TransAlta (TAC) Q3 2026 Earnings Call Transcript

TransAlta reported Q3 2026 adjusted EBITDA of C$264 million and free cash flow of C$93 million, and reaffirmed full-year guidance of C$950 million–C$1.05 billion in adjusted EBITDA and C$350 million–C$450 million in free cash flow. Performance held up despite Alberta spot prices averaging C$36/MWh, down from C$51/MWh a year earlier, with hedges and fleet optimization generating realized prices above spot. The company closed a US$1 billion acquisition of 318 MW of Colorado gas peaking facilities expected to contribute about US$110 million of annual adjusted EBITDA, and sold a 49% interest in its Oklahoma wind portfolio for US$305 million. TransAlta is advancing the Boreal data-center project, which could support up to 1.6 GW of load, while mothballing Battle River units for economic reasons and preserving the option to return them to service.

Analysis

TA’s near-term earnings resilience is more a hedge-and-dispatch story than proof that Alberta merchant economics have turned. That distinction matters: if weak spot prices persist, hedges support cash flow but do not eliminate the value risk in uncontracted generation or justify capitalizing the Boreal opportunity as contracted growth. Mothballing marginal units may reduce carrying costs now, while also tightening supply during peak periods if load arrives; that can raise the value of retained flexible gas and ancillary services. The payoff is asymmetric, however: data-center load is not yet a bankable substitute for today’s low utilization.

The key valuation catalyst is conversion of Boreal from site potential into enforceable customer commitments and an AESO-approved, reliable allocation of power. The gap between the public build-out ambition and the still-unresolved allocation/curtailment terms creates meaningful execution risk; the project should not be underwritten on headline capacity. Separately, Colorado adds contracted cash-flow duration but concentrates capital and raises the importance of financing discipline after the equity issuance and asset sale. Verify counterparty terms, acquisition funding, and whether the stated EBITDA contribution is incremental after financing and operating costs.

Over the next 1–3 months, AESO clarity, definitive customer agreements, and year-end permitting are the actionable catalysts. Over 6–18 months, the thesis depends on data-center construction, Alberta demand growth, and Centralia conversion permitting/FID. Contrarian point: strong quarter-level hedging may obscure continued softness in underlying spot economics; conversely, the market may discount the option value of existing grid-connected sites. Falsifiers include delayed or reduced Boreal allocations, adverse curtailment terms, sustained weak Alberta forwards, or guidance pressure from costs/capital requirements.

Key Decisions for Investors

• Treat TA as a catalyst-driven, moderate-conviction long rather than a clean power-price recovery trade. Prefer entry after volatility or once AESO allocation and definitive Boreal customer terms are disclosed; do not pay for the full announced site capacity before those milestones.
• Keep Alberta spot/forward spreads and TA’s realized merchant premiums on a watchlist. A sustained improvement in forward pricing would strengthen the case for retained flexible generation; persistently weak prices or falling realized premiums would weaken the thesis and argue against adding.
• Monitor the Colorado acquisition’s financing, counterparty contract protections, and contribution to free cash flow after interest and maintenance needs. Reassess if leverage rises without asset-sale proceeds or if reported cash generation fails to support guidance.
• Use Boreal milestones as explicit thesis tests: AESO capacity and curtailment terms, definitive agreements, master-site permits, and evidence of customer-funded construction. A material delay or smaller contracted load is a reason to reduce the data-center option value assigned to TA.