Kyle Harrison
concept

EBITDA

EBITDA

Earnings before interest, taxes, depreciation, and amortization — a profitability proxy that recurs across Kyle’s investing reading, almost always under scrutiny. The dominant frame in his notes is skeptical: Warren Buffett in the Berkshire Hathaway Annual Letters writes that references to EBITDA “make us shudder — does management think the tooth fairy pays for capital expenditures?”, calling it “bad terminology” that asks you to “unthinkingly accept concepts that are dangerously flawed.” Kyle annotates these passages “Evils of EBITDA.”

The Messy Marketplace (Brent Beshore) makes the practical version of the same point: “EBITDA is a proxy, but far from a perfect one. All businesses require reinvestment merely to maintain market position. Therefore, EBITDA never represents cash flow to owners” — which is why Beshore’s adventur.es prefers Owner Earnings (pre-tax profit less capex, operating interest, and normalized owner comp) as its valuation base. EBITDA still functions as the lingua franca of small-business pricing: traditional firms “routinely sell for EBITDA multiples of 4X to 5X,” and earnouts can be tied to it. On the buyout-economics side, Norwest is referenced via a proprietary deal data point — ~$3M+ EBITDA at first call, ~$7M average at close — anchoring the Private Equity frame.

Context: EBITDA is a non-GAAP earnings measure that strips out interest, taxes, depreciation, and amortization to approximate operating profitability; it is widely used in valuation and deal multiples but criticized for ignoring real capital-reinvestment needs.

Where this appears

  • Berkshire Hathaway Annual Letters — Buffett’s repeated attacks on EBITDA as misleading “bad terminology”; Kyle tags “Evils of EBITDA” and “Community-adjusted EBITDA”
  • The Messy Marketplace — flagged as an imperfect proxy that “never represents cash flow to owners”; small firms sell at 4–5X EBITDA; earnouts can be structured on it
  • Norwest — buyout-economics reference point (deal-size EBITDA thresholds, time-to-close)