Why Duke Won’t Raise Guidance Yet

Duke Energy posted a strong earnings beat, but management is not raising guidance yet, signaling caution around the near-term outlook. CFO Brian Savoy said AI data center demand is reshaping the power grid and highlighted Duke’s “speed to power” strategy, which can get customers online years faster while managing grid strain. The update is constructive for Duke’s growth narrative and underscores rising electricity demand from AI infrastructure.

Analysis

The market is still underappreciating how AI load growth changes utility economics from a slow-growth regulated asset to a capacity-constrained scarcity play. Duke’s advantage is not just incremental demand; it is the ability to monetize interconnection bottlenecks by compressing time-to-revenue, which should support a higher durability multiple than peers with longer queue backlogs or more exposed transmission dependencies. The second-order winner is the industrial ecosystem around grid buildout: transformers, switchgear, gas turbines, and EPC capacity should see a multi-year demand tailwind even if utility EPS growth remains mechanically capped.

The key investor mistake is to focus on guidance conservatism as bearish. In a supply-constrained utility, reluctance to raise outlook often signals management is preserving optionality until it can quantify customer additions and capital intensity, not that the demand is fragile. If AI-related load becomes visible in rate cases, the earnings power can re-rate over months rather than quarters because regulated returns compound off a larger asset base; the real upside is allowed base expansion, not near-term beat-and-raise optics.

Main risks are regulatory lag and execution drag. If data-center customers push too much spend onto the balance sheet before recovery is approved, equity can face temporary multiple compression even as fundamentals improve; the market will punish any sign of stranded capex or load requests that fail to convert into signed, bankable projects within 6-12 months. A broader risk-off move in rates could also mute the valuation benefit, since the bull case relies on both growth and bond-proxy stability.

Consensus is likely too focused on Duke as a steady defensive utility and not enough on it as an infrastructure bottleneck owner. The underpriced option is that AI demand forces a step-up in asset turns and service territory economics sooner than expected, creating a multi-year earnings runway without needing a dramatic macro inflection. That makes the setup better viewed as a slow-burn secular compounding story than a single-quarter earnings event.

Key Decisions for Investors

• Long DUK on pullbacks over the next 1-3 weeks; target a 6-12 month hold into visible data-center interconnection wins. Risk/reward favors a 10-15% upside rerate if the market starts pricing a higher terminal growth rate for regulated asset base expansion.
• Pair trade: long DUK / short a slower-growth regulated utility with less AI-load exposure over 3-6 months. The trade captures relative multiple expansion tied to faster queue monetization while reducing rate-beta.
• Buy calls on industrial grid beneficiaries (e.g., ETN or HUBB) for a 6-9 month horizon. If AI load growth accelerates, the supply chain spend is more immediate than the utility EPS translation, giving better convexity.
• Avoid chasing DUK after any short-term guidance-related pop; use earnings volatility to enter. The best entry is when investors misread prudent guidance as lack of demand, which should create a 2-4% dip that can be accumulated.
• For higher-conviction expression, consider DUK Jan-2026 calls financed partially by selling near-dated upside. This captures the multi-quarter rerating while limiting theta decay if the market takes time to price the AI backlog.