Should I use multiples or a DCF?

Both, for different jobs. The multiple is the market's answer and is hard to argue with directionally; the DCF makes your assumptions explicit and tests them. For SMEs the multiple usually leads and the DCF cross-checks. If the two disagree by more than ~30%, find out which assumption drives the gap before negotiating.

What data do I need to value a private company?

Minimum: three years of financial statements (P&L and balance sheet), the debt schedule, and an honest picture of owner compensation and one-offs. For Czech companies the statements are in the public commercial registry (Sbírka listin) — Misano pulls them automatically from the IČO.

What discount applies to a small private company versus public comps?

Stacked, not single: a size discount (structurally lower multiples for smaller EBITDA), an illiquidity discount (order of 15–30%), and a basis adjustment (private control transactions are priced pre-synergy). Together, comparing a €3m-EBITDA Czech company directly to US listed peers can overstate value by a factor of two.

What is total beta and when do I use it?

Damodaran's correction for undiversified owners: market beta divided by the correlation of the industry with the market. It prices the total risk a concentrated private owner actually bears, not just the market risk a diversified fund bears. Use it when valuing for a buyer or seller whose wealth is concentrated in the business — it materially raises the cost of equity and lowers the DCF value.

What terminal growth rate should I use?

At or below the risk-free rate, never above it — perpetual growth above the economy's nominal rate is a contradiction. Misano's engine hard-caps g at Rf. If your valuation only works with g above inflation plus real GDP growth, the valuation doesn't work.

How to value a private company: multiples and DCF, the Damodaran way

Almost every private-company valuation that holds up rests on two legs: a market multiple applied to normalized earnings, and a discounted cash flow as the intrinsic cross-check. Everything else — asset values, rules of thumb, 'strategic value' — is either a special case or a negotiating position. This guide walks through both legs the way Misano's valuation engine actually computes them.

Leg one: EV/EBITDA on normalized earnings

The mechanics are three lines. Enterprise value = normalized EBITDA × the sector multiple. Equity value = enterprise value − net debt. The price of a 100% stake is the equity value; a minority stake takes a further discount. The judgement hides in two words: 'normalized' and 'the multiple'.

Normalizing EBITDA means restating it as a rational buyer would run the business: market-rate compensation for the owner (many SME owners pay themselves too little or too much), one-off items removed (a property sale, a COVID subsidy, a lawsuit), related-party contracts repriced to market. In small companies the normalization adjustment is routinely 20–50% of reported EBITDA — larger than the disagreement about the multiple.

The multiple comes from the sector, and within the sector from quality. Misano publishes the full benchmark it works with — EV/EBITDA and EV/Sales bands (low / mid / high) for 82 NACE divisions and 62 curated niches at misano.ai/multiples. Mid is an average solid company in the sector; low is small, owner-dependent or cyclical; high is growing, recurring-revenue, defensible. Positioning the company inside the band is the valuation judgement — the band itself shouldn't be a matter of taste.

The discounts that make private companies different

Size: a business with €2m EBITDA trades at a structurally lower multiple than the same business with €50m EBITDA. Small-cap risk is real and priced.
Illiquidity: a private stake cannot be sold this afternoon. Depending on the study and the situation this costs on the order of 15–30% of value.
Basis: benchmark bands from private control transactions are pre-synergy control prices. Comparing an SME to liquid US public comps without these adjustments is the most common valuation error in practice.

Leg two: DCF — FCFF at WACC, parameters from Damodaran

The DCF answers a different question: not 'what does the market pay for businesses like this' but 'what are these specific cash flows worth'. Misano's implementation (Metodika C1.0) is deliberately orthodox corporate finance: free cash flow to firm, discounted at WACC, terminal value by Gordon growth.

01Project FCFF: revenue growth → EBITDA margin → subtract cash taxes, capex, and working-capital investment. Sector defaults for margins, capex/sales and NWC/sales come from Damodaran's industry datasets (Jan 2026 vintage, Europe with global fallback).
02Build the discount rate via CAPM: cost of equity Ke = Rf + β·ERP + CRP. The beta starts as Damodaran's unlevered industry beta, re-levered to the company's debt. For an owner whose wealth is concentrated in the company, use total beta — Damodaran's correction for undiversified private owners, which can double the cost of equity versus a diversified investor.
03WACC = equity and after-tax debt costs at target weights. A Czech SME in 2026 typically lands in the low-to-mid teens; a WACC below 10% for a small private company should make you suspicious of your own inputs.
04Terminal value by Gordon: TV = FCFF × (1+g) / (WACC − g), with g capped at the risk-free rate — a company cannot outgrow the economy forever. Misano hard-caps g at Rf in code.
05Sanity checks before you believe the number: what share of value sits in the terminal value (over ~80% means the projection carries almost no information), what exit multiple the TV implies (compare it to the sector band), and whether steady-state capex ≈ depreciation.

Reconciling the legs: value is a range

A defensible valuation is not a number, it is a range with a story: the multiples band says what the market pays, the DCF says what the cash flows justify, comparable transactions anchor both against reality. Misano draws all legs side by side against the asking price — the 'football field' — and the interesting cases are exactly the ones where the legs disagree, because the disagreement tells you which assumption carries the value.

For a Czech company the whole exercise starts from a company ID: Misano pulls three years of financial statements from the commercial registry by IČO, normalizes the numbers, applies the NACE or niche band, builds the DCF with Damodaran parameters, and writes up the thesis. Minutes, not days.