$TDS Kerrisdale Capital’s January 2026 report frames Telecom and Data Systems (TDS) as a post-divestiture infrastructure holding company whose equity valuation continues to reflect the pre-sale complexity and underperformance of the former US Cellular wireless business, despite a fundamental shift in asset mix, capital allocation, and near-term cash realizations. The central claim is a 2-layer valuation disconnect: 1) Array Digital Infrastructure (Array Digital, AD) is priced as though tower EBITDA is permanently depressed and announced spectrum monetizations will not be realized, and 2) the TDS parent is consequently priced as though TDS Telecom is a distressed legacy copper ILEC rather than a scaling fiber platform. The report’s sum-of-parts framework arrives at a $68 TDS fair value, implying 70% upside from a $40 reference price and supported by (a) imminent special dividends sourced from monetizing spectrum at AD (net $1.796b to AD, equating to $21 per AD share and $12 per TDS share given 82% ownership), and (b) a re-rating of TDS Telecom’s fiber platform toward $2,300 per 2028E fiber passing, consistent with cited fiber transaction precedents and a simplified DCF.
At the prevailing market prices at the time of this review (TDS $40.33; AD $54.37), the report’s framing remains current and directly testable against observable catalysts. The investment debate therefore reduces to a small set of underwriting questions: the probability-weighted timing and net proceeds of spectrum dispositions; the durability and realizable optimization of AD’s tower cash flows given a historically single-tenant tower base and forthcoming T-Mobile site selection; the achievable penetration, unit economics, and competitive resilience of TDS Telecom’s fiber builds (both expansion markets and E-ACAM-supported ILEC upgrades); and the persistence of a holding-company discount given family control, preferred equity, and ongoing elevated capex. Kerrisdale’s analysis is directionally coherent in highlighting that the current equity prices embed a very low implied value for TDS Telecom once AD is “fairly marked,” but the magnitude of mispricing is highly sensitive to assumptions around tower normalization (EBITDA run-rate and multiple), fiber monetization paths (public-market re-rating vs strategic sale), and the ultimate pace of free cash flow conversion given that fiber build programs can remain cash consumptive for extended periods even when NPV-positive.
SITUATION UPDATE AND CORPORATE RESET
The post-sale structural reset is real and externally corroborated. On 08/01/2025, the company formerly known as UScellular completed the divestiture of its wireless operations and select spectrum assets to T-Mobile for approximately $4.3b of total consideration after adjustments (approximately $2.6b cash proceeds and approximately $1.7b of assumed debt), and the retained entity was repositioned as Array Digital Infrastructure, retaining approximately 4,400 owned towers, noncontrolling investment interests, and spectrum holdings. ([https://t.co/RHzxWud1Rd][1]) The same announcement confirms a 15-year Master License Agreement (MLA) under which T-Mobile becomes a long-term tenant on a minimum of 2,015 incremental towers and extends the lease term on approximately 600 towers where it was already a tenant, creating a contracted anchor-tenant revenue stream. ([https://t.co/RHzxWud1Rd][1]) These facts anchor Kerrisdale’s argument that AD has transitioned from an integrated wireless operator with an internal tower function to a standalone tower and infrastructure vehicle whose primary operational mandate is now tower leasing optimization and cost normalization rather than subscriber retention.
Despite this simplification, 3 structural features continue to justify some degree of valuation friction versus pure-play infrastructure peers. 1) TDS remains a holdco with a large, publicly traded operating affiliate (AD) plus a wholly owned wireline operating company (TDS Telecom), introducing look-through complexity, double leverage optics, and potential for persistent conglomerate discount. 2) Governance remains controlled through a super-voting Series A structure; Kerrisdale notes the TDS Voting Trust controlled by the founding family holds approximately 95.6% of Series A shares (10 votes per share) and approximately 54% of total voting power, increasing the probability that capital allocation and strategic alternatives are optimized for long-duration control preferences rather than near-term price maximization. 3) Capital intensity remains elevated at TDS Telecom as fiber build continues, which can keep consolidated free cash flow depressed even if the underlying investment IRRs are attractive, and can delay a public-market re-rating that tends to reward visible free cash flow rather than “build metrics” alone.
