

Taiwan is often discussed as a geopolitical risk. That framing is incomplete. Taiwan is a systemic economic chokepoint.

More than 90% of the world’s most advanced chips are manufactured in Taiwan. These are not the chips in household appliances. These are the chips that power AI models, advanced data centers, modern defense systems, and high-end industrial equipment.

For perspective, before Russia invaded Ukraine, Europe relied on Russia for roughly one-fifth of its energy supply. Markets assumed that dependence was manageable. Semiconductor concentration today is significantly higher.

The U.S. is not dependent on Taiwan for all chips, but it is deeply exposed where it matters most. Nearly half of U.S. logic chip imports trace back to Taiwan once indirect supply chains are included. Logic chips are the brains of modern computing. There is no quick substitute.

South Korea completes the picture. At the cutting edge of manufacturing, the world effectively operates on a Taiwan–South Korea duopoly. Together, they account for nearly all leading-edge production capacity.

Japan is the quiet pillar many people overlook. Japan dominates critical semiconductor materials such as photoresists and silicon wafers. Advanced chipmaking does not function without them. If Taiwan is the factory, Japan supplies the ingredients.

This is why any disruption around the Taiwan Strait would not stay regional. It would ripple through global manufacturing, AI development, defense spending, and financial markets at the same time.

Markets tend to price geopolitics as binary events. Supply chains rarely behave that way.

2. The AI Capital Cycle

From earnings stories to balance-sheet stories

AI is no longer a hype cycle. It is a capital cycle.

Since late 2022, the largest technology companies have spent well over a trillion dollars building AI infrastructure. For most of this period, the story was simple and comforting. These investments were largely funded through operating cash flow. Strong earnings absorbed the cost.

Historically, that is unusual.

In past technology booms, leverage showed up early. This time, it arrived later.

What is changing now is how AI is being financed. Large data center projects are increasingly being funded through structured financing, project debt, and private credit vehicles rather than quietly absorbed into annual capex budgets.

First, it shifts AI from an earnings conversation to a balance-sheet conversation. Financing introduces sensitivity to interest rates, refinancing risk, and capital market conditions.

Second, it widens exposure. Once AI infrastructure looks more like utilities or energy projects, it attracts insurers, pension funds, and private credit at scale. The risk and opportunity spread well beyond Big Tech shareholders.

Capital cycles usually do not end because demand disappears. They end because financing conditions change. We are not there yet, but 2026 may be the year investors start paying closer attention to the plumbing behind AI growth.

3. Policy Whiplash

Tariffs, courts, and unexpected cash flows

Tariffs rarely work the way people expect.

Despite aggressive headline rates, the actual price impact of recent U.S. tariffs was muted. After supply chain rerouting, margin compression, and exclusions, import prices rose far less than advertised. A service-heavy economy helped cushion the blow.

The wildcard for 2026 is not new tariffs. It is old ones.

There is a real possibility that U.S. court rulings could force partial refunds of previously collected tariffs. If that happens, the amounts could reach into the tens of billions of dollars.

Whether that becomes an economic tailwind depends on where the money lands. If refunds stay on corporate balance sheets, the effect is muted. If they flow into prices, wages, or investment, it becomes a quiet stimulus.

Markets struggle to price legal outcomes because they do not fit neatly into economic models. But history shows that policy-driven cash flows can move cycles when they arrive unexpectedly.

The common thread across these three forces is concentration.

Semiconductor production is geographically concentrated.

AI investment is financially concentrated.

Policy risk is legally concentrated.

None of this guarantees disruption. But when multiple forms of concentration overlap, resilience matters more than prediction.

This outlook is not about calling the year perfectly. It is about understanding the terrain. Helping clients make decisions with context, patience, and perspective.