JPMorgan Global Macro Conference: Central banks may raise interest rates "faster and larger than expected," but "stock and bond yields" may "rise together."

According to TrendFocus, JPMorgan Chase's Strategic Research Department released the minutes of its Global Macro Conference, stating that the most crucial signal from the conference can be summarized in one sentence: stocks and US Treasury yields can rise simultaneously, at least for now. Global central banks, especially the Federal Reserve, may have to raise interest rates at a faster and larger pace than the market expects.

The report states that 15 macroeconomic and market speakers engaged in in-depth discussions on the US economic outlook, Federal Reserve policy, AI capital expenditures, and geopolitical risks, resulting in two highly confident assessments:

First, central banks—especially the Federal Reserve—may have to raise interest rates at a faster and larger pace than the market expects.

JPMorgan global economists have characterized the September FOMC meeting as the starting point of a new round of interest rate hikes in developed markets (DM), expecting the Federal Reserve, the European Central Bank, the Bank of Japan, and the Reserve Bank of Australia to act successively this month. Of the nine DM central banks tracked, eight are projected to raise rates before the end of 2026. JPMorgan economists believe that the equilibrium Taylor rule points to a further increase of approximately 100 basis points in DM policy rates, but the risks are clearly skewed towards a "faster and larger" rise .

Second, both stock and bond yields rose.

JPMorgan equity strategists set a year-end target price of 8,000 for the S&P 500, with current positions being light rather than extremely overweight, and valuations not yet at extreme levels. AI-driven capital expenditures and resilient earnings provide core support. No one at the event was bearish on the stock market; "bears" were described as an "endangered species in this cycle."

Furthermore, the rise in long-term yields is a global structural trend: fiscal deficits, government bond issuance, and term premiums are the core drivers , with the total US debt reaching $40 trillion.

The report also touched on AI bottlenecks, AI risks, geopolitics, and the upcoming US midterm elections. This conference summary is based on the JPMorgan Global Macro Conference held in New York on September 10, 2026, which brought together 15 macro and market speakers to discuss key topics such as the US economic outlook, Federal Reserve policy, the impact of AI investment on productivity, geopolitical risks, and the US midterm elections.

Central Bank Interest Rate Hike Path: Market Pricing May Significantly Underestimate the Tightening Intensity

The report states that the current moderate tightening of approximately 100 basis points in the developed market yield curve is essentially just a "withdrawal of the 2025 rate cuts," rather than the start of a larger and faster rate hike cycle. The market expects rate hikes of approximately 25 basis points per quarter, citing supply-side shocks rather than wage-driven inflation, weak labor bargaining power, and a lower-than-expected transmission of energy shocks.

However, the speakers at the conference clearly pointed out that the risks surrounding this baseline path are significantly skewed toward "faster and bigger," specifically in the following aspects:

There is a risk of misjudging the Federal Reserve's positioning of the neutral interest rate. The Fed's claim that the current policy rate is "well below neutral" does not align with the results of the New York Fed's primary dealer survey—which showed the nominal neutral rate to be between 3.0% and 3.25%. The gap between the policy rate and the neutral rate may be much smaller than the official narrative suggests. If the labor market strengthens further, the conflict in Iran continues, or price pressures spread, short positions remaining at the short end of the yield curve will be further reinforced.

The structural stickiness of inflation exceeds the ability of predictive models to capture it. Several speakers emphasized that no single inflation indicator can simultaneously reflect the level, breadth, and persistence of inflation. Commonly used indicators such as core PCE, trimmed mean, and sticky price inflation share a common limitation: they infer potential inflation solely from the inflation data itself, ignoring crucial signals from the labor market, wage growth, productivity, and inflation expectations.

It is worth noting that unit labor cost inflation has averaged 2% over the past four years and only 1.5% in the past year; average hourly wage growth is at its lowest level in seven years, while productivity growth has risen to its fastest pace in a decade. After adjusting for productivity, wage growth is in line with the 2% inflation target, meaning that underlying inflation is actually closer to the target than the surface data suggests.

The challenge, however, lies in the fact that many current sources of inflation are not sensitive to interest rates. The AI construction boom is creating near-term inflationary pressures, with demand being released in advance through large-scale capital expenditures on data centers, which is already reflected in PCE chip and software prices, capital expenditure data, and the credit market.

