Elsevier

Journal of Empirical Finance

Volume 34, December 2015, Pages 131-155
Journal of Empirical Finance

The information content of R&D reductions

https://doi.org/10.1016/j.jempfin.2015.07.006Get rights and content

Highlights

  • This paper studies the long-term performance following R&D reductions.

  • We find that large R&D cuts are associated with positive future stock returns.

  • This return drift cannot be explained by additional asset pricing factors.

  • R&D spillover and firm life cycle are two potential motives behind R&D reductions.

  • This paper supports the life cycle story.

Abstract

An extensive literature shows that R&D intensities and increases are positively related to firm performance, but little research examines the valuation of R&D reductions. This paper fills the void by studying long-term performance following R&D reductions. We find that, contrary to conventional wisdom, large R&D cuts are associated with positive future stock returns. This return drift cannot be explained by asset pricing factors, including R&D intensities and R&D increases. We explore two potential economic motives behind R&D reductions: R&D spillover and firm life cycle. We show that operating performance deteriorates immediately before R&D reductions but exhibits no abnormal pattern afterward. While firm growth falls substantially and variability in profitability reduces, firms with low or declining investment opportunities and mature firms outperform. These findings are inconsistent with the spillover hypothesis, but support the life cycle story that firms attempt to resolve overinvestment in R&D that arises over the course of firm life cycle.

Introduction

We examine R&D reductions and offer evidence of a positive impact on firm value in this paper. R&D reductions are potentially relevant events to investors: more than 30% of R&D firms reduced their R&D spending during the four decades between 1975 and 2012, and the average reduction in R&D expenditures represents a 23.3% decrease in their R&D investments.3 Although research seems to suggest a low valuation for R&D reductions given that higher R&D intensities or R&D increases generate economic benefits,4 we find that R&D reductions can enhance shareholder value. This paper adds to the literature by documenting long-run positive abnormal returns and examining economic motives associated with decreases in R&D expenditures.

The examination of R&D reductions is of interest for at least two reasons. First, investors do not have complete information to assess R&D reductions as R&D is poorly disclosed. Firms seldom voluntarily make public announcements about their R&D cuts.5 Thus, investors may not pay attention to R&D reductions. The US accounting treatment of immediately expensing R&D spending exacerbates the information asymmetry associated with R&D. In particular, there is no marking-to-market for R&D investments as R&D expenditures are expensed and hence no information on the value and productivity changes of R&D is reported (Aboody and Lev, 2000). The scant disclosure in R&D reductions hinders investors' full understanding of the investment policy in R&D activities. As a result, it is hard for investors to measure the magnitude of the impact of an R&D decline.

Second, unlike R&D increases, which are commonly considered good news as they indicate more investment opportunities or better future productivity (Eberhart et al., 2004, Eberhart et al., 2008), R&D reduction signals may not be straightforwardly evaluated. The existing literature suggests that R&D reductions can affect firms negatively (Ciftci and Cready, 2011, Hall, 1993a, Lev and Sougiannis, 1996, Lev and Thiagarajan, 1993). As a result, cutting R&D investments would lower firm productivity and send a strong negative signal about future deteriorating fundamentals. Moreover, previous research raises a concern that some managers are myopic and may sacrifice long-term profits by slashing R&D expenses in exchange for increases in short-term earnings (Graham et al., 2005, Stein, 1989). The positive relation reported in the literature between R&D investments and future economic benefits (Chan et al., 2001), plus potential managerial myopia, suggests a lower market valuation for R&D reductions.6 Yet, R&D reductions can be good news in our view. For example, firms may reduce their own R&D activities in anticipation of benefits from R&D investments by other firms, the so-called R&D spillover. To the extent that R&D spillovers increase productivity in general (Jaffe, 1986), R&D reductions will be followed by improved firm profitability. Further, firms may suffer overinvestment problems when they shift to a lower growth stage. Managers might choose to reduce R&D to return to an optimal level of R&D investments. Such an action could mitigate agency problems and help the firm regain value previously destroyed.

Overall, given different managerial incentives and their contradictory effects, there is no conclusive view about the information content of R&D reductions. Accordingly, it is an open empirical question whether reducing R&D investments creates a long-term negative impact (as the literature suggests) or enhances the value of firms.

