CIO Insights: AI runs hot, jobs run cool, and the Fed runs loose

10 minute read

As we head into the final quarter of 2025, investors are navigating a constant stream of market headlines – from AI euphoria to political shifts to trade flare-ups. That makes it all the more important to stay focused on the bigger picture: a cooling, but still-resilient global economy that’s being supported by capital expenditure and policy easing.

In this month’s CIO Insights, we unpack how these forces – hotter capex, cooler jobs, and a looser Fed – are shaping the global economy’s path into year-end, and why this macro backdrop remains broadly supportive for risk assets. 

Key takeaways

• The US economy is increasingly running on two tracks, with strong corporate investment and a softer jobs market. On the one hand, AI-driven capital expenditure is helping to underpin growth. At the same time, the labour market is cooling amid a broad-based cyclical weakening in demand, as well as public-sector layoffs, which weigh on hiring. This divergence between resilient business investment and cooling employment highlights the uneven nature of the current expansion. 
• In the near-term, shifting headlines could add volatility to otherwise resilient markets, but are unlikely to derail the Fed from its easing path. While AI continues to shape the longer-term macro and market environment, uncertainties surrounding policy and geopolitics are likely to add short-term noise. In particular, disruptions related to the US government shutdown and flare-ups in US-China trade tensions may trigger bouts of volatility even as underlying growth remains steady. For the Fed, this backdrop reinforces the case to proceed carefully with its signaled rate cuts.
• This macro backdrop – driven by sustained AI investment and easier policy – continues to be supportive for risk assets. With the Fed restarting its rate-cutting cycle and loose fiscal stances across major economies, policymakers appear willing to keep policy accommodative to sustain growth. This mix of easier monetary and fiscal conditions should allow the current regime of “inflationary growth” to continue – an environment that’s broadly positive for equities and other risk assets. For gold, despite this year’s sharp gains, its long-term demand drivers and role as a portfolio diversifier remain strong.

(See our Glossary at the end of the article for descriptions of terms used.)

A two-track US economy: stronger AI investment vs. softer jobs

At the start of the year, in our 2025 Macro Outlook: “FAT” is the new normal, we flagged AI as one of the structural forces set to reshape markets. What we’re seeing now is real: robust AI-driven investment pushing GDP higher, even as the US jobs market cools.

This split between strong business investment and softening employment underscores how uneven the expansion is. It also lays the groundwork for a more dovish Fed stance ahead to support the labour market.

Strong investment into AI is driving faster growth

Capital expenditure (capex) has been a key driver of US economic growth this year, and much of that strength has come from the build-out of AI infrastructure. 

The impact of this investment is already visible in the macro data. As shown in Chart 1, tech-related capex, which includes servers, software, communications equipment and R&D, contributed more than 4 percentage points (ppt) to total business-investment growth in Q2, offsetting weaknesses in other areas. Of that, the share coming from capex into servers and IT hardware (the yellow section of the bar) has seen a steady increase over the past several quarters.

For the economy as a whole, JPMorgan estimates that AI-related capex contributed 1.1 percentage points to US GDP growth in H1 2025 – nearly four-fifths of the total (1). Harvard economist Jason Furman found that boost could have been even larger, noting that GDP growth would have been close to zero – or just 0.1% annualised – without AI investment (2).

Most of the investment into AI is being driven by the large tech “hyperscalers”: Amazon, Microsoft, Google, Meta, and Oracle. Citi Research estimates that hyperscalers’ capital expenditure will reach nearly $400 billion this year, equivalent to about 1.3% of GDP. Over the next few years, this figure could rise sharply. As shown in Chart 2, McKinsey projects AI-related data center capacity could grow by more than 3.5x by 2030 (3). In dollar terms, that would require investment of about $1 trillion a year by 2030, or more than 3% of GDP.

The critical question now is whether this pace is sustainable or whether it risks overheating. So far, the hyperscalers have been able to fund their investments largely through their own cash flows – with an estimated $550 billion in operating cash flows versus $400 billion of capex in 2025 (4) – limiting the need for external financing. But for smaller players, access to credit and investor appetite matter much more. If liquidity tightens or sentiment shifts, the rate of spending could slow.

Still, even if the pace moderates, the scale of investment remains large enough to move the needle on growth. If AI-related capex holds near current levels – and assuming a reasonable share of the supply chain stays onshore – it could continue contributing roughly 0.3 to 0.5 ppt to annual US GDP growth. That’s a meaningful tailwind, even as other parts of the economy cool.

Finally, AI’s impact isn’t limited to capex. Its surge has also contributed to a wealth effect that could help sustain consumer spending. JPMorgan estimates that the rise in AI-related stocks has added over $5 trillion to household wealth over the past year, which could contribute about $180 billion to consumer spending, or 0.6% of GDP. While this is likely concentrated among higher-income households, this effect provides a meaningful cushion for consumption as income growth slows.

