The market has begun to doubt its own certainties: AI, the Fed and inflation complicate the end of 2026.
Wall Street enters the final weeks of the year facing an unusual combination of risks. Artificial intelligence continues to drive billions of dollars in investment, but investors are beginning to demand evidence of profitability. US inflation is accelerating again, and the Federal Reserve is considering further interest rate hikes. The result is a market that has become more sensitive to any disappointment, where good news for the economy can turn into bad news for the stock market.
For much of the past few years, markets had found a relatively straightforward way of interpreting the future. Artificial intelligence promised an economic transformation on an enormous scale, major technology companies appeared well positioned to lead it, and the Federal Reserve would eventually cut interest rates as inflationary pressures subsided. The narrative had its uncertainties, but it offered a recognisable direction.
November has begun to undermine that sense of comfort.
Artificial intelligence remains one of the decade’s biggest corporate bets. Nothing has demonstrated that this is about to change. However, the enthusiasm that justified paying increasingly high prices for certain companies is now having to coexist with rather less attractive questions: how much will it cost to build all this infrastructure, when will it generate sufficient profits, and what happens if financing becomes more expensive?
Meanwhile, US inflation has once again become a cause for concern. The Consumer Price Index rose by 4% year on year in October, up from 3.7% in September. The Federal Reserve, which during an earlier period had been able to focus on supporting a weakening labour market, is once again facing the risk that price pressures could persist.
This is no minor shift. The possibility of further interest rate hikes affects stock market valuations, increases corporate borrowing costs and changes the relative attractiveness of equities compared with bonds.
Concern has already become visible in market prices. In early November, the Nasdaq fell to almost 12% below its October 6 peak. In a single trading session, the S&P 500 technology sector dropped by 7.3%, while the Philadelphia Semiconductor Index lost more than 12%.
These figures are not enough to conclude that a prolonged bear market has begun. They do, however, show that investors have become less tolerant of bad news.
And that may be the defining feature of the final weeks of 2026: not so much the emergence of entirely new risks, but the fact that markets are beginning to demand answers to questions they had been willing to postpone for far too long.
Artificial intelligence enters a less forgiving phase
The first major uncertainty lies in the sector that has contributed most to sustaining expectations of corporate growth.
Artificial intelligence is no longer merely a technological promise. It is driving real investment in data centres, semiconductors, electricity grids, cooling systems and computing capacity. Companies need infrastructure to train and run increasingly demanding models, and that demand is spreading into industries that, only a few years ago, had a relatively indirect relationship with the digital economy.
The scale of investment is a sign of corporate confidence. It is also an economic commitment.
Building a data centre requires capital to be committed before there is any certainty about the revenue it will generate. Buying cutting-edge chips may provide a competitive advantage, but it does not guarantee that this advantage will translate into high margins throughout the equipment’s useful life. Financing projects through debt also introduces an additional vulnerability when interest rates rise.
During the most expansive phase of a technological transformation, markets tend to focus on the potential scale of the opportunity. Later, they begin to distinguish between companies capable of capturing that value and those simply taking part in the race to build it.
The fact that artificial intelligence improves the productivity of the economy does not mean that every company connected to it will deliver extraordinary returns to its shareholders. A technology can transform entire industries while also generating unprofitable investments, excess capacity or competition that ultimately erodes margins.
Economic history offers plenty of examples of this separation between innovation and financial returns. Railways, telecommunications and the internet transformed productive activity, but not every company that financed their expansion achieved the expected profits.
The question today is not whether artificial intelligence is a passing fad. That would be an oversimplification. The real question is how much of its future value has already been priced into the market.
On November 12, the International Monetary Fund called for caution, arguing that it is still too early to determine whether recent market movements represent a lasting reassessment of investment and the expected returns from AI.
A stock market correction may reflect a fundamental change in a sector’s prospects, but it can also result from a temporary decline in risk appetite, forced selling or adjustments to investors’ positions. Price movements alone do not explain their underlying cause.
What does appear to be changing is the level of scrutiny. Technology companies can no longer rely exclusively on the market rewarding every increase in spending on artificial intelligence. It will become increasingly important to demonstrate what revenue that spending generates, how much additional capital it requires and when it can be converted into cash flow.
The Federal Reserve is no longer where many investors expected it to be
The second source of tension is in Washington, although not at the White House.
The Federal Reserve is facing a situation that makes monetary policy decisions considerably more complicated. The US economy remains relatively strong, while inflation is still above its 2% target and shows signs of spreading across a wider range of goods and services.
In October, the central bank kept its benchmark interest rate within a range of 3.75% to 4%, following an increase in September. The decision allowed policymakers to observe how prices and economic activity were evolving before introducing further changes.
Christopher Waller, a member of the Federal Reserve Board of Governors, said on November 10 that he intended to support at least a quarter-percentage-point rate increase in December. Other policymakers have raised the possibility of acting more quickly if inflation continues to surprise on the upside.
