Aswath Damodaran's Session 22: Pricing Closure and Asset-based Valuation/Pricing: skim's analysis identifies 12 key moments. This finance lecture concludes relative pricing by demonstrating how to price young, money-losing firms and how to run market-wide multiple regressions while managing multicollinearity.
Category: Education. Format: Educational. YouTube video analyzed by skim.
skim AI Analysis
Credibility assessment: Rigorous Empirical Finance. The lecture draws on concrete market regression data, historical case studies like Amazon in 2000, Twitter in 2013, and United Technologies, and explicit accounting standards such as FAS 157. Claims about multiples and valuation mechanics are grounded in reproducible statistical methods and financial theory.
Bias assessment: Intrinsic Value Pragmatist. The speaker favors discounted cash flow analysis over relative pricing and openly criticizes accounting rules like FAS 157 for conflating price with intrinsic value. However, he remains pragmatic, teaching students how to execute relative pricing effectively because equity research and banking jobs demand it.
Originality: 85% — Original Empirical Models. Rather than repeating textbook formulas, the lecture presents original market-wide regressions across thousands of equities, custom valuation case studies for Amazon, Tesla, Twitter, United Technologies, and General Electric, and firsthand accounts of consulting with GAAP standard-setters.
Depth: 92% — Deep Quantitative Mechanics. The analysis dissects the statistical pitfalls of financial regressions, including multicollinearity and non-linearity, reconciles divergent intrinsic and relative valuations, and walks through three distinct methodologies for sum-of-the-parts valuation using divisional financial data.
Key Points (12)
1. Trade-Offs in Valuing a Company by Its Pieces
Timestamp: 00:02:50 to 00:06:28
Valuing a multi-business company like Disney in separate pieces increases precision by allowing distinct stories for each division, but demands far more granular data on margins, capex, and working capital. Most companies are worth more aggregated than broken up, or else markets would split them.
Significance (Medium): Frames the economic trade-off between analytical precision and data availability when deciding whether to model a conglomerate as a single entity or separate units.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
2. Pricing Dot-Com Stocks and Reconciling Amazon
Timestamp: 00:06:33 to 00:15:10
In 2000, regressing dot-com price-to-sales ratios against net margins yielded a 4% R-squared, whereas regressing against log revenues, revenue growth, and cash-to-revenues explained 32% of variation. Amazon looked overvalued intrinsically at $84 versus $35, yet underpriced at 25.4x sales versus a 30.42x predicted multiple.
Significance (High): Demonstrates how a stock can be simultaneously overvalued on cash flow fundamentals and underpriced relative to sector peers, making investment horizon the deciding factor.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
3. Adjusting Forward Multiples for Time and Risk
Timestamp: 00:15:10 to 00:19:40
When pricing young companies with negative current earnings using year-10 forward multiples, analysts cannot compare future terminal values directly to today's market cap. In a Tesla valuation 10 years ago, a $69 billion year-10 terminal value shrank to $8 billion in present equity value after discounting for risk, dilution, and failure risk.
Significance (High): Exposes a widespread equity research error where analysts overvalue growth stocks by failing to discount forward multiples for cost of capital, cash burn dilution, and survival odds.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
4. Pricing Twitter's 2013 IPO by User Count
Timestamp: 00:19:40 to 00:24:00
At Twitter's October 2013 IPO, intrinsic valuation yielded $18 per share, but across online advertising peers, user count explained 98% of market cap variation at roughly $100 per median user. Multiplying Twitter's 240 million users by $100 produced a $24 billion pricing that matched its first-day IPO market cap.
Significance (High): Illustrates how market pricing relies on empirical correlation rather than theory, evolving from raw user counts to engagement minutes and finally to revenues as a sector matures.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
5. Market-Wide PE Regressions and Multicollinearity
Timestamp: 00:24:00 to 00:33:30
Regressing start-of-2025 US stock PE ratios against earnings growth, payout ratio, and beta yields a 24% R-squared, with a 16.09 base constant and +0.52 PE per 1% growth. However, beta shows a counterintuitive positive coefficient because high-growth firms have high betas, causing multicollinearity where beta acts as a proxy for growth.
