Chapter 4

The Machine That Ran on Borrowed Shares

Defending a shrinking broadband base costs Charter 21.3% of revenue in capital spending against $94 billion of net debt. The question this act answers is what the company did with the cash the network threw off before that build — and the answer is singular. For a decade Charter returned capital exactly one way: it bought its own stock, and it borrowed to do it. It has never paid a cash dividend and says it does not intend to [1]. Since the buyback program began in September 2016, Charter has repurchased roughly 179.7 million shares and Charter Holdings units in the public market for approximately $78.8 billion [2]. That is the whole capital-return story of the last ten years, and it is now running in reverse.

The scale is easier to see against the cash the business actually generated. In the peak years the repurchase bill ran well above free cash flow; the gap was filled with debt.

Source: derived from reported financials, FY2016–FY2025 Forms 10-K, Consolidated Statements of Cash Flows [3].

Across FY2016–FY2025 Charter spent more repurchasing stock than the business produced in free cash flow, and it funded the difference by letting the balance sheet grow: net debt rose from $60.2 billion at the end of 2016 to $94.3 billion at the end of 2025, up $34.1 billion [4]. The purpose was arithmetic: with revenue flat to falling, per-share value had to come from a smaller denominator. It did. Diluted shares fell from 235.5 million in 2018 to 137.7 million in 2025, a cut of roughly 42% [5]. Charter shrank its share count while its debt climbed, and the two moved in lockstep because they were the same transaction.

Source: derived from reported financials, FY2016–FY2025 Forms 10-K; net debt is total debt less cash [6].

The 2017 bump — a share count that briefly jumps to 296.7 million — is the residue of the 2016 Time Warner Cable and Bright House deals that built modern Charter with equity. Everything after is the buyback pulling the count back down. This is the identity The Broadband Incumbent and The Broadband Fulcrum described from the outside: a company whose equity story was manufactured on the financing line, not the revenue line.

A Leverage Band, Not a Discipline

Management's framing of this is discipline. Charter runs to a stated target of 4.0 to 4.5 times net debt to adjusted EBITDA, "near the midpoint," and it has held there with unusual steadiness — 4.15 times at the end of 2025 on roughly $94.6 billion of debt principal [7]. The leverage ratio is the constant. What flexed was the buyback.

Source: derived from reported financials, FY2021–FY2025 Forms 10-K [8]. Capex/revenue rose 14.8% → 21.3% over the same span.

Repurchases fell from $15.4 billion in 2021 to $1.2 billion in 2024 — down 92% — while capital intensity climbed to a decade high [9]. Leverage barely moved through all of it. That is the tell: if the band is fixed and capex is rising, the buyback is not a return policy set by management's read of value — it is the residual, whatever cash the network build and the leverage target leave behind. In the fourth quarter of 2024 the residual went to nearly nothing. Charter repurchased just 292,000 shares for $113 million, an explicit pause held open pending the February 2025 Liberty Broadband shareholder vote, even as its leverage drifted down to 4.13 times [10]. Management's own stated order of priorities names the buyback last: free cash flow goes "for organic opportunities first and then accretive M&A and buybacks" [11]. For a decade the first two claims were small, so the buyback got almost everything. That ordering is about to be tested by the first large deal in ten years.

The Price of the Shares It Bought

Because the buyback was the equity story, the prices Charter paid are part of the record. The company discloses an average repurchase price each quarter, and the trajectory is stark: about $384 per share in the fourth quarter of 2024, roughly $262 in the fourth quarter of 2025 (2.9 million shares for $760 million), and $225 in the first quarter of 2026 [12] [13]. On 22 July 2026 the stock implied roughly $129 per share. The $78.8 billion the program deployed since 2016 was spent across a price range that runs from about $225 to well above $400, against an equity the market now values at about $17.8 billion in total [14]. The per-share accretion that justified borrowing to buy stock was real when it retired shares cheaply relative to their later value; measured against today's price, most of the capital was returned above where the shares now trade. The arithmetic is left for the reader — but it is the reason a decade of shrinking the count did not compound into a rising equity value.

The Machine in Reverse

For ten straight years Charter spent nothing on acquisitions — cash used for acquisitions was zero every year from FY2016 through FY2025. That decade of pure buyback is now reversing into the largest deal in the company's history, and the reversal is being paid for in the one currency the buyback existed to conserve: equity.

The Cox Transactions, announced 16 May 2025, combine Charter with Cox Communications' residential cable and commercial businesses. The cash piece is small and debt-funded — Charter will fund $4.0 billion of cash consideration with debt and assume roughly $12.6 billion of Cox net debt and finance leases [15]. The bulk of the consideration is stock: Cox Enterprises receives convertible preferred and common partnership units in Charter Holdings. The parallel Liberty Broadband Combination — announced 12 November 2024, whose principal asset is roughly 41.5 million Charter shares — is an all-stock unwind expected to close at the same time [16].

