How the Energy Industry Makes Money: The Business Model Guide

The energy industry makes money in a handful of distinct ways: selling oil, gas and power at market prices, charging fees to move and store them, earning a regulated return on network assets, signing long-term contracts, getting paid to be available, selling equipment, collecting subscriptions and monetizing policy credits. Each link in the chain, from the wellhead to the household meter, relies on a different mix of those engines. Durable profit goes to the companies that hold the scarce part of the chain at the moment it is scarce, and that have financed themselves to survive the bottom of the cycle.

A single week in Texas shows how unusual this business is. In February 2021, during Winter Storm Uri, the regulator held ERCOT's wholesale price at its $9,000 per megawatt-hour cap for days, against a normal average in the tens of dollars. A generator whose plants ran earned years of revenue in days. Vistra, the state's largest generator and retailer, had sold power forward that its frozen gas plants could not produce, and put its net loss at about $1.6 billion. Brazos Electric Power Cooperative received an ERCOT bill of about $1.8 billion for the week and filed for bankruptcy on 1 March 2021. The product was identical; the outcomes were opposite. What decided them was whether the plant physically ran and what each company had promised. That is the energy business in miniature, and the rest of this guide explains why.

What is the energy industry's business model?

There is no single energy business model. "Energy company" covers a national oil company, a pipeline operator, a regulated utility, a commodity trader and a rooftop-solar installer. The most useful way to see the industry is as a chain in which each link sells to the next and depends on the one before it.

Around the chain sit three things it cannot work without: capital (from utility bond buyers to tax-equity investors and venture funds), rules (from state utility commissions and FERC to OPEC+ quotas and tax credits) and resources (from the rock that holds the oil to lithium and copper).

What separates energy from wheat or copper is a set of four problems ordinary commodities do not have. Electricity must be produced in the same instant it is consumed, so shortages produce price jumps, not gentle rises. Almost every asset is paid for up front and earns back over decades, so the industry is a financing business as much as an engineering one. Energy is worthless until it reaches a buyer over a wire, a pipe or a shipping route, and most of those routes are monopolies or long contracts. And nearly every step needs a government's permission. The book's opening chapter on why energy is not just a commodity works through each of these.

Where does the money in an electricity bill go?

Follow one kilowatt-hour and the business model becomes concrete. Take a stylized household in a PJM state, where generators compete in a wholesale market and the local utility owns only the wires. It uses 900 kWh a month. The numbers below are invented round figures, not any utility's tariff, but the effective rate is close to the 2025 US residential average of about 17.3 cents per kWh (EIA).

Three things stand out. The commodity is less than half the bill; the wires take nearly half. Each line is priced by a different authority: wholesale markets, a grid operator's auction, a federal regulator, a state commission and legislators. And the lines move at different speeds, from monthly energy prices to rate cases every few years. When PJM's capacity price jumped roughly ninefold in 2024, the capacity line on bills like this one went from about $2 a month toward $20–25 with no change in usage. Our explainer on the PJM capacity auction shows how that happened.

A gallon of gasoline is the opposite. With WTI at $70 a barrel, crude is about $1.67 of a stylized $3.12 gallon, more than half the price, and every $10 a barrel adds about 24 cents. The refiner earns a spread, not a price, and retailers earn more from coffee and car washes than from fuel. Chapter 3, "Following the Money", traces both journeys in full.

The eight ways energy companies make money

Almost every energy business runs on one or more of eight revenue engines. Knowing which one pays is the first step in reading any company.

The engines subsidize each other. The household's kilowatt-hour above may have come from a plant financed partly by a tax credit sold to a bank, delivered over wires funded by bonds sold against an allowed return, with capacity priced by an auction that data-center demand pushed up. For any energy business, three questions follow: which engine actually pays, who sets its price, and does it still pay in the year the benchmark falls, the credit expires or the regulator says no?

Which parts of the energy value chain earn the best margins?

Margins settle unevenly along the chain, and the pattern explains most strategy in the industry.

The links with the steadiest margins own either the cheapest resource or a route others must use. The links that earn the most in a crisis earn the least afterward. The links closest to the customer carry the most risk for the least margin. So every link tries to become a bottleneck: producers buy terminals, utilities build generation, traders buy refineries, and data-center developers sign 20-year contracts for reactors.

How do oil and gas companies make money?

