Under the Radar: 3 Small-Cap Stocks Powering the AI Energy Boom

I keep reading about the enormous sums being spent on artificial intelligence, and I keep coming back to the electric bill.

All those expensive computers have to run on something.

Put enough of them in enough buildings, leave them working around the clock, and supplying the electricity becomes a pretty good business opportunity.

Some of the smaller companies helping to get that power delivered could make investors a lot of money.

Solar has my attention because demand is growing while plenty of investors remain thoroughly disgusted with solar stocks.

After watching some of these share prices, I can understand the frustration.

But a stock that has disappointed investors can still belong to a business with a promising future, especially when the price has become low enough to make the arithmetic interesting.

And right now, we have some very good reasons to spend time looking at this corner of the market.

The AI Boom Has an Enormous Electricity Bill

The International Energy Agency expects global electricity consumption to rise 3.6% in 2026 and 3.8% in 2027.

It also expects data center electricity use to roughly double between 2025 and 2030.

Somebody has to supply all that additional power.

And the demand reaches well beyond the companies building AI systems.

Families with rising incomes want air conditioning, refrigerators, and the other conveniences most of us take for granted.

Factories need electricity to make things.

Electric vehicles shift some of what we spend at the gas pump onto the power grid.

That gives the electricity business several sources of growth.

We can be enthusiastic about the opportunity without having to guess exactly how much the next AI model will add to a technology company’s profits.

Solar Is Becoming an Important Part of the Solution

Solar is already helping meet the growing need for electricity.

The world installed approximately 690 gigawatts of solar capacity in 2025, an increase of 15%.

U.S. installations jumped 45% in the second quarter of 2026, with large solar developments accounting for most of the growth.

The economics help explain why developers keep building.

The International Renewable Energy Agency (IRENA) puts the global average lifetime generation cost for new utility-scale solar projects installed in 2025 at approximately 4.4 cents per kilowatt-hour.

Costs can be considerably higher in the United States, and that figure does not include everything required to supply electricity around the clock.

But it helps explain why a customer facing an enormous electric bill might want solar in the mix.

Batteries make the proposition more useful.

Store some of the electricity generated during the day, and you can use it in the evening.

Add other generation sources and a connection to the grid, and customers have more ways to obtain the reliable service they need.

Amazon’s plans for future data centers in Nevada provide an example.

The arrangement includes:

• 600 megawatts of solar power
• 600 megawatts of battery storage
• 100 megawatts of geothermal power

Several businesses get paid to help keep those computers running.

Of course, we will have to put up with the usual frustrations.

Permits take time.

Grid connections can take even longer.

Politicians change the tax rules, and construction schedules get rearranged.

The latest U.S. industry forecast calls for roughly flat annual solar installations over the next five years.

Even that would approximately double the installed fleet as new projects are added year after year.

A smaller company does not need every industry forecast to be revised higher to build a very profitable business in that market.

That brings me to three companies worth a closer look.

Each trades in the United States, has a market value comfortably below $3 billion, and averages far more than 5,000 shares in daily trading.

They provide three different ways to benefit from the equipment and services required to build more electricity infrastructure.

Array Technologies (ARRY): A Beaten-Down Solar Stock Trading at a Bargain Valuation

Array Technologies gives us a particularly interesting combination of a growing business and a beaten-up stock.

The company makes tracking systems that move solar panels to follow the sun.

Getting more electricity out of a solar installation improves the economics for its owner, which makes the equipment worth paying attention to.

Array also supplies foundations and other equipment used in large solar projects.

It is a straightforward business serving a growing industry.

And the company’s order book suggests customers still need what it sells.

At the end of June, Array had approximately $2.5 billion in executed contracts and awarded orders, up 37% from a year earlier.

It booked more than $500 million in new orders during the quarter.

That is a substantial amount of business for a company whose common stock has a market value of approximately $525 million at recent prices.

Of course, those orders will take time to become revenue.

And the company’s expected growth is fairly modest.

Analysts project revenue growth of approximately 12% in 2026 and 13% in 2027.