CAPITAL RETURNS FROM SPECTRUM MONETIZATION
Kerrisdale’s near-term catalyst is the conversion of AD’s spectrum portfolio into cash special dividends. The report cites announced transactions totaling $1.796b net cash to AD after taxes and fees (AT&T $883m, Verizon $765m, T-Mobile $148m), equating to $21 per AD share and approximately $1.473b to TDS given its 82% economic interest, or $12 per TDS share (31% of the referenced TDS share price). The AT&T component has received FCC approval in early December 2025 (as characterized in the Kerrisdale report), which materially reduces binary risk on that tranche relative to the Verizon and remaining items. The FCC order adopted and released 12/03/2025 grants assignment applications for Lower 700 MHz and 3.45 GHz licenses from Array entities to AT&T and grants a waiver related to the 3.45 GHz aggregation limit, supporting the premise that this leg is past the primary regulatory gate.
Several technical points matter for investment committee evaluation of this “special dividend” catalyst.
First, special dividends are value-realizing, not value-creating, absent reinvestment or discount closure. AD’s equity price should mechanically adjust downward by the amount of the cash dividend on the ex-date, all else equal, and TDS’s look-through value similarly shifts from “embedded” assets to cash at parent. The economic benefit comes from reducing uncertainty, increasing balance sheet flexibility, and enabling repurchases or other accretive actions if the equity is trading below intrinsic value. Kerrisdale’s “getting $21 for free” framing depends on an additional assertion: that AD’s ex-dividend intrinsic equity value is approximately equal to (or above) today’s pre-dividend market price, implying the market is not capitalizing the dividend as part of the price. This is an empirical question and can be monitored via options-implied dividend pricing and spot behavior approaching record dates.
Second, the Verizon spectrum transaction is explicitly subject to regulatory approval and closing conditions. The FCC order for the AT&T assignment references that other spectrum transactions exist and that applications were filed; timing and any incremental conditions remain open-ended and can shift based on FCC composition and policy posture. The risk is less about ultimate non-consummation (given credible counterparties and spectrum utility) and more about timing slippage, incremental divestiture conditions in specific CMAs, or tax/fee leakage exceeding assumptions. Timing matters because Kerrisdale’s thesis uses “within 9 months” framing for aggregate dividends, which is a key component of near-term IRR.
Third, the remaining spectrum valuation is a large residual component of AD’s sum-of-parts and is marked conservatively in the Kerrisdale framework at a 10% discount to book for the primarily C-band portfolio (report cites $1.445b remaining spectrum value, with $1.314b attributed to C-band marked at $0.80 per MHz-pop and $131m for other bands). The critical underwriting variable is the forward clearing price for this spectrum given carrier balance sheet constraints, network needs, and the growing role of fiber backhaul plus densification. Kerrisdale points to management commentary that bids did not exceed book during the strategic review and uses that as an “observable pricing signal.” This approach avoids speculative upside but also implies that any monetization above the assumed mark could be incremental value. The counterpoint is that book is not an economic floor; if carrier demand weakens or if spectrum is more geographically fragmented than buyers want, clearing can fall below book, especially when sellers are not compelled but the opportunity cost of holding increases as buildout obligations approach.
Fourth, buildout obligations for 3.7 GHz (C-band) are a medium-term constraint rather than a near-term one, supporting the report’s claim of no forced timetable before 2029. The FCC’s 47 CFR 27.14 construction requirements for 3700-3980 MHz reference interim and final benchmarks (45% of population by 8 years and 80% by 12 years, with a possible 4-year extension for the final benchmark), consistent with a 2029 interim deadline for licenses granted around 2021 and a 2033 final benchmark absent extension. This matters because it lowers the probability of distress selling but introduces an eventual capex or leasing/partnering requirement if the spectrum is retained longer.
TDS TELECOM FIBER PLATFORM: ECONOMIC ENGINE OR CAPITAL SINK
Kerrisdale’s valuation gap argument at the parent level relies on treating TDS Telecom as a fiber-forward platform and asserting that the market-implied valuation is inconsistent with the economics of building and operating fiber at scale. The report’s mechanistic claim is that if AD is marked at “fair value” and forthcoming special dividends are included, the implied value assigned to TDS Telecom is approximately $1.228b, equating to 3.5x 2027E EBITDA and $738 per 2028E fiber passing, which is framed as “distressed copper ILEC” valuation rather than fiber valuation. The implied $738 per passing is positioned as illogical given fiber build costs cited at $1,100+ per passing plus $500 cost-to-connect, implying negative implied value creation if the public market’s valuation were “true.”