Upstream producers sell a commodity at a price they do not set, so their business is a cost curve. A producer whose barrels are profitable at $100 and loss-making at $60 does not have a low-cost asset; it has a high price. The winners hold barrels near the bottom of the curve. Saudi Aramco's cost position and its export routes are why it earned about $93.4 billion in 2025. In 2026, after traffic through the Strait of Hormuz was effectively halted from early March, Brent reached a reported high of $126.41 on 30 April, and producers with routes that bypassed the strait were worth more than producers with cheaper barrels and no way out.

US shale added a twist. Shale wells decline steeply, often 60–70% in their first year in many plays (reported), so holding output flat requires constant drilling. Through the 2010s the sector as a whole spent more than it earned from operations, funded by high-yield debt and equity. The busts of 2015–16 and 2020, including Chesapeake's Chapter 11 on 28 June 2020, ended that growth model. Survivors adopted "capital discipline" and consolidated, most notably in ExxonMobil's purchase of Pioneer Natural Resources for about $59.5 billion, completed on 3 May 2024.

Because a producer's value rests on what is still in the ground, the evidence matters. Reserves are forecasts, and the rules for booking them shift with prices; our guide on how to read an oil and gas reserves report explains what to check.

How do pipelines and LNG terminals make money?

Midstream's model is a toll road: earn a fee for every unit moved or stored, ideally under a contract that pays whether or not the capacity is used. The best example is Cheniere. Its Sabine Pass terminal was built to import gas, then stranded when shale made US gas cheap. Between 2011 and 2013 Cheniere signed 20-year agreements under which buyers paid 115% of the Henry Hub price for gas plus a fixed fee per unit of capacity, about $2.25 to $3.00 per MMBtu, whether or not they lifted a cargo (reported). Lenders financed the plant against those fees.

The fees did their job in both directions. In mid-2020, when overseas prices fell below the cost of shipping US gas, buyers cancelled more than 100 cargoes and kept paying (reported). In 2022, when Europe's TTF price hit a record of about €340/MWh, the buyers holding the contracts earned the windfall. The plant owner earned roughly the same fees in every scenario, which is why a contracted LNG terminal trades like infrastructure rather than like a gas producer. The weak point is trust in the contract itself, a theme our article on how power purchase agreements work takes up for electricity.

How do utilities make money?

A regulated utility does not set its own price and does not earn a margin on electricity. It holds a monopoly over the wires, and often over generation, in exchange for an obligation to serve and a regulator who sets its revenue. The formula is simple: operating costs, depreciation and taxes are recovered at cost, plus a return on the capital invested in the network, the rate base, at an allowed return on equity. US allowed electric ROEs averaged about 9.74% in 2024, about 9.77% in 2025 and about 9.79% in the first half of 2026 (S&P Global Regulatory Research Associates, reported).

That formula means a utility grows earnings by growing its rate base, so management is paid to invest. US investor-owned utilities planned more than $1.1 trillion of capital spending over 2025–29 (Edison Electric Institute). The limit is the customer's bill. Every capital plan is also a rate-increase plan, and "affordability" became the dominant word in 2025–26 rate cases and state elections. Two incumbent Georgia commissioners lost in November 2025 in a race fought largely over bills and the cost of the Vogtle nuclear expansion (reported), the story told in our case study on Vogtle's cost overrun. To see the arithmetic from the inside, read our guide on how to read a utility rate case.

How do power generators make money?

About two-thirds of US electricity load is priced in organized wholesale markets such as PJM, ERCOT, MISO and CAISO. Every five minutes the operator stacks generators' offers from cheapest to dearest and pays every dispatched plant the same clearing price. That uniform price creates an inframarginal rent, the gap between the clearing price and a cheap plant's own cost, and it is how nuclear, wind, solar and efficient gas plants recover their capital. In most hours gas plants set the price, so owners of nuclear and renewable plants are, in effect, long gas.

Markets that pay only for energy struggle to keep the plants a grid needs on its worst days, which run only a few hundred hours a year. ERCOT answers with scarcity prices; PJM and other Eastern markets pay separately for capacity. Whether a state uses markets at all is itself a choice: our comparison of regulated and deregulated electricity markets sets out the trade-offs.

From 2024 the market repriced generators. Data-center demand, plant retirements and accreditation changes sent PJM's capacity price from about $28.92 per MW-day for 2024/25 to about $269.92 for 2025/26 and about $329.17 for 2026/27. Owners of existing nuclear plants signed long contracts: Constellation with Microsoft for the roughly 835 MW Crane restart (September 2024) and with Meta for Clinton's roughly 1,121 MW from June 2027. Vistra, which had lost about $1.6 billion in Uri, became one of the best-performing stocks in the S&P 500 in 2024 (reported). Its plants had not changed. What they owned, firm power with a grid connection, had become scarce. Our case study of the 2021 Texas power crisis shows the same company on the other side of that coin.

