The loan sales started quietly. Then they didn't.

Banks with exposure to heavily-indebted broadband fiber operators have begun selling their loans to distressed debt funds at a discount - not a temporary liquidity shuffle, but the first honest price discovery in a market where every other metric was marketing.

Any astute infrastructure investor would have seen this coming. The per-home-passed math never worked once you layered private equity leverage, fiber overbuild pricing wars, and the subscriber acquisition subsidies propping up take rates.

The build was real. The economics were fiction.

The numbers look impressive if you don't decompose them. The Fiber Broadband Association reported 11.8 million new U.S. homes passed with fiber in 2025 alone, pushing total FTTH passings to 98.3 million - fiber now reaches more than 60% of U.S. households, up 13% year-over-year.

Those are construction metrics, not revenue metrics. And construction costs tell you why this land grab was economically toxic from the start.

Median fiber deployment costs in 2025 were $18 per foot for underground builds and $8 per foot for aerial builds. Run the per-home-passed math: a typical suburban run of 300 feet underground costs $5,400 before you've even installed the ONT or secured a subscriber. Aerial is cheaper at $2,400 but limited to older neighborhoods with existing poles. In urban infill with multiple incumbents already present, the cost per activated subscriber stretches to $8,000–$12,000 because you can't build to 100% take rate - most operators land between 25% and 45%.

At $62 average monthly broadband ARPU, that gives a raw payback period of 14 to 26 months - and that's before debt service, maintenance, marketing, churn, and the regulatory overhead of BEAD compliance.

Now layer in the private equity leverage that made the build possible.

Wall Street has been churning out billions in complex bonds and leveraged loans to bankroll this fiber construction since at least 2023. Eighty-four percent of private equity buyouts in 2024 were financed by private debt firms, not traditional banks. The broadband sector was a primary beneficiary - PE firms acquired regional cable and fiber operators, loaded them with debt, and funded fiber greenfield builds on the borrowed capital.

That structure works only if subscriber growth compounds faster than interest. It hasn't.

Overbuilding turned the market into a negative-sum game.

By mid-2025, nearly 1 in 10 U.S. households had multiple fiber options - a condition that was once considered economically unviable. Fiber overbuilding is now a defining market dynamic, not a fringe edge case.

Here's what happens when three fiber providers serve the same zip code: nobody builds margin. Everyone builds price.

Charter Communications lost 120,000 broadband subscribers in Q1 2026 - nearly double the 59,000 it lost the prior quarter. The incumbent cable companies that spent decades building DOCSIS infrastructure are watching their subscriber base drain to fiber overbuilders who, ironically, can't monetize the customers they steal at anything close to the original cable ARPU.

The promotions make it worse. Charter added 1.5 million lines in 2025, partly through a one-year free line promotion that's rolling into paying customers this year. Free lines are the accounting equivalent of throwing a match: you gain the subscriber, you lose the margin, and you train the customer to leave again when the next competitor offers a better deal. S&P Global has explicitly warned that competitive pricing pressures will not ease through 2026 or 2027.

The incumbent-incompetence attribution works here, but only partially.

It is tempting to say that fiber overbuilders succeeded because cable incumbents collapsed on their own - the Graviton-to-Intel playbook. Cable companies did misread the fiber transition. They underinvested in DOCSIS 4.0 upgrades, let customer service rot, and priced themselves as monopolies for too long. That created the opening.

But the fiber overbuilders didn't close the opening with superior unit economics. They closed it with cheaper debt and cheaper ARPU. The result is a market where everyone is losing margin simultaneously.

Why banks are selling now, and what it means

Banks don't sell loans to distressed funds at a discount because they're short on balance sheet liquidity. They do it because they're recognizing that the underlying credit quality has deteriorated faster than the covenant stack can hide.

The U.S. technology sector is approaching a $330 billion debt maturity wall across high-yield bonds and leveraged loans. Broadband fiber operators financed through private equity carry a disproportionate share of that wall. When the first maturities hit - likely in the second half of 2026 and into 2027 - refinancing will be at higher rates on lower subscriber multiples because the overbuild competition has permanently depressed the ARPU floor.

S&P Global already downgraded Cable One to BB- in May 2025, citing weak operational execution. That was the canary. The bank loan sales are the rest of the flock.

The cross-currents:

  • Downside: Subscriber ARPU continues to erode as overbuild density increases. Debt maturities hit refinancing in a higher-rate environment. Multiple regional operators face covenant breaches simultaneously, triggering distressed restructurings that compress valuations further.
  • Upside: Consolidation through acquisitions creates a rationalized market with fewer competitors per metro. Survivors that built in lower-cost aerial markets first retain a per-home-passed cost advantage. BEAD funding (despite its June 2025 policy changes) still subsidizes builds in cost-prohibitive areas, deferring the unit economics reckoning in rural markets.
  • What resolves the ambiguity: The Q2 2026 earnings season. If Charter's subscriber losses accelerate beyond Q1's 120,000 and ARPU compresses further, the thesis is confirmed - this is a structural margin collapse, not a cyclical promotion hangover.

Directionally, the evidence points to sustained margin erosion. The bank loan sales are not a temporary liquidity event. They are the financial system acknowledging what the per-home-passed math showed from day one: the broadband land grab was funded on borrowed time and borrowed money, and neither is renewable.

You decide which was marketing fluff and which one was analysis.