Shentel: A Perfect Quarter, a Broken Multiple, and a Hyperscaler That Never Called Back

Glo Fiber just posted its best operating quarter while the stock fell 31% — and the data center story quietly vanished from the script

SHEN · Earnings Call · 2026-10-28

The quarter that should have worked

On the operating line, Shenandoah Telecommunications just posted the best quarter of its fiber life. Glo Fiber data customers reached roughly 109,000, up 32% year-over-year, on a record 8,600 net additions — 48% above the year-ago quarter. Glo Fiber crossed 501,000 passings with 25,000 added in the period, and the company is now 97% complete with the Glo Fiber expansion and 99% complete with the incumbent broadband build. Penetration reached 21.7%, up 108 basis points year-over-year, and the 2019–2020 cohorts are averaging 38.5%, ahead of the 37% mid-term target. Combined Glo Fiber and Commercial Fiber revenue grew 19.5% and now represents 52% of the consolidated business. Total revenue was $93.9 million (+4.6%) and adjusted EBITDA $33.2 million (+11.8%), with an incremental adjusted EBITDA margin of 85% and 200 basis points of margin expansion in the quarter, 300 year-to-date. EBITDA growth is finally outrunning revenue growth — the exact operating leverage the equity story has promised for years. The stock, meanwhile, is down 30.7% over the last 90 trading days and 37.3% from its June 29 high of $15.62, on a $543 million market cap. That gap is the story of this report.

we're disappointed where the stock is sitting today. We think there's some macro issues driving that more so than our fundamentals. But ironically, if you do the calculations, we're trading at roughly 10x today, whereas 5 months ago, we were trading roughly at 12x EBITDA. And we're significantly derisking our business plan by completing the construction phase and going free cash flow positive.
James Volk, Executive (likely CFO) · 2026-10-28

That 12x EBITDA reference is the CFO's own framing, and the multiple history backs him up. The market once paid 10.4x revenue for this build story at the 2021 peak; it is 2.3x today. The re-rating happened on the way in and did not reverse on the way out.

The capex paradox

Here is where the quarter gets genuinely awkward. In the same breath as "substantially complete," management raised 2026 capital expenditure guidance to $255–275 million from $220–250 million. The decomposition is mundane and worth spelling out — “about $10 million of the increase related to the 7,000 additional passings. About $15 million relates to the government grant subsidies, which is really just a timing issue.” — James Volk, Executive (likely CFO) · 2026-10-28 The remaining $5 million is a grab bag. So $15 million of the raise is grant cash sliding into the first quarter of 2027, not new spending. But the optic lands badly: the free-cash-flow-inflection narrative and a higher capex number, in the same print. Net capital expenditures were actually $193 million for the first nine months after $22 million of grants collected, down 9% year-over-year — the right direction, obscured by the headline. The scale of what is being unwound is the whole investment case. Capex was $70 million against $93 million of revenue in the last filed quarter. Guidance implies ~70% for 2026 and management targets mid-20% capital intensity thereafter. Free cash flow was negative $48 million last quarter, improving 28% year-over-year and up from the negative $80 million trough in 2025Q4. That is the shape of a build turning over, not a business deteriorating.

The keyword that vanished: the hyperscaler

In the second quarter of 2026, data center was the single highest-ranked keyword in this company's own transcript, on talk of roughly 20 data centers sitting near Shentel fiber across eight states and a master service agreement with a major hyperscaler. In the current quarter's keyword set, it has fallen out of the top 30 entirely. That is not an accident of NLP. It is three consecutive calls of the same answer: “it takes time to negotiate contracts with these major hyperscalers. Our discussions are ongoing regarding multiple opportunities... I just don't have anything additional to share right at this moment.” — Edward McKay, Executive (likely CEO or President) · 2026-10-28 In July it was “we don't have anything specific to report at this time” — Edward McKay, Chief Executive Officer (CEO) · 2026-07-29 — after an MSA that "will enable them to potentially buy services in the future." An MSA that has produced no disclosed revenue across three quarters is optionality, and the market has stopped paying for optionality here. The global tape agrees. AI data center exposure is now a 90-day decliner across the market — 6 advancers against 50 decliners in the recent window, after a year in which it was the single strongest theme. Shentel's own data center interconnects pitch was a beneficiary of that 360-day enthusiasm; it is now swimming against it. What replaced it in the narrative is telling. Asked about the internal rate of return on hyperscaler deals, the CFO answered instead with customer economics: “we tend to do more of a customer lifetime value calculation... for Glo Fiber, that gives us -- the value of those customers are worth a couple of thousand dollars.” — James Volk, Executive (likely CFO) · 2026-10-28 lifetime value is now the company's top-ranked keyword, against roughly $1,500 per passing of build cost. That is a much more defensible argument than a hyperscaler maybe — but it is also the argument of a company that has run out of new things to promise.

ARPU: the pivot that keeps getting deferred

The soft underbelly is price. broadband ARPU in Glo Fiber fell about 1% year-over-year to just under $76 (the five-year price guarantee working as designed), while incumbent ARPU fell 3% to roughly $80 and incumbent data churn rose to 1.76%. The offset is speed mix: 83% of new residential customers took 1 gig or better. The drag is rural: households where inflation is squeezing budgets, plus wired overbuilders in part of the footprint. This has been deferred repeatedly. In May, management said “in our Glo Fiber business, we're expecting fairly flat ARPU in the near term.” — Edward McKay, CEO · 2026-05-01 In October the language is that ARPU is "just about at the point that it starts to pivot back up," with a new incumbent rate card in early 2027 and sub-1% declines thereafter. Meanwhile the July promise of “300 to 400 basis points a year for the next couple of years” — James Volk, Chief Financial Officer (CFO) · 2026-07-29 of margin expansion and a 40% consolidated margin target is being delivered — the 2027 guide reiterates another 300 to 400 basis points — but it rests on a full year of savings from the announced reduction in force, not on pricing power. The satellite scare, at least, is dead. It peaked in Q1, when “they started offering in the first quarter was $15 off for 4 months as a promotion” — Edward McKay, CEO · 2026-05-01, and has decayed every quarter since — churn impact from satellite is now under 10 basis points in incumbent markets. Yet an analyst noted the stock still trades down on SpaceX headlines, which is as good a diagnosis of the 90-day drawdown as any model. The balance sheet is the reason the market is unforgiving. Effective net cash is negative $674 million against $760 million of debt, with leverage guided to 5.7–5.8x at year-end, below 5x by end-2027 and below 4x by 2029. No maturities until 2029 and 75% fixed-rate is genuine protection, but a levered small-cap with negative free cash flow in a market that has stopped paying for long-duration promises is exactly where the multiple goes to 10x. There are still 600 DSL customers left to migrate — a rounding error that is nonetheless the last artifact of the legacy business. The question for 2027 is whether a company that finally owns its network, at 21.7% penetration against a 37% target, gets paid for the harvest it has spent six years planting. Management's own answer, on the record, is that it should be worth more than this.