Home Alone: inflation and the new Fed chair; investing in China’s AI ecosystem; Prediction markets | Michael Cembalest | J.P.Morgan Asset & Wealth Management · Nuggets
Monetary Policy//8 min read/am.jpmorgan.com
Home Alone: inflation and the new Fed chair; investing in China’s AI ecosystem; Prediction markets | Michael Cembalest | J.P.Morgan Asset & Wealth Management
Disclaimer: Orignal content owned by or sourced from third parties. It does not represent the views of 'Nuggets' platform or it's team. AI is used extensively across this platform including for summaries. Accuracy is not guaranteed, there can be mistakes. Any info or content on this platform is not a financial, legal, or investment advice. Do your own research. Refer for complete disclosures:- Terms of Use · Full Disclaimer
Michael Cembalest compares incoming Federal Reserve Chair Kevin Warsh to the movie character Kevin McCallister in Home Alone. Warsh faces a lonely vigil where he must survive and manage economic pressures until "the adults get home".
Trimmed PCE Biases and Political Pressure
The Policy Dilemma: Warsh has highlighted trimmed PCE (Personal Consumption Expenditures) inflation, which currently sends the clearest signal that it is acceptable to cut interest rates. However, this specific calculation suffers from inherent biases and lags. Just three years prior, trimmed PCE failed completely to highlight the "Biden inflation" era, which the GOP justifiably highlights as a major policy mistake. Relying on it now simply because it offers an "all-clear" signal is highly inconsistent.
White House Demands: Warsh faces immense pressure from President Donald Trump, who has argued for months that the U.S. should have the lowest policy rates in the world. In December 2025 (described as four months prior to the podcast), Trump stated that interest rates should be at 1% or potentially lower within a year.
Key Macroeconomic Challenges
Warsh must navigate three critical, long-term economic realities:
Commodity Price Surges: Despite U.S. energy independence, various land commodities, fertilizers, and energy products have experienced 50%, 60%, 70%, 80%, to 100% year-over-year increases. Natural gas is the only commodity trending downward.
Inflation Sensitivity: Over time, the inflation sensitivity of the business cycle has been growing, meaning that for given shifts up or down in the cycle, inflation responds more rather than less.
The 2031 Fiscal Crossover: By 2031, federal tax revenues are projected to be entirely consumed by entitlements and interest on the national debt, leaving zero margin for a monetary policy mistake.
Historical Fed Vectors vs. Current Data
J.P. Morgan plotted four critical business cycle variables to compare historical Fed actions against current conditions:
Employment conditions.
Prices Paid Index in the manufacturing sector (measuring price pressures).
Supplier Deliveries Index (measuring supply chain tightness).
The output gap (estimating growth relative to potential).
The Findings: Current values (marked as yellow dots on J.P. Morgan's charts) map significantly closer to historical conditions when the Fed was raising interest rates rather than cutting them. Easing rates right now goes entirely against historical Fed behavior. Furthermore, J.P. Morgan technicians and economists using alternative rules of thumb (including Taylor Rules and alternative $R^*$ calculations) find that the Federal Funds Range should currently sit at 4.0% to 4.8%, rather than the actual current range of 3.5% to 3.75%.
Labor, Wages, and the Immigration Buffer
Inflation Expectations: Derivations from household/consumer surveys and Treasury Inflation-Protected Securities (TIPS) markets have ticked slightly higher this year (by roughly half a percent), consistently staying above the 2% threshold, but are not yet catastrophic.
Tame Wage Trends: Clearer signals are found in the wage and labor markets. Indicators like the Atlanta Fed wage tracker and the National Employment Index look stable, with unit labor costs staying below 2%, showing very limited signs of a wage-price spiral.
The Immigration Shock Absorbent: Despite a massive drop-off in active immigration, wage inflation hasn't spiked in low-wage, immigration-sensitive sectors (childcare, cleaning, construction, food prep, and home healthcare). Brookings estimates that net migration (combination of documented and undocumented) for 2026 will hit minus 1 million people. However, the historic Biden-era surge peaked at plus 3 million people in 2023, compared to the normal baseline trend of 1 million per year. This massive leftover "stock" of workers continues to buffer the labor market, and it will take two to three years of Trump administration policy to erode it.
Shelter Inflation: Rents from actual rental properties and owner-occupied imputed rental growth show nothing remarkable; they are down significantly from 2023 levels and are hovering roughly at pre-COVID baselines.
CPI vs. PCE and Producer Price Warnings
The Index Split: Core levels for both CPI (Consumer Price Index) and PCE remain above 2%. Right now, PCE looks more elevated than CPI because it places a much higher weight on software and certain business applications, while CPI weighs housing more heavily.
Producer Price Pipeline: The primary risk for Warsh and the FOMC is producer prices, which are rising sharply across headline, core, and intermediate goods. This warning shot is driven by rising domestic transport costs (via energy), high copper/aluminum demand (for solar panels, EVs, and energy storage), AI infrastructure costs (electronic components, memory chips, application software, GPUs), new U.S. tariffs, and Chinese export controls on rare earths and medical equipment.
Productivity: The Supply-Shock Debate
Warsh vs. Powell: Former Fed Chair Jerome Powell viewed the AI boom as short-term inflationary due to massive capital expenditure demands and their ripple effects. Warsh views AI as a potential deflationary supply shock that would allow the Fed to maintain steady rates or execute cuts despite pipeline pressures.
