Dan Rasmussen
About
Founder of Verdad Advisers, author of 'The Humble Investor', value investor and researcher on private equity and capital allocation
Cast within
No topic-region cast yet — this appears once Dan Rasmussen's compiled claims are aligned into a topic region's argument tree.
Claims by Dan Rasmussen (20 of 72)
Being a 'futurist' (crediting speculative long-term technology predictions, as one might from reading popular science magazines and betting on speculative innovations) has been a profitable investment strategy for the past 10-15 years, while skepticism about valuations or technology hype has been consistently wrong, creating path-dependent conditioning where investors assume speculation will continue to be rewarded.
AI is the first technology innovation since fiber to be highly capital-intensive, with big tech companies having shifted from 1/3 the capital intensity of traditional US industrial companies to 3x the capital intensity, creating risks where depreciation schedules (5-year vs 1-year Nvidia server lives) materially impact valuations and could result in massive losses if capex intensity proves unsustainable.
US small-cap value returns have been severely dragged down by concentrated exposure to three underperforming sectors: (1) energy (which peaked in 2014-15 then collapsed and has been a 'wasteland' as shale fracking companies appear to be 'capital incineration machines'), (2) regional banks (which have not reached sufficient scale through M&A and faced a major crisis with First Republic), and (3) biotech (which is down ~60% from peaks and is 'annihilated'), making US small-cap value a poor equity exposure unlike international small-cap value which is more diversified.
Private equity has been the 'darling of investor eyes' for the past decade, with surveys showing that 90%+ of institutional investors believe PE will outperform public equity by a median of 200 basis points per year net of fees—representing correlated bullish beliefs with no differentiated view.
Institutional investors including pension funds (15% allocation), college endowments (30% allocation), and elite college endowments (40% allocation) are massively overallocated to private equity relative to its actual market size, and this overallocation is unjustified given that PE companies are the smallest, lowest-margin, lowest-quality businesses with the highest debt loads.
The Yale Model of endowment investing, pioneered in the 1980s by David Swenson and others, was originally a rational response to the 1970s stagflation crisis that destroyed both stock and bond returns, and it did generate outperformance through venture capital and private equity in early decades when these markets were underpenetrated and inefficient (PE traded at 40% discounts to public comps), but this advantage has disappeared as trillions of capital have flowed in.
PE distributions have collapsed from a historical ~30% of NAV per year to approximately 10% of NAV currently, which is the lowest level since 2008, despite the fact that stock markets are reaching all-time highs and the economy is not in recession, indicating severe distress in PE exit channels.
PE volatility is artificially suppressed in reported NAVs due to illiquidity and mark-to-fantasy accounting, but publicly-listed PE funds on the London Stock Exchange (like Harbourvest funds) reveal that true PE volatility is approximately 24% annualized, driven by discounts to NAV that vary with investor sentiment, making PE roughly as volatile as the Russell 2000 (small-cap) index with a beta of ~1.6.
Institutional investor sentiment toward PE has shifted dramatically in the past 7-8 months from euphoria (where PE was 'the apple of their eyes' and all conversation focused on new PE fund commitments and co-invest deals) to concern, with endowments and large investors now openly discussing being 'overallocated' to PE—a statement Rasmussen has never heard from them before.
Small-cap value in the US worked well from COVID through the release of ChatGPT (approximately 2020-2022), but has not worked since, and appears to only work during early-stage cyclical recoveries when cyclical stocks outperform, suggesting it has become a tactical cyclical play rather than a persistent value premium.
AI queries incur massive marginal energy costs, making AI a worse business model than advertising-based or subscription-based software (like Salesforce), because Salesforce subscriptions have zero marginal cost and don't require capex recovery, while AI providers must spend significant capex and recoup it through per-query revenue.
Learning and feedback loops in investing are more effective with frequent feedback (monthly returns) than with infrequent feedback (annual or longer cycles), because investors with more cycles in their career (more learning events) develop better intuition about what actually predicts returns.
It's possible that both the bullish narrative on tech/AI and the skeptical narrative are simultaneously true: tech innovations are genuinely transformational, but valuations may also be excessive and result in poor forward returns, creating a scenario where innovation thrives but investors suffer losses.
My Notes
Loading notes...