The next meeting of the Federal Open Market Committee is scheduled for December 8 and 9. It will be one of the defining events of the final weeks of the year, although subsequent market movements will depend not only on the interest rate decision itself, but also on the message accompanying it.
If the Fed maintains an overly restrictive policy for too long, it could slow investment, make credit more expensive and weaken economic activity. If it eases financial conditions before inflation is under control, it risks allowing price increases to become entrenched.
On November 10, Michelle Bowman, the Federal Reserve’s Vice Chair for Supervision, explained that improvements in the labour market and persistent inflation had led her to pay greater attention to the second risk. She also noted that financial conditions had recently tightened as a result of falling share prices, higher private-sector interest rates and a stronger dollar.
Investors generally welcome signs of economic strength because they suggest higher corporate revenues and profits. But when inflation remains elevated, those same figures can reinforce the possibility of further interest rate increases.
An economy growing faster than expected can trigger a stock market decline if investors believe that such growth will force the Fed to keep borrowing costs high for longer.
Inflation can no longer be explained solely by oil prices
The war in the Middle East and the crisis surrounding the Strait of Hormuz have placed considerable pressure on energy markets. Higher oil prices and logistical difficulties are spreading across numerous sectors, from transport to industrial production.
However, attributing the rise in US inflation solely to crude oil prices would be insufficient. The latest figures show price increases extending beyond energy. October’s report recorded pressure across services and consumer goods, fuelling concerns that inflation may prove more persistent than expected.
A temporary increase in energy prices can push the price index higher for a limited period. If oil prices subsequently fall and the rest of the economy does not incorporate that increase into its costs and wages, inflation may moderate. The problem arises when the initial increase spreads to other activities, changes the expectations of businesses and consumers, or coincides with new sources of pressure.
Investment in artificial intelligence is one of them. Building data centres requires vast quantities of equipment, energy, land, materials and electricity grid capacity. If demand grows faster than available supply, the prices of these resources may rise.
Bowman noted that tariffs, energy costs and pressures associated with investment in artificial intelligence had made a significant contribution to keeping core inflation above target. She also suggested that the adoption of AI could increase productivity and reduce costs in the future.
The same technology can generate inflationary pressures while it is being built and disinflationary effects once it is put to use. What remains unclear is how long each phase will last.
Simply buying technology is not enough to become more productive. Companies must reorganise tasks, train workers, adapt systems and establish whether the savings generated outweigh the implementation costs.
The risk of technology and interest rates reinforcing each other
Until recently, artificial intelligence and monetary policy could be analysed as two relatively separate stories. The first promised corporate profits and economic growth. The second determined the cost of money and the relative attractiveness of different financial assets.
Now the two are becoming increasingly interconnected. Investment in artificial intelligence can increase demand for energy, equipment and financing. If it contributes to keeping inflation elevated, the Federal Reserve may be forced to maintain higher interest rates. Those rates, in turn, make precisely the technological projects whose profitability depends on future revenue more expensive.
This creates an uncomfortable relationship: the investment driving growth may also contribute to the financial conditions that make it harder to sustain.
That does not mean an inevitable vicious cycle exists. Many companies have substantial internal resources and are not exclusively dependent on debt. Moreover, genuine improvements in productivity can increase profits and offset some of the financial costs.
However, vulnerability increases when stock market valuations incorporate particularly demanding expectations. A share price reflects, among other things, the profits investors expect a company to generate in the future. When the interest rates used to value those future earnings rise, their present value may fall.
The market does not need to conclude that a company is going to fail in order to significantly reduce its valuation. It may be enough to believe that its profits will arrive later, that it will require additional investment or that the cost of capital will be higher.
Wall Street discovers that not all good news is good for its stocks
November’s volatility has an economic component, but it is also driven by expectations.
During periods of optimism, markets tend to interpret economic data favourably. Strong growth means higher profits. Moderate inflation means lower interest rates. Rising technology investment means new opportunities.
When sentiment changes, these relationships no longer work as easily. A strong economic activity reading may reinforce expectations of an interest rate increase. Higher technology spending may raise doubts about its profitability. A fall in oil prices may improve the inflation outlook while simultaneously being interpreted as a sign of weaker demand.
The same data can have different implications depending on the context. Financial prices incorporate expectations about the future, not just information about the present. A development can be positive in absolute terms and still disappoint if the market was expecting something even better.
The problem becomes more pronounced when investment positions are highly concentrated. If a relatively small group of major companies accounts for a significant share of the performance of market indices, a correction in their valuations can spread across the broader market, even if many other companies maintain relatively stable prospects.
This concentration does not prove that a bubble exists. But it can amplify market movements when investors revise their expectations at the same time.
