Capital Investment, Innovative Capacity, and Stock Returns

• Author: PRAVEEN KUMAR,DONGMEI LI
• Date: 01 October 2016
• Published date: 01 October 2016
• DOI: http://doi.org/10.1111/jofi.12419

THE JOURNAL OF FINANCE •VOL. LXXI, NO. 5 •OCTOBER 2016

Capital Investment, Innovative Capacity,

and Stock Returns

PRAVEEN KUMAR and DONGMEI LI∗

ABSTRACT

We study the dynamic implications of capital investment in innovative capacity (IC)

on future stock returns, investment, and profitability by modeling the unique effects of

IC investment on uncertain option generation/exercise and postexercise revenue. The

model highlights the diverse effects of IC investment on expected returns in different

postinvestment regimes and yields the novel prediction that, under the neoclassical

assumption of nonincreasing revenue returns, IC investment is positively related to

subsequent cumulative stock returns with a lag. The model also predicts a positive

effect of IC investment on future investment and profitability.We find strong empirical

support for these predictions.

WHAT ARE THE IMPLICATIONS OF capital investment for asset pricing? The tradi-

tional real options view is that firms undertake capital investment to convert

growth options into assets-in-place (AIP), with an attendant negative rela-

tion between capital investment and subsequent short-run expected returns

(e.g., Berk, Green, and Naik (BGN) (1999), Carlson, Fisher, and Giammarino

(CFG) (2006)). This view is an apt description of technologically mature in-

dustries. However, in many industries with rich possibilities for innovation,

firms undertake capital investment to develop innovative capacity (IC) to gen-

erate and commercialize potential future innovations (see Furman, Porter, and

Stern (2002)). For example, setting up long-range research facilities and ac-

quiring patents can help to generate growth options because innovation-related

∗Praveen Kumar is with the C.T. Bauer College of Business at the University of Houston and

Dongmei Li is with the Moore School of Business at the University of South Carolina. We thank two

anonymous referees, an Associate Editor,and Ken Singleton (the Editor) for very helpful comments.

We also thank Heitor Almeida; Jonathan Berk; Jeffrey Brown; Louis Chan; Jaewon Choi; Prachi

Deuskar; Phil Dybvig; Wayne Ferson; Slava Fos; Paolo Fulghieri; Joao Gomes; Andrew Grant; Eric

Ghysels; Richard Green; Dirk Hackbarth; David Hirshleifer; Gerard Hoberg; Elvis Jarnecic; Dirk

Jenter; Charles Kahn; Nisan Langberg; Michael Lemmon; Jun Liu; Mark Loewenstein; Graham

Partington; Neil Pearson; George Pennacchi; Gordon Phillips; Josh Pollet; Jeffrey Pontiff; Michael

Roberts; Laura Starks; Sheridan Titman; Selale Tuzel; Neng Wang; Joakim Westerholm; Toni

Whited; Jianfeng Yu; Lu Zhang; and Guofu Zhou; as well as seminar participants at University

of Illinois (Urbana-Champaign), the USC Conference on Financial Economics and Accounting,

University of Sydney, University of South Carolina, Texas A and M, Georgia Tech, University of

California Riverside, and Southwestern University of Finance and Economics for useful comments.

We hav e read the Journal of Finance disclosure policy and have no conflicts of interest to disclose.

DOI: 10.1111/jofi.12419

2059

2060 The Journal of Finance R

activities are often sources of new ideas and opportunities (Schumpeter (1942)

and Maclaurin (1953)).

In this paper, we model unique features of IC investment and empirically

test refutable predictions on the dynamic effects of IC investment on future

stock returns, investment, and profitability. Distinct from traditional option-

exercising investment, IC investment helps to develop innovations (or new

growth options), but with significant uncertainty in their generation and exer-

cise given the nature of the innovation discovery and commercialization pro-

cess. Moreover, IC investment enhances expected revenues by allowing the

firm to generate higher quality–based sales when innovations are generated

and exercised. These unique features are consistent with the innovation and

industrial organization literature, and we exploit them to derive dynamic im-

plications of IC investment.

