Lumen Looks Cheap After Its AI Run — But the Equity Is a Leveraged Bet on a Backlog That Hasn't Converted Yet

Lumen Technologies' stock spent the past two years as one of the market's more dramatic AI trades. It sat near $1 in the summer of 2024, then climbed more than tenfold to a 52-week high of $11.95 before fading to about $6.94 today — roughly 40% off that peak. The question lobbed at a stock that big and that fallen is usually the same: is this a cheap second chance to buy the story, or confirmation that the market was right the first time?

The honest answer requires separating two very different things: the genuine business shift underneath the run, and how little of it has actually reached the income statement yet.

A run that priced a fiber empire

The catalyst was real, not froth. In August 2024 Lumen announced $5 billion in new business from companies scrambling to lock down fiber for AI. By early 2026 that pipeline had grown to nearly $13 billion in private custom fiber (PCF) contracts — long-haul routes built to order for hyperscalers such as Microsoft, Amazon, Google and Meta. The company then sold its consumer fiber business to AT&T for $5.75 billion, used the proceeds to cut debt, and earned credit-rating upgrades from all three major agencies. Its February 2026 investor day declared the turnaround complete.

That is a legitimate transformation story. A beaten-down, debt-heavy legacy telecom with a declining copper network repositioned itself as the essential connective tissue for AI.

The number the run glossed over

Here is the gap at the center of the trade: nearly $13 billion of announced contracts produced just $116 million of recognized revenue in all of 2025 — roughly nine-tenths of one percent of Lumen's $12.4 billion in total revenue for the year. The run priced the backlog as a near-term earnings machine. The reality is that the money appears only as fiber routes are physically built and billed, one network at a time, over many years. The stock's slide is the market slowly repricing that timeline.

Meanwhile the legacy business keeps shrinking. In the June quarter, total revenue fell 9% year over year to $2.8 billion on steep declines in the sold-off Mass Markets unit, and the remaining main business was still down 2% — even as the faster-growing "strategic" slice grew 14% and reached 53% of business revenue.

Why "cheap" is mostly a leverage illusion

The superficial case for cheap is easy to build. Lumen trades at about 6.4x enterprise value to EBITDA, roughly in line with AT&T at 6.6x and below Verizon at 7.9x. But the comparison flatters it. Look past the multiple: the equity is worth about $7 billion, against an enterprise value near $18.5 billion. The roughly $11 billion difference is net debt, and shareholders' equity is negative — the company has lost money in aggregate for years.

That structure is why a modest EV/EBITDA does not mean the stock is cheap. On a leveraged balance sheet, equity is what is left after the bondholders get paid. If the backlog converts exactly as hoped, the shares gain the most; if it converts more slowly than legacy revenue fades and interest accrues, the residual shrinks fast. Owning the shares is closer to owning an option on the backlog than to owning the whole network.

What a real turnaround would look like

None of this makes Lumen a fake story. The axle of the thesis is genuinely specific: its long-haul network holds spare conduit built by Level 3 twenty-five years ago, giving it a real cost and time-to-market edge on routes hyperscalers now want badly. That is a moat in the concrete sense, not a buzzword. AInvest's aggregate signal labels the stock a Hold, and its composite analysis rating is weak, but that is positioning, not a verdict on the network itself.

From a growth-at-a-reasonable-price frame, though, the honest conclusion is that this fails the quality gate before valuation even matters. Revenue has not inflected. Margins are thin on a reported basis, growth is still negative, and the balance sheet is levered. That is not a clean contrarian buy — it is a high-risk, high-leverage turnaround in which the cheapness is the byproduct of debt, not of confirmed earning power.

What would flip it: the main business starting to grow — management guides for that in 2028 — and PCF revenue ramping from tens of millions toward hundreds of millions a year, fast enough to outpace the legacy fade and the interest bill. Until that shows up in the income statement, the market's reset from $12 toward $7 is not obviously wrong.