Key Takeaways
- For clues on the stock market’s ability to extend its rally, investors should watch the pace of spending on AI infrastructure and its impact on semiconductor company profits.
- The bond market could see a huge influx of debt sold to finance the AI infrastructure buildout, which could ripple through to the government bond market in the form of a bias toward higher yields
- Tariff- and housing-related inflation could open or close the door to rate cuts from the Fed.
As investors look ahead to 2026, the forces shaping stocks, inflation, and bond yields are already familiar. We have artificial intelligence spending, Federal Reserve policy, and stubbornly high housing costs. What’s less clear is how these drivers will play out across markets.
Stock investors came into last year in an AI-buying frenzy, but they grew more selective as concerns mounted about how the massive technology buildout will be funded. Meanwhile, tariffs may have faded from investors’ day-to-day concerns since the market convulsions of March, but with inflation stubbornly high, the ripples will keep spreading in 2026.
We spoke to analysts about what data they will be watching to determine how companies are funding the AI buildout, the knock-on effects of that spending on credit markets and energy costs, and the lingering inflationary pressures from tariffs and housing. Six datasets made our final list. Together they offer a clearer view of market trends of 2026.
- AI accelerator revenue
- Hyperscaler capital expenditures
- AI-driven bond issuance
- Electricity costs
- Tariffs
- Rental costs
Here’s a closer look at these trends.
AI Accelerator Revenue: How High Can It Go?
Perhaps no trend has dominated the stock market more than the AI technology boom; it’s certainly led in terms of returns. In 2025, some 60% of the market’s total 17.4% gain came from technology and communication services stocks. Of the 21.4 percentage points gained by the Morningstar US Technology Index, 11.9 points were from semiconductor stocks.
Can the AI rally continue? According to Morningstar senior equity analyst Brian Colello, a key data point is revenue from AI accelerators, which are specially designed hardware (generally advanced semiconductor chips) that speed up the processing of these programs. Nvidia NVDA and Broadcom AVGO are leaders in designing accelerator chips.
“The AI industry is compute-restrained, and the early revenue and profits from AI are coming from the chip and infrastructure makers supporting the AI buildout, led by Nvidia,” Colello explains. Morningstar expects revenue at Nvidia to rise 46% in 2026 following an estimated 55% gain in 2025. At Broadcom, revenue is seen rising some 124% in 2026, following a roughly 50% estimated gain in 2025.
Hyperscaler Capital Expenditures: Driving the Revenue Boom
Whether the tidal wave of profits at semiconductor companies can continue depends on how much hyperscalers (large cloud computing companies and other tech giants) spend to build out the data centers required to train and run AI computer models.
Hyperscalers include mega-cap tech companies Alphabet GOOG, Amazon AMZN, Meta Platforms META, Microsoft MSFT, and Oracle ORCL. These firms are investing huge amounts of capital in data centers, and this spending has skyrocketed in the past two years. Morningstar estimates that hyperscaler capital expenditures more than doubled from 2023 to 2025, and that number is expected to nearly triple again by 2029.
How hyperscaler capex plays out will be critical to the outlook for both semiconductor and cloud computing stocks. Given the huge weight these stocks have in key indexes, the trajectory of the market itself may also be at stake.
Colello thinks the torrent of spending should persist: “We don’t foresee spending on Nvidia’s AI GPUs, Broadcom’s AI XPUs, or AMD’s GPUs slowing down anytime soon.” However, “future AI spending is uncertain. It’s possible spending comes in lower if funding is tighter, energy constraints mute the buildout, AI provides fewer benefits to users, or some breakthrough techniques occur that make AI far more efficient. This would lead to lower AI accelerator revenue and lower hyperscaler capex.”
Those stumbling blocks could make it difficult for semiconductor companies to fully meet expectations, “but we think they’ll get close,” Colello says. “There’s more upside in these names if none of these issues emerge.”
AI-Driven Bond Issuance: The AI Wave Hits the Bond Market
The AI buildout is going to cost a lot, and all the money has to come from somewhere. “The global data center and AI buildout will be an extraordinary and sustained capital markets event,” wrote JP Morgan analysts in November. They estimate that the cost of global data centers, AI infrastructure, and power supplies could total $5 trillion.
For the giant hyperscalers, cash flows will cover a big portion of the bill. But many companies are expected to tap the debt markets heavily in coming years. Meta sold $30 billion of bonds in late October to finance its AI efforts—its largest bond sale ever.
