future promise and present pressure

The investment landscape for 2026 is marked by three powerful interconnected forces that mark a new market frontier: artificial intelligence (AI), global fragmentation and inflation uncertainty.

JP Morgan has produced an interesting report explaining how to invest in a world where the promise of AI-powered productivity growth collides with the pressure of more volatile inflation and a fractured world order.

Generative AI has captured investors’ attention since late 2022 and continues to gain momentum. The potential of AI is profound, as it could drive the cost of experience to near zero, transform the labor market, and boost business productivity and profit margins.

The models’ capabilities have improved rapidly and it is estimated that they could reach human-level performance by spring 2026:

Technological progress has led to a massive increase in infrastructure investment. Large US technology companies have tripled their annual capital spending (capex), going from $150 billion in 2023 to possibly more than $500 billion in 2026.

AI-related investment has contributed more to US GDP growth in 2025 than consumption. Despite this boom, Investment in AI is currently around 1% of US GDP, so it could even more than double, since in other revolutionary cycles in history investment peaks of 2% to 5% of GDP were reached.

AI has sparked a debate about whether we are in a bubble. Bubbles are usually characterized by excess capacity not justified by demand.

According to JP Morgan, there is no excess capacity in the current AI sector, data center vacancy rates are at an all-time low of 1.6%, and three-quarters of the capacity under construction is already pre-leased.

Furthermore, listed AI companies have generated their returns entirely through earnings growth. Over the past three years, the P/E multiple of publicly traded AI stocks has declined, while earnings per share estimates have doubled.

To achieve scaling (more and better AI), computational infrastructure and architecture are required. It will always be necessary to invest in more and better capacity.

Regarding circularity, the agreements involve visionary companies with big plans. We are talking about titans with a mentality startupthat is, the best of both worlds. These are firms with their founders still very involved and with a tremendously innovative way of thinking.

Only these giants have the financial capacity to sustain a $20 trillion investment cycle without debt. This circularity of agreements is designed to strengthen the monopolistic position of these companies, generating a new, totally different balance.

The current bull market is the most hated rally in history. In my opinion, much of the bubble narrative is attributed to analysts and managers who have been left out of the cycle by refusing to concentrate their portfolios on the few dominant stocks.

Disruption

If today’s technological disruption is comparable to or greater than that of the 1990s, the bull market could last a long time. Tom Lee predicts 10 more years of bullish stock market supercycle.

Let’s now look at this chart tracking the performance of recent IPOs. Although euphoria is building, the report suggests we are not at the peak of a bubble right now:

So far, the biggest winners in the market have been the hyperscalersdata center and electrical infrastructure companies, and energy providers. Companies expected to benefit from AI-driven productivity gains have been left behind:

The most pressing limit to the expansion of AI is energy. In the US, there is a five-year delay in adding new power generation to the existing grid. Data centers require reliable, affordable power, making natural gas a critical baseload source, although renewables (which take less time to build) will also help.

He agua for data center cooling is also a factor investors should keep an eye on.

JP Morgan recommends a four-pillar strategy to capture value in AI:

1. Focus on large cap leaders– It is recommended to maintain focus on large hyperscalers (Microsoft, Amazon, Google, Meta) due to their rapid profit growth and significant incremental AI revenue generation (estimated at $25 billion quarterly).
2. AI Supply Chain Opportunity: invest in technology enablers: energy, semiconductors (demand outstrips supply, Nvidia’s Blackwell chip expected to sell out in 12 months), connectivity and cooling systems.
3. Identify “smart” corporate AI users: look for companies that successfully implement AI to increase sales and profits. Current winners will likely increase their profits exponentially due to faster and more efficient integration of AI.
4. Venture capital exposure: Private markets are crucial because companies remain unlisted longer (the average tech IPO occurs at 14 years now, up from eight years in the 1990s). Most of the future value of AI is expected to accrue in the platform and application phases. In the latter, venture capital played a relevant role in the previous cycle (internet and cloud):

Global fragmentation is replacing globalization, prioritizing resilience and security over efficiency. This change manifests itself in wars, tariffs and the formation of trade blocs. The energy factor is central to the fragmentation and rise of AI. Europe has pivoted to liquefied natural gas instead of Russian gas, prioritizing safety over cost.

Tariffs are here to stay, affecting almost 70% of US goods imports.

The White House appears to be closing a North American trade bloc. The following graph compares the change in the 2025 effective tariff rate: the increase is significantly smaller for Canada and Mexico than for China and the rest of the world:

Decoupling between the US and China continues, with China’s share of US imports falling from 22% in 2017 to 12% today.

Even so, China has known more than to compensate for this situation with its exports to the rest of Asia.

However, this export power is not yet translating into substantially stronger corporate profits:

China and the US enjoy a clear advantage in the cost of electricity compared to other economiesespecially Europe and Japan. This advantage is critical given the immense electricity demand of AI data centers. It should be noted that the situation in Spain is benign:

JP Morgan’s evidence is clear: We are not in a bubble, but at the beginning of a supercycle driven by massive investment in AI infrastructure. Those companies that embrace this disruption will not only invent the future, but will dominate the present.