Those are respectable numbers, but they are not what makes Array especially interesting to me.

It is the valuation.

Paying Less Than Five Times Projected Adjusted Earnings

At approximately $3.40 per share, Array trades for roughly 4.5 to five times management’s projected 2026 adjusted earnings of $0.68 to $0.75 per share.

That is a low multiple for a company with a substantial order book and expectations for continued revenue growth.

A business does not have to grow at a spectacular rate to reward shareholders who buy at that kind of earnings multiple.

Continued growth and a little more confidence from other investors can do a great deal of the work.

If the company executes well, improves its financial position, and convinces investors that its earnings are sustainable, the combination of higher profits and a higher valuation multiple could produce a substantial gain.

But there is a reason we need to do more than look at the earnings multiple.

The Balance Sheet Gets a Vote

Array’s June financial figures included approximately:

• $658 million in debt
• $307 million in cash
• $506 million in preferred stock liquidation preference

The lenders and preferred shareholders have claims ahead of us as common shareholders.

That matters.

A company can have a growing business and still leave common shareholders with disappointing returns if too much of the value goes toward servicing debt or satisfying preferred claims.

I want to see Array turn its orders into cash, meet its obligations, and improve what is left for common shareholders.

That is how a low stock price can turn into a worthwhile investment.

Array gives us an established supplier with substantial customer orders and an opportunity to grow alongside a market that needs more electricity.

If management executes well and the company’s finances improve, today’s valuation could prove very attractive.

For someone willing to do the balance sheet work, ARRY deserves a close look.

Shoals Technologies (SHLS): Strong Sales Growth With Room for Margin Improvement

Shoals Technologies gets paid for another necessary part of the job.

Connecting everything.

Shoals makes electrical equipment used in solar projects, battery storage installations, and data center power systems.

Its products help customers install and operate those systems more efficiently.

A developer can spend a fortune on panels and batteries, but the electrical connections still have to work.

That may not make for a particularly glamorous business description.

I can live with that if the company keeps selling more equipment and making money.

And the recent sales numbers are encouraging.

Revenue Growth Is Already Showing Up

In the second quarter, Shoals reported 47.4% revenue growth to $163.4 million.

Backlog and awarded orders increased 19.4% to $801.4 million.

Analysts expect approximately 32% sales growth and 22% adjusted earnings growth in 2026.

The company also earned a profit in the second quarter.

Those figures tell us that demand for its products is translating into actual business.

And unlike a company that depends entirely on a future industry forecast, Shoals already has substantial orders and revenue growth to examine.

But there is another part of the story that could make the earnings picture more interesting.

Improving Margins Could Make a Big Difference

Shoals’ gross margin declined to 30.3% from 37.2%.

The company has been dealing with a factory transition, product mix changes, and quality-related expenses.

Those problems have weighed on profitability.

They also create an opportunity for improvement.

If management can address those issues, Shoals could keep more of each additional dollar of revenue.

That is the kind of situation I like to investigate.

We have customers ordering more equipment, while management has specific operating problems it can work on.

If sales continue growing and the factories become more efficient, earnings could improve meaningfully.

Of course, the improvement is not guaranteed.

Manufacturing transitions can take longer than expected, product mix can remain unfavorable, and quality expenses can persist.

We need to see the progress in the financial results.

But the opportunity is there.

Shoals has an equity market value of approximately $1.3 billion.

That is still a relatively small business compared with the amount of solar, storage, and electrical infrastructure that needs to be built.

More orders, better margins, and stronger earnings provide a reasonable path toward a much more valuable company over time.

SHLS offers a combination of strong sales growth and potential operating improvement.

That makes it worth further research.

Ameresco (AMRC): Winning Major Data Center Power Projects

Ameresco gives us a third way to participate in the growing demand for electricity.

And this company has already started winning substantial business from customers trying to solve their power problems.

Ameresco develops energy infrastructure, combines generation and storage, and modernizes buildings and other facilities.

A customer that needs more reliable electricity may need several of those services.

