Chevron Found a New Way to Profit from the AI Boom. What a 20-Year Microsoft Contract Means for CVX

Below are the specific parameters for the Chevron trade idea. The ticker for trading via RoboForex MobileTrader and MT5 on RoboForex is CVX.

Chevron has spent most of its history doing one thing: extracting oil and gas, refining it, and selling fuel. That model is not going away, but on 22 June 2026 the company added something new to it.

Chevron’s subsidiary Energy Forge One signed a 20-year power supply agreement with Microsoft to supply electricity to a data centre in Pecos, Texas. Under Project Kilby, Chevron will build a gas-fired power plant with capacity rising in stages to 2.67 GW.

The deal works for both sides. Microsoft gets a dedicated power source for its cloud and AI infrastructure for two decades, with no dependency on an already strained regional grid. Chevron gets a single large customer and guaranteed revenue stretching to 2046. Power deliveries are planned to start in 2028. The final investment decision is expected before the end of 2026.

The power plant will sit next to Microsoft’s data centre near Pecos, Texas, feeding it electricity directly without touching the regional grid. The construction is planned in phases: first power in 2028, then a gradual expansion to the full 2.67 GW.

The 20-year contract length matters for Chevron’s finances. A two-decade commitment gives the company a clearer picture of future cash flows and makes the project far easier to finance. Chevron plans to finalise the investment decision before the end of 2026, locking in the construction cost, financing structure, and technical specs.

GE Vernova will supply the core equipment for the plant. Solar Turbines, a Caterpillar subsidiary, will provide additional generating capacity. Investment firm Engine No. 1 is also a project partner.

The Permian Basin is the largest oil-producing region in the US. When you pump oil, you also bring up large quantities of natural gas alongside it. That gas needs somewhere to go, and the pipeline infrastructure in West Texas often cannot move it all out of the region fast enough.

The result: gas prices at the local Waha Hub regularly collapse to near-zero, and sometimes turn negative. Producers effectively pay to get rid of gas they cannot sell. For most companies, this is a problem. For Chevron, it is an input cost advantage.

Instead of selling that gas at a deep discount, Chevron can burn it in its own power plant and sell the output (finished electricity) to Microsoft at a contracted rate. The closer the power plant is to the gas source, the less pipeline capacity matters. That is exactly what Project Kilby is designed to exploit.

A standard power company buys fuel from someone else, generates electricity, and sells it. Chevron’s position is different: the company already owns the gas in the ground. Project Kilby lets it earn at multiple points in that chain:

The same cubic metre of gas that Chevron once had to sell at whatever the market offered now generates revenue at each of these steps.

Chevron’s earnings have always tracked oil and gas prices closely. When prices are high, cash flows are strong. When they fall, profits shrink fast. This is the nature of commodity businesses, and investors who own CVX shares understand it comes with the territory.

Project Kilby introduces a different kind of income. A 20-year fixed-term contract with Microsoft means Chevron knows, years in advance, how much electricity it needs to deliver and roughly what it will be paid for it. If the agreement includes capacity payment structures or fuel cost pass-through clauses, the revenue from this project becomes largely insulated from swings in the oil market.

Think of it as the difference between selling wheat at whatever price the market sets each harvest, and owning a bakery with a long-term supply contract. Both use the same raw material, but one produces far more predictable income. Chevron has stated publicly that Project Kilby is designed to generate cash flow that is less sensitive to the commodity cycle. A series of similar plants, once built, would gradually shift more of Chevron’s revenue into this steadier category.

Chevron and Microsoft have not disclosed the electricity price in their agreement. To estimate the project’s revenue potential, we use a range of 70 to 100 USD per MWh, built from publicly available benchmarks.

Average wholesale electricity prices in the ERCOT market (the Texas grid) are lower than this range, but they reflect trading on the existing grid, which is a different product. Project Kilby is a new plant built specifically to supply one customer around the clock, every day. According to Lazard’s estimates, the all-in cost of electricity from a new gas plant in the US runs at 65 to 67 USD per MWh, which sets the floor. On top of that base, a long-term contract, dedicated capacity, and the ability to bypass grid congestion all justify a price premium. That gives us 70 USD per MWh as the low scenario, 85 USD as the base case, and 100 USD as the high end.

