Kai Wu

Kai Wu is the founder and chief investment officer of Sparkline Capital, a quantitative investment firm, and is best known for his research on intangible assets — the brands, human capital, network effects, data and know-how that dominate modern company value but sit awkwardly in traditional accounting and factor models. He co-hosted Ritavan‘s appearance on Excess Returns.

His work argues that conventional value investing struggles partly because it measures the wrong things: book value captures factories and inventory but misses the intangibles where much of today’s value lives. He builds systematic, data-driven methods to quantify those intangibles across large universes of stocks and long histories.

Core positions

• Intangibles carry a J-curve. Spending on R&D or other intangible investment depresses near-term earnings and pays off over a decade — a profile many investors wrongly read as a negative when it is often a positive.
• A moat is contingent on the game. Code was a moat when coding was scarce; once AI writes code cheaply, that moat can evaporate. The investor’s task is to identify the dominant paradigm, ask what is scarce within it, and find assets mispriced for the old paradigm.
• The technology is not the point — the harness and the setting are. What made tools like Claude Code work was less raw model quality than the instrumentation around it and a closed-loop setting (software) where the tool could check itself.
• Base rates over stories. In disruptions, the market shoots first; historically many presumed losers recover, so the durable question is the empirical success rate and what is already priced in.

In the wiki

• Ritavan on System Gambits and Finding Leverage — Excess Returns, July 2026; co-host