JPMorgan economists have raised their core inflation forecasts for DM by an average of 0.5 percentage points. They expect the Federal Reserve to raise its 2027 core PCE forecast by 0.5 percentage points from the SEP level in December last year to 2.6%, while the European Central Bank will simultaneously raise its 2027 core inflation forecast from 1.9% to 2.6%.

In terms of the pace of interest rate hikes, history has never seen a cycle that truly ends with a single rate hike. Several speakers pointed out that if the Federal Reserve publicly discusses rate hikes, it should be prepared for sustained action across multiple meetings. Furthermore, the upcoming revision of the BEA's core PCE calculation methodology has led some speakers to favor waiting until December to take action, in order to avoid the risk that a September rate hike might be offset by subsequent data revisions.

Stock and bond yields rise simultaneously: Why does this logic hold true?

The core narrative of this conference was a systemic challenge to the conventional wisdom that "rising interest rates inevitably suppress the stock market." The report pointed out that the speakers at the conference demonstrated the sustainability of the simultaneous rise in stock and bond yields from multiple perspectives.

The interest rate transmission channel has been structurally weakened.

Compared to previous interest rate hike cycles, higher policy rates are being transmitted to the US real economy with less force, which is the most important structural change in this cycle. AI, healthcare, and services account for a larger share of growth and capital expenditure and are relatively insensitive to interest rates. The constraints of traditional interest rate channels have been significantly weakened, and the impact of Fed policy changes on corporate behavior is far less than in previous cycles.

Therefore, the critical point for US Treasury yields to disrupt the stock market balance may be much higher than previously expected, potentially ranging between 5.5% and 6.0%. The current market's "fear threshold" has shifted from 5% to 5.5%, with 6% considered a "truly frightening level," even though the S&P 500 rose sharply during the 1990s when interest rates were between 6% and 7%. Analysis of historical data from the past 70-80 years suggests that a yield of approximately 5.5% is compatible with the current environment of strong earnings growth.

The stock market's supporting logic is solid, and the position is not extreme.

JPMorgan equity strategists have set a year-end target of 8,000 for the S&P 500, citing strong Q3 earnings prospects and a light rather than overweight position—a stark contrast to some comments that suggest the market is already in a state of extreme optimism. When the S&P 500 was at 7,600, investors generally preferred to "buy on dips" rather than "chase rallies." Valuations of semiconductor and AI "bottleneck" related stocks have not yet reached extreme levels in terms of price-to-earnings ratios.

AI capital expenditure is a core theme of the current earnings season. Consensus forecasts indicate that AI capital expenditure will reach approximately $900 billion by the end of 2026 and exceed $1.2 trillion by the end of 2027. Hyperscale cloud computing companies are expected to account for approximately 87% of total AI capital expenditure in 2026 and 2027. In 2026 alone, the capital expenditure guidance of the five largest hyperscale companies in the United States exceeds $750 billion, and this figure is expected to exceed $1.1 trillion in 2027. The total scale of AI capital expenditure is expected to reach $5.5 trillion by 2030.

The pace of interest rate increases is more crucial than their absolute level.

The report states that speakers repeatedly emphasized that the speed and volatility of interest rate increases are more destructive than their absolute levels. An orderly rise of 100-200 basis points can be absorbed without disrupting the AI theme, but a rapid increase will put pressure on risk assets and force a reassessment of capital expenditures. An increase of 50-75 basis points in long-term interest rates may have a less significant impact on overall consumption than a sharp correction in the stock market; pressure on consumption will only become more pronounced after an increase of more than 100 basis points.

As an example, the 10-year US Treasury yield has risen by 90 basis points since April, but because this process took about six months and was relatively orderly, the market has largely absorbed it. In contrast, several historical instances of rapid short-term increases (above 50 basis points) coupled with geopolitical shocks or a stronger dollar have truly triggered sustained capital outflows.

Rising long-term yields are part of a global fiscal narrative.

This round of rising long-term yields is essentially a reflection of global fiscal expansion, rather than simply an AI or inflation narrative. European long-term yields have risen in tandem, while Europe lags significantly behind in AI innovation, refuting the explanation that "AI is the main driver." A survey of primary dealers by the New York Fed shows that the term premium has risen from negative levels a decade ago to approximately 125 basis points, a change primarily driven by market concerns about long-term fiscal sustainability, including the massive $40 trillion in nominal US Treasury bonds.