At 2.67 GW of capacity running at roughly 90% utilisation, the plant would produce around 21 TWh of electricity per year.

In the most favourable scenario, Project Kilby could generate over 2 billion USD in annual revenue for Chevron. Relative to the company’s total scale, that is a meaningful but not dominant figure. The real importance of this number is what it signals: if one plant can reach 2 billion USD, a portfolio of similar plants changes the financial picture significantly.

Project Kilby did not come out of nowhere. Back in 2025, Chevron, Engine No. 1, and GE Vernova announced a programme to build plants with combined capacity of up to 4 GW for US data centres. At 2.67 GW, Project Kilby is the biggest piece of that programme.

If the Kilby plant delivers, on time, within budget, and with reliable power, Chevron has a working proof of concept to take to other buyers. Amazon, Alphabet, Meta, and Oracle are all spending tens of billions of USD on data centre infrastructure, and all of them face the same problem: securing enough power. Chevron can walk into those conversations with a completed project as evidence that the model works.

Chevron’s position in this market is genuinely hard to replicate quickly. The company already owns gas in the Permian, has the engineering expertise to run complex energy assets, has access to capital for large projects, and has existing relationships with the equipment manufacturers building these plants. A tech company wanting to copy this approach would need years and billions of USD to build the same starting point.

If AI infrastructure spending continues to grow, powering data centres could become a standalone segment of Chevron’s business, one that earns from the AI boom through electricity, without having to build AI models or take on the risks that come with them.

The honest answer for the short term: not much. Construction spending starts before revenue does, and the plant will not be generating power until 2028. CVX shares will continue to be driven by the factors that have always mattered most: oil and gas prices, Permian Basin output, progress on Guyana projects, results from the Hess acquisition, dividend policy, and buyback pace.

What the deal changes is the longer-term story. Once Project Kilby starts delivering power and generating cash, it adds a revenue stream that behaves very differently from the rest of Chevron’s business. At full capacity, the plant could add several hundred million USD of cash flow per year. The exact number depends on Chevron’s final ownership stake and the contract terms, but the direction is clear.

The more interesting question is what comes after Kilby. If Chevron closes two or three more contracts like this one over the next few years, the company starts to look meaningfully different: part oil major, part infrastructure operator with a contracted revenue base. That shift, if it happens, is what could gradually lift how the market values CVX.

On the daily chart, CVX shares are holding above the 200-day Moving Average. This is one of the most widely watched indicators in stock analysis: when a price stays above it, the long-term trend is considered intact. CVX is doing exactly that.

At the same time, the Stochastic indicator has dropped into oversold territory. This does not mean the stock will automatically bounce, but it does suggest that selling pressure has run most of its course and a recovery is increasingly likely.

The setup points to a specific entry trigger: a confirmed breakout above resistance at 180 USD. If the price closes above that level, it signals that the correction is over and the upward move is resuming. The first target from there is 199 USD, and the primary target is 213 USD, the historical high for CVX.

Below is a sample trading strategy for CVX shares. This example is for educational purposes only and does not constitute investment advice. Investors should assess their own risk tolerance independently.

A risk/reward ratio of 1:3 sits comfortably in the acceptable range for position trading. Keep in mind that markets are volatile: CVX shares can move both for and against an open position.

Project Kilby is a compelling opportunity on paper. Before acting on it, it is worth understanding what could go wrong.

The Microsoft deal does not transform Chevron overnight, revenue from Project Kilby is two years away at minimum, and oil prices will still be the main driver of CVX for the foreseeable future. But the deal adds a layer to the investment case that did not exist before: a contracted, long-term income stream that is structurally separate from the oil cycle.

Chevron is turning a local problem (too much gas, not enough pipelines in West Texas) into a long-term revenue stream from AI infrastructure. If the plant performs, the company has a template to replicate with Amazon, Alphabet, Meta, and others. That is the part of this story that analysts are likely tracking closely.

On the charts, CVX is holding above the 200-day Moving Average, with the Stochastic in oversold territory. A breakout above 180 USD confirms the correction is over, with the first target at 199 USD and the main target at 213 USD.