Variant Perception

Where We Disagree With the Market

The market is reading Charter as a Cable One-trajectory cigar butt; the report's evidence says it is a leveraged consolidator whose per-share math compounds whether or not pricing power rebounds. Consensus has compressed CHTR's multiple to a 10-year low (5.3x EV/EBITDA, ~4x P/FCF, 25% FCF yield) on the back of one quarter — Q1 2026 — where residential ARPU printed -1.4% YoY for the first time in 20+ years while Internet subscribers shrank another 120K. We don't dispute the Q1 print; we dispute three of the inferences the market has stacked on top of it. The optical ARPU break was meaningfully distorted by a $171M video-revenue reallocation to programmer streaming apps, the Cox + Liberty Broadband combination is being priced as one leveraging event when the Liberty leg is a mechanical ~33% share-count retirement with no cash leaving Charter, and the Cable One analog the bears are anchored on has none of Charter's mitigants. The single resolving event is the Q2 2026 ARPU print on July 24 — if it backs out the reallocation and prints flat-to-modestly-negative, the bear's load-bearing data point evaporates inside 60 days.

Variant Perception Scorecard

Time to Resolution: 2–6 months (Q2 print Jul 24, CPUC ruling Jul 16 / Aug 13 fallback, Cox/LBRD close before Sep 15).

The variant strength score is held at 65 — not higher — because the deepest disagreements are about the interpretation of disclosed evidence (Q1 reallocation, LBRD mechanics, Cable One non-analogy) rather than about hidden financials. Consensus clarity is genuinely high: analyst price targets cluster bimodally with Goldman at $185 Sell and Benchmark at $435 Buy on the same data, the stock prints at a 10-year-low multiple, and 6 of 22 sell-side analysts now carry Sell ratings. Evidence strength is meaningful but conditional: each disagreement points to a concrete, near-dated, observable signal — the Q2 print (Jul 24), the CPUC ruling (Jul 16 / Aug 13 fallback), and the Cox/LBRD close (before Sep 15 DOJ deadline) — that resolves the variant view inside six months rather than over years.

Consensus Map

The five issues above are the load-bearing planks of the bear thesis as the tape is currently writing it. Three are testable on a single date (Q2 earnings July 24); two require a longer horizon (capex glide, leverage). The disagreement work below focuses on the testable three — those are where the variant view earns the right to be different.

The Disagreement Ledger

Disagreement #1 — The ARPU regime change is partially an accounting artifact. Consensus has internalized "first negative residential ARPU print in 20+ years" as proof that fixed wireless has capped cable's installed-base pricing power forever. Our reading is narrower: Charter disclosed in the Q1 release that $218M of Q1 2026 video revenue was reallocated to programmer streaming apps versus $47M in Q1 2025 — a +$171M optical drag worth approximately 80-100 basis points of residential ARPU on the relevant revenue base, exactly the size of the YoY change. If the bear were right, we would have to concede that pricing power is permanently capped and that the long-term thesis Driver #3 (broadband demand keeps growing at stable price) is broken; what we would be conceding is one quarter of disclosed accounting reclassification doing most of the heavy lifting in the headline. The cleanest disconfirming signal is a Q2 2026 ARPU print at -1.5% or worse with the reallocation backed out — that would prove the underlying pricing has cracked, not the optics.

Disagreement #2 — Cox and Liberty are being collapsed into one number when they move per-share value in opposite directions. Consensus is treating both transactions as "Charter does big M&A → leverage goes up → equity gets riskier." The actual mechanics differ sharply. Cox is a $34.5B headline acquisition that adds $12.6B of assumed debt plus $4B in cash, pushing pro-forma leverage briefly above 4.5x in exchange for ~12M legacy-Cox customers and a $500M (raised to $800M per Q1 transcript) synergy target. Liberty Broadband is a separate, all-stock combination that retires approximately 41.5M Charter shares — roughly 33% of FY2025 float — with no cash leaving Charter. If we are right, the per-share FCF math compounds mechanically: a 33% lower share count on a flat $5B FCF lifts FCF/share by ~50% without needing ARPU stabilization, Cox synergies, or capex glide. What the market would have to concede is that the bull case at $300 doesn't require the bull operating scenario — it requires only that the deals close. The disconfirming signal is a refile (CPUC slips past Aug 13 → DOJ HSR expires Sep 15 → deals push into 2027); that defers but does not break the mechanic.

