Talen Energy (TLN)

One-line summary: Independent power producer with 2 GW nuclear + 6 GW natural-gas baseload, pitched as a below-replacement-cost play on PJM power scarcity — trades ~$25B EV vs ~$45B replacement cost, with an Amazon data-center contract rolling up and optionality to build new capacity.

What it is

Talen Energy (Nasdaq: TLN) is a US independent power producer (IPP) operating ~2 GW of nuclear and ~6 GW of natural-gas baseload generation, concentrated in PJM (Pennsylvania-Jersey-Maryland). Its flagship asset is the Susquehanna nuclear plant, which anchors a behind-the-meter data-center co-location arrangement with Amazon (AWS). Pitched at the All-In Best Ideas competition (2026-06-12) by Dan Dreyfus of Bornite Capital as the cleanest below-replacement-cost way to own scarce PJM baseload.

Why it matters to stock-market

TLN is the merchant-IPP peer to constellation-energy (CEG) in the nuclear-baseload-for-ai-data-centers thesis — but with a sharper replacement-cost angle and a different risk profile (more interest-rate sensitive via long-dated PPAs). Where CEG is the regulated-scale nuclear operator, TLN is the levered, asset-value play: buy hard infrastructure below replacement cost when new capacity is needed, sell at a premium when the market wakes up (the Sam Zell playbook). It is a direct beneficiary of the same PJM-power-scarcity forcing function — 106 GW of new PJM demand forecast over 10 years — that drives the CEG/ferc-large-load-to-dc-gridscale-construction chains.

Key facts

From 2026-06-12-podcast-all-in-podcast-all-in-s-best-ideas-pitch-competition-4-investors (Dan Dreyfus / Bornite Capital pitch):

• Assets: 2 GW nuclear + 6 GW natural-gas baseload.
• Valuation: ~$25B enterprise value vs ~$45B replacement cost; "the equity value just to get to replacement cost is more than a double from where it's trading today." Stock in the "high $300s," ~7x FCF vs ~15x for good US infrastructure assets.
• The three-scenario FCF ladder (Dreyfus): (1) "do absolutely nothing… let their Amazon data center contract roll up → $50/share FCF" (~2x at a 15x multiple); (2) more co-location / front-of-meter deals or higher power prices → $70/share → ~$1,050 at 15x; (3) build ~4 GW of the needed 100 GW → $100+/share FCF.
• Thesis does not require behind-the-meter to clear: dan-dreyfus: "It's the latter [clean firm baseload is scarce enough] … The $50 a share of earnings, nothing has to happen. You just sit and double your money."
• PJM demand: grid-operator forecast of "106 gigawatts of new power in the PJM … over the next 10 years" — "the size of what Japan consumes today for one little part of the US"; coal retirements won't happen because the replacement can't be built in time.
• The Microsoft/Constellation TMI template (cited as proof of the model): Microsoft paid Constellation "$100 [/MWh] a year for 20 years minimum price" to restart Three Mile Island vs a ~$50/MWh spot — the behind-the-meter PPA premium TLN can replicate.
• Regulatory friction: regulators pushing back on taking power off-grid for data centers ("how are we going to heat the homes of our customers"); "crunch time."

Strengths (thesis-input)

• Hard assets below replacement cost in a structurally short market; "data center = refinery" ($50B/GW) with power as the scarce input.
• Amazon contract roll-up gives a floor-ish FCF ramp without needing new deals.
• Pole position to do utility+DC "go build power in a room" deals the regulator is now encouraging.

Weaknesses / risks (thesis-input)

• Interest-rate sensitivity — david-sacks: "more interest rate sensitive than MGM… the power purchase agreements really are where a lot of the revenue comes from… if interest rates shoot up you actually get margin compression from that 15x outlook." The build-out scenario is "eight years out."
• AI/regulatory risk — gavin-baker: "everything in AI is going to need to grapple with increasing… regulatory risk… the big negative externality for talent is nothing to do with talent" (i.e. exogenous AI-power-policy risk).
• Merchant-IPP-to-contracted-infrastructure re-rate is not guaranteed; terminal multiple debated.

Open questions

• talen-energy-replacement-cost-pjm-scarcity

Related

• nuclear-baseload-for-ai-data-centers
• constellation-energy — regulated-scale nuclear peer (CEG)
• pjm-capacity-prices-to-nuclear-premium
• ferc-large-load-to-dc-gridscale-construction
• ai-capex-to-power-and-materials-cascade
• dan-dreyfus

Sources

• 2026-06-12-podcast-all-in-podcast-all-in-s-best-ideas-pitch-competition-4-investors — Dan Dreyfus (Bornite) pitch: 2GW nuclear + 6GW gas, ~$25B EV vs $45B replacement, $50/$70/$100 FCF-per-share ladder, PJM 106 GW/10yr; Sacks + Baker risk commentary.