America's grid is being asked to absorb a load it was not built for.
- The North American grid operator's own watchdog is sounding the alarm. NERC's January 2026 Long-Term Reliability Assessment found that 13 of 23 North American assessment areas face elevated or high resource-adequacy risk over the next five years. MISO, PJM, ERCOT, SERC-Central and two WECC regions are flagged as high risk, where planned resources point to energy shortfalls beyond adequacy targets.
- The demand forecast is being revised up violently. NERC now projects summer peak demand growing by 224 GW over ten years — 69% higher than the 132 GW it projected just one assessment cycle earlier. New AI and digital-economy data centers account for most of that increase.
- Gas can't fill the gap on the timeline AI needs. Only three companies on earth build large-frame gas turbines, and all three are sold out for years. GE Vernova closed Q2 2026 with a 116 GW gas equipment backlog and is taking reservations for 2031 delivery. Siemens Energy sits on a 69 GW backlog; Mitsubishi Heavy is booking deliveries into 2028–2030. Lead times have stretched from 2–3 years to 5–7 years.
- Nuclear is the only source that delivers what a data center actually needs: firm, around-the-clock power. Per the Department of Energy, nuclear plants run at a capacity factor above 93% — roughly twice as reliable as coal or gas units and about three times wind or solar.
And the hyperscalers have already voted with their checkbooks. Constellation is restarting Three Mile Island to sell the output to Microsoft under a 20-year agreement, and signed a separate 20-year PPA with Meta for 1,121 MW from its Clinton plant beginning June 2027.
That is the trade. Here are the three companies we think are best positioned — and, critically, what Wall Street's own analysts are saying about each one, by name.
Stock #1 — Constellation Energy Corp. (NASDAQ: CEG)
The pure-play owner of America's largest nuclear fleet.
Constellation operates the biggest fleet of nuclear plants in the United States. When a hyperscaler wants carbon-free, 24/7 power at scale today — not in 2032 — Constellation is one of a very small number of phone calls they can make. It is the company that signed both marquee AI-nuclear deals of this cycle.
Analyst consensus: BUY
| Metric | Value |
|---|---|
| Price (Sep 17, 2026) | $263.18 |
| Consensus rating | Buy (22 analysts) |
| Average 12-mo. price target | $348.30 |
| Implied upside | +32.3% |
Recent ratings from named analysts:
| Firm | Analyst | Rating | Target | Date |
|---|---|---|---|---|
| Exane BNP Paribas | Moses Sutton | Buy | $374 | Aug 19, 2026 |
| Morgan Stanley | David Arcaro | Buy | $364 | Aug 21, 2026 |
| DBS | Pei Hwa Ho | Buy | $350 | Aug 18, 2026 |
| Bernstein | Sunaina Ocalan | Buy | $296 | Aug 17, 2026 |
The catalysts:
- 20-year PPA with Meta for the full 1,121 MW output of Clinton Clean Energy Center, starting June 2027 — a market-based deal that replaces Illinois' expiring ratepayer-funded ZEC subsidy and funds a 30 MW uprate.
- 20-year agreement with Microsoft tied to the restart of Three Mile Island as the Crane Clean Energy Center.
- Every additional data center PPA re-rates the whole fleet, because the marginal megawatt is being repriced from wholesale power rates to what a hyperscaler will pay for firm clean power.
The bear case, stated honestly: not everyone is on board. Mizuho's Anthony Crowdell carries a Hold with a $290 target (Aug 12, 2026), and Bernstein's $296 Buy target is only modestly above the current price. The stock already embeds a lot of AI optimism.
Stock #2 — Cameco Corp. (NYSE: CCJ)
You cannot run a reactor without fuel. Cameco sells the fuel.
Every restarted plant, every license extension, every SMR that eventually gets built needs enriched uranium. Cameco is the largest publicly traded uranium producer in the Western world, and it also owns 49% of Westinghouse, giving it a second claim on reactor construction and servicing economics. This is the “sell the picks and shovels” position in the nuclear trade.
