Underappreciated Investment Edge: Company Trademark Filings

Not part of financial reporting, trademark activity predicts stock returns

Hedge funds and the broader institutional investor class employ armies of Ph.D.s tasked with creating the next great arcane algorithm that can exploit a market or trading inefficiency for profit.

Research published in Management Science provides a reminder that, sometimes, smart analysis of seemingly prosaic public data is all it takes to find an investing edge.

The Overlooked Trademark Factor

National Tsing Hua University’s Po-Hsuan Hsu, University of South Carolina’s Dongmei Li, Hong Kong Polytechnic University’s Qin Li, UCLA Anderson’s Siew Hong Teoh and National Taiwan University’s Kevin Tseng mined the database of the United States Patent and Trademark Office to study the predictive value of new trademarks on a public firm’s stock valuation. All approved trademarks are reported weekly in the USPTO’s Official Gazette.

In the ecosystem of intellectual property protection, trademarks are the less sexy — and thus more overlooked — cousin of patents. A patent is a claim on innovation, a hallowed ingredient of capitalism. A trademark is merely the protection of branding built around a product or service.

Yet the researchers find that firms that take out the most trademarks each year — relative to their total assets — may be delivering an intriguing tell: Their stock performance, on average, is higher in the next 12 months than companies that weren’t as active with new trademarks.

The researchers are among the first to organize this large volume of trademark data that, as it exists, is not in a user-friendly format. (They hope the reconfigured dataset can useful going forward for innovation researchers and investors.)

Using data from more than 300,000 trademark registrations filed with the USPTO from 1976 to 2014, the researchers devised a hedged portfolio that took long positions in the top-third of companies that are the most active trademark registrants in a 12-month period and shorted the companies whose trademark activity ranked in the lowest third.

The annualized return of the hedged trademark portfolio is an average of 5.2%. Adjusting for industry performance, the hedged portfolio still managed a 3.7% annualized return. The trademark factor also “predicts significantly higher” return on equity and return on assets.

The authors note that the bulk of the outperformance was driven by the long side of the hedge portfolio, suggesting that new trademarks serve as a compelling signal of promising innovation that leads to higher sales.

Yet the outperformance edge wanes after the first 12 months. The authors posit that the initial performance edge is effectively a result of analysts and other investors not reading the trademark tea leaves.

Hiding in Plain Sight

U.S. companies are not required to treat intellectual property as an asset accorded its own valuation in their balance sheet reporting. Nonetheless, a firm is likely to provide guidance on its patent pool and how that translates into revenue prospects and, ultimately, earnings potential. (think: big pharma.) Trademarks? Not so much.

And that creates a blind spot for analysts, and the investors who rely on them. The researchers find that across their database, the average analyst forecast error was significantly higher among companies that were in the high-trademark activity tercile compared with companies that were less active in registering trademarks.

The researchers note that the level of undervaluation of new trademarks correlates with the degree of analytical difficulty in sizing up a given company. The more complex a business, the bigger the R&D budget (there will be hits and misses) and the bigger the analyst forecast dispersion, all of which lead to greater undervaluation of new trademarks.

The authors also drop two intriguing crumbs for the research inclined. The magnitude of undervaluation of new trademarks was larger when a firm’s trademarks started showing up in new USPTO categories (seemingly a hint of moving into new fields) and when the new trademarks triggered more opposition from competitors (businesses can oppose a trademark application before it is approved by the USPTO).

The authors position their research as evidence that firms should be encouraged/required to provide more guidance on trademark valuation in their reporting. But that would entail the Securities and Exchange Commission providing safe harbor protection.

In the meantime, enterprising investors might consider the research as a crude instruction manual of sorts to analyze the potential mispricing upside of companies in heavy trademark mode.