Telecom

An analysis on the current and predicted future state of top public companies in the telecom sector.

5.12.26 - TY

Genesis: Telecom companies, such as cell service and internet service providers, source monthly recurring revenue directly from most Americans and many businesses. In this sector, are also infrastructure companies that build and maintain cell towers that providers use to distribute internet to consumers. From a high-level, these sound like fantastic businesses to be in: predictable revenue from millions of citizens who cannot live without your services. Yet, there’s mixed performance across competing companies, which warrants a deeper look. The goal of this write-up is to clearly understand what’s working in telecom, how that’s been reflected in share prices, and what companies we expect to be watching going forward.

Background: In the US, 92% of Americans purchase home internet, and 98% have a phone that requires cellular service.^[1] After decades of consolidation and regulatory capture, telecom providers mainly belongs to a standard list of players: T-Mobile, Verizon, AT&T, Comcast, Charter, and Echostar. These businesses cover a mix of wired internet (fiber/cable), wireless internet (called fixed wireless access) and cellular service. All of the aforementioned companies are converging to offer unified connectivity to the most customers. The difference comes down to distribution, debt, and the historical story of each of these companies. On the infrastructure side, spending on new cell towers is down in the US after the 5G build out and stock prices for the only established players, American Tower Corp and Crown Castle, are depressed. The final set of companies to introduce here are the pivot kings, Lumen and Cogent, who are internet service providers that are trying to make a play for AI using their fiber optic assets.

Here are some numbers to start with - last year share price move, grouped by the type of company:

Multi-connectivity (wireless, fiber, owned cell network): T-Mobile (TMUS): -20% / AT&T (T): -7% / Verizon (VZ): +11%

Cable internet: Comcast (CMCSA): -24% / Charter (CHTR): -64%

Infra: AMT: -14% / CCI: -9%

Pivots: LUMN: +97% / CCOI: -68% / Echostar: +541% (Boost mobile, sling TV, dish TV)

From this, we can first see evidence of a clear divergence in customer appetite for legacy vs. modern internet technology. Whereas Comcast and Charter offer primarily cable internet, modern providers like T-Mobile and Verizon offer fully wirless internet, using their 5G networks. We should explore how durable this shift is and how we expect CMCSA/CHTR to respond.

Second, while in the same group, TMUS has the story of a growth stock compared to VZ and T. It’s “uncarrier” brand led it to skyrocket above their peers which appears to have since compressed. The internet has suggested for companies laden with debt (like asset-heavy telecom companies are), we use the EV/EBITDA multiple to compare them. TMUS has come down from 12x last year to ~8.7x where T/VZ sit at ~6.5-7x. We should consider whether TMUS gets readjusted by investors for its story or if we expect VZ/T to further catch-up and surpass TMUS.

This leads to the further question of which company is best positioned for the convergence to unified connectivity. After it’s acquisition of Lumen’s fiber business, AT&T sits at ~12.5M fiber subscribers. Verizon gets to ~10M after its acquisition of Frontier and Fios, while T-Mobile sits last with low single-millions. An additional consideration to the price adjustment question for TMUS is whether we expect that to be further dramatically readjusted in favor of AT&T and VZ who appear better positioned in the early days of the fiber race.

Third, is a deeper inquiry into Echostar. Several gap-ups get them to the gain they’ve posted over the last year which includes selling previously discounted spectrum assets (essentially internet capacity) for billions. A key story is how SpaceX has paid for over half of their $18B spectrum deal in stock, making them a proxy for the expected upcoming SpaceX IPO.

Finally, Lumen (LUMN) and Cogent (CCOI) are both undertaking business model pivots towards AI with differing results so far. LUMN has led the way, selling its household fiber routes, paying down debt, and focusing on leasing “dark-fiber” or dedicated, dense routes of fiber optic networks between data centers for AI training, for which they have ~$13B in backlog. CCOI has had a harder time shedding its losing fiber business and also has differnet assets to LUMN for the application. A recent purchase of T-Mobile’s fiber business gave Cogent thousands of miles of long-haul fiber optic networks, but also the liability of the rest of a dying business. T-Mobile in fact paid Cogent $700M for them to take on the business and its assets, showing there was known and strategic reasoning for Cogent to take on that debt. What’s more, Lumen has the density to offer a more custom and private service than Cogent can, which is offering shared capacity on its routes. Still, that business is growing 91% YOY and is a clear need in an AI market with one player. A hidden line item also shows they own ~40M IPv4 addresses which they make ~$65M/year on by leasing out for a business that’s growing 25% YOY (AWS recently started charging for these IPs too because of how much they’re used in AI workflows). If CCOI can pay down its debt and transition out of the legacy Sprint fiber business, we can reasonably predict they have the seeds for a growing AI-infra business like LUMN has demonstrated.

This brings us to the following actionable monitors for the next 12 months:

• Short-term: AT&T vs. TMUS » monitor AT&T vs. TMUS earnings for signs of acceleration on either side. TMUS re-accelerating will re-rate it to its premium, AT&T accelerating gives it a chance to breakthrough.

• Short-term: LMN » while a lot of the current upside is priced in, further expansion of TAM could push the price further out of its current resistance. If we see breakthrough after additional backlog announcements, there’s further upside possible.

• Long-term: CCOI » monitor earnings for signs of restructuring legacy business and debt. If data center fiber continues to grow and cost structure improves, there’s a significant opportunity for a true runner.

More to come.

Sources

1. https://www.pewresearch.org/internet/fact-sheet/internet-broadband/

2. https://about.att.com/story/2025/echostar.html

3. https://www.cnbc.com/2025/08/26/echostar-stock-att-wireless-musk.html

4. https://ir.echostar.com/news-releases/news-release-details/echostar-announces-spectrum-sale-and-commercial-agreement-spacex

5. https://ir.echostar.com/news-releases/news-release-details/echostar-agrees-sell-full-unpaired-aws-3-spectrum-license

6. https://about.att.com/story/2026/att-lumen-deal-close.html

7. https://www.fierce-network.com/broadband/fiber-gets-spotlight-earnings-verizon-t-mobile-att

8. https://finance.yahoo.com/news/comparative-telecom-analysis-verizon-better-140003828.html

9. https://www.datacenterdynamics.com/en/news/lumen-outlines-multi-billion-dollar-ai-network-expansion/

10. https://thepopularinvestor.substack.com/p/cogent-owns-the-ai-spine

11. https://ipbnb.com/blog/ipv4-address-price-2026