The labour market is softening amid weaker demand, government layoffs

While AI-related capex is helping to support growth, the US labour market is clearly cooling. Over the past three months to August (the latest available data before the US government shutdown – more on that later), job creation has averaged about 29,000 per month. That’s down from around 82,000 over the same period last year. The unemployment rate has also been edging higher – from 4.0% at the start of the year to 4.3% as of the latest data in August.

As shown in Chart 3, the rise in unemployment over the past year has been most pronounced in sectors that tend to move closely with overall demand: leisure and hospitality, manufacturing, and retail trade. Meanwhile, unemployment has also risen among government and education and health services workers, reflecting public-sector layoffs and the unwinding of pandemic-era support. 

We also note that, despite concerns that AI may be contributing to rising unemployment, studies including recent research from MIT, find that its impact on jobs has so far been limited (5). At the same time, there’s still little evidence that AI investment is creating large-scale hiring just yet. For example, the data-centre build-out is labour-intensive during construction, but once facilities are up and running, they require relatively few workers to operate (6). Together, this highlights the current two-speed nature of the economy – strong investment, but slower job growth.

In the near-term, news flow around policy and politics could add to market volatility

We believe this two-track economy will remain the dominant driver for markets in the months ahead. Still, headline risks could contribute to shorter-term volatility in markets. Two key sources for the final quarter stand out: the ongoing US government shutdown and flare-ups in US-China trade tensions.

US government shutdown creates data gaps, risk of lasting job losses

The ongoing US government shutdown – now the longest since the 35-day episode in 2018-19 – is creating a direct but likely short-lived drag on growth. Each week of closure is estimated to shave 0.1 to 0.2 percentage points from Q4 GDP, as up to 750,000 federal employees remain furloughed and agencies suspend key functions.

Ultimately, this shutdown – as with previous episodes – will eventually come to an end. But there are some differences this time around that investors should be aware of:

1. Labour market impact. President Trump’s plan to permanently cut 300,000 federal jobs by year-end suggests that temporary furloughs could become lasting job losses and would be a clear departure from past shutdowns (7). As shown in Chart 4, federal government employment has already fallen by about 94,000 since the start of the year and, as mentioned above, is adding further pressure to an already cooling labour market. That said, hitting that year-end target would require eliminating roughly 200,000 additional positions in just three months, a significant challenge given both the compressed timeline and the scale of reductions implied.
2. Data disruption. The shutdown has created an economic data blackout for policymakers and markets. With key statistical agencies closed, and official releases delayed, the Federal Reserve has been left with limited visibility into economic conditions. In this environment, it’s unlikely to deviate from its signaled policy path – currently, 50 basis points of cuts by year-end. On top of that, distortions from the government shutdown could also nudge the unemployment rate higher in the near term, potentially reinforcing the central banks’ bias toward gradual easing.

US-China trade tensions and tariff uncertainty are likely to continue

At the same time, flare-ups in US-China trade tensions and broader uncertainty over US tariffs remain potential sources of market volatility:

• Earlier in October, for example, China tightened export controls on rare earths and other critical minerals, prompting a US response that included the threat of tariff increases of up to 100% on Chinese goods (8). That exchange triggered a market sell-off – with the S&P 500 falling 2.7% on 10 October – followed by a rebound after President Trump backtracked a few days later. The episode underscores how sensitive markets remain to trade frictions, even as the broader economy holds up.
• More broadly, November is shaping up to be a busy month, with several key events on the horizon – including a potential meeting between Trump and Chinese President Xi Jinping in South Korea at the APEC summit (31 Oct-1 Nov), a US Supreme Court review on the legality of existing tariffs (5 Nov), and the expiry of the US-China trade truce (10 Nov).

The current environment signals a continued regime of “inflationary growth” – supportive for risk assets

The cooling labour market and continued uncertainties are prompting the Fed and US government to continue to run easier monetary and fiscal policy. This easing bias, combined with strong AI investment, should continue to support growth in the US. Similar trends are also unfolding globally:

• In China, strong investment in AI and advanced tech is supporting growth as Chinese tech firms compete to keep up with the US. (Read more: CIO Insights: China AI – Build Now, Profits Later)
• In Europe, fiscal policy has turned more supportive as governments prioritise infrastructure, defence, and energy security. (Read on: CIO Insights: Will Europe’s fiscal spark ignite real growth?)
• And in Japan, newly elected Prime Minister Sanae Takaichi has been a proponent of fiscal stimulus, particularly in strategic areas like defence and tech.

Looking forward, a stimulative fiscal environment should help to keep growth steady, but also keep inflation elevated. This suggests that we are likely to remain in a regime of “inflationary growth”, which is broadly supportive for equities and other risk assets. This can be seen in the past quarter’s “everything rally” – one which our General Investing portfolios powered by StashAway have also benefited from. 

