Measuring Corporate ‘Dark Matter’ With an ETF

Episode Summary

Executive Summary: The episode examines Kai Wu’s Sparkline Intangible Value ETF and the case for treating intangibles—IP, brand, human capital, and network effects—as a modern source of value. The discussion argues traditional book-value metrics miss much of today’s economy, and that NLP and alternative data can help systematically identify undervalued innovative companies.

Main Topics: What intangible value is (Priority: 5/5): Kai Wu defines intangible value as a company’s non-physical assets: intellectual property, brand equity, human capital, and network effects. Why traditional value metrics fall short (Priority: 5/5): The hosts and guest argue that book value and price-to-book were built for an industrial economy and undercount modern assets like R&D, software, and brand. How the ETF is built (Priority: 5/5): Sparkline’s strategy uses traditional accounting plus alternative data and NLP to score companies on intangible investment and output, then constructs a systematic portfolio. Relationship to growth, quality, and innovation (Priority: 4/5): The fund is framed as neither pure growth nor traditional value, but as an innovation tilt that seeks cheapness relative to intangible capital while avoiding some classic growth-factor baggage. Factor analysis of QQQ and ARKK (Priority: 4/5): Wu says both funds have positive loading on intangible value, helping explain their long-run outperformance versus the S&P 500, though with volatility from other factor exposures. Portfolio construction and implementation (Priority: 4/5): The strategy holds 150 names, uses a modified market-cap weighting, and balances concentration with diversification to avoid overreliance on idiosyncratic risks. Active vs. passive in smart beta (Priority: 3/5): The hosts argue smart beta is effectively active management embedded in a rules-based structure, with the real decisions made at model design time.

Key Arguments: Modern corporate value is increasingly driven by intangibles, so value strategies that rely mainly on book value are structurally outdated. Value investing was popularized in an economy dominated by factories and railroads; today’s largest firms derive earnings from brand, IP, software, and ecosystems. Accounting rules capitalize physical investment but largely miss intangible investment, creating a distorted picture of book value. Alternative data such as patents, LinkedIn, Glassdoor, job postings, trademarks, and media signals can make intangible assets measurable. A stock can be cheap relative to its intangible capital, just as it can be cheap on earnings or dividends; the strategy seeks that relationship. The fund is not simply an innovation or growth fund; it is designed to capture undervalued innovators without fully abandoning value or quality exposure. QQQ and ARKK appear to outperform partly because they are exposed to intangible-value companies, though ARKK also carries negative exposure from early-stage unprofitable tech. Quality and intangible value are complementary but distinct: quality reflects firms with moats today, while intangible value seeks firms investing toward future moats. The long-only ETF structure preserves equity beta, which many advisors want, while long-short versions are better suited for institutions seeking low correlation. The strategy’s methodology is systematic and data-driven, with no discretionary “finger on the scale” after construction.

Data Points: Intangible pillars: 4 - Sparkline defines intangible value through IP, brand equity, human capital, and network effects. ETF ticker: ITAN - Sparkline Intangible Value ETF discussed as the strategy being analyzed. Original market focus: Top 1,000 largest U.S. stocks - Current product universe for ITAN. Planned expanded universe: ~9,000 stocks - Wu says the framework is being expanded globally toward MSCI ACWI IMI-level coverage. Portfolio size: 150 stocks - Current cutoff chosen to balance concentration and diversification. Correlation: Zero - Wu says intangible value has zero correlation with quality, and also low correlation with traditional value. Correlation with traditional value: Low - Bloomberg’s Chris Kane notes the correlation between traditional value and intangible value is low. Value spread after intangible adjustment: Reduced from about 2 standard deviations to about 0.5 - Wu says accounting for intangibles materially narrows the apparent value-growth spread. Launch date: June 2021 - Wu notes the ETF launched right before a tech selloff. Relative factor universe: 6 factors - Wu describes adding intangible value as a sixth factor alongside the Fama-French style framework. Short book backtest: Top 15% long / bottom 15% short - Wu says the long-short version historically works well in backtests. Internal benchmark: Factor-neutralized stock performance - Wu says the team evaluates results relative to an internal factor-neutral benchmark. Intangible value contribution to outperformance: Positive loading in QQQ and ARKK - Wu says both funds loaded positively on intangible value in factor decomposition. Value ETF category size: Hundreds of billions, potentially trillions including active - Wu characterizes value as a major investment category.

Pivotal Quotes: "the killer app of AI within investing is the natural processing NLP toolkit" — Kai Wu: Explaining why unstructured data and NLP were necessary to make intangible-value investing feasible. "What is quality today? It's what is the modern moat" — Kai Wu: Describing how intangible assets underpin profitability and competitive advantage. "the universe is comprised of it. Yes. This is why the cues are the cues. It's this dark matter of intangible value" — Joel Weber / Eric Balchunas: Metaphor used to explain how invisible intangible assets may help explain QQQ’s persistent strength.

Implications: If intangibles are a real factor, value investing may need a broader definition of book value. Investors could gain a more modern way to identify durable innovators, and factor models may keep evolving as data and AI reveal new sources of alpha.