Skip to main content

High-ROIC Compounders

Updated Jun 24, 2026 at 2:35pm

Research Draft High 1,164 words

A high-ROIC compounder is a business that earns a return on invested capital well above its cost of capital and can reinvest a large share of its earnings back into the business at those same high rates — so intrinsic value snowballs over many years. The core tension is twofold: high returns on capital invite competition and tend to fade toward the cost of capital over time (mean reversion), and even a genuine compounder only rewards shareholders if it has a long reinvestment runway. A wonderful business that throws off cash but cannot redeploy it at high rates is a "cash cow," not a compounder. The whole thesis lives or dies on durability and reinvestment, not on a single year's ratio.

How it's calculated

The defining metric is Return on Invested Capital (ROIC):

> ROIC = NOPAT ÷ Invested Capital

  • NOPAT (net operating profit after tax) = EBIT × (1 − tax rate). It is pre-financing, so it reflects operating economics independent of capital structure (Wall Street Prep).
  • Invested Capital is the net assets the business needs to generate NOPAT. Two equivalent constructions: (a) operating approach — net working capital + net fixed assets + acquired intangibles/goodwill; or (b) financing approach — total debt + total equity − excess cash − non-operating assets (Wall Street Prep; Corporate Finance Institute).

Value is created only when ROIC > WACC (weighted average cost of capital); when ROIC < WACC, growth destroys value (CFI). The compounder math is then driven by two levers: the spread (ROIC − WACC) and the reinvestment rate (the fraction of NOPAT plowed back). Sustainable earnings growth ≈ ROIC × reinvestment rate. A business at 25% ROIC reinvesting 60% of earnings grows intrinsic earnings power ~15% a year — that is the compounding engine.

Note ROIC is distinct from look-alikes: ROE is levered and distorted by buybacks/debt; ROA uses total (not invested) capital; gross profitability (Novy-Marx) is a related academic quality proxy but not the same construct.

How it's used in practice

Practitioners (Buffett, Munger, Terry Smith, Mauboussin's framework) use the concept as a screen-then-judge process:

1. Screen for sustained high ROIC — commonly cited rules of thumb treat >15% as high-quality and >25% as elite, and demand multi-year consistency, not a single peak year (Sure Dividend; Compounding Quality). These thresholds are heuristics, not validated cutoffs. 2. Diagnose the moat — high ROIC is the symptom; the analysis is why it persists (brand, switching costs, network effects, scale, low-capital business models). Asset-light, low-capex models tend to show the highest ROIC. 3. Assess the runway — can the company reinvest at high rates for years? This is the single most important and hardest judgment. 4. Hold, don't trade — the thesis is multi-year. Munger's logic: over decades a stock's return converges on the business's return on capital, so a high-ROIC compounder bought at a fair (even "expensive-looking") price beats a mediocre business bought cheap (USC 1994 talk, widely cited).

This is a long-horizon, fundamentals-driven philosophy — not a trading signal.

Adoption, debate & evidence

The concept is mainstream in quality/growth investing and is the explicit framework of Counterpoint Global (Mauboussin) and managers like Fundsmith. The supporting academic pillar is the profitability/quality factor: Novy-Marx (2013) showed gross profitability predicts the cross-section of returns with power comparable to book-to-market, and AQR's Quality-Minus-Junk (Asness, Frazzini, Pedersen) found high-quality firms earn higher risk-adjusted returns and, notably, that the market underpays for quality (Novy-Marx; NBIM Quality Factor note). Important nuance: Novy-Marx himself argues profitability is the real driver and that other "quality" metrics mostly proxy for it.

The central honest caveat is mean reversion. Mauboussin's "Death, Taxes, and Reversion to the Mean" (2007) found ROIC strongly reverts toward the cost of capital, and warned much of the apparent pattern is statistical noise in a high-randomness system. Yet persistence beyond chance does exist: of companies starting in the top ROIC quintile, 41% were still there nine years later (vs. ~20% expected by chance) — though fewer than half of those (under ~4% of the whole sample) stayed top-quintile every year, so the in-between path is highly volatile — with business-model differences the most promising explanation (Mauboussin summary, Hedge Fund Alpha; Greenbackd). So: persistence is real but the base rate is against any given high-ROIC firm staying high — the edge is in identifying the durable minority.

Folklore vs. measured: the slogan "high ROIC = great stock" is incomplete. The factor evidence supports profitability, but the compounder thesis adds the unproven-at-the-firm-level bet that a specific company defies fade and sustains reinvestment.

Strengths & limitations

Works when: the moat is genuine and the reinvestment runway is long; ROIC is high and stable across a cycle; the price paid doesn't already discount decades of perfection.

Fails when:

  • ROIC fades (the base-rate outcome) — paying a premium multiple for reversion-to-mean economics is the classic trap.
  • No runway: high ROIC but nowhere to reinvest → it's a cash cow; returns then depend heavily on payout and entry price.
  • Accounting distortions: goodwill/acquired intangibles inflate invested capital (understating ROIC); conversely, expensed R&D/brand-building (not capitalized) can overstate ROIC for asset-light firms. ROIC is sensitive to definition — always check how invested capital was built.
  • Valuation indiscipline: the #1 misuse is treating "high quality" as license to ignore price. Munger's "even if you pay an expensive-looking price" is frequently quoted to justify overpaying; it has limits, and a high-quality business at a bubble multiple can still deliver poor returns for a decade.

Sources

Dispute flagged: ROIC persistence is genuinely contested — Mauboussin shows mean reversion dominates and much apparent persistence is statistical noise, while a real minority of firms persist. Reinvestment-runway and accounting-definition risks are the main practical caveats. Threshold numbers (15%/25%) are popular heuristics, not validated.