Currently, interest payments in the United States account for about 14% of the fiscal budget and continue to rise. Historically, sovereign debt crises usually occur when interest payments reach 20%-25% of the fiscal budget. There is still some buffer room at present, but the direction of the pressure is clear.

The real bottleneck for AI is the quality of licensing, approval, and execution, not insufficient demand.

The report argues that for AI infrastructure, demand is not the problem; construction is. Energy supply, licensing, and local political resistance are becoming more real limiting factors than a collapse in demand.

At the grid level, the current problem lies in transmission and licensing approvals, not in generation itself: a grid designed for approximately 100 basis points of load growth per year is now facing a growth rate of approximately 300 basis points, with both planning and capital allocation failing. The last approximately 1,000-mile interstate high-voltage transmission line took 18 years to obtain licensing approvals.

Key risk signals for data center investments include: "facilitated termination" clauses, leases without tariffs or commodity protection, and non-investment grade tenants. A typical construction cycle involves two large buildings requiring approximately 40,000 tons of copper and 100,000 tons of steel, and takes about three years from signing to stable operation.

It's worth noting that 98% of newly built data centers (by megawatt) are single-tenant, meaning that tenant credit quality and construction quality will determine the long-term value of the asset. Data centers are likened to submarines—"good when they're on, bad when they're off; there's no middle ground."

Furthermore, the report points out that AI has significantly enhanced the offensive capabilities of cyberattacks, turning defense into an "arms race." For highly regulated institutions such as banks, the cybersecurity and fraud risks brought about by AI mean that they must invest heavily in defensive spending rather than generating new revenue.

Geopolitical risk premiums are underestimated, and high oil prices may continue into the first half of 2027.

The report states that the Strait of Hormuz remains a critical chokepoint for global energy, with historical traffic volumes of approximately 20-23 million barrels per day, and existing alternative infrastructure is far from sufficient to replace it. Even though the strait's direct importance to the United States has declined somewhat, it still constitutes a significant bottleneck for global crude oil, refined oil, and LNG trade.

Iran's strategic goal is to establish a credible deterrent against future attacks from the United States or Israel, and control of the Strait of Hormuz could provide this leverage. Asian oil-consuming nations appear willing to accept a transit fee of approximately $1 per barrel, which translates to a potential revenue of about $40 to $50 billion annually for Iran.

In a scenario of "permanent conflict," JPMorgan's commodities strategy team estimates that the average price of Brent crude oil in 2027 may be only $87 per barrel, compared to $64 per barrel in their baseline scenario (where the world returns to peace in early 2027). Even if crude oil stabilizes around $90, continued tightness in refined product supply and rising geopolitical risks will maintain energy inflation and commodity market volatility.

One speaker predicted that the geopolitical risk premium would persist until at least 2027, with the core logic being that the root causes of regional tensions would remain regardless of whether the Iranian regime weakens after the conflict (leading to further instability) or survives (continuing its support for groups such as Hezbollah, Hamas, and the Houthis).

"Affordability politics" will shape fiscal and regulatory pathways.

Polls show that only 25% of Americans are satisfied with the country's direction, down from the historical average of 33%. According to a Gallup survey, high cost of living is the most important family financial issue for 31% of respondents, a figure particularly high among younger groups.

The report states that speakers at the JPMorgan Global Macro Conference believe a Democratic House "flip" is almost a certainty (with a roughly 90% chance of winning), but the Senate still leans towards Republican control. Regardless of the midterm election results, the theme of "affordability" will continue to shape fiscal and regulatory paths, including discussions on a wealth tax and the resurgence of universal basic income policies.

JPMorgan Chase believes the market may be overly optimistic about the policy environment following the midterm elections. Potential risks of a "lame duck" period include:

Trump's executive orders have intensified the implementation of reciprocal tariffs, bringing the global effective tariff rate back from 5%-6% to 17%-18%; and a third round of reconciliation budget of at least $300 billion to $500 billion, including about $15 billion in agricultural aid and more than $150 billion in additional defense spending, which will further suppress US debt and prompt rating agencies to reassess the fiscal outlook.

Stanford economists estimate that, under the latest oil price scenario, the average family may need to pay an extra $857 for gasoline for the remainder of the year, which would almost offset the personal tax cuts from the "Big Beauty Act".

~~~~~~~~~~~~~~~~~~~~~~~

The above content is from Zhuifeng Trading Platform .

For more detailed analysis, including real-time updates and firsthand research, please join the [ Trading Channel Annual Membership ].

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.