Is Intangibles Talk Informative about Future Returns? Evidence from 10-K Filings

We construct a measure of intangible intensity — intangibles talk — based on textual analysis of discussions on intangibles in a firm’s 10-K filings. This measure is based on firms’ discussion of ex-post outcomes of intangibles investments. Our measure is correlated with, but carries orthogonal information to, prior measures of intangibles that are largely based on capitalization of initial investments. This distinction between expected outcomes of initial investments, as measured in prior studies, and ex-post outcomes, we examine, is important because intangible investments often have uncertain lottery-type payoffs. Thus the value of outcomes could differ from capitalized historical costs, if it was permitted in accounting. Managers are likely to discuss those positive outcomes in the 10-K documents. We test the informativeness of our measure about future returns. Returns from long and short portfolios based on high and low values of intangibles talk, respectively, outperform traditional book-to-market value strategy. Our strategy generates an average annual alpha of 3.26% from 1995 to 2020 in the four-factor (three Fama and French factors plus momentum) model. Our alphas are higher than those generated from portfolios sorted on other indicators of intangible intensity shown in the literature. We also generate positive returns by taking a long positions on our measure and taking a short position on intangibles-enhanced value portfolio studied in prior literature. Positive alphas are concentrated in stocks with higher arbitrage risk, proxied by idiosyncratic volatility, suggesting that investors misprice stocks with higher intangible intensity as described in 10-K documents