Disagreement #3 — The Cable One analog is the wrong analog. Bears are using CABO's collapse from high-30s EBITDA margin to 8.7% as the visible warning, and the market has accepted that frame. CABO is a ~1M-customer pure-cable operator in rural markets, with no MVNO, no DOCSIS 4.0 chip co-investment, no scale density advantage, and ran capex at 19.2% of revenue with no announced step-down. Charter has all four mitigants — 11.8M mobile lines on a Verizon MVNO with 89% WiFi offload, joint Broadcom DOCSIS 4.0 chipset investment with Comcast, 58M passings at the highest gross-margin density in the peer set, and a stated path back to under-$8B capex post-2027. If we are right, the appropriate cross-section comparator is Comcast at 5.2x EV/EBITDA with 17.7% FCF margin — not CABO at 23.5x on collapsed EBITDA — and Charter's current 5.3x deserves to converge toward CMCSA's level (not CABO's) as the capex hump rolls off. The disconfirming signal is that Cox legacy passings, when disclosed in the first post-close 10-Q, exhibit CABO-style ARPU and customer decay rather than Charter-style stability.

Disagreement #4 — Credit and equity disagree about going-concern, and credit usually wins. Charter bonds yield 6.87-7.15% on 10-30 year maturities — a wide spread to investment-grade benchmarks (150-200 bps) but a normal one for a high-BB / low-BBB infrastructure issuer with 4.4x leverage. S&P rated the operating sub BBB- in November 2024. The equity, however, trades at multiples (5.3x EV/EBITDA, ~4x P/FCF, 25% FCF yield) that normally signal a market consensus the business will not exist in ten years. When senior-claim holders price normal credit and residual-claim holders price distress, the residual claim is usually overweighting tail risk that the senior claim isn't seeing — because the bondholders have the cleaner read on FCF coverage and have $96B of capital at stake to get it right. The disconfirming signal is a credit-spread widening at the next refinancing tranche or an S&P outlook change to negative; absent either, the cross-asset asymmetry favors the equity catching up to credit, not the other way around.

Evidence That Changes the Odds

How This Gets Resolved

The resolution calendar is unusually dense for a name this size: three of the six signals resolve inside a 60-day window between July 16 (CPUC preferred) and September 15 (DOJ deadline). That density is itself useful — the variant view does not have to wait 18 months for evidence to accumulate. By the end of September 2026 we will know whether the Q1 ARPU print was reallocation-driven (Disagreement #1), whether the LBRD share retirement mechanic crystallizes on schedule (Disagreement #2), and whether the credit market's read of normal-credit was vindicated through the deal close (Disagreement #4). Only Disagreement #3 (Cable One analog) requires longer, since Cox legacy customer behavior is not disclosed until the first post-close 10-Q.

What Would Make Us Wrong

The cleanest way to break the variant view is for Disagreement #1 to fail at the Q2 print. If residential ARPU YoY comes in at -1.5% or worse with management unable to credibly attribute the move to the video-to-programmer-app reallocation — or, worse, if non-promotional cohort ARPU also prints negative — the bear's frame is correct: pricing power on the installed base has been re-priced by FWA at $50, the "first negative print in 20+ years" was a regime change rather than an accounting artifact, and the cable franchise is on the Cable One trajectory we argued was the wrong analog. The disagreement about the CABO frame loses its weight at that point because the data starts to confirm it; the disagreement about LBRD-mechanical-accretion stops mattering because the franchise being acquired-into is the franchise that is impaired.

The second break point is the Cox-and-LBRD close itself. If the California PUC misses August 13 and the deals refile beyond September 15, the per-share compounding mechanic on which Disagreement #2 rests is deferred by 12-18 months. Charter does not get the share-count retirement; the buyback stays paused or throttled by the Liberty side-letter; the market re-prices the entire deal optionality lower. The variant view does not break here — the mechanic is delayed, not killed — but the time-to-resolution stretches from "this year" to "next year," and the FCF/share marker of $50+ that anchors the bull arithmetic gets pushed out of any reasonable underwriting horizon.

The third break point — slower but the most existential — is the credit market itself moving against the variant view. If S&P moves outlook to negative on pro-forma Cox leverage and bond spreads widen 50+ bps on the next CCO Holdings issue, credit will have stopped pricing normal-credit and started pricing the same tail risk the equity already prices. At that point the cross-asset asymmetry argument in Disagreement #4 collapses; both senior and junior claims are agreeing that the franchise is impaired and the variant view has to defend itself only on the operating disagreements (#1 and #3), with less margin for being wrong on either. The honest read is that this is the slowest of the three break points to develop and the one most under management's direct control — so it sits as the structural risk to monitor rather than the near-term resolving signal.

A red-team observation worth naming: the strongest piece of evidence the variant view leans on (CEO and director open-market buying at $172-$174) is not, by itself, large enough to move a $17B market cap. CEO Winfrey already owns ~$175M of equity; adding $1.2M is rounding error against his existing exposure, and the buys do not change the alignment math — they signal conviction at price, not commitment of new capital. We have leaned on this signal because it is the cleanest insider read against an aggressively negative tape, not because the dollar size moves the equity story. If the variant view is right, the insider buys will be vindicated as the under-the-radar tell; if wrong, they will look like a value-investor's anchor in a value trap.

The first thing to watch is the Q2 2026 residential ARPU print on July 24, with the management commentary on whether the $171M video-to-programmer-app reallocation accounts for the bulk of the Q1 -1.4% headline.