Charter Communications, Inc

Charter fails the return on capital and debt tests, and its per-share value model cannot be relied on yet.

Charter Communications sells monthly subscriptions for Internet, mobile, video and voice, mostly as bundles, over its Spectrum fibre network in 41 states. Revenue was $54.8B in the latest year, and the 15 years of filings we reviewed show no loss year. Returns are the weak point: return on invested capital was 9.0% in the latest year against 6.7% on average, and the average return on equity of 25% is lifted by $95.7B of borrowing. At 3.0× earnings the shares look cheap on paper, but owner earnings have fallen at -17% a year, so our research status for this company is hard to predict. Our per-share value model returned a negative figure in every case, with -$434.59 in the base case, so it gives no usable estimate of what the shares are worth.

Charter sells monthly subscriptions for Internet, mobile, video and voice services, mostly bundled, to households and businesses across 41 states over its Spectrum fiber-powered network. Its filing says the monthly fees are the core of the revenue, and advertising, device sales and fees make up the rest. Revenue was $54.8B in the latest year, and the company's returns depend on keeping subscribers and raising revenue per customer while paying heavily for programming and network upgrades.

• Monthly subscription fees for bundled Internet, mobile, video and voice
• Mobile line growth sold alongside Internet
• Programming costs that rise faster than the prices customers pay
• Ongoing network upgrades and rural construction spending
• Advertising sales that move with political and economic cycles

Inside the circle? The subscription model is recurring, which helps. However, the company faces fiber, fixed wireless and satellite competitors, rising programming costs it cannot always pass on, and a large debt load, so the cash it can take out has fallen in recent years. A careful owner would find it hard to estimate the economics ten years out with the confidence this lens asks for, and the filings offer no clear evidence that the moat is widening.

What protects Charter's profits is mainly its fibre-powered network and the bundles it sells over that network. The filing says the company can upgrade capacity at a lower incremental capital cost than its competitors, and that cost edge would matter if it holds. Pricing power is only partial: the filing allows gradual price increases at the end of promotional periods, but it also says it may not pass on all programming cost increases. Returns on invested capital of 9.0% in the latest year, against 6.7% on average, and owner earnings falling at -17% a year do not show a moat that earns high returns on the capital it uses. At best we see a narrow advantage in the network and the bundle, and we cannot say how long it will last. The model's per-share values are negative (-$434.59 in the base case), so we do not rely on them here.

Charter has spent much of its cash on buybacks and pays no dividend, so the share count has fallen by -5.8% a year. Buybacks are only part of the picture. Total debt stands at $95.7B, and the company expects to keep adding borrowing as Adjusted EBITDA (earnings before interest, tax, depreciation and amortisation) grows. Book value per share, which is the net worth on the balance sheet for each share, has fallen by -4.2% a year, and the retained earnings test reads -1.3×. On candour, the filing states its leverage plans and buyback limits plainly, but the business section leans on promotional phrases. Our reading is that capital allocation is questionable under these principles.

• Charter has reduced its share count by -5.8% a year on average and pays no dividend (average payout 0.0%).
• Buybacks have run alongside rising borrowing. The filing says the proceeds of the 2025 senior secured notes were used, among other things, to fund potential buybacks, and capital spending is funded by operating cash flow and borrowings.
• Book value per share has fallen by -4.2% a year, and owner earnings growth is -17%, so the cash kept in the business has not built a larger claim per share.
• The base case value per share is -$434.59, which is negative. The model cannot give a usable estimate at this debt level, so the margin of safety of 25% cannot be tested against the price.

• FailsHigh returns on tangible capitalFails

  After-tax return on invested capital averaged 6.7% over 10 years and was 9.0% in the latest year.

  Why this test
• FailsLittle or no debtFails

  Total debt of $95.7B would take 39.7 yrs of normalised owner earnings to repay; operating income covers interest 2.6×. Both tests set this result: payback over 7 years and interest cover under 4x.

Several of Buffett's principles point the same way here. The test of durable advantage and pricing power is only partly met. The filing says the company has 'the ability to upgrade capacity at a lower incremental capital cost relative to our competitors', but it also warns 'We may not have the ability to pass on to our customers all of the increases in programming costs'. Buffett's tests for little debt and the $1 retained test both fail: debt is 6.0× equity, interest is covered only 2.6×, and book value per share has fallen by -4.2% a year. The owner earnings test passes on the normalised measure, which stands at $2.4B, and free cash flow has averaged 134% of reported profit. Capital spending has averaged 1.0× depreciation, which suits a business that doesn't eat capital only if the rural and network spending earns its keep, and the filing does not yet show that. Buybacks have cut the share count by -5.8% a year, and Buffett would count them as value creation only if they were made below intrinsic value, which we cannot measure reliably. On interest rates, the owner earnings yield of 16% sits well above the 10-year Treasury yield of 5.1%, but a high yield on shrinking cash flow is a warning sign. The circle of competence test is also hard to meet, because fibre, fixed wireless and satellite rivals and the rural build all make the future difficult to estimate.

• Refinancing of debt due in the coming years, and the interest cover ratio

  The filing says some borrowings carry variable rates, and new senior notes were issued in January 2026 partly to repay older debt. The cover ratio shows whether interest costs are getting easier to carry.

• Owner earnings and free cash flow in the 2026 annual report

  Owner earnings have fallen for several years. The filing credits lower cash taxes and interest for the 2025 cash improvement, and we need to see whether that help repeats.

• Wording on passing programming cost increases to customers

  The risk section says the company may not pass on all programming cost increases. Watch whether video pricing and margins change in later filings.

• Returns from the rural build, stated as a figure

  The filing expects 'infrastructure-style returns' from the rural build but gives no return figure yet. A stated return would test that claim.

• Buyback activity and the suspension of A/N's repurchase arrangement

  The filing says A/N suspended its repurchase arrangement until the Cox transaction closes or ends. Buybacks only add value if they are made below intrinsic value.

Questions the filings couldn't answer

• Which free cash flow figure is right? Our measure differs from the company's own reconciliation in the filing, and the excerpts do not settle the gap.
• Why does our per-share value model turn negative? The output is not a usable estimate, and the excerpts do not show whether the inputs or the method are at fault.