Constructing an Intangible Asset Based Value Strategy with Kai Wu

Episode Summary

Executive Summary: Kai Wu argues that markets increasingly misprice intangibles—brand, human capital, network effects, and IP—because accounting undercaptures them. Sparkline uses NLP, social-network data, patents, and hiring data to quantify these assets, then applies a value framework by ranking companies on intangible value relative to price. The result is an active, tax-efficient ETF that aims to identify modern moats at reasonable valuations.

Main Topics: The rise of intangible assets in the economy (Priority: 5/5): Wu explains that modern value creation has shifted from tangible capital to intangibles such as software, brands, talent, and platforms, making traditional value metrics less informative. Why accounting misses modern sources of value (Priority: 5/5): He argues that accounting treats intangible investment inconsistently or omits it entirely, reducing the explanatory power of book value and earnings for market value. How Sparkline measures the four pillars of intangibles (Priority: 5/5): The discussion walks through Sparkline’s framework for intellectual property, brand equity, human capital, and network effects, using machine learning and alternative data to score companies. Combining intangibles with value investing (Priority: 5/5): Wu emphasizes that the goal is not to buy the most intangible-rich companies, but the cheapest companies relative to their intangible strength, creating an intangible-based value strategy. Portfolio construction and implementation (Priority: 4/5): He describes the ETF’s universe, stock count, weighting approach, industry exposure, and rebalancing logic, highlighting trade-offs between concentration, diversification, and benchmark tracking. Launching and operating the ETF (Priority: 3/5): Wu shares lessons from running an ETF, including operational simplicity, liquidity mechanics, and tax efficiency, arguing the structure is well suited to active factor strategies. Future data sources and strategy evolution (Priority: 3/5): He points to improving data on consumer behavior, employee quality, lobbying, and venture capital as ways to enhance intangible measurement in the future.

Key Arguments: Modern markets are increasingly driven by intangibles, so classic tangible-heavy metrics like book value miss a growing share of economic value. Accounting rules expense many intangible investments—like R&D and brand building—rather than capitalizing them, which distorts valuation analysis. Brand, human capital, network effects, and intellectual property can be measured with alternative data and NLP, even if imperfectly. A company’s intangible strength should be evaluated relative to price, not in absolute terms; the strategy seeks undervalued intangibles, not just high-intangible firms. Cross-sectional normalization and pillar-level aggregation help combine many correlated signals into a usable ranking system. The strategy is intended to be universal and adaptable across market caps, sectors, and eventually geographies or asset classes. ETF structure provides tax efficiency and liquidity advantages that can make active factor investing more scalable than traditional funds.

Data Points: Intangible share of corporate balance sheet/market cap: roughly 50% to 80% today - Wu says intangibles have risen from near zero in 1980 to a dominant share of value today. R-squared of book value and earnings explaining market cap: about 90% in 1950, less than 50% today - He cites research from The End of Accounting to show declining explanatory power of traditional accounting. Four pillars of intangibles: 4 - Sparkline’s framework: brand, human capital, network effects, and intellectual property. Patent data history: back to 1790 - Wu notes patents offer the longest historical time series among the data sources discussed. Investor universe size: top 1,000 largest U.S. stocks - The discussed ETF strategy starts with the Russell 1000 universe. Portfolio holdings: 150 names - Wu says the strategy holds the top 15% of the universe after ranking. Portfolio turnover: around once per year - He says the blended intangible fundamentals lead to roughly annual full portfolio turnover. Intangible-heavy market share: 80% of S&P market cap - Wu states intangible-intensive businesses represent about 80% of the S&P’s market value and are growing. Platform company share: over half, around 60% of S&P market cap - He says platform companies now make up a majority of market cap. PhD hiring signal: more PhDs, especially from top programs, tends to outperform - He says firms with higher PhD intensity have shown better factor-adjusted results. MBA hiring signal: underperforms - Wu mentions an unpublished result that firms hiring lots of MBAs underperform. Monthly turnover of positions: 5% to 10% - Used as an example of gradual portfolio turnover, even with monthly trading.

Pivotal Quotes: "The four largest companies today by market value do not need any net tangible assets." — Kai Wu: Used to illustrate how value creation has shifted from physical assets to intangibles. "It’s not how many patents you have, it’s about owning innovative patents." — Kai Wu: Explains that Sparkline measures quality and relevance of IP, not just quantity. "What I want to do is I want to own companies with strong brands, strong human capital, network effects, and IP obtained at a reasonable price." — Kai Wu: Summarizes the strategy’s core philosophy: intangible strength plus value discipline.

Implications: Investors should expect more value to reside in intangible-rich firms and more old-school valuation methods to mislead. The future likely favors strategies that combine alternative data, ML, and valuation discipline to capture modern moats efficiently.