JP Morgan predicts that investment-grade bond markets could see more than $300 billion of AI- or data-center-related debt in 2026, part of a record $1.81 trillion in investment-grade bond issuance for the year. That would include roughly $120 billion from hyperscalers, along with another $100 billion or so tied directly to data center and power buildouts. Over the next five years, JP Morgan forecasts a need for $1.5 trillion of AI/data center funding from the bond market.
For John Briggs, head of US rates strategy at Natixis, the key is to see this in the context of the overall US bond market, especially in terms of a rising cost of capital for both corporations and governments. At the same time, some of the corporate AI issuers (most notably Microsoft) are seen as exceptionally safe credits and treated almost on par with government debt.
“There’s a little bit of a worry that all these issues … will mean upward pressure on interest rates in general,” Briggs says. They come at a time of ballooning budget deficits that the US government needs to finance, along with rising government borrowing needs in Europe. Briggs says the question is whether the flood of AI issuance could effect the government bond market, weighing on prices and elevating yields. “From a macro perspective, these are dollars that need to be found … Will the marginal yield need to go up?”
Electricity Costs: Shocking the Consumer
Still another layer to the AI wave has been the big rally in utilities stocks, especially those servicing big data centers. In 2025, the Morningstar US Utilities Index gained nearly 20%, following its 26.7% gain in 2024.
Utilities stocks have historically been sleepy and dividend-focused, but the group was transformed by AI. The mega-sized data centers that train AI models require huge amounts of electricity to run their servers, storage, and networking equipment, as well as their cooling systems.
While many massive data center projects are still years away from coming online, seem say the demand from tech companies training their AI models has already had a knock-on effect: higher electric bills for consumers. Notably, this has happened even as overall energy costs have dropped.
Utilities operate in a heavily regulated marketplace, often under the oversight of elected officials keenly sensitive to the impact high rates have on their constituents. Analysts say rising electricity costs constitute one potential source of backlash against the AI data center buildout. Such a backlash could also cut off the utility stock rally.
“The price people pay for electricity has gone up more than 10% since the beginning of 2024,” says Morningstar senior equity analyst Travis Miller. “At a high level, this has resulted in faster earnings growth for utilities and the big sector rally the last two years. It’s also a risk in 2026-27. If prices go too high, regulators might push back on price increases and earnings growth will slow.”
Tariffs: A Lingering Impact but Uncertain Future
Rising electric bills could be one part of the inflation story for 2026. Another important variable is the continued role tariffs play in driving goods prices higher for consumers. While investors’ worst-case scenarios for President Donald Trump’s tariffs did not play out in 2025, higher import levies bled through to the US economy and kept inflation well north of the Fed’s 2% target.
“Businesses have passed little of the tariffs onto US consumers so far, but that could change in 2026,” says Morningstar senior US economist Preston Caldwell.
Crucially, the story on tariffs isn’t over yet. In the coming months, the Supreme Court will rule on the legality of Trump’s tariffs under the International Emergency Economic Powers Act. But even if the court rules against the president, there’s no guarantee that tariffs will come back down and bring consumers relief.
“On the one hand, there is a litany of alternative statutes that the Trump administration could use in lieu of the IEEPA, which could keep the average tariff rate in the low teens or higher,” Caldwell explains. “On the other hand, public opinion shows increasing impatience with stubbornly high prices, so perhaps the administration will lose interest in tariffs.”
Rental Costs: Some Good News on Inflation?
While tariffs could keep inflation higher than it otherwise would be, one factor that had been driving consumer costs higher could start to provide some good news and help keep the Fed on track to cut rates in 2026.
Throughout 2024 and for much of 2025, housing inflation stayed high. However, many economists saw this as an issue with the data, rather than a reflection of real-world inflation. The housing component of the Consumer Price Index reflects trends in rental costs, which tend to lag actual market rents. (House prices are not part of the CPI.)
Owing to this difference, “a gap opened between market rents and housing CPI over 2021-22,” Caldwell says. “That kept housing CPI inflation high over 2023-25, even after market rent growth slowed.” However, housing inflation dropped sharply in the final months of 2025. “Because the gap has now closed, we expect housing CPI inflation to slow further in 2026,” Caldwell says. “That will have a meaningful impact on overall inflation, given housing’s large weight.”
Caldwell thinks that taken together, tariffs and housing inflation could influence the Fed’s decisions on interest rates. “Both relief on tariffs and falling housing inflation could allow the Fed to cut another 50 basis points (as the market expects) or more in 2026,” he says. “However, if tariffs go higher, or if they merely stay at current levels and we get significantly more pass through into consumer prices, then inflation could come in higher in 2026 than the Fed is currently projecting, which could take further rate cuts off the table.”