Productivity Data: Since the launch of GPT in Fall 2022, data processing, the information sector, and general corporate productivity metrics have trended upward. It remains unproven but highly indicated that this is AI-related.
Market Risk: Equity Risk Premium
U.S. 30-year Treasury yields have recently broken above 5% for the first time consistently since the early 2000s.
This yield spike, paired with expensive stock valuations, has compressed the equity risk premium (the return equity investors earn over risk-free Treasuries) to its lowest level since the late 1990s/early 2000s dot-com era. This leaves equity markets highly vulnerable to an interest rate shock if monetary policy is perceived as too weak or overly accommodative. Normalization of energy prices and an end to the Iran war are critical factors to watch.
Historical Precedent: The Arthur Burns Failure (1972)
Cembalest warns that Warsh must not replicate the failure of Fed Chair Arthur Burns. During his 1972 re-election campaign, President Richard Nixon demanded unemployment be pushed down to 3.5% (the level seen at the end of the 1960s).
Burns resisted initial calls from the FOMC to raise rates, backed inflationary wage and price controls, and oversaw a money supply expansion that peaked at 13%—the highest on record until the Fed's COVID actions in 2020. This resulted in high inflation, high unemployment, and a decade of 0% real returns on both stocks and bonds.
White House Intimidation: White House tapes and records show Burns capitulated under extreme duress. When he resisted, Nixon's team planted false stories in the press claiming Burns was greedily demanding a large personal pay raise (when he had offered to take a cut) and floated plans to dilute Fed power by doubling the number of board members. Nixon warned Burns on a recorded call that he could "destroy the last conservative administration in Washington" and stated that the Fed would bear 100% of the blame if the economy failed to expand. By 1971, Chief of Staff H.R. Haldeman stated, "We have Arthur Burns by the balls on the money supply."
2. Investing in China's Homegrown AI Ecosystem
Aggressive sanctions spanning both the Trump and Biden administrations have effectively compelled China to adopt a "home alone" strategy, building a completely parallel technology ecosystem from lithography machines to high-bandwidth memory and semiconductor fabrication.
GPU Self-Sufficiency: At the time of GPT's launch in late 2022, China was only 20% GPU self-sufficient. That figure has risen to 40% and is projected to exceed 80% by 2030.
Narrowing the Tech Gap: While Nvidia still leads in raw performance, Chinese frontier AI model scores are improving on a near 1-to-1 trajectory with U.S. advancements over the last two years.
Total Cost of Ownership (TCO): Chinese private sector uptake is driven by cost-effectiveness, not just state coercion. When evaluating total cost of ownership (combining power, installation, hardware, and software) and cost-per-token, domestic chips from companies like Huawei and Capricorn are considerably cheaper than or roughly equal to Nvidia's H100, H200, and H20 lineup.
Exploding Market Presence: Based on token API calls monitored via OpenRouter across the top nine models, Chinese AI models were nearly invisible 12 months ago. Today, their aggregate usage is higher than the top nine U.S. models.
Investment Basket: J.P. Morgan Asset Management portfolio managers identified a basket of 32 leading Chinese AI stocks. This basket has kept pace with the 42 AI-driven stocks in the U.S. S&P 500, offering a distinct investment opportunity.
3. The Reality of Prediction Markets
Cembalest warns retail participants that prediction markets (e.g., Polymarket) are heavily weighted against individual traders due to structural imbalances and soft governance rules.
Algorithm Domination
Prediction markets suffer from extreme profit concentration that far exceeds the top 1% margins found in online poker, retail options trading, equity day trading, horse racing, or traditional sportsbooks:
Top 1% of accounts on Polymarket earn between 75% and 80% of all platform gains.
Top 0.1% of accounts earn 67% (two-thirds) of all profits.
This is driven by high-speed algorithms utilized by a tiny subset of accounts to reprice prediction contracts based on breaking information faster than humanly possible.
Governance and Resolution Manipulation
Unlike institutional financial markets that utilize formal International Swaps and Derivatives Association (ISDA) committees to resolve contract disputes (e.g., credit default swaps), platforms like Polymarket utilize a decentralized, third-party token-based voting system.
Any individual—even non-Polymarket users—can buy resolution tokens directly from cryptocurrency exchanges to cast votes during a dispute. This introduces severe conflict-of-interest risks, as bettors can buy tokens to manipulate the outcome of a resolution they have actively bet on.
Case Study of Manipulation: A Polymarket contract betted on whether Ukraine would agree to a critical mineral deal with the U.S. "before April 2025". The deal was signed at the very end of April, sparking an explicit dispute over whether "end of April" met the parameter of "before April". A single voter purchased enough tokens to cast 25% of the total votes, unilaterally swaying the decision to "Yes" and entirely altering the financial payout for all contract participants.
Upcoming Content
The next edition of Eye on the Market (scheduled for June 2026) will be a 250th anniversary of the United States special piece, analyzing the U.S. grip on global financial markets and its status as the world reserve currency.
1. Executive Briefing TL;DR The Core Thesis: The global macroeconomic landscape is shifting away from hyper crowded mega cap AI trades toward a broadening cyclical recovery, marked by positive European earnings revisions, expanding space i…