Valuing Young, Start-up and Growth Companies

Author: Aswath Damodaran (Stern School of Business, New York University) · Published: May 2009 (working paper, 67 pages) · URL: https://pages.stern.nyu.edu/~adamodar/pdfiles/papers/younggrowth.pdf

One-line: Young companies are hard to value, but the standard venture-capital shortcut (forecast a few years out, apply an exit multiple, discount at a 50-70% target rate) hides that difficulty instead of dealing with it. Value them properly and handle failure risk explicitly rather than burying it in the discount rate.

Summary

Full title: Valuing Young, Start-up and Growth Companies: Estimation Issues and Valuation Challenges. Damodaran’s claim is that the tools analysts use for young companies either fail or “yield unrealistic numbers,” and that “the venture capital approach to valuation that is widely used now is flawed and should be replaced.”

Why young companies are different. He places them on the early life cycle: idea companies (no revenue, operating losses), start-ups (small revenue, growing losses) and second-stage companies (growing revenue, heading toward profit). What they share: almost no history; little or no revenue and operating losses; dependence on private equity (founders, friends and family, then VCs); multiple classes of equity with different claims on cash flow and control; illiquidity; and a high chance of failure. For that last point he cites Knaup and Piazza’s Bureau of Labor Statistics data on businesses founded in 1998: only 44 percent survived four years and 31 percent survived seven. In the information sector (which includes technology), only about 25 percent made it to year seven, against almost 44 percent in health services.

Why the standard tools break. Intrinsic (DCF) valuation struggles with all four of its inputs:

• Existing assets are tiny, and the costs of chasing future customers are mixed in with operating costs, so SG&A can run three or four times revenue.
• Growth assets carry most of the value, but with no revenue history analysts fall back on management’s forecasts, “with all the biases associated with these numbers,” and current return on capital is usually negative.
• Discount rates normally come from market prices (betas from regressions, bond yields), which don’t exist for a private company. The owners (fully invested founders, partly diversified VCs) will also want to be paid for firm-specific risk, not just market risk.
• Terminal value can be 90 percent, 100 percent or “even more than 100%” of today’s value. That rests on three hard questions: will the company survive to stable growth, when, and what will it look like then?

Relative valuation doesn’t get around any of this. There’s nothing sensible to scale to (earnings are negative, book value is tiny, revenue is minuscule), the comparable young companies aren’t traded, and the questions of survival, differing equity claims and illiquidity all remain.

“The dark side of valuation.” He names five shortcuts: forecasting only revenue and earnings with nothing in between; cutting forecasts off at three to five years; bolting a relative-valuation exit multiple onto an intrinsic valuation; loading every uncertainty, failure included, into the discount rate; and making ad hoc adjustments for differences in equity claims. All five come together in the venture capital method: (1) forecast earnings or revenue two to five years out, to the planned exit; (2) apply a current public-company P/E or EV/sales multiple; (3) discount at a target rate of return; (4) add the new money to get post-money value, and divide the investment by post-money value to get the investor’s stake.

His table of typical VC target rates: start-up 50-70%, first stage 40-60%, second stage 35-50%, bridge/IPO 25-35%. The rates are that high because they have survival risk built in. Actual returns show it: as of 2007, early/seed VC earned 21.4 percent a year over 20 years (“well below the 50-70% target returns”), and all VC earned 16.9 percent, against 9.2 percent for the NASDAQ and 8.0 percent for the S&P over the same period.

His four objections to the method:

1. Because value rises with projected revenue, founders inflate forecasts and VCs push them down, so “the projected value becomes a bargaining point between the two sides rather than the subject of serious estimation.”
2. A multiple three years from now depends on the cash flows after that, so “not estimating those cash flows … does not mean that the uncertainty has gone away.”
3. An equity target rate is wrong for an enterprise-value multiple, and building failure into one fixed rate implies the risk never changes as the company matures.
4. Adding new capital to pre-money value is only right if the money stays in the company; if some of it cashes out existing holders, that portion shouldn’t be added back.