This argument is intuitively compelling but incomplete without explicitly addressing 4 offsetting realities: (1) fiber build economics are heavily back-end loaded, so passings and even penetration can rise while consolidated free cash flow remains deeply negative; (2) “per passing” precedent valuations often reflect assets at a different maturity stage, with different competitive contexts and different embedded capex remaining; (3) TDS Telecom is not a pure-play fiberco; it includes legacy copper and cable revenues that can continue to decline and consume operating attention, and those declines can partially offset fiber growth in near and medium horizons; (4) valuation in public markets is frequently anchored to near-term cash generation and perceived capital discipline, not only to long-term IRR math.
EXPANSION MARKETS STRATEGY AND COMPETITIVE POSITIONING
The report describes a “first-to-fiber” strategy targeting approximately 100 Tier-2 and Tier-3 markets where TDS can be the first scaled gig-capable provider and face muted incumbent response. The logic is that smaller markets are less likely to attract simultaneous overbuild from large national fiber builders, and incumbents (often cable operators) may be slower to invest in aggressive competitive responses, allowing penetration curves to follow a predictable trajectory (Kerrisdale references approximately 40% by year 5). The “muted response” assumption is plausible but not guaranteed. Cable’s response function is not purely based on market size; it is based on churn risk, return-on-upgrade calculations, and the availability of network upgrade paths (DOCSIS upgrades, targeted node splits, promotional pricing). In addition, fixed wireless access (FWA) has become a more salient competitive force in precisely the kinds of non-urban markets TDS serves, because FWA economics can be attractive where spectrum holdings (especially mid-band) allow capacity at acceptable speeds without the fixed cost of fiber. Public disclosures from national carriers show aggressive FWA customer acquisition over recent years, and industry reporting has linked cable broadband net adds pressure partly to FWA substitution. Even if fiber is superior on latency and reliability, competitive pricing and “good enough” speeds can cap penetration or force ARPU concessions.
Kerrisdale’s operational evidence for progress includes: residential fiber connections nearly doubling in 2 years; fiber service addresses surpassing 1.0m (up 93% from 2022); 42,000 new marketable fiber addresses in 3Q25 (up approximately 30% from 3Q24); and more than 80% of capex directed to fiber. These are meaningful output metrics, but the investment committee-level question is input-normalized unit economics: cost per passing, cost to connect, CAC, ARPU, churn, and incremental margin as penetration scales. The report’s DCF uses $1,300 cost per passing, $500 cost-to-connect, CAC $350 (net $300 after $50 recapture), 3% annual pricing increase, 1% churn, and long-run EBITDA margins increasing from 45% to 60% by maturity. These inputs are not obviously unreasonable for a fiber build, but each is a high leverage variable.
Penetration: A 40% penetration by approximately year 5 can be achieved in some markets, but it is not automatic. Markets with strong cable incumbents, aggressive promotional offers, and entrenched bundling can flatten penetration curves. Penetration under E-ACAM may be higher if competition is structurally limited, but in expansion markets the competitive set is broader and includes cable and FWA.
ARPU and price increases: The DCF uses $85 starting ARPU and assumes 3% annual lift. Sustained 3% annual pricing power is sensitive to competitive intensity, regulatory scrutiny, and the product mix (standalone broadband vs bundled). Fiber ARPU can be resilient, but broadband markets have been competitive and promotional pricing is common, particularly when FWA offers lower entry price points.
Margins: The DCF’s movement toward 55%-60% EBITDA margin presumes meaningful operating leverage, lower maintenance costs versus copper, and a mature footprint with lower selling expense as the base stabilizes. The key risk is that competitive churn, promotional activity, and ongoing edge-out builds keep sales and support costs elevated, delaying margin normalization.
Capex trajectory: Kerrisdale expects capital intensity to decline as expansion markets mature and E-ACAM build progresses, enabling free cash flow inflection. The company-level financial summary in the report still shows large capex through 2028E (TDS Telecom capex of $567m, $608m, $584m, $494m for 2025E-2028E), which implies that equity re-rating may require a demonstrated path from capex spending to sustained EBITDA and ultimately to free cash flow rather than incremental “build” milestones alone.
