Visible Alpha

Visible Alpha broker models via S&P Xpressfeed · 19 brokers · 441 line items · freshest revision 2026-07-27.

Broker Models

Brokers do not model growth here. Revenue and EBITDA fall in every modeled year to FY-2028 and broadband subscriber losses widen. What the models do carry is a cash inflection: capital expenditure rolls off as the build program ends, and the diluted share count falls by roughly a quarter, so modeled free cash flow per share more than doubles. The genuine disagreement in this feed is over how much stock gets retired — not over the income statement.

The model is a cash story, not a growth story

Source: derived from vendor data.

Those four tiles are the whole argument. Revenue in FY-2028 is modeled below FY-2025; capital expenditure is roughly a third lower; and the share count is roughly a third lower again. Free cash flow per share is the only headline number in the set that goes up materially, and it does so because the denominator shrinks alongside the cash outflow.

The capex cycle turns in FY-2027, and that is the entire inflection

Source: derived from vendor data.

Capital expenditure is modeled flat through FY-2026 and then drops in each of the next two years, and by FY-2028 free cash flow has closed most of the gap to it. The split view shows this is not one program ending. Line extensions — the subsidized rural build — come down every year from FY-2025. Everything else actually rises into FY-2026 before falling, which is why the cash benefit lands in FY-2027 rather than immediately.

Free cash flow per share more than doubles, and the buyback does much of the work

Source: derived from vendor data.

The mean path runs $35.42, $39.93, $54.88, $81.53 — an inflection with almost nothing to do with trading performance. Underneath it are two separate estimates, and only one of them is agreed.

Source: derived from vendor data.

Here is where the estimates come apart. Through FY-2026 the models sit within a few dollars of each other on cash per share and within a few million shares of each other on the count. By FY-2028 the most aggressive model has retired enough stock to leave a share count barely over half the most conservative one, and the per-share cash figure inherits that fan directly: the top model carries more than half again the free cash flow per share of the bottom one. Brokers broadly agree on the cash; they disagree on how many shares are left to divide it by.

Broker dispersion sits in capital allocation, not the income statement

Source: derived from vendor data.

Read the table by relative width rather than absolute dollars. FY-2028 revenue is a tight consensus and EBITDA nearly as tight. Share repurchases are the outlier: the highest model buys back close to three times what the lowest one does, from a cash pool the same brokers size within a much narrower range. Headline P&L consensus and its revision history live on the CapIQ tab; the value added here is that the argument sits one line below it.

Mobile is the only product adding customers — and it is decelerating

Source: derived from vendor data.

Mobile carries the volume story and the models fade it every year. Broadband moves the other way: internet losses widen from FY-2025 into FY-2027 before easing slightly, and the quarterly path is the more uncomfortable read — the modeled loss deepens in almost every quarter through to 2QFY-2027. Video and voice losses shrink, but from a smaller base. The crossover is the point worth holding: on the mean paths shown, mobile line additions cover the combined internet, video and voice losses in FY-2025 and FY-2026, and stop covering them from FY-2027.

Mobile's economics are the thinnest-covered numbers in the set

Source: derived from vendor data.

The direction is consistent — the mean mobile EBITDA path rises every year — but the level is barely agreed. For FY-2026 the four models span roughly a three-fold range on mobile EBITDA while sizing mobile revenue within about a quarter of each other. In other words, the disagreement is about mobile's cost base, not its scale. Given how much of the equity story rests on mobile, that is the least well-covered line that matters.

Broadband loses volume, not price; video loses both

Source: derived from vendor data.

Residential internet pricing is modeled essentially flat — $71.19 in FY-2025 against $71.31 in FY-2028 — so the entire modeled decline in internet revenue is subscribers walking out, not discounting. Video is the opposite: monthly revenue per video subscriber falls from $94.88 to $81.73, and the subscriber base falls too. Mobile line ARPU inches up, which is why mobile revenue grows faster than its already-decelerating line additions.

Source: derived from vendor data.

Mobile service is the only large line the models grow meaningfully, and it does not grow by enough: the modeled gain there is smaller than the modeled loss across internet, video and voice combined. Commercial revenue is modeled as near-static — a segment with over seven billion dollars of revenue that the models expect to do almost nothing for three years, which is itself a claim worth testing.

Leverage barely moves, because the cash goes to shareholders

Source: derived from vendor data.

Net debt to EBITDA improves only from 4.2 times to 3.94 times across four modeled years, and part of that is arithmetic working against the models — EBITDA is falling, so holding the ratio flat requires paying debt down. The models are not deleveraging Charter; they are running it at roughly constant leverage and directing the growing free cash flow at the share count. That is a coherent plan and it is what the per-share numbers on this page depend on, which is why the share-count dispersion above is the line to watch.

Headline P&L consensus, momentum and beat/miss live in the CapIQ tab.