The bond market has something to say again
During periods of very low interest rates, much of the appeal of equities came from the scarcity of alternatives offering meaningful returns. That environment has changed.
When yields on US government debt rise, investors can obtain higher income from assets that, in general terms, carry less credit risk than corporate shares.
The comparison is not straightforward. Bonds can also lose value when interest rates rise, and their real returns depend on inflation. Nevertheless, higher yields change capital allocation decisions.
Following the release of October’s inflation figures, the yield on the US 10-year Treasury bond stood at around 4.99%. The move added pressure to stock prices and to sectors most sensitive to borrowing costs.
The impact extends far beyond Wall Street. US government bond yields influence international financial conditions, the cost of financing for companies and governments, and asset valuations in other markets.
For emerging economies and companies with dollar-denominated debt, this environment can be particularly challenging. The bond market is therefore becoming a central factor in the final weeks of the year.
Europe is not immune, although its problems are different
The debate tends to focus on the United States because that is where the major technology companies are concentrated and because Federal Reserve decisions have global effects. European markets, however, are also exposed to these pressures.
The first channel is financial. Major stock markets are closely interconnected, and changes in international risk appetite can spread rapidly between regions.
The second is economic. European companies depend on financing conditions, international trade and energy prices. Persistent inflation in the United States can influence the dollar, capital flows and investment decisions.
The third relates to artificial intelligence. Europe participates in the technology supply chain through industrial companies, equipment suppliers, energy infrastructure, digital services and specialised manufacturers.
For Spain, there is also a particularly relevant dimension: the cost of financing and its impact on investment and housing. Federal Reserve decisions do not directly determine Spanish mortgage rates, which depend on European monetary and financial conditions. However, a major international disruption can influence debt markets, business confidence and the availability of credit.
Fear is not an economic indicator, but it changes behaviour
Talking about fear in financial markets may seem like an indulgence in headline-driven language. Stock markets do not have emotions. The people making decisions within them do.
Risk aversion influences how much investors are willing to pay for an asset, what losses they are prepared to tolerate and how quickly they react to unexpected news.
When uncertainty increases, appetite may decline for investments whose returns depend on particularly favourable scenarios. Attention also shifts towards liquidity, balance-sheet strength and companies’ ability to finance themselves without constantly relying on new resources.
The VIX index, commonly used as a benchmark for expected short-term volatility in the S&P 500, reached levels in early November not seen for approximately seven months. This signals stronger demand for protection against stock market movements, although it cannot, on its own, predict the market’s future direction.
High volatility does not necessarily mean that share prices will continue to fall. It indicates that the market is assigning greater importance to substantial price movements.
Fear can also have real economic consequences. If companies believe that financing will become more expensive or that their share prices are losing value, they may reconsider investment projects. If households see a significant decline in their financial wealth, some may reduce consumption.
December will be a test of expectations
The Federal Reserve meeting on December 8 and 9 is attracting much of the market’s attention. But reducing the end of the year to a single interest rate decision would mean overlooking part of the problem.
The evolution of prices, the employment situation, signs of corporate investment and the ability of major technology companies to justify their spending plans will also matter.
If inflation begins to moderate and economic activity maintains reasonable growth, pressure on the Fed could ease and perceptions of risk may improve. If prices continue to surprise on the upside, the central bank will have less room to adopt a more flexible monetary policy.
There is also a more complicated scenario: technology investment could slow sharply while inflation remains elevated because of energy or supply-side factors. In that case, the central bank would face a combination of weaker economic momentum and persistently high prices.
These are scenarios, not predictions. The available information does not make it possible to assign reliable probabilities to them without an explicit model and assumptions that could change rapidly.
The real test begins when patience runs out
Markets have spent years trying to anticipate the economic value of artificial intelligence. During that time, the technological promise has coexisted with an enormous supply of capital, growth expectations and a considerable willingness to finance projects whose results will take time to materialise.
The environment is now less favourable. US inflation has accelerated, the Federal Reserve is considering further interest rate increases and technology valuations are facing more demanding scrutiny. The war in the Middle East adds uncertainty over energy costs and trade routes.
There is no sufficient reason to conclude that the expansion of AI has come to an end or that markets are destined for a prolonged decline. The US economy continues to demonstrate its capacity for growth, and technology investment may deliver genuine productivity improvements.
But neither is there any reason to assume that share prices should continue rising simply because technology has a promising future.
The difference between a major economic transformation and a good financial investment often becomes clear when profits are compared with the money spent to generate them.
In the coming weeks, Wall Street will have to confront that comparison while trying to anticipate the decisions of a Federal Reserve that can no longer take inflation control for granted.
For a long time, the dominant question was how much the economy could gain from artificial intelligence. Now another question is beginning to take precedence, one that is less spectacular but probably more important to investors: how much are they willing to pay today for profits that have yet to materialise?











Comments