To The Moon

I've heard that the end of AI is energy. But it was a shock to see Wells Fargo's estimate. AI electricity demand will grow from 8 terawatts today to 652 terawatts in 2030. Just looking it up, 1TW is equivalent to 1 trillion watts.

As data centers' hunger for electricity continues to grow, the power industry is embracing a new growth cycle driven by AI technology. In key data center regions like Texas and the Mid-Atlantic, the surge in power demand is already significantly impacting electricity prices.

Three Power stocks hit new highs

Take $Vista Energy(VIST)$ for example. In their recent earnings call, they revealed that forward electricity prices in northern Texas have jumped 13% since last November. Based on this price increase, Vistra expects its adjusted Ebitda to exceed $6 billion by 2026, a whopping 24% higher than analysts' pre-earnings call estimates, indicating a significant acceleration in growth.

$NRG Energy Inc(NRG)$ is also optimistic, forecasting gross profits of $420 million in Texas over the next few years, a 27% jump compared to 2024 expectations.

As the largest nuclear power plant operator in the US, $Constellation Energy Corp(CEG)$ expects its adjusted operating profits to grow an average of about 10% annually from now through 2028, thanks to the inflation adjustment mechanism in the federal nuclear tax credit policy.

Constellation's CEO, Joseph Dominguez, pointed out in the earnings call:

The power demand from data center customers has hit a 20-year high. This high demand for instant power is even prompting some companies to consider using carbon-emitting energy sources.

Vistra reported that potential customers have expressed interest in buying power from their natural gas plants. NRG said a data center customer wants to triple the capacity of its existing facilities in the next three years.

Shares of these three power companies have soared significantly after their earnings releases, joining Nvidia and Super Micro Computer as the top five performers in the S&P 500 over the past 12 months. Vistra's stock has surged nearly 273% in the past year, while Constellation and NRG have gained 173% and 150%, respectively.

The four waves of opportunities brought by AI

The First Phase: Nvidia's Prosperity

As $NVIDIA Corp(NVDA)$ , the AI leader, consolidates its gains these days, $Goldman Sachs(GS)$ released a research report. If we consider NVIDIA's meteoric rise as the first phase of the "AI boom" in US stocks, then AI investments should unfold into three more extensive phases after NVIDIA.

Goldman Sachs sums it up: the first phase of AI hype in US stocks was characterized by "Nvidia's Prosperity." While the stock has nearly sextupled since 2023, its rolling P/E ratio hasn't changed much compared to early 2023 due to revenue explosions.

However, the focus of Goldman Sachs' equity strategy team in this report is not NVIDIA, but rather which industries and concepts are likely to become market darlings as the "AI trade" gradually extends.

The Second Phase: AI Infrastructure

Goldman Sachs points out that the second phase of the AI boom will spill over to infrastructure companies involved in further developing AI technology.

According to Goldman Sachs' framework, this phase includes companies in semiconductors, cloud service providers, data center REITs, hardware and equipment companies, security software developers, and utility companies. Goldman Sachs emphasizes that while many second-phase companies have already seen valuation expansions, there are significant differences in earnings revisions.

Especially utility companies (essentially power stocks), Goldman Sachs believes that the increased power demand brought by AI development and potential rate increases will be positive for the industry. Looking at share price performance over the past year, US utility stocks have generally been in a slight decline.

At this stage, Goldman Sachs has listed 93 companies that meet the "AI Infrastructure" criteria, including $ARM Holdings Ltd(ARM)$ $Taiwan Semiconductor Manufacturing(TSM)$ $Cisco(CSCO)$, and virtually all power companies. In addition to these household names, companies like $Synopsys(SNPS)$ $Power Integrations(POWI)$ $KLA-Tencor(KLAC)$ $Vertiv Holdings LLC(VRT)$ also made the list.

The Third Phase: AI Enablers

Goldman Sachs said that this phase is about companies and industries that stand to benefit from AI, but not as directly as server makers or power companies. They're gonna ramp up revenue by commercializing AI tech.

At this point in the AI adoption cycle, software and IT services are sitting pretty. Many company execs are already talking about how their products will help others harness AI in the future.

For this phase, Goldman Sachs has named 31 US-listed stocks, including software companies like $ServiceNow(NOW)$ $Datadog(DDOG)$, and IT service providers like $Snowflake(SNOW)$ $MongoDB Inc.(MDB)$.

These stocks in the third phase basket have seen returns of nearly 8% this year. Sure, AI isn't the only reason for their rise, but it's a sign that some investors are already trading on this third-phase wave.

The Fourth Phase: Productivity Gains

Now, we're entering the sweet spot of AI boom — companies that can achieve significant productivity gains by adopting AI.

Goldman Sachs is pointing to software services, business, and professional services as the industries with the most potential to boost labor productivity with AI. These industries have high labor costs, and a good chunk of those wages could be impacted by AI automation. Plus, in the recent earnings season, 30% of the execs talking about AI-driven productivity gains were from these three industries.

It includes retail giants like $Wal-Mart(WMT)$ $Costco(COST)$, healthcare companies like $Illumina(ILMN)$ $Tenet Healthcare(THC)$, and even The New York Times, which is currently in a legal battle with OpenAI. These are the companies that are really riding the AI wave and reaping the benefits!