How do renewable developers and manufacturers make money?

Sunlight and wind are free, so a solar or wind farm's cost is almost entirely capital. That turns the business into a contest over the cost of money. In a stylized 200 MW onshore wind project in the book's research, the price the project needs under a long-term contract rises from about $42 per MWh when debt costs 5.5% to about $52 when it costs 8%, a 24% increase from financing alone, and to about $66 without the production tax credit. Value in renewables has therefore flowed to whoever has cheap capital, such as NextEra and Brookfield, and to the lowest-cost manufacturer, usually Chinese.

The contract that pays the developer is usually a power purchase agreement with a utility or a corporate buyer. The 2022–23 rise in interest rates broke fixed-price contracts signed before it: Ørsted cancelled its 2.2 GW Ocean Wind projects off New Jersey on 31 October 2023 and booked about DKK 28.4 billion of impairments for the year (company). Chapter 14 on project finance and the cost of capital explains why the financing is the price.

Manufacturing is a different game: cost is learned on the factory line. Module prices fell to about $0.10–0.20 a watt for Chinese modules in 2023–25 (reported), and battery packs from more than $1,000 per kWh around 2010 to about $108 in 2025 (BNEF). First Solar survived by spending two decades on its process before scaling. Europe's battery champion did the opposite, as our case study on what happened to Northvolt explains.

Why do so many energy companies fail?

Energy failures follow recognizable patterns. A boom funds capacity that overshoots. Debt is sized to peak prices against assets that earn for decades. A fixed-price promise meets a cost or price shock. A licence is lost through an accident, a fraud or a change of government. The list of casualties runs from Enron in 2001 through Solyndra, SunEdison, Westinghouse, PG&E twice, Chesapeake and Northvolt. Our roundup of the biggest energy bankruptcies and what they teach names the factor each one turned on.

The common thread is that the losses fall hardest on the thinnest balance sheets. In Uri, Vistra and NRG lost heavily and survived. A non-profit cooperative, Griddy and Just Energy did not. Risk in energy is correlated: the Texas gas system and power system failed together, each making the other worse, at the moment demand peaked.

The power equation: four factors that multiply

The book behind this guide, The Power Equation, argues that an energy company creates durable value only when four things are true at once, and that they multiply rather than add. If any one is zero, the result is zero.

1. Supply at a cost the market will pay. A resource, plant, machine or service that delivers energy below the price buyers will pay across the cycle, not just at the top of it.
2. A route to market. An offtake contract, pipeline, grid connection, rate base or customer base that turns production into revenue. Stranded supply is worth nothing.
3. Capital matched to the asset. Financing whose cost, term and structure fit an asset paid for up front and earning over decades, strong enough to survive the low point.
4. A licence to operate. The permits, approvals, safety record and public consent to build and keep running, met continuously rather than owned.

Two forces sit around the equation like weather: commodity prices and politics. Neither decides whether a project is sound. Both decide how long an unsound one survives, and both can bankrupt a sound one that was financed for fair weather only.

The equation explains the cases above. Cheniere's story turned on the route to market: cheap gas was worthless to a terminal facing the wrong way. Vogtle turned on cost. Northvolt had customers and capital but never a competitive cost per good cell. PG&E's monopoly and capital were never in doubt; its licence was, after its equipment started the 2018 Camp Fire that killed 85 people. Uri is the rare case in which all four failed at once. Most failures come from treating one factor as proven because another was: cheap capital mistaken for low cost, a subsidy mistaken for a market, a licence assumed to be permanent.

How to analyze an energy company

A short screen, applied before any deeper work, catches most problems. Test each factor at the bottom of the cycle, not at today's prices.

Then name the revenue engine that pays, who sets its price, and what happens when that price is at its low. The glossary defines the terms used here, from capacity factor to rate base.

The takeaway

The energy industry makes money through a small set of engines, commodity sales, fees, regulated returns, long contracts, capacity payments, equipment, subscriptions and policy credits, spread across a chain that runs from the wellhead to the meter. Durable profit goes to whoever owns the scarce link at the moment it is scarce, whether that is a low-cost reserve, a contracted terminal, a grid connection or a running reactor. The bottleneck moves, and the companies that thrive are the ones whose cost, route, capital and licence all hold together when prices and politics turn. For the full framework and the cases behind it, start with Chapter 1 of The Power Equation.