We decipher the information content of R&D reductions by asking two questions: (a) How does the market respond to an R&D cut, positively or negatively? (b) What drives a firm's decision to reduce R&D and accordingly explains the post-R&D reduction stock performance? We answer the first question by examining long-run stock returns following significant R&D reductions. We use a long-horizon approach because R&D investments are generally long-term in nature, and an extended period of time may pass before firms recognize the total costs associated with R&D reductions. Moreover, as mentioned earlier, firms seldom voluntarily report R&D reduction. Thus, it is difficult to conduct an event study with a meaningful sample size and examine its short-run market reaction. To address the second question, we examine various consequences of R&D reductions, such as operating performance, analysts' forecast revisions, firm growth, variability in profitability, and changes in the cost of capital. As there might exist many different reasons for why firms cut R&D, we intend to focus on only managerial motives that could potentially account for the abnormal market reaction we find in R&D-cut firms.

We extract 2834 observations between 1975 and 2012 for firms with significant and unexpected reductions in R&D expenditures. We find that, contrary to conventional wisdom, firms experience significantly positive abnormal returns over the five years following large R&D reductions. The abnormal returns amount to 0.48% to 0.62% per month, after controlling for size and book-to-market effects. We also control for a variety of return anomaly variables, such as past returns, accruals, net share issuance, asset growth, and profitability. We find that these variables cannot fully account for our results. We perform various robustness checks to verify the positive abnormal returns associated with R&D reductions. We find that our results are distinct from the R&D level effect (Chan et al., 2001, Lev and Sougiannis, 1996), the return predictability of large R&D increases (Eberhart et al., 2004), the capital investment effect (Titman et al., 2004), and the return drift following repurchases (Ikenberry et al., 1995). Finally, we examine industry effects by dropping some industries or controlling for industry R&D changes. None of these tests alter our conclusions. All these results, along with our control of various anomaly variables, mitigate the endogeneity concern that some factors incentivize managers to reduce R&D spending while these same factors contribute to the return drift following R&D reductions.

To identify managerial motives that incentivize managers to cut R&D and can potentially account for a positive return drift following R&D reductions, we explore the two most prominent hypotheses: R&D spillover and firm life cycle. First, the R&D spillover effect suggests that one firm's R&D can make other firms more productive (Bernstein and Nadiri, 1988, Jaffe, 1986, Megna and Klock, 1993). Managers may cut R&D spending in anticipation of the benefits of R&D spillovers (Arrow, 1962, Jones and Williams, 1998).7 Reducing R&D investments lowers expenses and thus increases current and future earnings, resulting in superior stock performance. Therefore, the spillover hypothesis predicts an improvement in firm performance following R&D reductions. Moreover, this spillover effect would be weaker for high R&D firms in an industry because they have likely already acquired the R&D knowledge by their own intensive R&D investments and will not enjoy extra benefits due to spillovers. However, our empirical evidence is not consistent with these predictions of the spillover hypothesis. We find that firms generate poor earnings immediately prior to R&D reductions and exhibit no significant improvement in operating performance following R&D reductions. Further, we do not find that sample firms with low R&D perform better than their high R&D counterparts. Analysts appear to be indifferent in revising forecasts between these two R&D groups. As a result, we fail to find empirical support for the spillover hypothesis.

The second rationale behind R&D reductions is related to changes in firm life cycle. As firms experience changes in life cycle, we would expect to see slower firm growth and stable profitability (DeAngelo et al., 2006, DeAngelo et al., 2010, Mueller, 1972). In this transition to a lower growth stage, firms may face an overinvestment problem by undertaking projects that were once profitable but now generate negative NPVs. To mitigate this problem, firms can reduce their R&D outlays to remove unprofitable projects (Grullon and Michaely, 2004, Jensen, 1986, Jensen, 1993). Moreover, R&D projects can be viewed as growth options that are riskier in general than assets in place in firm valuation. As firms reduce R&D investments and thus reduce their options to grow, assets in place play a bigger role in determining firm value, leading to lower firm risk and a lower cost of capital (Barth et al., 2013, Berk et al., 1999, Carlson et al., 2004). Accordingly, the life cycle hypothesis predicts a lower cost of capital after R&D reductions.8

We find that the growth rates of both assets and sales fall remarkably, especially in the year prior to R&D reductions. Analysts also revise earnings forecasts downward at the same time. When firms experience slower growth, their R&D investments do not adjust accordingly, leading to an abnormally high level of R&D. These results, along with poor earnings in the year before R&D reductions, suggest an overinvestment in R&D activities. After the R&D cut, as indicated by the life cycle hypothesis, firms indeed generate more stable cash flows and their profitability is less volatile. We also find a reduction in cost of capital after R&D reductions. Additional analyses show that firms that are more likely to incur overinvestment problems (low growth firms or firms with declining growth potential) and firms with large reductions in the cost of capital significantly outperform. Old firms that tend to be mature in their life cycle also outperform their counterparts. Overall, our results are consistent with a life cycle explanation in that managers reduce R&D investments to resolve an overinvestment problem that arises over the course of firm life cycle, but investors seem to underestimate the decline in cost of capital.