(See more here: StashAway’s Q3 2025 Returns)

One standout in this “everything rally” has been gold, which has risen by as much as 66% this year. As shown in Chart 5, gold’s performance this year has been the second strongest in the past century, trailing only the surge following the 1979 oil shock. With such outsized gains and positioning extremely extended, short-term volatility is likely to emerge – as seen in the pullback over the past week. Even so, we believe the long-term case remains intact, as structural drivers – including central-bank demand, fiscal uncertainty, and gold’s role as a portfolio diversifier – continue to support the asset. Gold remains a valuable hedge within diversified portfolios, particularly in an environment where inflation stays elevated and policy remains loose.

Looking ahead, we see a balance of opportunities and risks as markets enter the final stretch of 2025. The combination of strong investment momentum, steady but uneven growth, and easier policy provides a constructive backdrop for risk assets – though bouts of volatility are likely as politics and trade remain in focus. For long-term investors, staying diversified across asset classes remains the best way to tune out the noise and stay focused on the long-term signal.

Authors

Stephanie Leung, Chief Investment Officer

Stephanie and her team oversee the full spectrum of investment products and portfolios offered at StashAway. She brings more than two decades of investment expertise across multiple asset classes. Prior to joining StashAway in 2020, she managed investment portfolios at institutions such as Goldman Sachs and multi-billion dollar family offices in the region.

Justin Jimenez, Head of Macro and Investment Research

Justin has more than a decade of experience in economic and investment research, and contributes to shaping the investment office's views on the global economy and asset classes. Prior to joining StashAway in 2022, he was an economist at Bloomberg.

Glossary

Policy easing

When policymakers act to boost economic activity – whether by making credit cheaper through rate cuts or by increasing government spending and reducing taxes to put more money into the economy. 

Capital expenditure

The money companies spend on fixed assets such as buildings, equipment, and infrastructure. In the AI context, this includes spending on data centers and the chips needed to train and run models.

Hyperscalers

The world’s largest technology companies that operate massive cloud and AI infrastructure. Their investments in data centres, chips, and networks underpin much of today’s digital economy and are driving the current AI spending boom. Major players include Amazon, Microsoft, Google, Meta, and Oracle. 

Wealth effect

The tendency for people to spend more when the value of their assets – such as equities or real estate – increases. Rising wealth boosts consumer confidence and supports economic growth, while falling asset prices can have the opposite effect.

Headline risks

Events or developments that can cause sudden and significant market reactions, often driven by news headlines. These risks can affect investor sentiment and lead to volatility in financial markets.

Inflationary growth

An economic regime where the overall price level rises alongside expanding economic output. This means the economy is growing, but so are prices.

Fiscal stimulus

Government measures aimed at boosting economic activity, usually implemented through increased public spending or tax cuts. The goal is to encourage consumption and investment to support growth during downturns.

References

1. J.P. Morgan Asset Management. (2025). Is AI already driving US growth? Retrieved from: https://am.jpmorgan.com/us/en/asset-management/adv/insights/market-insights/market-updates/on-the-minds-of-investors/is-ai-already-driving-us-growth
2. Fortune. (2025). Data centers accounted for all of US GDP growth in early 2025, says Harvard’s Jason Furman. Retrieved from: https://fortune.com/2025/10/07/data-centers-gdp-growth-zero-first-half-2025-jason-furman-harvard-economist/
3. McKinsey & Company. (2025). The cost of compute: A $7 trillion race to scale data centers. Retrieved from: https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-cost-of-compute-a-7-trillion-dollar-race-to-scale-data-centers
4. Morningstar. (2025). This is an AI boom, not a bubble, says Alger Funds CEO Dan Chung. Retrieved from: https://www.morningstar.com/markets/this-is-an-ai-boom-not-bubble-says-alger-funds-ceo-dan-chung
5. MIT Sloan School of Management. (2025). How artificial intelligence impacts the US labor market. Retrieved from: https://mitsloan.mit.edu/ideas-made-to-matter/how-artificial-intelligence-impacts-us-labor-market
6. The Wall Street Journal. (2025). AI data centers create jobs — just not as many as you think. Retrieved from: https://www.wsj.com/tech/ai-data-center-job-creation-48038b67
7. Reuters. (2025). US government shed 300,000 workers this year, Trump’s HR chief forecasts. Retrieved from: https://www.reuters.com/legal/litigation/us-government-shed-300000-workers-this-year-trumps-hr-chief-forecasts-2025-08-14/
8. Bloomberg. (2025). China tightens exports of rare earths and related technology. Retrieved from: https://www.bloomberg.com/news/articles/2025-10-09/china-tightens-exports-of-rare-earths-and-related-technology