AT&T at Goldman Sachs conference: fiber push drives growth plan

AT&T reaffirmed a target for double-digit EPS CAGR through 2028 and mid- to high-single-digit advanced-connectivity EBITDA growth, supported by fiber-wireless convergence, improving service revenue and operational leverage. Convergence reached roughly 45%, up 5 percentage points in two years, while gross adds among converged customers in acquired Lumen markets are nearly 50% above the prior owner's level. Management also targets removal of $6 billion in legacy costs through copper retirement and central-office closures, while maintaining its dividend, pursuing buybacks when undervalued and reducing leverage to about 2.5x net debt/adjusted EBITDA.

Analysis

The investable issue is not the stated earnings algorithm but whether AT&T can convert network overlap into lower acquisition cost and lower churn without conceding too much broadband ARPU. A household-level optimization strategy can lift lifetime value while obscuring near-term product margin pressure; the next two quarterly reports need to show service-revenue growth accelerating faster than promotional expense and fiber gross-add costs. If that occurs, T deserves a rerating from a levered yield proxy toward a converged-connectivity compounder; if not, the long-duration EPS target is unlikely to receive valuation credit.

The most underappreciated earnings lever is the fixed-cost unwind, but it is also execution- and regulation-dependent. Copper retirement should generate a multi-year EBITDA/cash-flow tailwind through real-estate, power, maintenance and labor reductions, while scrap proceeds can partially fund removal costs; however, savings will arrive unevenly and likely require upfront restructuring cash. Watch disclosed legacy-cost run-rate reduction, cash restructuring spend, and net-debt/EBITDA rather than treating the full cost pool as near-term free cash flow.

Competitive risk has migrated to broadband, where T's willingness to price individual products below stand-alone economics makes cable operators the more exposed party only if T maintains household profitability. Charter (CHTR) and Comcast (CMCSA) face the clearest regional share and retention pressure in fiber-overbuild markets, while TMUS and VZ can counter with fixed wireless where fiber economics are weakest. The contrarian point: satellite is less relevant to the core wireline market than feared, but its value as a bundled coverage feature reduces differentiation among national wireless carriers rather than creating a new profit pool.

Near term, this conference commentary alone is unlikely to change estimates materially. The 1-3 month catalyst is evidence that acquired fiber markets sustain higher converged gross adds without a deterioration in postpaid or broadband ARPU; the 6-18 month catalyst is measurable legacy-cost removal alongside deleveraging. A reversal would be signaled by elevated churn, broadband revenue growth lagging fiber footprint expansion, or leverage failing to trend toward the stated target because capex, integration, or promotions absorb incremental cash flow.

Key Decisions for Investors

• Accumulate T on weakness over a 6-18 month horizon, sized as a value/carry position rather than a near-term catalyst trade. Add only if quarterly service-revenue growth and wireless/fiber net adds remain intact while net debt/adjusted EBITDA trends down; exit or reassess if leverage rises for two consecutive quarters or broadband ARPU declines materially without offsetting churn improvement.
• Consider a 3-6 month pair: long T / short VZ, with equal dollar beta-adjusted exposure. T has a more visible fiber-convergence and cost-removal rerating path, while VZ has less direct fiber ownership and comparable capital-intensity constraints; stop the spread if VZ demonstrates superior postpaid growth without materially higher promotional cost or T misses its next operating KPIs.
• Use CHTR or CMCSA as a watch-list hedge rather than an immediate short. Initiate only after market-level evidence shows fiber-driven broadband share loss translating into weaker cable broadband revenue or rising retention spending; absent that data, cable valuation and buyback support can overwhelm the structural thesis.
• Do not underwrite material upside from ECHO until transaction economics, closing timing, spectrum deployment obligations, and incremental capex are independently disclosed. Treat any delay, altered regulatory condition, or capex increase as a trigger to reduce T exposure because the spectrum thesis is partly embedded in the convergence-growth narrative.