The global economy's path in a fragmented world

What’s new?

In our Central Macro Scenario for 2025 CMA we incorporate:

		New socio-economic paths to include higher geopolitical fragmentation
		More fragmented and delayed climate transition
		More granular assessment of AI impacts at the country level

With the world on the brink of significant structural changes, projecting the long-term macroeconomic scenario for 2025-2050 presents unique challenges. It is crucial to identify emerging trends while evaluating the acceleration of existing ones. Key factors to consider include shifts in trade relationships shaped by geopolitical developments, a more fragmented global landscape marked by rising nationalism, and major demographic changes such as declining fertility rates and a shrinking working-age population. These trends may be counterbalanced by renewed migratory flows, increased labour intensity through later retirement ages, and enhanced productivity driven by a technological revolution.

Compared to the 2024 Capital Market Assumptions (CMA), we have revised the macro long-term scenarios to factor in important expected changes in terms of socio-economic developments, the climate transition, and the impact of artificial intelligence (AI) on growth.

We introduced a new combination of socio-economic scenarios, moving away from “a world that follows a path in which social, economic, and technological trends do not shift markedly from historical patterns”¹, adding instead complexity through different emerging paths. Regional rivalry becomes the dominant narrative, as nations increasingly prioritise domestic and regional issues over global cooperation. This resurgence of nationalism, driven by concerns over competitiveness and security, leads to a focus on achieving energy and food security within regional confines, often at the expense of broader development goals. Consequently, inequality rises as resources are concentrated within regions, leaving marginalised communities behind. In this starkly divided landscape, investments in human capital are unevenly distributed, further exacerbating disparities in economic opportunity and political power. This stratification is evident both within and between countries, where high-tech sectors thrive while local environmental policies primarily benefit affluent areas. Meanwhile, the energy sector diversifies, balancing investments in carbon-intensive fuels with low-carbon alternatives, yet the costs and benefits are not equitably shared. Lastly, in a world where fossil-fuelled development coexists with rapid technological progress, local environmental challenges like air pollution should be effectively addressed, and geo-engineering is considered a viable option for sustainable development.²
 

We factor in a more fragmented and delayed transition. In line with findings from previous years, the 2025 CMA is characterised by a disorderly Net Zero transition, with both transition costs and especially physical ones on the rise over the long term, although transition costs for the next decade will be lower compared to last year’s assumptions, as they are deferred to the future.

The Network for Greening the Financial System (NGFS) framework remains the starting point for our scenario.

Changes in NGFS scenarios: The newly released NGFS scenarios are similar to last year's, but the narrative indicates a generally more disorderly transition, introducing a new damage function that significantly impacts GDP levels due to climate risk. All information is retrievable on the NGFS website
 

Last year we explored how Artificial Intelligence (AI) could impact the major production factors in the economy (labour, capital, and productivity). Here, we extend our analysis by focusing on differentiating macroeconomic productivity gains across countries. We use the IMF’s AI Preparedness Index and its sub-components (digital infrastructure, innovation and economic integration, human capital and labour market policies, and regulation and ethics) to differentiate the starting point and adoption of AI across economies. Productivity, capital investment, and capital depreciation are then modelled to estimate the impact of productivity on potential GDP growth. According to our simulations, AI adoption should progressively boost productivity at the global level, reaching its peak in ten years' time. Broader adoption of more commercial applications of AI should allow for similar patterns across regions, although for some countries and regions economy-wide benefits might appear sooner in light of their preparedness, for which the DM vs. EM divide does not apply. If the first years see some differentiation, AI broadening adoption will help the ‘laggards’ to catch up on productivity as we reach the middle of the next decade. Ongoing competition and innovation in the field of AI is key to broaden  the benefits of the new technology across the economy (by lowering barriers to entry, accelerating adoption, and creating new opportunities) and to make the benefits spread from the microeconomic level (firm, industry, sector) to the macroeconomic one. Breakthroughs in efficiency gains that make a technology more affordable will be pivotal as they make AI investments and benefits affordable to many small companies, previously cut out. Lower costs for AI models could lead to faster adoption by both corporates and households, higher spending, and aggregate investment for AI, boosting aggregate productivity.

Our central scenario has some important macro implications:

A generally more disorderly and delayed transition extends over a longer timeframe, deferring economic costs and softening inflation peaks.

Enactment of emission policies is more gradual, translating into modest initial increases in carbon prices and investments in infrastructure to ensure energy transmission and grid stability are challenged by shifting policy agendas that have to cope with the new geopolitical environment.

While transition-driven inflation may take a backseat, the technological transformation, amid national security concerns, might represent a structural support for the demand for commodities, potentially driving up prices. Food security policies, coupled with increasing climate-related concerns, might cause food prices to come under pressure, with significant impacts especially on Developed Markets (DM).

For food exporters, a rise in global food prices may trigger more protectionist measures by governments aiming to favour domestic markets over external ones to shield consumers, ultimately contributing to a further escalation of protectionism and supply chain disruption. Moving towards 2040, as pressures for the transition mount, we see inflation deviating from targets in a more pronounced way, which will only be partly offset by increases in productivity and technological progress. On average, we therefore expect the inflation regime to remain supported above central banks’ targets by a relay of factors over the next three decades (from 2025 to 2055), with temporary bouts of volatility, but remaining under control overall.
 

Long-term growth patterns remain driven by demographic and productivity trends. Asian tech leadership will make it a larger contributor to global growth in the future, while the demographic dividend will benefit Africa the most.

Against a general backdrop of older and shrinking populations across several developed and emerging countries, the positive effects of the productivity gains generated by adopting AI temporarily limit the deceleration for the most demographically challenged countries, especially around the mid-2030s to mid-2040s. Only those countries enjoying a positive demographic dividend such as India and Sub-Saharan countries are able to grasp both the positive effects of younger and more productive economies, enjoying higher sustainable potential growth for longer. 

Asia's growth premium is likely to persist, fuelled by technological advancements that are reshaping economies and creating substantial investment opportunities. With a strong commitment to innovation and a young, tech-savvy population, the region is positioned as a leader in the global tech landscape. Asia is a powerhouse in the technology sector, leading in various fields: Manufacturing and Supply Chain Technology (China and Vietnam), Semiconductor Production (Taiwan and South Korea), Fintech innovations and financial inclusion (India), and E-Commerce. In 2023, Asia accounted for 76.6% of World IT Goods Exports and for 33% of World IT Services Exports, with China and India respectively leading the two segments.

In the medium to long term, insufficient early efforts to mitigate climate risks will lead to increased chronic physical costs across regions. This, combined with the waning impact of AI on productivity, will ultimately shift potential growth to be primarily driven by demographic factors. Consequently, we will see a deceleration in potential growth across developed markets and China, while Asian countries like India will play an increasingly significant role.
 

1 Refer to SSP2 Shared Socioeconomic Pathaways  (SSPs) called “Middle of the Road”.
2 We have also included part of the SSP3 (“Regional Rivalry”), SSP4 (“A Road Divided”) and SSP5 (“Fossil-fuelled Development”).