E-ACAM: THE REGULATED-RETURN SUBSIDY ENGINE
The report’s most differentiated bull point versus generic “fiber build” narratives is the emphasis on the FCC’s Enhanced A-CAM (E-ACAM) program as a second engine that converts rural ILEC copper footprint into a fiber franchise with limited competitive overlap. Kerrisdale characterizes E-ACAM as delivering 15 years of predictable, high-margin wholesale revenue of over $1.2b total to fund upgrades of approximately 300,000 rural copper locations, with penetration targets in the 65%-75% range because TDS is the only eligible participant in these markets.
Externally, E-ACAM is a 15-year high-cost support mechanism with defined performance obligations. USAC describes Enhanced A-CAM offers as providing support over 15 years and targeting deployment of 100/20 Mbps service to locations that would otherwise lack it, with total program-level support in the tens of billions over the term. The macro implication is that a meaningful portion of TDS Telecom’s rural fiber build can be funded with predictable support, reducing the risk of uneconomic overbuild and improving the stability of wholesale-like cash flows.
However, E-ACAM underwriting requires careful separation of (1) revenue certainty vs (2) cost certainty. Support is contractually visible, but actual construction costs, labor availability, permitting delays, and supply chain can introduce variance. Additionally, “limited competitive overlap” is directionally true in deeply rural footprint, but it is not absolute. FWA is a credible competitor in rural settings, and state/local municipal initiatives or other subsidized builds (depending on mapping and program rules) can create pockets of overlap. The key diligence items are the exact eligibility and overlap protections in the awarded E-ACAM geographies, and whether TDS’s planned build sequences prioritize areas with strongest subsidy-to-cost economics.
ARRAY DIGITAL TOWERS: UNDER-EARNING PLATFORM OR STRUCTURAL DISCOUNT WARRANTED
Kerrisdale treats AD’s tower business as a classic “under-earning infrastructure platform” with large embedded operating leverage. The reported tower base is 4,449 owned towers as of 3Q25, with a tenancy rate of 1.02 and 4,517 colocations (excluding T-Mobile interim sites), contrasted against 2.2x-2.6x tenancy at mid-scale peers. The report argues that the tenancy gap is primarily under-marketing rather than asset quality (“not inferior steel”), because for 2 decades the towers served primarily 1 customer (UScellular). Early evidence cited includes 3Q25 site leasing revenue of $46m (including $5m interim T-Mobile rent), approximately 10% organic site rental revenue growth, and a 125% acceleration in new colocation applications year-to-date relative to 2024 after insourcing sales/marketing.
If correct, the embedded value is significant because the incremental margin on additional tenants is typically very high once fixed costs and ground rent are covered, particularly in rural towers where zoning and scarcity can be meaningful constraints. The investment committee-level question is whether AD’s “low tenancy” is actually a reversible marketing problem or whether it reflects structural realities of its footprint (rural towers in areas where incremental carrier demand is lower) and a future where carrier network spending prioritizes densification in higher-traffic corridors.
T-MOBILE MLA SITE SELECTION: PRIMARY IDIOSYNCRATIC RISK
The most important non-consensus variable in AD tower economics is the T-Mobile MLA site selection process. The company has disclosed that T-Mobile has until January 2028 to finalize which 2,015 committed sites remain under the MLA, and AD expects 800 to 1,800 “naked” or tenantless towers based on that selection. This introduces 3 intertwined risks:
1. Revenue volatility during the transition from interim arrangements to final committed sites, with potential for churn in certain towers not selected or rationalized.
2. Cost leakage from maintaining uneconomic towers with high ground rent and low tenant prospects.
3. Potential for tower portfolio “shrink” via demolition of uneconomic towers, which can be value-neutral or value-destructive depending on salvage and contract exit costs.
Kerrisdale argues these are standard tower company decisions (market to third parties, renegotiate ground leases, evaluate demolition after colocation upside is exhausted) and forecasts EBITDA ramp from approximately $60m next year to approximately $100m by 2028E with long-term 45%-50% EBITDA margins. The critical scrutiny is whether AD can replicate the playbook of scaled towercos given subscale operations, rural weighting, and the absence of the “multi-tenant metro” demand engine that often drives rapid tenancy ramp. The cited cost normalization opportunity (approximately 40% of 3Q25 SG&A reflecting legacy wireless wind-down, legal contingencies, tax matters, spectrum administration, and strategic review costs) is credible in concept and provides a visible pathway to margin improvement independent of leasing success. Even so, cost normalization alone does not guarantee peer-like multiples; the market will apply lower multiples if growth is uncertain, churn is elevated, or if the tower footprint lacks long-duration contracted escalators typical in larger tower portfolios.
