Finance

White House Shake-Up, AI Funding Cracks and Gold Buying: Repricing Capital in a Policy Cycle

Updated · 2026-11-08 11:09 · 6 sources cited

White House Shake-Up, AI Funding Cracks and Gold Buying: Repricing Capital in a Policy Cycle

The policy reshuffle after the U.S. midterm elections is converging in the same window with financing pressure on AI capital expenditure, corporate product-line adjustments and central-bank reserve diversification. White House personnel changes signal continuity in the policy line, while the credit market is pricing the huge AI infrastructure buildout more cautiously; capital is looking for a new balance between the technology cycle and safe-haven assets.

White House Shake-Up: Same Direction, Narrower Business Channels

After the U.S. midterm elections, Trump announced that longtime loyalist and current Deputy Chief of Staff Stephen Miller will become White House chief of staff, succeeding Susie Wiles, with the appointment taking effect on Monday. A White House official said 'he will not change direction.' With Republicans having lost control of both chambers of Congress and Democrats about to lead the House and Senate and planning large-scale investigations, the personnel move is seen as a signal that Trump will stick to his core positions rather than seek compromise.[1]

Miller, 41, is one of the few core aides who has continued from Trump's first term to now. He is best known for pushing aggressive immigration enforcement, including a daily arrest target of 3,000 and sending masked immigration enforcement agents into communities across the country; an Ipsos poll showed only 36% of respondents supported the policy, while 54% were clearly opposed. On domestic policy, he was also closely associated with the dismissal of tens of thousands of civil servants in 2025.[1]

Wiles submitted her resignation on Thursday evening and will move to Trump's political operation. She was not only the first female White House chief of staff in U.S. history, but also the main liaison between the business community and the White House; several well-known corporate executives said Wiles and Treasury Secretary Bessent were the two key figures they believed were willing to listen to business. With Wiles leaving, communication channels between business and the White House face a marked narrowing, and with Miller taking over, relations between the White House and Congress and the business community are expected to become more confrontational.[1]

AI Capex: Credit Spreads and Loan Sales Flash Warnings at Once

Jefferies strategist Christopher Wood warned in a latest report that financial markets have already signaled a possible peak in the AI capital-expenditure cycle through wider credit spreads and sustained declines in semiconductor stocks. According to the Jefferies report, in the four quarters through the third quarter of 2026, U.S. AI-related real capital spending contributed 1.39 percentage points to year-on-year real GDP growth, accounting for 67% of the overall 2.1% growth; investment in information-processing equipment and software rose to 5.35% of nominal GDP, surpassing the 4.46% peak during the 2000 internet bubble.[4]

Credit-market cracks are more direct. The 10-year corporate bond spreads of Amazon, Alphabet and Meta have widened by 16, 16 and 35 basis points respectively from recent lows; the 5-year CDS spreads of Oracle and CoreWeave have surged to 329 and 1,163 basis points respectively from early-June lows. Jefferies believes this means there is material downward-revision risk to the market consensus forecast that the four major hyperscalers will spend a combined about $1.04 trillion in capex in 2026.[4]

Latest quarterly earnings have further intensified doubts. The four major hyperscalers' combined third-quarter capex rose 108% year on year to $202 billion, but free cash flow at Amazon and Alphabet has turned negative, to -$35 billion and -$9 billion respectively. Remaining performance obligations at Microsoft, Amazon, Alphabet and Oracle have risen to $2.5 trillion, and much data-center financing is being arranged off balance sheet; Microsoft announced a voluntary departure program for about 60,000 eligible employees, more than a quarter of its global workforce.[4]

Financing pressure has extended from bond spreads to loan sales. ING is seeking to sell a data-center loan portfolio of about $500 million in one block, listing about 12 loans, with bids due on Nov. 19 and a minimum bid threshold of $10 million per loan. The portfolio includes financing related to the 'Stargate' AI infrastructure project jointly pursued with OpenAI, SoftBank and Oracle, including a $25 million loan tied to a large data center in New Mexico.[5]

The sale comes after the midterm elections, amid market fears of stricter tech regulation; chip stocks and emerging cloud-computing companies were hit hard, and S&P 500 constituents lost more than $3 trillion in combined market value. More closely watched, Oracle has issued a force majeure notice to the developer of the New Mexico project, meaning it is preserving legal room to potentially exit related obligations. ING's loan portfolio also covers debt of data-center operators such as Vantage Data Centers and CyrusOne. Analysts note that one core part of the AI trade logic is financing-supported compute expansion; once loan buyers start demanding discounts, financing costs for new data centers will rise, and marginal projects face the risk of delay or even being shelved.[5]

Chip Supply Chain: TSMC Evaluates 3D DRAM

According to a Friday report by The Information, TSMC is evaluating the feasibility of producing advanced memory chips, a move prompted by Nvidia's request, with three-dimensional dynamic random access memory (3D DRAM) as the core option under consideration. TSMC is already the main foundry for Nvidia's AI processors; if it takes on memory-chip production as well, its coverage in the AI hardware supply chain would extend from logic chips to memory, helping customers such as Nvidia gain more comprehensive visibility and control over future chip supply.[2]