His running example, a pre-revenue antivirus start-up called Secure Mail, shows the method at work: $300 million of projected year-3 revenue at a 2.275x EV/sales multiple (the average of Symantec and McAfee), discounted at 50 percent, gives a $202 million pre-money valuation and 12.91 percent for a $30 million investment.

“The light side of valuation”: doing it properly. Most of the paper is a step-by-step replacement:

• Cash flows. Forecast either top-down (total market, then market share, then a target operating margin taken from mature peers with a “pathway to profitability,” then the reinvestment needed to grow) or bottom-up from capacity (his second example is Healthy Meals, a one-owner storefront business). Market definition drives everything: Amazon in 1998 counted as a book retailer faced a market under $10 billion. Owners naturally forecast optimistic growth and margins, but neither is “delivered for free,” so the reinvestment behind the growth has to be modelled. On detail, his rule is that “in valuing young companies, less (detail) is often more (precision).”
• Discount rates. Use sector-average betas from public companies, then adjust for how diversified the owner is with a total beta (market beta divided by the correlation with the market). For Secure Mail: 1.20 / 0.40 = 3.00 for an undiversified founder, falling to 2.40 once a VC holds it and 1.60 as the investor base widens. Assume the cost of capital falls as the company matures.
• Survival, handled separately. Value the going concern, then weight it: expected value = going-concern value × (1 − probability of failure) + distress-sale value × probability of failure. Estimate the probability from sector survival averages, a probit model or a simulation. Secure Mail’s $177.56 million going-concern value, at a 40 percent failure probability and no salvage value, becomes $106.54 million. He also shows a key-person discount: value the business with and without the founder.
• Equity claims and illiquidity. Value preferred rights, control and anti-dilution protections instead of applying rules of thumb, and make illiquidity discounts depend on the company rather than using the common fixed discount taken from restricted-stock studies.
• Relative valuation, done right. Use forward revenue or earnings, adjust the multiple for the company’s characteristics at the forecast date, discount back at a rate that falls over time, and then adjust separately for survival, lack of diversification and illiquidity.
• Real options, used sparingly. An option to expand is worth a premium only with exclusivity, meaning learning that competitors can’t use, such as Microsoft’s control of the operating system when it built Office, or Apple’s brand from the iPod before the iPhone. Otherwise analysts “double count the value of growth.” Secure Mail’s option to build a database product, valued with Black-Scholes, is worth about $56 million.

Conclusion. Short histories, operating losses and a high probability of failure push analysts toward “forward multiples and arbitrarily high discount rates.” The better process asks for inputs that are hard to pin down, but it is still worth doing because it forces you to “confront the sources of uncertainty, learn more about them and make our best estimates.”

Full text

Archived privately against link rot: the original PDF is younggrowth.pdf (67 pages), with a searchable text extraction at younggrowth.md.

Connections

• Aswath Damodaran: the author. The paper sets out formally his later public argument that VCs “play the pricing game” instead of valuing businesses.
• Valuation: the most complete treatment in the wiki of valuing a company with no earnings, no history and a real chance of going to zero.
• Valuing High-Tech Companies: McKinsey’s shorter case for the same position (DCF over multiples, start from the future market). Damodaran adds explicit survival weighting, total beta and the valuation of separate equity claims.
• How Startup Valuation Works — Illustrated: the founder’s guide to the comparables-and-target-return method this paper criticizes.
• Discounted Cash Flow: the method rebuilt for young companies. On his account, terminal value can be more than 100 percent of today’s value.
• Venture Capital Returns: the gap between the 50-70% target rates and the 21.4% early-stage VCs actually earned over 20 years.
• Survivorship Bias: why failure has to be priced explicitly. Only about a quarter of 1998 information-sector businesses survived seven years.
• Total Addressable Market: the top-down cash-flow method starts from market size, and the Amazon book-retailer example shows how much the market definition matters.
• Dilution: pre-money and post-money math, and the reminder that secondary proceeds don’t belong in post-money value.