An important implication of our model is that the effects of IC investment

on future expected returns depend on the resolution of the uncertainty

regarding the generation and exercise of options. Specifically, these effects

are generally ambiguous before innovation is generated and exercised, but

become unambiguously positive after the option is exercised as long as there

are nonincreasing revenue returns from IC investment. We also generate

implications of IC investment for future investment, profitability,and standard

risk factor loadings. Because IC investment facilitates the generation and

exercise of growth options (due to its positive revenue effect following exercise),

the prediction is that IC investment increases expected future investment (as-

sociated with exercising the potential options). Furthermore, combining these

effects on future expected returns and investment with the decomposition

of the market-to-book equity ratio with clean surplus accounting (see Fama

and French (2006)), we predict that, for firms with nonincreasing revenue

returns, IC investment increases expected profitability. These novel dynamic

implications of IC investment also have implications for the evolution of stan-

dard risk factor loadings (Fama and French (1993) and Carhart (1997)) after

IC investment. For example, we expect investing firms’ loading on the value

factor to be increasing and their loading on the size factor to be decreasing

over time as firms generate and exercise new options and become bigger.

To test the predictions of the model, we take IC firms to be R&D-active

firms because being R&D-active is widely viewed as being innovation-driven

(Rogers (1998)). We mainly use the total asset growth (AG) (Cooper,Gulen, and

Schill (2008)) of R&D-active firms to identify IC investment since AG is a gross

measure of capital investment. Building on a variety of empirical perspectives

(see Section II.B), we use large firm size in addition to R&D activities to capture

firms that are likely subject to nonincreasing revenue returns of IC investment.

We find strong empirical support for the implications of our conceptual frame-

work. For large IC firms, the effects of AG on subsequent cumulative returns

are insignificant over the first five years, but become significantly positive as

of the sixth year. In contrast, we observe no such patterns for small IC (and

R&D-inactive) firms. In addition, IC firms show significantly higher future

investment, and IC investment is positively related to future profitability for

Capital Investment, Innovative Capacity, and Stock Returns 2061

large IC firms. We also verify the predictions regarding the post-AG evolution of

standard risk factor loadings. Our results are robust to other measures of capi-

tal investment that have been used in the literature (Polk and Sapienza (2009)

and Lyandres, Sun, and Zhang (2008)) and alternative ways of identifying IC

firms.

By modeling the unique features of capital investment in IC, we uncover

both theoretically and empirically rich dynamic patterns following capital in-

vestment. Specifically,we identify the uncertainty of option generation/exercise

and the post-exercise operating leverage (OL) effect (under neoclassical restric-

tions on the revenue returns) as the key aspects of IC investment that drive

the dynamic effects of investment on subsequent returns. This qualitative in-

tertemporal heterogeneity in the relation of (IC) investment to future expected

returns is novel because extant literature has typically not considered the im-

plications of uncertain option generation/exercise on the effects of investment

on expected returns. In particular, as we mention at the outset, existing real

options literature views capital investment as exercising available growth op-

tions and typically predicts a negative relation between capital investment

and subsequent short-run expected returns. Furthermore, the key aspects of

IC investment above generate novel predictions on IC investment with respect

to future investment, profitability, and evolution of risk loadings, which we

empirically verify.

In a related vein, our results are also distinct from the literature that con-

siders the relation between capital investment and subsequent stock returns

using behavioral and agency-theoretic perspectives. For example, Titman, Wei,

and Xie (2004), among others, who find a negative relation between capital in-

vestment and subsequent abnormal stock returns—the so-called “investment

anomalies” literature—argue that investors underreact to empire building by

managers. However, since IC investments can be associated with empire build-

ing, this view also predicts a negative relation between IC investments and

subsequent stock returns, especially for firms with higher investment discre-

tion. Meanwhile, Cao, Simin, and Zhao (2008) use the Galai and Masulis (1976)

model to posit a positive association between growth options and idiosyncratic

volatility (IVOL), and suggest that more levered firms have greater incentive

to undertake projects with high idiosyncratic risk because these risks are borne

by debt holders. This perspective would also imply a negative relation between

investment and subsequent returns given the well-documented negative IVOL-

return relation. Notably, this literature does not generate the dynamic post-

investment patterns that we predict and find empirically for IC firms, which are

apparently masked by looking at the average responses of the universe of firms.

We organize the rest of the paper as follows. Section Ipresents the model and

develops empirical predictions. Section II describes the data and the empirical

framework. The empirical results are presented in Sections III,IV,andV.

Section VI concludes. All proofs and robustness checks are presented in the

Internet Appendix.1

1The Internet Appendix may be found in the online version of this article.