Return on Capital

Return on Capital

Judged by the reader's filter — a durable moat and strong ten-year cash flows — Molina does not clear it: its revenue is rebid and its 2025 free cash flow was negative. This chapter looks at the company on the terms that filter sets aside. On them Molina is a capital-light, high-return reinvestment machine: mid-20s-percent return on equity on almost no capital through 2024. That capital efficiency is real — and earned without a moat, which 2025 shows unwinds fast.

What the screen sets aside

Two prior chapters have already settled how Molina scores against this reader's core tests. It has no customer lock-in — Medicaid revenue is awarded by competitive bid on three-to-five-year cycles and can be rebid away (Revenue Durability). And its cash flow does not stand still: operating cash flow was negative $535 million in 2025 and free cash flow was negative $636 million, so the free-cash-flow-yield screen fails outright (Cash and Capital). A durable-moat, strong-and-steady-cash investor stops reading here.

That stopping point is itself a choice about what kind of business to own. The moat-and-cash filter is built to find companies whose returns are protected by a structural advantage and harvested as distributable cash. It is not built to see a different archetype: the capital-light operator that earns very high returns on a thin sliver of equity and turns those returns back into more business rather than into a dividend. Molina is that second kind of company. Evaluating it on its own economics — as if the moat-and-cash preference were not held — is the purpose of this chapter.

A high return on very little capital

The pattern is clearest in return on equity. From 2019 through 2024, it ran between 25% and 38%, and it did so while the company spent almost nothing on physical capital — capital expenditure has held around a quarter of one percent of revenue for years.

Source: derived from reported financials, FY2019–FY2025 10-Ks; FY2025 results per the FY2025 Annual Report (Form 10-K) [1].

Sources: FY2024/FY2025 ROE derived from reported financials; capex intensity from the FY2025 Annual Report (Form 10-K), Consolidated Statements of Cash Flows [2].

The reason so little capital produces so much return is structural to health insurance. Molina collects premium in advance and pays medical claims in arrears; the medical claims payable and the amounts owed to states sit on the balance sheet as float, funding the assets in place of shareholder capital. Equity is therefore small relative to the business it supports — the company runs roughly $45 billion of revenue on about $4 billion of equity [3]. A modest margin on that throughput, geared by the float, lands as a high return on equity. This is a genuine form of capital efficiency, and it is the quality a pure cash-yield screen does not price.

The corollary is that Molina compounds by reinvestment, not distribution. It pays no common dividend; retained earnings and the float have carried book value from about $2.0 billion in 2019 to $4.5 billion in 2024, before the 2025 combination of lower profit and continued buybacks pulled it back to $4.1 billion.

Source: derived from reported financials, FY2019–FY2025 10-Ks; FY2025 equity per the FY2025 Annual Report (Form 10-K) [4].

Where the equity is not paid out or bought back, it is redeployed the way the business grows — into the statutory capital that new premium requires, and into acquisitions and contract wins. In 2025 alone the company reported over $9 billion of incremental annual Medicaid premium from new RFP wins, a sole-source Florida contract worth roughly $6 billion in annual premium, and the closed ConnectiCare acquisition [5]. The negative free cash flow that fails the reader's screen is, in part, this reinvestment: writing more premium obliges the company to hold more statutory capital in its regulated subsidiaries, and management has committed to funding that capital as needed [6]. Cash consumed to grow a high-return book is a different thing from cash lost.

Where the return comes from

Breaking the return into its parts shows both the strength and the fault line. Return on equity is the product of net margin, asset turnover, and balance-sheet leverage. For Molina the leverage and turnover terms are stable; the return moves almost entirely with the margin.

Source: derived from reported financials, FY2021–FY2025 10-Ks (net margin = net income / total revenue; turnover = revenue / total assets; multiplier = total assets / equity) [7].

Turnover held near 2.3–2.9 times and the equity multiplier near 3.5–4.6 times across all five years. The 2025 halving of return on equity — from 26.2% to 11.6% — came almost entirely from net margin falling from 2.9% to 1.0%, as the consolidated medical care ratio moved from 89.1% to 91.7% (Margin Reset). The return is set by a spread of one to three cents on the premium dollar, and that spread is what broke.

The advantage that produces the spread is not a moat but a cost position. Molina has described itself since its 2003 IPO as a low-cost operator — the prospectus claimed its "administrative efficiency is among the best in our industry" [8] — and it still frames its vision as distinguishing itself as "the low-cost, most effective and reliable health plan" [9], built on a singular focus on government programs that lets it "identify and implement efficiencies" [10]. The 2025 general-and-administrative expense ratio of 6.6%, down from 6.7%, is the visible edge of that discipline [11]. A low cost of administering each premium dollar is what lets a bidder win state contracts and still clear its return threshold — an operational advantage that stands in for the structural one the reader looks for.

Why the screen is not merely fussy

Looked at fairly, the same evidence explains why the moat requirement is not idle caution in this case. Three features cut against treating Molina as a durable compounder.

First, a return earned without a moat is a return the system can reclaim. Molina's rates are not set by the company; states and CMS set capitation rates that regulators require to be "actuarially sound," with the plan bearing the risk and left to advocate for adequacy [12]. A payer that sets prices to be adequate — not generous — for the incumbent is a mechanism for competing excess returns away over time. The high historical return on equity is what such a system permits when costs run below the rate assumption; it is not a right.

Second, the cost advantage is conditional on scale and reverses when volume falls. The company's own risk disclosure is explicit: a loss of members brings not just lower dollar margins but lower percentage margins, through "loss of cost efficiency or cost leverage, and the resulting stranded administrative costs" [13]. Operating leverage that flatters returns on the way up works in reverse on the way down — and 2026 premium is guided to fall for the first time in years (Revenue Durability).

Third, the reinvestment is only as good as the returns it earns, and those returns are now visibly lower. Redeploying capital at a 26% return on equity compounds value quickly; redeploying it at 2025's 11.6% does not, and the company chose to return $1 billion through buybacks in each of 2024 and 2025 rather than fund still more growth (Cash and Capital). The float that finances the model also traps it: statutory rules limit how much of the subsidiaries' earnings can ever reach the parent to pay stockholders [14].

A calibrated read

The deprioritized angle earns its hearing. Molina is a real capital-light compounder — high returns on a thin equity base, near-zero capital intensity, value created by reinvestment and buybacks rather than a coupon. Judging it on free-cash-flow yield alone understates a business whose negative cash flow is, in material part, the cost of holding statutory capital against a growing, high-return book. An investor who did not require a moat could reasonably own the compounding.

The same look, though, supports the core of the reader's caution rather than dissolving it. The returns are earned without protection, so the rate-setting mechanism and periodic rebids can reclaim them; the cost edge depends on scale that is now contracting; and 2025 is the live demonstration — a 260-basis-point move in the medical care ratio cut return on equity by more than half in a single year. That fragility is exactly what a durable-moat, steady-cash filter is designed to avoid.

What would tilt the read toward the compounder case: return on equity re-normalizing toward the low-20s as the spread recovers, on the roughly 2–3% pre-tax margin management now targets, with the ratio's stability restored across quarters. What would vindicate the screen: return on equity settling near 2025's low-teens level and staying there, confirming that a rate-taker's returns, however high in good years, are not the durable cash-generative kind — in which case a filter that passed on Molina will have been right to.