The net effect on the share count runs directly against the last decade. Management sized it plainly: at close Charter will issue the equivalent of just over 46 million shares to Cox Enterprises, partly offset by a reduction of about 6.8 million shares from the Liberty combination, so a standalone as-converted, as-exchanged share count that stood near 137.7 million would rise to about 179 million [17].

Source: FY2021–FY2025 shares as reported [18]; ~179M as-converted count at close per management [19].

Twenty-eight percent of the shares the company spent a decade retiring come back at close, issued to a family. The freed cash the roll-off is meant to produce — management frames the fall in capital spending from about $11.7 billion in 2025 toward below $8 billion by 2028 as over $28 of annual free cash flow per share — now has three claims on it rather than one: the buyback at a depressed price, a promised de-levering to a new 3.5 to 3.75 times target within three years of close, and integrating Cox [20]. How that fork is resolved is the next allocation decision, and it will be made under different owners than the ones who ran the buyback.

The Handover

Charter has always been a controlled company; the question the deal settles is by whom. Today two related holders, not the public float, decide the board. As of 31 December 2025 Liberty Broadband held about 29.22% of the vote and Advance/Newhouse about 13.12% — roughly 42% between them — under a stockholders agreement that fixes the board at thirteen directors, lets Liberty designate up to three nominees and A/N up to two, and gives each a seat on every board committee [21]. On paper the board is highly independent — twelve of thirteen directors are NASDAQ-independent, with CEO Christopher Winfrey the only insider and Eric Zinterhofer, founder of Searchlight Capital, as non-executive chairman. But five of those seats are affiliate-designated — three for Liberty, two for A/N — and those five are independent under NASDAQ rules yet not SEC-independent for audit-committee purposes [22].

The Cox and Liberty closings redraw this. Liberty Broadband's governance layer terminates and its three designees resign; A/N's rights are modified and preserved; and Cox Enterprises arrives as the new controlling shareholder with about 25.1% of the combined company's diluted shares [23]. Alexander C. Taylor, chairman and CEO of Cox Enterprises, becomes chairman of Charter's board for an initial three-year term; Zinterhofer steps down to lead independent director; and the amended stockholders agreement caps Cox acquisitions at 30% and A/N at 19%, with voting above the cap (30% for Cox, 15% for A/N) forced to mirror the public float [24].

Source: FY2025 10-K, control-structure and Cox governance disclosures [25] [26].

Winfrey said it himself on the call that opened the pause: the cable industry is "largely made up of family-owned or family-controlled entities making the decisions on potential combinations" [27]. Charter is completing that sentence about itself. The overhang the market watched for years — Liberty's stake and the eventual question of what it would do with it — is being resolved not by dissolution into the float but by transfer to a new family in the chair.

Whose Skin

The people running the company hold very little of it. CEO Winfrey owned 38,385 Charter shares in April 2026, worth about $5 million at the implied price; CFO Jessica Fischer held 2,546 [28]. The large economic stakes sit with the holder entities — Liberty Broadband alone held about 38.6 million shares [29]. Management's exposure is not owned equity that falls when minority holders' does; it is upside optionality.

Winfrey's amended employment agreement, approved in December 2025 and running to December 2028, makes the design explicit. On a base salary of at least $2.5 million and a target bonus of 300% of base, he receives annual stock-option grants with a grant-date fair value of at least $23 million from 2027, three-year cliff-vested, plus a $6 million top-up option in January 2026 [30]. These are ordinary options struck at grant-date price: they pay on any absolute recovery of the share price and are not indexed to how cable peers perform. The incentive is keyed to the same per-share number the buyback engine was built to lift — and it pays on a rebound whether Charter earns it or the sector simply re-rates.

The capital-return machine and insider liquidity are also plumbed together. Under a 2016 letter agreement, A/N sells shares back to Charter every month on a pro-rata basis alongside the public buyback, at the average price Charter paid that month — 1.0 million units for $373 million in 2025 alone [31]. A separate tax receivable agreement obliges Charter to pay A/N half of the tax benefit it realizes when A/N exchanges its Charter Holdings units [32]. And Charter booked about $39 million of FY2025 carriage revenue from HSN and QVC, whose parent is chaired by former Charter director Gregory Maffei [33]. None of this is hidden; all of it is disclosed. But it means part of what looked like a return of capital to all shareholders doubled as a managed sell-down channel for the largest ones — the buyback that shrank the float was also the exit ramp for the holders now handing over control.

The capital identity, then, is a company that borrowed to retire more than 40% of its shares at prices mostly above where they trade today, ran the repurchase as the flex variable beneath a fixed leverage band, and is now issuing 46 million new shares to install a new controlling family — while paying its CEO almost entirely in options on the price that whole machine was built to move. What remains unsettled is whether the capital freed by the promised capex roll-off flows back to buybacks, to the new de-levering target, or to integrating Cox, and under whose incentives that choice gets made. That is where the record of promises meets the price the market has put on them.