The View from the Office. 

I met up with Kai Wu, the founder and chief investment officer at Sparkline Capital at the BP Cafe in Bryant Park. I had a salad. He had the tacos. 

Kai was on his way to the New York Public Library to study some of the old filings by Warren Buffett’s Berkshire Hathaway that haven’t been digitized and are only available on microfilm. 

There’s probably no better metaphor for Kai’s commitment to data than hunting down records on microfilm. 

In this project, Kai is analyzing how Buffett’s investment style and view of “value” has evolved over the years.

Part of the goal is to understand how Buffett looks at intangible assets. 

Kai believes the financial industry hasn’t kept up with the fact intangible assets like brand and intellectual property rights matter more now.

He argues that as much as 80% of the value in the modern economy comes from intangible assets. Apple and Google are good examples. 

After graduating from Harvard, Kai started his career at GMO (Jeremy Grantham’s firm). He focused on equities and options. 

He later moved to a hedge fund with another GMO colleague before launching his own firm.

Kai has been using natural language processing (NLP) and large language models for almost a decade. He has been trading based on his own models since 2019.

He uses NLP to quantify the value of  intangible assets including brand, IP, human capital and network effects.

He says they are under-appreciated because they are harder to quantify. 

In 2021, Sparkline launched its first ETF. It now manages about $50 million in two publicly traded ETFs. 

The first ETF traded under symbol ITAN (U.S. stocks, more IP and Human Capital-driven) and the second symbols DTAN (non-US/developed markets, more brand-driven).

The research on Buffett is a passion project. While not part of his day job, Kai hopes it yields valuable insights. 

He’s already mined the easy-to-get digitized Buffett data back to 1995. 

Among other things, it helped quantify Buffett’s shift to quality companies with low price-to-book ratios, high and stable profit margins.

On the day we met, he was hoping to collect data further back in time.

He expects it will illustrate a move from deep value to higher-quality businesses.

You can connect with Kai via LinkedIn or DM me for a warm intro.