Asia's capex super-cycle under higher interest rates and oil prices Report Interpretation

Overview

Morgan Stanley examines whether higher interest rates and oil prices could derail Asia's capex super-cycle. Its central conclusion is that these factors create downside risks to growth but should not end the business cycle under its base case, because borrowing conditions are not broadly restrictive, leverage is contained, and structural investment demand remains broad-based.

Core views

Morgan Stanley frames a typical end to a business cycle as a sequence in which excessive private-sector leverage and central-bank tightening push the cost of capital into restrictive territory, slow income growth, and trigger defaults. It argues that this sequence is not currently evident in Asia. Real policy rates in Asia have risen 50bp from the February 2026 trough, but remain 160bp below their February 2025 peak. The firm expects further Asian rate hikes, largely as counter-cyclical responses to stronger demand, inflation, or foreign-exchange pressures, rather than as tightening that makes capital broadly restrictive. The report's second support for continued expansion is that the capex cycle is still relatively early and has not displayed broad excess, particularly in Asia ex China. Capital-goods imports, a high-frequency capex indicator, are described as growing at their strongest pace since the mid-2000s. Morgan Stanley expects investment to generate jobs and income growth, while inflation and current-account indicators are not signalling macro instability. It argues that the cycle is supported by structural drivers, so it should be less sensitive to cyclical headwinds than a conventional late-cycle investment boom. Healthy balance sheets are the key buffer in the analysis. Corporate debt-to-GDP in Asia ex China stands at 73%, below its pre-Covid level; household debt ratios are also below pre-Covid levels. For comparison, US corporate debt-to-GDP is 76% and likewise below its pre-Covid level. Morgan Stanley characterizes the present expansion as unusual for its sixth year because the 2020 recession was an external shock followed by fiscal support, rather than a leverage-driven cycle in which households and corporates needed to borrow heavily. In its view, that leaves fewer balance-sheet vulnerabilities that could turn slower growth into widespread defaults. The firm identifies slower AI adoption as an endogenous risk. If adopters cannot generate meaningful productivity gains or take longer to deploy AI, compute demand could decelerate, reducing near-term AI capex. That could weaken Asia's technology exports and spill over to semiconductor manufacturers' investment. However, Morgan Stanley estimates that AI and semiconductor-related segments contribute about 11% of incremental Asian capex through 2030, so the bulk of the capex super-cycle rests on other structural demand drivers. It therefore sees a possible deeper slowdown under weaker AI spending, but not an end to the cycle while balance sheets remain healthy and the longer-term AI investment story remains intact. Oil is the principal exogenous risk. Morgan Stanley considers an oil-price rise to about US$110/bbl manageable because Asian governments can use subsidies, price caps, and other smoothing measures to share the burden with households. The report instead flags US$130-150/bbl sustained for six months or more as a threshold likely to produce a meaningful growth slowdown. Such an outcome would raise the oil burden, reduce consumers' purchasing power, strain public finances, and force central banks to address an inflation impulse while growth is weakening, creating pro-cyclical tightening. Even then, Morgan Stanley does not expect the business cycle to end solely because private balance sheets are not heavily levered. The oil-market backdrop is tighter in the firm's view. Middle East exports were 13mb/d in the week ended September 6, below roughly 18mb/d before the conflict, while inventories and strategic reserves have fallen and previous offsets from stronger US exports and weaker Chinese imports have become less supportive. Morgan Stanley's oil strategist expects a crude-market deficit in 4Q26 and 1Q27, with Brent at US$100/bbl in 4Q26 and US$95/bbl in 1Q27. At year-to-date average oil and gas prices of US$88/bbl and US$17/mmBTU, respectively, the estimated Asian oil burden rises to 4.5% of GDP from 3.6% in 2025. Under the firm's 4Q26 assumptions of US$100/bbl oil and US$25/mmBTU gas, it reaches 4.7% of GDP, equal to the 20-year average. Fiscal policy has so far limited household pass-through. Local-currency oil prices in Asia have risen roughly 34% since late February, while domestic gasoline and diesel prices rose about 13% and 17%, respectively, on a nominal dollar GDP-weighted basis. Since early July, Asia's gasoline and diesel prices rose 7.5% and 9.6%, compared with 14.2% and 37.3% in the US. Morgan Stanley expects subsidies or price caps to remain important, with Malaysia and Indonesia likely to hold subsidized fuel prices unchanged through year-end, while India and Japan may allow greater pass-through if the increase persists. The country analysis shows that the monetary-policy response depends on domestic growth, inflation, currency conditions, and the degree of fuel-price pass-through. The firm expects Fed hikes to affect Asian central banks mainly through foreign-exchange moves rather than automatic policy-rate matching. It sees upside rate risks for India, Japan, Korea, the Philippines, and Indonesia. At oil around US$105-110/bbl for six months or more, Morgan Stanley expects additional inflation and potential rate pressure in several economies; at US$120/bbl for a prolonged period, it expects growing stagflation and growth concerns to constrain further tightening in some markets.

