UNIT · Uniti Group Inc. · Stock research

Current Thesis

The leg on offer is unchanged from June: a post-Windstream-merger levered fiber operator being re-cast as an AI-datacenter landlord. The 2026-07-30 Q2 print made the operational half of that case harder to argue with. Fiber Infrastructure booked a record $2.2M of monthly recurring revenue, roughly 30% above the prior record, and full-year Fiber Infrastructure guidance was raised to a $1.0B revenue and $575M contribution-margin midpoint on what management attributed to hyperscale and AI-related activity.

The equity did not pay for it. The 2026-08-14 close was $10.01 — 21.5% under the $12.75 52-week high, -9.4% over three months, RSI(14) at 40.3 — with sell-side consensus sitting at $10.31 across 9 analysts at Hold, effectively level with the market price. A story that delivers a segment guidance raise and then loses ground for two weeks has stopped being carried by the story.

The narrative is maturing. Dating it: the narrative's proof points are still arriving (2026-07-30 fiber guide raise, record bookings, 2026-07-15 Kinetic ABS Series 2026-2 at $1.14B), while the flow that re-rated the name from roughly $7.80 in early June to $11.07 on 2026-06-26 has stopped extending and the two post-print target revisions both moved toward spot rather than away from it. It is not saturated — 9 covering analysts is thin coverage — and the structure has not broken, since the $9.80 area that has floored the June leg is still intact.

Bullish and bearish views on Uniti Group Inc.

The model's bull view on Uniti Group Inc. (UNIT), in brief: Segment guidance moved up at the 2026-07-30 print. The bear view: Capex is guided above EBITDA. FY2026 midpoints put consolidated net capex at roughly $1.525B against adjusted EBITDA of roughly $1.475B. On the company's own guidance the build outspends cash operating profit before interest, which keeps the equity dependent on the ABS window… Both cases follow in full.

Bull Case

• Segment guidance moved up at the 2026-07-30 print. FY2026 Fiber Infrastructure midpoint raised to $1.0B revenue / $575M contribution margin. The segment printed $234.1M revenue and $121.8M contribution margin in Q2 at roughly 52% margin.
• Bookings set a record. Q2 Fiber Infrastructure bookings of $2.2M in MRR, ~30% above the previous record — contracted revenue that has not yet reached the P&L.
• Top line cleared the estimate. Q2 consolidated revenue $909.7M with adjusted EBITDA $357.1M at ~39% margin; coverage of the print flagged a revenue beat alongside an EPS miss.
• Consumer build accelerating. The 2026 construction target was lifted to 475,000–525,000 homes, with consumer fiber revenue guided $635M–$655M and Kinetic contributing $539.0M of Q2 revenue at ~42% contribution margin.
• The securitization channel is functioning. Kinetic ABS Series 2026-2 closed 2026-07-15 at $1.14B, ~6.18% weighted coupon, following $960.1M in January 2026 — each deal terms out corporate debt below the unsecured stack and funds the build.
• The hyperscaler motion is recurring, not a single headline. Q1 2026 recognized $68.7M of incremental sales revenue from one large-hyperscaler dark-fiber IRU, against a funnel management has sized near $1.5B of total contract value with hyperscalers at roughly 40% of pipeline.

Bear Case

• Capex is guided above EBITDA. FY2026 midpoints put consolidated net capex at roughly $1.525B against adjusted EBITDA of roughly $1.475B. On the company's own guidance the build outspends cash operating profit before interest, which keeps the equity dependent on the ABS window staying open.
• The loss line widened and is guided wider. Q2 net loss $155.9M; FY2026 net loss guided $480M–$530M.
• The raise came in the smaller segment. Kinetic's FY midpoint was only maintained ($2.145B revenue / $905M contribution margin) while Fiber Infrastructure was raised — and Kinetic was $539.0M of Q2 revenue versus $234.1M for Fiber Infrastructure. The AI-adjacent piece is not yet the revenue engine.
• Targets compressed toward the price on the print. TD Cowen cut to $11 from $12 on 2026-07-30 while keeping a Buy; BofA lifted its Sell-rated target to $7.50 from $7 on 2026-07-31. Consensus $10.31, Hold.
• Target dispersion is wide. The published range runs $6.00 to $16.00 — a spread that reads as disagreement about the capital structure rather than about the fiber demand.
• Counterparties remain unnamed. Disclosure still describes the demand side as "a large hyperscaler" and "a neo-cloud provider." A named MSFT/META/GOOGL/AMZN dark-fiber or IRU headline is the un-modeled event, and it has not arrived.

