Variant Perception

Where We Disagree With the Market

The market is pricing Charter as if cable's unit economics have broken. They have not. Residential broadband ARPU has held at roughly $119/month for three consecutive years even as the company shed 393K Internet customers in FY25 and another 120K in Q1 2026 — a pattern that contradicts the "Altice destination" framing now embedded in a 5.4× EV/EBITDA multiple. The structural-decay thesis requires both volume and price to crack; only volume has. Two other gaps follow from the same misread: the market is pricing a perpetual controlled-company governance discount that has a hard expiry at the Liberty Broadband Combination close in summer 2026, and it is under-pricing the value transfer to Cox Enterprises now baked into a deal struck at a $353 VWAP and closing with the stock at $171. None of these is the same debate as Stan's — they are mispricings that survive whether Q2 broadband net adds print -60K or -120K.

Variant Perception Scorecard

The 68/100 variant score reflects three real but partially measurable disagreements rather than a single one-shot bet. Consensus on the "value trap" framing is unusually crisp — JPMorgan cut its target from $370 to $215, Goldman maintains Sell at $185, the stock fell 25% on a single broadband print — which makes the disagreement falsifiable. The four-month resolution window covers the July 24 Q2 print (volume vs. value test), the summer Cox/Liberty close (governance + dilution test), and the first post-close trending schedule.

Consensus Map

The cleanest consensus signal is the symmetric pessimism between Goldman ($185 Sell) and Wells Fargo ($180 Underweight) on one side and the post-Q1 reset of JPMorgan and Citi cutting targets 30-40% in a single week. Where consensus is not clear is on the Cox transaction's per-share value to Charter holders specifically (rather than the combined entity), and on whether the governance overhang has a known expiry.

The Disagreement Ledger

Disagreement #1 - ARPU stability disproves structural decay

A consensus analyst would say "Charter shed 393K Internet customers in FY25 and another 120K in Q1 2026; the unit economics of cable broadband are deteriorating because fiber and FWA offer comparable speed at lower price." Our evidence disagrees because the test of "deteriorating unit economics" is not net adds, it is price. Residential ARPU has held at roughly $119/month for three consecutive years; programming costs per video customer have fallen as the company sheds the most subsidy-dependent customers. If the market is right, Charter has to concede that the customers it is keeping are increasingly price-shoppable - which would show up as ARPU compression. The cleanest disconfirming signal is residential ARPU per customer relationship in the Q2 trending schedule: a break below $116 would validate the bear; sustained $119+ alongside any sub print falsifies the structural-decay frame.

Disagreement #2 - Cox deal economics now favor Cox holders

A consensus analyst would say "Cox closes in summer 2026 with $800M of synergies; the combined entity is bigger and better." Our evidence disagrees because the Cox consideration was struck on a 60-day VWAP of $353.64 and Charter is now $171 - meaning Cox Enterprises receives 23% of the pro-forma equity priced at deal-time levels, while Charter holders pay for it in stock now worth roughly half. The market would have to concede that the equity portion of the consideration is no longer "neutral to per-share value" but actively dilutive to standalone Charter holders' claim on combined-entity FCF. The cleanest disconfirming signal is the first post-close 10-Q: if pro-forma leverage prints at or below the 3.9x target with $800M synergy run-rate visible, the value transfer is offset; if it slips above 3.9x or transition costs absorb meaningfully into Adjusted EBITDA "Other, net," the dilution is uncompensated.

Disagreement #3 - Governance discount expires this summer

A consensus analyst would say "Charter is a controlled company with a structural governance discount - Liberty/A&N take cash through pro-rata buybacks at their dictated pacing." Our evidence disagrees on time horizon: the Liberty Broadband Combination closes contemporaneously with Cox in summer 2026, terminating the $100M monthly buyback floor and retiring three Liberty-affiliated board seats. The People tab explicitly states this alone would move the governance grade from C+ to B/B+. The market would have to concede that the controlled-extraction frame loses its forward-looking weight 4-6 months from now. The cleanest disconfirming signal is the first post-close repurchase authorization and the Q3 2026 trending schedule: if buyback pacing visibly decouples from Liberty's prior monthly floor, the variant is correct; if Cox Enterprises (the new largest holder) replicates the affiliated-buyback structure in a successor side letter, the discount migrates rather than disappears.

Disagreement #4 - Q1 -120K already inside revised consensus

A consensus analyst would say "Q1 broke the stabilization narrative; the multiple now reflects acceleration." Our evidence disagrees more narrowly: the post-print sell-side walk-down was 5-6% on FY26 EPS, not 25%. Sell-side analyst Houdlik publicly modeled 400K FY26 broadband losses - identical to FY25's 393K. The market priced confirmation of a step-change that analyst estimate cuts do not actually corroborate. The market would have to concede that the price reaction over-extended relative to the model revision. The cleanest disconfirming signal is the Q2 print itself - anything in the -60K to -90K range matches the consensus modeled run-rate and undoes the panic; -100K or worse confirms the step-change.

Evidence That Changes the Odds

The strongest single piece of evidence is the ARPU stability over three years of subscriber bleed. The most fragile is the Q4 YoY-improvement comparison, which Q1 2026 promptly disrupted. The Cox VWAP-vs-price math is the most quantifiable: the deal economics are publicly disclosed and the dilution arithmetic is checkable from the merger proxy.

How This Gets Resolved

Five of seven resolution signals fall inside a 90-day window (May 22 - August 31, 2026). The single most resolving signal is residential ARPU in the Q2 trending schedule because it tests the structural-decay frame independently of the binary Q2 net-add print. ARPU is the volume-vs-value test in a single line item.

What Would Make Us Wrong

The variant view is most exposed on ARPU compression we cannot yet see. Stable headline residential ARPU could mask a deteriorating mix: if Charter is replacing departing customers with promotional-rate customers and the back-book ARPU is declining while introductory-rate ARPU props up the average, the structural-decay frame is right and we are reading the wrong number. The Life Unlimited brand relaunch, the "$1,000 savings guarantee," and the Q1 2026 commentary on "low move rates" all point in this direction. A clean test is residential broadband-only ARPU vs. bundled ARPU, but Charter does not break this out cleanly. If Q2 prints show stable headline ARPU but management commentary references "competitive promotional intensity" or "retention spend," the variant is fragile.

The variant on Cox deal economics depends on the equity-component value transfer not being offset by synergy delivery. If the $800M synergy run-rate is delivered cleanly within 18 months of close and Cox EBITDA contribution prints in line with the merger proxy ($4.5B pro-rata), the dilution to per-share FCF is roughly compensated. The variant fails if synergies arrive but the Cox base brings worse subscriber trajectory than standalone Charter - in which case the dilution compounds the operating drag.

The variant on governance expiry is fragile if Cox Enterprises - which becomes the largest single holder post-close - negotiates a successor side letter that replicates Liberty's monthly buyback floor. The Cox merger documents do not appear to include such a clause but the Stockholders Agreement structure is well-precedented and could migrate. If Cox replicates the Liberty floor at any size, the controlled-extraction frame migrates rather than disappears.

The variant on Q1 being a one-quarter overshoot is the weakest of the four because the sample size is one. If Q2 2026 prints another -100K-or-worse Internet net loss, Houdlik's 400K FY26 estimate revises up, the EBITDA estimate cuts get serious (5-10% rather than 5-6%), and the panic pricing turns out to have been correct. The variant is one quarter from confirmation or refutation.

The first thing to watch is residential broadband ARPU in the July 24 Q2 trending schedule - a stable $119+ alongside any net-add print is the single observation that most resolves the variant view; a print below $116 collapses it.