The Macro Regime: Where Are We Now?

Growth has been more resilient than expected

The post-pandemic economy has repeatedly surprised investors with its ability to absorb shocks. Supply-chain disruption, aggressive monetary tightening, geopolitical uncertainty and higher trade barriers have all created reasons to expect a more pronounced slowdown, yet economic activity has generally proved more durable than those risks initially suggested.

Part of that resilience comes from the unusual strength of investment associated with artificial intelligence. The construction of data centres, expansion of electricity infrastructure, demand for advanced semiconductors and the capital expenditure programmes of the major technology companies have created a substantial investment cycle. The BIS argues that this AI investment boom helped offset some of the drag from higher tariffs during 2025, while the IMF estimates that technology investment related to AI contributed around 0.5 percentage points to US GDP growth during the year.

This matters for investors because the AI cycle is no longer simply a story about technology stocks. It has become part of the macroeconomic cycle itself. Semiconductor manufacturing, data-centre construction, power generation, networking equipment and industrial infrastructure all sit somewhere in the investment chain, meaning that a continuation of AI capital expenditure can support activity well beyond the companies developing the underlying models.

The uncertainty lies in how long that investment impulse can continue and, eventually, how much of the spending translates into measurable productivity and corporate cash flow. Large capital expenditure programmes can support GDP in the short term even before the assets being built generate an adequate economic return. That distinction is becoming increasingly relevant as the scale of investment grows.

The bond market deserves more attention

For much of the post-2008 period, investors became accustomed to an environment in which inflation was low, policy rates were unusually subdued and long-duration assets benefited from declining discount rates. That backdrop created a powerful tailwind for both government bonds and growth equities, particularly companies whose expected cash flows were concentrated far into the future.

The current environment is less forgiving. Even when central banks are no longer aggressively tightening policy, long-term government bond yields can remain elevated because investors are pricing in higher inflation risk, stronger nominal growth, larger fiscal deficits or simply a greater supply of government debt that the private sector needs to absorb.

That distinction between short-term policy rates and long-term borrowing costs is important. A central bank can reduce its policy rate while long-dated yields remain stubbornly high if investors believe that inflation will remain above target, fiscal policy will remain expansionary or the supply of government bonds will continue to increase.

The BIS has increasingly focused on this fiscal-financial interaction. Its 2026 report points to near-record public debt and a growing role for non-bank financial institutions in sovereign debt markets, arguing that the relationship between public finances and financial stability can amplify the transmission of market stress.

For investors, the practical implication is that the bond market should not be viewed simply as a place to observe interest-rate expectations. Long-term yields influence mortgage costs, corporate financing, equity valuations and the relative attractiveness of cash versus risk assets. A sustained increase in real yields can therefore change the investment landscape even when corporate earnings remain healthy.

Fiscal policy is becoming harder to ignore

The other major shift is the growing importance of fiscal policy. Government borrowing is not a new issue, but the combination of elevated debt levels, ageing populations, higher defence spending and significant investment requirements makes fiscal policy increasingly relevant to asset prices.

In an environment of low inflation and abundant demand for government bonds, large deficits can sometimes be absorbed without creating much immediate pressure on yields. The adjustment becomes more difficult when investors begin to demand greater compensation for holding long-duration sovereign debt. At that point, fiscal policy and monetary policy become intertwined through the bond market.

This does not necessarily imply an imminent fiscal crisis. In developed economies, the adjustment mechanism is usually considerably slower and more complicated than the language surrounding sovereign debt sometimes suggests. Governments have substantial taxing and spending powers, debt is typically denominated in domestic currency, and the investor base is deep.

The concern is more gradual: higher debt servicing costs can reduce fiscal flexibility, making governments more sensitive to movements in interest rates. If yields rise at the same time that economic growth weakens, policymakers can find themselves facing conflicting objectives. Supporting demand may require more fiscal stimulus precisely when borrowing becomes more expensive and investors are becoming less comfortable with additional debt issuance.

That is one reason the direction of long-term yields deserves close attention even if the immediate inflation data appears benign.

AI is both an economic opportunity and a financial risk

Artificial intelligence adds another layer of complexity because the technology has the potential to improve productivity while simultaneously encouraging an unusually large investment boom.

The IMF's 2026 analysis describes AI as capable of transforming productivity, investment and labour markets, but also highlights the possibility that expensive, increasingly debt-financed investment could produce disappointing returns. If expectations become embedded in asset prices faster than the underlying economic benefits materialise, the eventual adjustment could affect both corporate valuations and financial conditions.

There is already evidence of how large the investment cycle has become. The BIS estimates that the five largest hyperscalers were set to spend more than $1 trillion on AI-related capital expenditure across 2025 and 2026, with commitments in some cases running ahead of earnings and free cash flow.

None of this means that the AI investment thesis is wrong. In fact, the economic potential may be considerable. The challenge for investors is that a technology can be genuinely transformative while individual investments made in anticipation of that transformation still produce poor returns. The history of technological investment contains plenty of examples where the technology ultimately changed the economy but the capital allocated during the early boom was not always deployed profitably.

That distinction between technological progress and investment returns is likely to become increasingly important as the AI ecosystem matures.

What would change the current picture?

Macro regimes rarely change because of one data release. They usually evolve through a combination of forces, and investors are better served by watching the relationships between variables than by attempting to predict individual economic numbers.

For the current environment, several relationships appear particularly important. One is the connection between long-term bond yields and fiscal expectations. Another is the relationship between AI capital expenditure and the revenue, earnings and productivity gains generated by that spending. A third is the interaction between economic growth and inflation: if growth remains strong enough to support demand but productivity improves rapidly, the economy may be able to sustain stronger nominal activity without producing the same inflationary pressure that would otherwise accompany it.

There is also a less obvious risk on the other side. If AI investment becomes sufficiently large that it accounts for an important share of corporate capital expenditure and economic growth, a slowdown in that spending could eventually have a disproportionately large macroeconomic effect. The same investment cycle that currently supports growth could become a source of weakness if expectations change abruptly.

That does not require an outright collapse in AI demand. A moderation in growth expectations, a reduction in capital expenditure plans or a reassessment of the returns available on new data-centre capacity could be enough to change the marginal direction of markets.

The regime is more complicated than the headline numbers suggest

The current macro environment therefore sits somewhere between the familiar post-pandemic narratives. Inflation is no longer behaving as it did during the initial reopening shock, but neither can investors assume that the low-inflation environment of the 2010s has simply returned. Growth has remained resilient, but some of that resilience is being supported by an investment cycle whose sustainability is still being tested. Monetary policy remains important, yet long-term bond yields and fiscal expectations increasingly determine financial conditions independently of central-bank decisions.

For investors, this creates an environment in which the interaction between markets may be more important than any individual asset class. A move in Treasury or gilt yields can change equity valuations; changes in equity valuations can influence corporate investment; a shift in AI investment can affect semiconductor demand, industrial spending and energy requirements; and changes in fiscal expectations can feed back into bond markets.

The result is not necessarily a hostile environment for risk assets. It is simply one in which the old assumption that strong economic growth automatically supports higher asset prices is less reliable. Valuations, financing conditions and the durability of the investment cycle matter alongside the direction of GDP.

The next phase of the cycle will probably become clearer as investors gain more evidence about whether today's enormous AI investment programme is translating into sustained productivity and cash-flow growth, while the bond market continues to test how much additional government borrowing it is prepared to absorb at current yields. Those two questions are increasingly connected, and together they may tell us more about the macro regime than the next monthly inflation print on its own.