That gives Ameresco multiple ways to earn a place on a project.

The company’s second-quarter awards certainly got my attention.

$1.8 Billion in New Project Awards

Ameresco reported $1.8 billion in new project awards during the second quarter.

Approximately $1.2 billion was associated with data centers.

Total backlog reached $6.73 billion, up 32%.

Those are substantial numbers.

Especially when recent stock prices value the common equity of the entire company at approximately $1 billion.

We have a relatively small company winning some very large projects.

That is exactly the sort of situation that gets me interested in a business.

But we have to give those figures their proper meaning.

Backlog becomes revenue over time.

Awarded work may still have contracting and development steps to clear before it becomes revenue.

And a large project award is not the same thing as a large profit.

Still, the numbers provide evidence that demand for more data center power is already reaching Ameresco’s business.

This is not simply a forecast about what customers might spend several years from now.

The company is already winning work.

Earnings Growth Is Promising, but Cash Flow Matters

Analysts expect approximately 10% revenue growth and 36% adjusted earnings growth in 2026.

That is an attractive earnings growth forecast.

However, part of the expected improvement reflects tax benefits and a change in accounting for transferable investment tax credits.

I will be watching cash flow alongside the earnings statement.

That distinction is important.

Accounting earnings can improve without an equivalent increase in cash available to shareholders.

And Ameresco operates a business where financing and project execution matter a great deal.

The company has meaningful financial obligations.

It needs to finance projects on terms that leave a worthwhile return for shareholders.

A big award becomes much more exciting when the company can build the project, collect the money, and earn a good profit after paying the financing costs.

That is the challenge.

It is also the opportunity.

With a backlog of $6.73 billion, Ameresco has a substantial amount of potential business to execute.

Successful project delivery, stronger cash generation, and more profitable work could make today’s approximately $1 billion equity valuation look very small.

AMRC gives us direct exposure to the growing demand for data center power infrastructure, with substantial awards already in hand.

The key will be converting that opportunity into cash and sustainable profits.

Three Different Ways to Profit From Growing Electricity Demand

What I like about these companies is that we can explain how they might make more money.

Array Technologies (ARRY) helps solar installations generate more electricity by supplying tracking systems and related equipment.

Its low adjusted earnings valuation is especially interesting to the value investor in me, although its financial obligations require close attention.

Shoals Technologies (SHLS) supplies the electrical connections and equipment needed to make solar, storage, and other power systems work.

Its strong sales growth and potential margin recovery give us two possible sources of earnings improvement.

Ameresco (AMRC) develops and delivers energy infrastructure projects, including work associated with the growing demand for data center electricity.

Its substantial backlog creates opportunity, but project execution, financing, and cash generation will determine how much of that opportunity reaches shareholders.

These are different businesses.

They also have different financial risks.

But they are connected by one important trend.

The world needs more electricity, and customers are spending money to obtain it.

That spending will be spread across several technologies.

Solar and batteries can win business alongside natural gas, geothermal, and nuclear power as customers figure out how to obtain dependable electricity at an acceptable cost.

There is plenty of room for companies that perform a useful job and do it profitably.

Finding Opportunity Beyond the AI Headlines

The biggest technology companies can spend billions building the next generation of artificial intelligence.

They can buy more chips, construct more data centers, and announce increasingly ambitious plans.

All those computers still need electricity.

That creates opportunities for companies that most investors are not talking about nearly as much as the big AI names.

Array has a low earnings valuation and substantial customer orders.

Shoals has strong revenue growth and an opportunity to improve profitability.

Ameresco is already winning major projects tied to the need for more power.

None is without risk.

We need to pay attention to balance sheets, cash flow, project execution, and whether growth actually produces attractive returns for common shareholders.

But these are the kinds of smaller businesses I want to keep digging into.

Bought at sensible prices, they could deliver very large gains if growing electricity demand translates into better earnings and stronger finances.

The big technology companies can keep announcing their plans for more computers.

I will be keeping an eye on the smaller companies that could get paid to help supply the electricity.