The potential move comes amid persistent supply tightness. The chip shortage has squeezed profit margins at major global equipment makers and pushed up retail prices of consumer electronics from Apple's iPhone to Sony's PlayStation. TSMC shares fell sharply this week along with Taiwan's benchmark Taiex, which at one point plunged nearly 9% on Friday intraday, entering bear-market territory and posting its worst weekly performance since 2001. TSMC has not yet made a production commitment; with mainland China leader CXMT also planning to expand capacity, the global memory-chip market could see a new competitive variable.[2]

Cybertruck Curtain Call: A Three-and-a-Half-Year Controversy Cycle

Musk said in a post on X on Friday that Tesla will gradually 'wind down' Cybertruck production at its Texas Gigafactory around the middle of next year and shift to expanding capacity for the robotaxi Cybercab. That means the model's actual sales cycle from first deliveries in late 2023 to the end of production will be only about three and a half years. One day before the production halt was announced, Tesla raised Cybertruck prices again, with the top-end Cyberbeast priced at $102,490.[3]

Cybertruck's controversy began on launch day. At the November 2019 unveiling, Musk staged a shatterproof-window demonstration, throwing a metal ball at the glass, and both windows cracked; four years later at the delivery event, Tesla's chief designer demonstrated again, and this time the glass remained intact. Reservation data once excited observers, with 250,000 reservations within five days of the launch, but the $100 refundable deposit mechanism made actual conversion value questionable, and some reservation slots were listed for sale on eBay.[3]

Production and pricing were also bumpy. Tesla initially promised to start production by the end of 2021, then delayed the plan to 2022 and 2023; the starting price at the 2019 event was $39,900, and by the time deliveries began in late 2023 it had risen to $60,990. The $59,990 dual-motor all-wheel-drive version launched in February 2026 later rose to $69,990 in March and $74,990 in August; before the production halt news, that configuration was priced at $82,490, the Premium at $92,490 and the Cyberbeast at $102,490, nearly double the lowest 2019 price.[3]

Quality problems followed closely. According to National Highway Traffic Safety Administration data, since its launch in late 2023 Tesla has issued 11 recalls for Cybertruck, with some vehicles requiring physical repairs and the rest fixed through over-the-air software updates. The broadest recall occurred in January 2024, when a warning-light font issue affected more than 2.19 million Tesla vehicles; in April 2024, 3,878 Cybertrucks were recalled because the accelerator pedal pad could come loose, and in March 2025 another 46,096 were recalled because the roof light bar could delaminate and fall off.[3]

Reserve Diversification: China's Central Bank Buys Gold for 24 Straight Months

Data released on Nov. 7 showed that China's gold reserves at the end of October were 78.30 million ounces, up 830,000 ounces from 77.47 million ounces at the end of September, expanding gold reserves for 24 consecutive months. Over the same period, foreign-exchange reserves in October were $3.405178 trillion, a modest rebound of about $4.9 billion, or about 0.14%, from $3.400251 trillion in September; but total official reserve assets fell to $3.777561 trillion from $3.789425 trillion in September, a month-on-month decrease of about $11.86 billion, mainly dragged down by lower gold valuations.[6]

Gold's book value and holdings diverged. Affected by international gold-price volatility, the value of gold reserves fell to $306.411 billion at the end of October from $323.517 billion at the end of September, a drop of about $17.1 billion; but physical purchases did not stop. Gold holdings were 74.19 million ounces at the end of January and rose to 78.30 million ounces by the end of October, a cumulative increase of 4.11 million ounces, or about 5.5%. On a monthly pace, buying clearly accelerated in the second half, with a combined increase of 2.22 million ounces from July to October, accounting for more than half of the full-year increase.[6]

China's continued buying is in line with the global central-bank trend. The World Gold Council's Gold Demand Trends report for the third quarter of 2026, released on Nov. 3, showed that global central banks and official institutions net bought 321 tonnes of gold in the quarter, up 42% year on year, the highest quarterly record since the fourth quarter of 2024; cumulative purchases in the first three quarters were 675 tonnes, up 6% year on year. The report also showed that gold ETFs had net inflows of 211 tonnes in the third quarter, with global gold ETF holdings rising to a record high of 4,256 tonnes; the average international gold price in the third quarter was $4,262 per ounce, and the total value of gold demand rose 22% year on year to $188 billion.[6]

Review: Policy Risk and the AI Financing Cycle Resonate

Putting these threads together, two mutually reinforcing main lines emerge: the political reshuffle after the U.S. midterm elections has raised uncertainty over regulation and investigations, while financing for AI capex is already showing stress through credit spreads, loan sales and project force majeure notices. TSMC's evaluation of memory chips, Tesla's contraction of Cybertruck and pivot to Cybercab are all corporate moves to adjust resource allocation within the same macro and technology cycle; the Chinese central bank's 24 consecutive months of gold buying echoes the backdrop of global central-bank net gold purchases rising to the highest quarterly level since the fourth quarter of 2024.[1][2][3][4][5][6]

Three points to watch: the outcome of ING's roughly $500 million data-center loan portfolio after the Nov. 19 bid deadline, whether TSMC makes a production commitment on 3D DRAM, and how the new White House chief of staff interacts with a Democrat-led Congress. If AI-related financing costs continue to rise, the risk of delays to marginal data-center projects will draw more attention; whether central-bank gold buying and gold ETF inflows can continue will also affect the pricing logic for safe-haven assets.[2][4][5][6]

Sources

← All Finance stories