Array Digital's 90% Revenue Growth Is a Liquidation in Disguise

The earnings recap just hit your feed. Array Digital Infrastructure Q2 2026 Earnings Call Highlights: Strong Revenue Growth. Revenue up 90% year-over-year. Adjusted EBITDA up 56%. EPS up from $0.17 to $3.86. If you read that headline and filed this under "tower infrastructure story playing out," you've been reading the press release instead of the income statement.

Array Digital is not growing. It is being liquidated and the parent company is trying to buy back the pieces it doesn't own anymore.

The 90% That Isn't Recurring

AD reported Q2 2026 revenue of $54.1 million, a 90% year-over-year increase, up from $28.5 million in Q2 2025. Site rental revenue alone climbed 95% year-over-year to $53.2 million. These are the numbers that get highlighted. They are also numbers that look better than they are.

The company sold its wireless operations to T-Mobile on August 1, 2025. The Q2 2025 comparison period was a hollowed-out shell transitioning out of the wireless business. Growing 90% from a shrinking base is mathematically easy — and strategically irrelevant. The real question is whether the tower business that remains can generate growth once the comparison period normalizes.

Here's what the quarter-over-quarter data says. Revenue fell 13.8% from Q1 to Q2. The competitor headline buried that because it was comparing to last year, not this one.

Annual growth from a shrinking denominator tells you what the company stopped doing, not what it's starting to do.

The GAAP Numbers Are Spectrum Sales, Not Tower Business

The $3.86 EPS figure in Q2 2026 looks spectacular compared to $0.17 in Q2 2025. It is also almost entirely a one-time accounting event. $409.8 million of gain from spectrum license sales flowed through the quarter. AD sold 700 MHz licenses for $74.8 million, 600 MHz licenses for $86.4 million, and then closed the Verizon deal for $1 billion on June 1. These proceeds boosted operating income to $399.3 million on $54 million of revenue.

That ratio — $399 million of operating income on $54 million of revenue — does not happen in a tower leasing business. American Tower, which actually runs a tower leasing business, trades at 23.6 times trailing earnings with a $80 billion market cap. Crown Castle trades at 37.3 times. AD trades at 14.6 times trailing earnings — but only because those earnings are being inflated by asset sales that will not repeat.

The market-cap comparison exposes the mismatch instantly. AD has a $3.05 billion market cap despite running a tower portfolio of 4,456 sites across 19 states. American Tower's 230,000-plus towers are worth 26 times more. That's not an outrage — it's arithmetic. AD is a tiny regional player pretending to be a national infrastructure story.

The Tower Business Is Facing Three Headwinds

Strip out the spectrum sales and look at what the tower business is actually doing. Three problems are visible, and management acknowledged all of them on the call.

Problem one: DISH is gone. DISH Wireless filed for bankruptcy in June 2026. AD took a full reserve for outstanding balances and removed DISH co-locations from its tenancy metrics. Management now excludes DISH from the reported tenancy ratio, which sits at 0.98 (up sequentially from 0.96 in Q1). Excluding a failed tenant from your utilization metric makes the remaining utilization look better. It doesn't replace the lost revenue.

Problem two: T-Mobile integration creates a tenantless tower problem. AD now estimates that 1,000 to 1,700 towers could be left without tenants following T-Mobile's Master Lease Agreement integration. T-Mobile's committed minimum covers 2,015 sites, but up to 1,800 interim sites — which are currently generating revenue — are not guaranteed beyond a 30-month window. Management calls this "ground lease optimization." The investor should call it revenue at risk.

Problem three: free cash flow collapsed. TTM free cash flow fell 96.4% year-over-year. Operating cash flow is $65.4 million against $31.4 million in capex, which sounds functional until you note that ROIC is 1.5% — essentially the cost of doing nothing. This is a capital structure that generates rental income but doesn't grow it.

When three of the top four tower risks — tenant bankruptcy, integrator exit, and cash flow collapse — land in the same quarter, the growth headline stops being credible.

TDS Doesn't Think This Is a Growth Story Either

Here's the detail that changes the framing entirely. Telephone and Data Systems, which owns approximately 82% of AD, submitted a non-binding proposal in May 2026 to acquire the remaining 18% of shares. An independent special committee is evaluating the offer.

TDS doesn't own AD to let the tower business compound as a standalone infrastructure play. TDS owns AD to extract the remaining spectrum value and collapse the public company into its parent. The $23-per-share special dividend in August 2025, the $10.25-per-share dividend in February 2026, and the $11-per-share dividend paid in June 2026 add up to $44 per share in distributions over 10 months. That is more than the current stock price of $35.24.

The parent company has already returned more in dividends than the market currently thinks the stock is worth. That is not a growth thesis. That is a liquidation schedule.

The strategic alternatives evaluation that management flagged on the call — complete with its own elevated costs — exists because TDS wants to take AD private. The "strong pipeline of colocation applications" that management highlighted is the public company's remaining job: keep the tower revenue flowing while the spectrum auction concludes and the buyout plays out.

The Cross-Currents

Where this stock goes from here depends on three forces that pull in different directions:

• TDS buyout premium (bullish). If TDS moves forward with acquiring the remaining 18%, there's likely a premium over the current $35 price. That would compress the downside and potentially deliver a quick pop.
• Guidance raise (moderately bullish). Management lifted 2026 guidance from $50–$65 million to $60–$75 million, and adjusted EBITDA from $200–$215 million to $220–$235 million. The tower business is performing better than the low-end base case, which helps the TDS valuation floor.
• Tenantless tower overhang (bearish). 1,000 to 1,700 potentially empty sites, DISH write-offs, and a QoQ revenue decline mean the growth story is structurally fragile. If TDS delays or the buyout fails, the public market has to price in the reality of a tiny, declining tower operator.

Directionally, the buyout premium is the dominant force. The stock has been beaten down 34% year-to-date and 31% over the last 120 days — a move that makes sense for a growth story but is too severe for a company about to be acquired by a $12 billion parent. The market is pricing this as a standalone growth company that's running out of growth, when it's actually a held-for-sale asset with a motivated buyer.

What the Earnings Recap Got Wrong

The headline "Strong Revenue Growth" describes a company that is selling its way to quarterly targets while the parent liquidates the remaining assets. The 90% revenue jump comes from comparing to a quarter when AD was mid-divestiture. The GAAP earnings number is driven by $1.16 billion in spectrum sales. The tower business — the only thing left to grow — declined sequentially, is facing a DISH vacancy problem, and has a 1,000-1,700 tower overhang from T-Mobile integration.

TDS knows this. It's why TDS is buying back the rest.

The investment case for AD right now is not about whether the tower business grows. It's about whether and at what price TDS completes the buyout, and whether the public market prices that event correctly. At $35, after $44 in dividends have already been returned, the remaining share count is priced as if it has perpetual growth optionality. It doesn't. It has an expiration date.

You don't buy a liquidation play on a growth headline. You buy it on the math of what's left.