AI, geopolitical shifts, and the new cost of capital

Investors have grown accustomed to navigating a wall of worry. This year alone, markets have absorbed geopolitical conflict, volatile energy prices, fiscal concerns, changing expectations for central banks, and persistent questions over whether AI is fueling a productivity revolution—or the next bubble.

Yet through it all, economic growth and corporate earnings have remained remarkably resilient. 

The challenge for investors therefore goes beyond spotting the next source of uncertainty. It requires identifying which developments are temporary and which represent genuine regime change. 

On that score, one conclusion stands out: the cost of capital matters again.

A broader set of forces

Two powerful forces have shaped 2026: AI and geopolitics. 

One is driving an extraordinary investment and productivity cycle. The other is accelerating fragmentation, energy insecurity, defense spending, and economic nationalism. Increasingly, the two intersect. 

Hyperscalers are expanding investment in data centers, power, networking, and compute capacity, while governments are simultaneously borrowing to fund defense, energy projects, infrastructure, and industrial policy.

Corporates and sovereigns are thus competing for the same pool of capital, with important consequences for interest rates and asset valuations. 

Much of the post-2008 period was defined by abundant liquidity, quantitative easing, and exceptionally low rates. Today’s environment is very different. 

Long-term government bond yields are now influenced by a much broader set of forces, from fiscal deficits and persistent inflation to stronger nominal growth, changing international capital flows, and the financing demands of AI infrastructure. 

Put simply, there is no single culprit behind higher yields. 

That means investors should spend less time fixating on the next 25-basis-point move from a central bank and more time considering the full range of factors shaping the yield curve. 

Inflation’s long shadow

Geopolitics is reinforcing this shift. 

The U.S.-Iran conflict has demonstrated how quickly energy security can move from a background concern to a major market driver. Even when the initial shock to crude oil fades, the effects could linger in refined products, transport costs, and supply chains. 

The longer-term consequences may prove more important than the immediate price move. 

As a result, countries are reassessing strategic autonomy across critical sectors and supply chains. Greater redundancy can make economies more resilient, but it isn’t the cheapest solution. Higher defense spending, reshoring, and duplicate supply networks all require capital and can add to fiscal and inflationary pressures. 

At the same time, global monetary policy is becoming less synchronized. 

The world’s economies now face very different combinations of growth, inflation, and fiscal constraints. That points to greater central bank divergence, less predictable correlations between asset classes, and potentially more interest rate volatility. 

Of course, this environment wouldn’t be entirely negative. Greater dispersion across countries, yield curves, and sectors can create more attractive relative-value opportunities for skilled investors. 

Is AI a bubble—or a revolution?

Nowhere is the tension between opportunity and risk more visible than in AI. 

There is little dispute that the underlying technology is transformational. Investment is spreading beyond semiconductors into memory, networking, power, cooling, and data centers, with capital increasingly channeling through credit markets (Fig. 1).

Yet whether the extraordinary amount of investment will generate sufficiently attractive returns is still up for debate. 

Headline market valuations appear expensive by long-term standards. On the other hand, the composition of the U.S. market has changed significantly. Technology companies now account for a much greater share of earnings, and margins have structurally increased, making valuation more nuanced than a single price-to-earnings multiple might suggest. 

The bigger risk may instead lie in the cycle itself. 

The largest tech firms cannot afford to fall behind in a major technological shift. But while accelerating AI investment can be rational for an individual company, it can still lead to excess capacity at an industry level. 

That is where the next phase of AI becomes more selective. The practical question investors should ask is: Where will the next dollar of AI investment earn an attractive return? 

From concentration to selectivity

The AI opportunity should not be treated as a single trade. 

Bottlenecks migrate. What is scarce today can become tomorrow's overcapacity. Investors should therefore be nimble in rotating within AI segments as the opportunity set evolves—from memory, networking, optics, and power to beneficiaries in areas such as cybersecurity and biotech, as well as businesses able to convert AI adoption into stronger revenues, margins, and productivity. 

Markets are already becoming more discerning, demanding evidence of monetization, cash generation, and return on capital. 

The same logic applies across asset classes. 

In fixed income, yields remain far more attractive than they were for most of the previous decade, even if credit spreads look relatively tight. However, higher sovereign issuance and greater rate volatility argue for selectivity around duration and credit risk. 

In equities, resilient earnings remain supportive, but elevated valuations leave less room for disappointment. Diversification beyond the largest U.S. technology companies—into emerging markets, small-caps, and second-order beneficiaries of the investment cycle—looks increasingly important. 

Conviction in a changing regime

Markets may be moving from chaos toward greater clarity, but clarity does not mean certainty. 

The investment regime is becoming more capital intensive, more geopolitically fragmented, and potentially more inflationary. That should create greater dispersion between economies, companies, and asset classes. 

The implication for portfolios is straightforward: stay invested, stay diversified, and stay active.

In a world where the price of capital matters again, conviction will come less from predicting every headline and more from identifying the structural shifts beneath them.