Significance (High): Explains why financial regressions often produce wrong-signed risk coefficients while remaining usable as a complete package to predict company multiples.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
6. Applying Market Regressions to Disney and PEG
Timestamp: 00:33:31 to 00:41:32
Plugging Disney's growth, beta, and payout into the start-of-2025 US market regression predicts a PE of 34 versus its actual 40x multiple, signaling it is overpriced relative to the market. Separately, tracking the PE growth coefficient against the equity risk premium reveals market regime shifts, while PEG regressions require natural logs of growth.
Significance (Medium): Connects classroom multiple regressions to quantitative hedge fund strategies, showing how predicted multiples identify relative mispricing across global regions.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
7. Four Steps to Select and Apply Pricing Multiples
Timestamp: 00:41:32 to 00:46:08
Proper relative pricing follows four steps: choosing between equity and enterprise value based on sector and leverage differences, selecting a denominator scalar, matching timing consistently across all peers, and defining the comparable group. Analysts should choose whichever multiple retains the largest sample size and yields the highest regression R-squared.
Significance (Medium): Provides a practical checklist for equity research and banking roles to prevent inconsistent multiple definitions and sample bias.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
8. Separable Assets in Sum-of-the-Parts Valuation
Timestamp: 00:46:08 to 00:50:27
Asset-based or sum-of-the-parts valuation can disaggregate a firm by geography, business unit, or subscriber base, such as valuing a US Netflix subscriber at five times an Indian subscriber who pays 200 rupees ($2.50) monthly. The method works best when assets are separable with standalone cash flows and active markets, like real estate properties.
Significance (Medium): Clarifies why sum-of-the-parts valuation is straightforward for standalone real estate portfolios but difficult for integrated businesses like Disney.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
9. Liquidation Discounts and FAS 157 Fair Value Flaws
Timestamp: 00:50:27 to 00:54:44
Liquidation valuation requires pricing individual assets with a discount for urgency, which for a distressed company like Weight Watchers yields only a fraction of balance sheet book value. Meanwhile, GAAP's FAS 157 defines fair value accounting as an arm's-length market participant price, yet forces accountants to reverse-engineer DCF models to justify a pricing number.
Significance (High): Highlights the conceptual confusion in FAS 157 fair value accounting, arguing that forcing DCF models onto market pricing tasks creates distorted valuations.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
10. United Technologies Sum-of-the-Parts Case Study
Timestamp: 00:55:39 to 01:06:38
Valuing United Technologies across its six divisions yields $61.7 billion under naive median EV/EBITDA pricing, $74.2 billion under regression-adjusted pricing, and $80.25 billion via divisional DCF models. Subtracting $4.6 billion for the present value of $48 million in unallocated headquarters expenses leaves $75.7 billion, beating its $71 billion aggregated value and $52 billion market cap.
Significance (High): Shows step-by-step how to execute rigorous sum-of-the-parts pricing and DCF valuation while accounting for unallocated corporate overhead and divestiture catalysts.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
11. How GE Capital Delayed GE's Breakup by a Decade
Timestamp: 01:06:54 to 01:10:30
In a 2018 sum-of-the-parts valuation of GE's eight businesses totaling $204 billion intrinsically ($212 billion priced), GE Capital was worth $27 billion but carried $84 billion in debt, draining $57 billion in equity value. Because GE Capital financed customers of GE Aviation and other units, shutting it down was impossible and delayed GE's breakup for nearly a decade.
Significance (High): Warns corporate strategists and investors that captive finance arms create deep operational entanglements that can turn into toxic liabilities during a downturn.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
12. Why Private Company Valuation Depends on Motive
Timestamp: 01:10:30 to 01:17:25
Valuing private businesses is complicated by the lack of a market price for cost-of-capital weights, betas, and sanity checks, alongside shorter histories, non-standard accounting, mixed personal expenses, and unpaid owner salaries. Furthermore, private valuation depends heavily on motive—whether for divorce, taxes, or sale—and whether the buyer is an individual, private equity firm, or corporation.
Significance (High): Establishes that private company valuation cannot produce a single blanket number without adjusting for owner compensation and the specific buyer's risk profile.
Sources in support: YouTube: Aswath Damodaran (NYU Stern Valuation Course Lecturer)
This analysis was generated by skim (skim.plus), an AI-powered content analysis platform by Credible AI. Scores and classifications represent the platform's AI-generated assessment and should be considered alongside other sources.
