The largest nonlinear risk in global markets: interest rate hikes, oil prices, and AI slowdown

A few months ago, a supply shortfall of one million barrels per day could still be partially offset by drawing from inventories, rerouting through alternative ports, or increasing refinery output. But with global inventories now down by over 500 million barrels, the next supply disruption of similar magnitude will require significantly higher prices to force further demand destruction. Diesel is especially worth watching: the Gulf region and Russia have long been major global diesel exporters, yet global refineries are already operating near full capacity. Rising diesel prices will eventually feed into trucking, agriculture, construction, and factory logistics; portions that businesses cannot absorb will gradually pass through to final goods and service prices. Should the Saudi East-West Pipeline, Fujairah terminals, or Red Sea shipping lanes face further disruptions, even the alternative routes originally intended to bypass the Strait of Hormuz would lose some of their effectiveness. The European economy may already be slowing, yet energy-driven inflation continues transmitting into other prices—meaning the European Central Bank’s first stress test will deteriorate further. The third stress test: AI slowdown As of August, Claude was already leading approximately 26% of Anthropic’s AI R&D tasks—up from under 1% in February. A McKinsey survey this year found that once enterprises move from localized pilots to company-wide deployment, AI spending nearly quadruples; 93% of surveyed companies reported budget overruns. In another survey, roughly two-thirds of firms observed productivity improvements, yet only 20% reported actual revenue growth attributable to AI—and 74% are still waiting for future revenue realization. Anthropic is also preparing for an IPO, prompting capital markets to scrutinize the company through the lenses of revenue quality, profitability, and cash burn. The company recently told investors that its adjusted operating profit has turned positive for two consecutive quarters, with annualized revenue growing rapidly—but frontier model training, inference, and data center operations still require massive investment, while open-source models and competing commercial offerings continue driving down pricing. Here, “AI slowdown” carries two meanings. On the R&D side, as AI increasingly participates in its own development, model iteration speed is beginning to approach the limits of safety review and human validation capacity. On the commercial side, deeper enterprise adoption intensifies scrutiny over cost efficiency and tangible returns. If enterprises’ expenditures on compute, integration, and governance grow faster than gains in revenue or labor savings, AI companies will ultimately need to answer a fundamentally traditional question: what profits will support such high valuations and capital outlays? My concern about nonlinear risks What worries me now is that several familiar self-correcting mechanisms in markets may simultaneously weaken. Under normal conditions, slowing European growth leads to falling inflation, allowing central banks to cut rates and easing corporate financing pressures. This time, however, if growth slows due to persistently high oil prices, core inflation may remain elevated even amid weakening economic activity—leaving the ECB unable to cut rates promptly. Oil markets previously relied on inventories to bridge supply gaps, but with over 500 million barrels already drawn down, the next comparable disruption will place greater adjustment burden on price. Meanwhile, U.S. AI investment may continue absorbing capital from global financial markets, keeping long-term funding costs elevated—even as the European economy decelerates. The AI industry itself is undergoing similar shifts: continued advances in model capability no longer automatically translate into commensurate corporate profit growth; as adoption penetrates core business processes, compute costs, data governance, human oversight, and system overhauls all begin appearing directly on the income statement.

September CPI releases for China and the U.S.: two divergent outcomes under identical oil prices

On the interest-rate front, one set of figures looks highly anomalous. China’s one-year Loan Prime Rate (LPR) stands at 3.0%, while the U.S. federal funds rate ranges from 3.75% to 4.0%—superficially suggesting looser policy in China and tighter policy in the U.S. Yet subtracting CPI changes paints a different picture: China’s 3.0% minus 0.9% yields a real policy rate of 2.1%; the U.S. midpoint of 3.875% minus 3.7% gives roughly 0.2%. China’s real interest rate is thus ten times higher than America’s. In a country where CPI stands at just 0.9%, borrowers bear the world’s most expensive real funding costs. This is the arithmetic of debt deflation: borrowing at 3% while prices rise only 0.9% means real debt burdens automatically increase each year. So why hasn’t the LPR moved for eight consecutive months—even though room for cuts clearly exists? It’s not reluctance—it’s that upstream input prices are rising; cutting rates now would effectively stoke the fire of cost-push inflation. The same logic applies to 10-year government bonds. The U.S. yield stands at 5.291%, China’s at 1.66%—a nominal spread of 363 basis points, seemingly attractive for arbitrage. But adjusting for inflation yields a very different picture: the U.S. 10-year real yield is 1.59%, China’s is 0.76%, narrowing the real spread to just 83 basis points. The genuine allure for hot money is far weaker than nominal spreads suggest. Here, I must pour cold water on my own analysis: strictly speaking, real interest rates should be calculated using inflation expectations—not backward-looking YoY CPI. My calculation is a rough approximation—directionally sound, but imprecise. On the external front, further Fed rate hikes and tightening global liquidity are pressuring our asset prices. Following the U.S. CPI surprise, the probability of a rate hike at the October 27–28 meeting rose to 54%, and the likelihood of at least one more hike before year-end exceeded 96%. New York Fed President John Williams stated that one additional hike this year is appropriate—and that none will be needed in 2027; Cleveland Fed President Loretta Mester argued instead for tighter policy to prevent entrenched inflation expectations. From the same data, two entirely different policy narratives emerge. That asset prices will face pressure is unquestionable. The uncertainty lies with the RMB—and the outcome runs counter to intuition. By conventional interest-rate differential logic, with China expected to cut rates and the U.S. expected to hike, the RMB should depreciate. Reality tells a different story: USD/CNY stands at 6.6996; over the past 12 months, the RMB has appreciated 6.03% against the U.S. dollar—and also strengthened against the euro, pound, and yen. Offshore RMB trades at 6.6995, hovering near its strongest level in 52 weeks (range: 6.69–7.00). The reason? The trade account has overwhelmed the capital account: September exports surged 36.3% YoY to a record $447.18 billion, generating a trade surplus of $133.53 billion—up sharply from $88.7 billion a year earlier. The RMB’s strength harbors another overlooked dimension: it functions as a distribution mechanism. The benefit is clear: imports of oil, copper, and soybeans become cheaper in RMB terms, buffering China’s PPI from spiking higher due to oil prices. The cost, however, falls squarely on exporters: they earn dollars but pay RMB-denominated costs—RMB appreciation directly slices into their gross margins. In other words, exchange-rate appreciation transfers the squeeze of imported inflation—from importers to exporters—while exports happen to be the strongest segment of China’s current economy. Looking further ahead, RMB depreciation would effectively discount China’s manufacturing capacity. Germany, Japan, and South Korea simply lack the industrial capacity to absorb that volume—a weapon others cannot wield, and therefore one China won’t deploy lightly. So while Fed hikes weigh on asset prices, they cannot break the RMB. Its resilience stems not from interest-rate differentials, but from trade surpluses and manufacturing capacity.

Strong Earnings Mask Growing Struggle With Inflation, Oil Prices

1 day ago

S&P 500 earnings growth is driven by energy and tech, but rising oil prices and inflation are pressuring margins in materials and consumer sectors. Companies like International Paper and Sherwin-Williams face challenges from higher costs and tepid demand.

bloomberg

2026-10-30

Strong Earnings Mask Growing Struggle With Inflation, Oil Prices

Strong Earnings Mask Growing Struggle With Inflation, Oil Prices

FAQ