Analysis framework

Morgan Stanley compares current conditions with its framework for a leverage-driven business-cycle downturn, then tests whether rates, capex conditions, and private-sector debt meet those conditions. It combines real-rate trends, debt-to-GDP ratios, capital-goods imports, capex forecasts, oil-market supply-demand conditions, oil-burden calculations, fuel-price pass-through, and country-specific fiscal and monetary-policy scenarios.

Methodology notes

• MacroeconomicsCredit and Debt Cycle

  Leverage and restrictive financing conditions as the mechanism through which an expansion can turn into a downturn.

  The report assesses whether private-sector leverage and real policy rates are sufficiently elevated to create a conventional debt-driven downturn, concluding that current balance sheets and rates do not meet that condition.

• Industry AnalysisSupply-demand framework

  Global crude-oil supply-demand balance and its implications for oil prices.

  Morgan Stanley links lower Middle East exports, depleted inventories, and less favourable demand-supply offsets to its expectation of a crude-market deficit in 4Q26 and 1Q27.

• Industry AnalysisUpstream-Midstream-Downstream Transmission

  Transmission from AI adoption and compute demand to AI capex, Asian technology exports, and semiconductor manufacturing investment.

  The report explains how slower AI adoption could reduce compute demand and then affect Asia's technology and semiconductor-related capex, while noting that these segments are only part of total incremental capex.

Key data

• Asia real policy rates+50bp from the February 2026 trough; 160bp below the February 2025 peakMorgan Stanley argues that rates are still not restrictive for most Asian economies.
• Asia ex-China corporate debt-to-GDP73%Below the pre-Covid level.
• US corporate debt-to-GDP76%Also below the pre-Covid level.
• AI and semiconductor-related share of incremental Asian capex through 203011%Supports the view that most capex drivers are not directly dependent on AI.
• Oil risk thresholdUS$130-150/bbl for six months or longerMorgan Stanley's scenario for a meaningful growth slowdown.
• Morgan Stanley Brent forecastUS$100/bbl in 4Q26; US$95/bbl in 1Q27Based on an expected global crude-market deficit through those periods.
• Asia oil burden4.5% of GDP at year-to-date average energy prices; 4.7% in 2026 under Morgan Stanley's 4Q26 assumptionsCompared with 3.6% in 2025 and a 20-year average of 4.7%.
• Asia domestic fuel-price pass-through since late FebruaryGasoline +13%; diesel +17%Local-currency oil prices rose roughly 34%, indicating partial pass-through.

Impact & implications

The report maintains that Asia's capex-led expansion can withstand moderate rate increases and oil near US$110/bbl because fiscal cushioning and sound private-sector balance sheets reduce the shock's transmission to spending and defaults. A more persistent and substantially higher oil-price scenario, or weaker-than-expected AI adoption, would slow growth and capex but is not viewed as sufficient by itself to end the cycle.

Risks

• AI adoption could progress more slowly than markets expect if users cannot generate meaningful productivity gains, reducing compute demand, AI capex, Asian technology exports, and semiconductor investment.
• Oil prices at US$130-150/bbl for six months or longer could materially slow growth by raising the oil burden, weakening consumer purchasing power, straining fiscal positions, and forcing pro-cyclical monetary tightening.
• A prolonged period of higher oil prices could increase fuel-price pass-through in India and Japan and raise stagflation risks in affected Asian economies.
• Further Fed tightening could pressure Asian currencies and create upside risks to policy-rate paths in several Asian economies.

What to watch

• The pace of AI adoption, productivity gains, compute demand, and resulting AI and semiconductor capex.
• Whether oil prices remain near US$105-110/bbl or rise into the US$130-150/bbl range for six months or more.
• Middle East export recovery, global inventories, and the oil-market balance through 4Q26 and 1Q27.
• The extent of Asian fuel subsidies, price caps, and pass-through to households and companies.
• Foreign-exchange movements following Fed hikes and resulting Asian central-bank policy decisions.