Earnings call transcript: SK Telecom Q1 2026 beats expectations, boosts stock

SK Telecom reported Q1 2026 EPS of $0.5206 versus $0.5166 expected and revenue of $3.02B versus $2.98B expected, while operating income reached KRW 537.6B, topping KRW 500B for the first time since the cybersecurity incident. The company resumed dividends at KRW 831 per share, and AI data center revenue rose 89% year over year, supporting a recovery narrative. Shares rose 2.68% after earnings and added another 1.16% in premarket trading.

Analysis

The cleaner read-through is not “telecom beat,” but “post-shock monetization reset.” Management is explicitly shifting from growth-at-all-costs to high-LTV subscriber harvesting and AI infrastructure with pricing power, which should support margins even if gross adds normalize. That matters because the stock can rerate on evidence of sustained operating discipline faster than on absolute revenue growth; the key variable is whether the recovery in customer trust proves sticky over the next 2-3 quarters.

The second-order winner is the domestic AI infrastructure ecosystem, but not all beneficiaries are equal. SKM’s data-center expansion and AI-RAN work increase demand for power, cooling, optical, and network equipment, yet the company’s comments imply a staged adoption curve, so vendors tied to immediate buildout may see more benefit than those exposed to speculative standards work. For NVDA, the signal is more nuanced: South Korea’s telco stack is becoming a reference customer for AI networking, but the stock remains vulnerable to any broad de-risking of AI capex if investors extrapolate too much from a single carrier’s recovery story.

The contrarian angle is that the market may be underpricing the durability of cash returns and overpricing the immediacy of AI contribution. If dividend normalization is sustained, SKM can attract income-driven capital and compress the equity risk premium even before AI meaningfully moves EPS. The bear case is a 1-2 quarter fade in subscriber momentum or a re-acceleration of security spending, which would hit the market’s confidence in both the telecom recovery and the dividend reset.

For NVDA-linked sentiment, this is mildly negative in the very short term because the market may read AI capex as more incremental than transformative here. But on a 12-24 month horizon, the more important effect is that carriers and data-center operators are validating AI traffic and edge inference as real workloads, which supports a broader demand base. The stock reaction can still be noisy around macro AI risk-off, but the structural takeaway remains constructive for networking-adjacent GPU demand.

Key Decisions for Investors

• Long SKM for 3-6 months on dividend reinstatement and operating recovery; target a re-rating toward the 52-week high with downside protected by the still-low expectations post-cyber incident.
• Buy SKM on any 2-3% pullback from current levels; add if management confirms stable subscriber gains in the next quarter, since the market is likely to pay up for visible cash-return normalization.
• Pair trade: long SKM / short a basket of higher-multiple AI capex proxies over 1-2 months; thesis is that this quarter proves cash returns are more immediate than AI monetization, while AI enthusiasm remains priced ahead of fundamentals.
• For NVDA, use the weakness as a hedge, not a structural short: hold a smaller tactical short only against a broad AI basket if the market starts extrapolating telco commentary into delayed networking spend.
• Consider a call spread on SKM for the next two quarters if implied volatility remains elevated; the convexity is favorable if dividend credibility and subscriber recovery continue without a security setback.