Analyst consensus: BUY — and notably, not a single sell rating. As of August 2026, the ratings split was 36% Strong Buy, 48% Buy, 16% Hold, 0% Sell.
| Metric | Value |
|---|---|
| Price (Sep 17, 2026) | $92.73 |
| Consensus rating | Buy (22 analysts) |
| Average 12-mo. price target | $129.23 |
| Implied upside | +39.4% |
Recent ratings from named analysts:
| Firm | Analyst | Rating | Target | Date |
|---|---|---|---|---|
| Jefferies | Laurence Alexander | Buy | $138 | Sep 3, 2026 |
| Bank of America Securities | Lawson Winder | Buy | $133 | Aug 6, 2026 |
| Truist Financial | Christopher Souther | Buy | $130 | Aug 12, 2026 |
| Stifel Nicolaus | Ralph M. Profiti | Buy | $130 | Aug 4, 2026 |
| Scotiabank | Orest Wowkodaw | Buy | $126 | Aug 5, 2026 |
That is five consecutive Buy ratings from five different institutions, with targets clustered in a tight $126–$138 band. Tight clustering across independent shops is a stronger signal than one outlier scream.
The bear case, stated honestly: uranium is a commodity, and Cameco's earnings ride the spot and term price of that commodity. The low end of the analyst target range is $84.96 — below today's price. Mining is also capital-intensive and subject to operational and geopolitical disruption.
Stock #3 — BWX Technologies, Inc. (NYSE: BWXT)
The company that actually manufactures the reactors.
BWXT is the sole-source supplier of nuclear reactors for the U.S. Navy's submarines and aircraft carriers — a franchise essentially impossible to replicate, because the qualification barriers and security clearances took decades to build. That government business funds the option on the commercial side: SMR components, fuel, and medical isotopes. If small modular reactors become the answer for data-center power, somebody has to forge the pressure vessels. There are very few somebodies.
Analyst consensus: BUY — with the largest implied upside of the three.
| Metric | Value |
|---|---|
| Price (Sep 17, 2026) | $146.64 |
| Consensus rating | Buy (18 analysts) |
| Average 12-mo. price target | $223.63 |
| Implied upside | +52.5% |
Recent ratings from named analysts:
| Firm | Analyst | Rating | Target | Date |
|---|---|---|---|---|
| Northland Securities | Jeff Grampp | Buy (maintains) | $250 | Aug 27, 2026 |
| BTIG | Andre Madrid | Buy (maintains) | $235 | Aug 27, 2026 |
| Guggenheim | — | Buy (initiates) | $192 | Sep 14, 2026 |
| Jefferies | Laurence Alexander | Buy (initiates) | $181 | Sep 3, 2026 |
Two major firms initiated new coverage with Buy ratings in the last two weeks — Guggenheim on September 14 and Jefferies on September 3. New initiations matter more than maintained ratings: they represent fresh institutional capital building a position and a thesis from scratch.
The fundamentals are backing it up. Q2 2026 delivered 18% revenue growth, strength in both the commercial and government segments, and backlog up 40% year over year. Management raised full-year 2026 guidance on revenue, EBITDA and EPS.
The bear case, stated honestly: this is the most contested name of the three. Wells Fargo's David Strauss carries only a Hold (upgraded from Underweight) at $170, and Deutsche Bank trimmed its target earlier in the year. The analyst target range is enormous — $170 to $290 — which tells you the Street genuinely disagrees about what the SMR opportunity is worth. Wide dispersion means real uncertainty, not just upside.
Summary table
| CEG | CCJ | BWXT | |
|---|---|---|---|
| Role in the trade | Owns the fleet | Supplies the fuel | Builds the reactors |
| Price (Sep 17, 2026) | $263.18 | $92.73 | $146.64 |
| Consensus | Buy | Buy | Buy |
| Analysts covering | 22 | 22 | 18 |
| Avg. price target | $348.30 | $129.23 | $223.63 |
| Implied upside | +32.3% | +39.4% | +52.5% |
| Dissenting voice | Mizuho (Hold, $290) | Low target $84.96 | Wells Fargo (Hold, $170) |