Setup & Price Structure

• The June advance ran from roughly $7.80 in early June to $11.07 on 2026-06-26. The 2026-08-14 close of $10.01 sits inside that leg, above the $9.80 area that has marked its lower boundary through the post-print drift.
• 21.5% below the $12.75 52-week high on the adjusted series; three-month return -9.4%. RSI(14) at 40.3 leaves the name mid-range — no oversold reading to mean-revert from, no reclaim of the late-June high.
• Positioning observables. Consensus $10.31 versus a $10.01 close is roughly 3% of implied upside as published by the aggregator; only the high-end marks sit meaningfully above spot. Both post-print revisions (2026-07-30, 2026-07-31) moved toward the market price from opposite directions. Coverage breadth is 9 analysts. No insider transaction and no equity issuance appears in the record reviewed for this note; absence of a filing is not evidence of absence, and the next proxy/10-Q disclosure is the place to check.
• What re-accelerates it. A weekly close back above $11.07 attached to fiber-segment newsflow would argue the June leg resumed rather than expired.

Catalyst Calendar (next 30 days)

• ~2026-11-03 (est.) — Q3 2026 results, per the consensus reporting calendar. Outside the window; the FY2026 revenue and adjusted EBITDA ranges are not re-tested before it.

Elapsed catalysts

• No confirmed dated catalyst falls inside the 30 days from 2026-08-15. That is the setup's central feature right now: nothing scheduled resolves the guidance question before November. (passed 11d ago)
• ~2026-09 (est., exact date not confirmed in the 2026-08-04 announcement) — TD Cowen 12th Annual Communications Infrastructure Summit, CEO Kenny Gunderman presenting. The only scheduled management appearance identified between now and the Q3 print; funnel sizing and counterparty framing are what would move on it. (passed 22d ago)
• Undated — the next Kinetic ABS series. Whether it prices tighter or wider than the ~6.18% weighted coupon of the 2026-07-15 deal is the cleanest available read on funding cost, and it can arrive without notice. (passed 42d ago)

What Would Change Our Mind

The structure that defines this read is the June re-rate shelf. Losing it is the break: a weekly close below $9.80 forfeits the hyperscaler-funnel leg and puts the $7.50–$7.80 pre-run zone back in play, and on a credit where guided net capex exceeds guided adjusted EBITDA, a shelf break carries more information than it would on an unlevered name.

Second condition: the Q3 print (~2026-11-03) taking FY2026 revenue below the $3.63B lower bound or adjusted EBITDA below $1.45B, or walking back the Fiber Infrastructure midpoint that was just raised to $1.0B / $575M. A third: the next Kinetic securitization pricing materially wider than the ~6.18% weighted coupon of the 2026-07-15 series, which would mark the funding channel repricing against the build.

The upside change-of-mind is narrower and specific: a named hyperscaler counterparty on a dark-fiber or IRU agreement, or a weekly close above $11.07 with fresh Fiber Infrastructure bookings behind it. Absent one of those, the September conference passing without a funnel update leaves the name with no scheduled event until November.

Correlation Notes

• Trades on a US exchange as a REIT-structured fiber operator; the drivers are credit-market, not hardware-cycle. The observable proxies are high-yield telecom spreads and the ABS market that priced the 2026-07-15 Kinetic deal at ~6.18%.
• Correlation to hyperscaler capex headlines is real but second-order: the revenue is contracted IRU and lease dollars recognized over time (Q1 2026's $68.7M IRU contribution being the visible example), so it responds to capex announcements with a lag rather than tick-for-tick. That timing gap is an inference from the revenue model, not a measured beta.
• Screening comparisons to optical and networking equipment names mislead here. This is a $909.7M-quarterly-revenue operator guiding to a $480M–$530M FY2026 net loss, with the equity behaving as the residual claim on a levered balance sheet.
• Headline valuation multiples on trailing net income remain merger-accounting artifacts and produce a false cheap signal on screens.