Our paper contributes to the literature in several ways. First, to the best of our knowledge, we are the first to report positive abnormal returns following significant declines in R&D investments. While the extant literature suggests a lower market valuation for R&D reductions, our result is in sharp contrast with this view. We show that a significant reduction in R&D spending, on average, is an indication of overinvestment mitigation, and shareholders gain via this value-enhancing action.

Second, previous research shows that R&D investments represent growth options and are highly correlated with the uncertainty of the stock (Berk et al., 2004). R&D expenditures also significantly contribute to information asymmetry (Aboody and Lev, 2000). While R&D investments play a crucial role in shaping firm risk, we find that the market appears to be slow in adjusting to the fewer options to grow following R&D reductions. As R&D constitutes a vital intangible component of firm value, we provide a piece of evidence to support the notion that future stock returns are related to intangible information (Daniel and Titman, 2006), and propose that investors' misestimation of changes in the cost of capital accounts for this result.

Third, our finding implies that there is more than one force driving the relation between R&D and future returns. While past papers focus on signaling (Chan et al., 1990, Eberhart et al., 2004), we propose a different viewpoint that life cycle of the firm may also explain R&D stock performance.

Finally, it has been established that R&D increases generate significantly positive abnormal returns in the long run (Eberhart et al., 2004, Eberhart et al., 2008). Our results of positive return drifts following R&D reductions, combined with superior stock performance associated with R&D increases, suggest a nonlinear (U-shape) relation between R&D changes and future stock returns. It would be interesting to know whether the nonlinear relation exists in reality and why investors do not recognize R&D changes at a timely fashion. Our paper opens up new avenues for future research.

The remainder of this paper is organized as follows. Section 2 describes our data, the definition of R&D reductions, and summary statistics. We present long-run return results in Section 3. Section 4 discusses the spillover and life cycle hypotheses. In Section 5, we test the spillover hypothesis by examining operating performance and analysts' forecast revisions. In Section 6, we test the life cycle hypothesis and examine changes in firm growth, variability in profitability, and changes in the cost of capital. We offer discussions of alternative explanations in Section 7. Section 8 concludes.

Section snippets

Data

Our sample formation begins with firms covered in both the Center for Research in Security Prices (CRSP) and annual Compustat during the period of 1974 to 2013.9 We focus on annual

Long-run stock returns

We begin by examining long-run abnormal stock returns following R&D reductions. We rely on the calendar-time portfolio approach combined with Fama and French's (1993) three factor model to examine long-run returns following R&D reductions.14

Hypotheses and predictions

Two hypotheses, R&D spillover and firm life cycle, may account for the abnormal returns following reductions in R&D investments. Both hypotheses offer economic reasons why firms significantly reduce their R&D investments. We provide predictions to verify each hypothesis empirically.

Do spillovers explain the return drift?

We test the spillover hypothesis by examining stock returns, operating performance, and analysts' forecast revisions.

Firm growth and overinvestment

The first prediction of the life cycle hypothesis is a sharp decline in firm growth around R&D reductions. Fig. 2 reports time series plots of R&D intensity, R&D growth, asset growth, and sales growth around R&D reductions. To reduce the impact of outliers, we present median values only. All three growth measures are stable during years − 5 to − 3. From year − 2, there is some evidence of slowdown and both asset growth and sales growth start to fall. However, the most severe decline in growth

Asset growth

As shown in Fig. 2, asset growth becomes negative prior to the R&D decreases. To the extent that asset growth is negatively related to future returns (Cooper et al., 2008), one potential concern for our main findings is that the positive return drift following R&D reductions is driven by the prior asset reduction.33

Conclusion

The literature is rich in reporting that R&D investments enhance shareholder value. However, very little research examines the valuation and impact of R&D reductions. We add to the literature by investigating stock returns following R&D reductions. Because previous research shows a positive relation between R&D expenditures and future stock returns, conventional wisdom suggests a negative signal for R&D reductions. Contrary to this view, based on data from 1975 to 2012, our results show that

References (82)

  • H. DeAngelo et al.

    Seasoned equity offerings, market timing and the corporate lifecycle

    J. Financ. Econ.

    (2010)
  • D. Engel et al.

    Firm‐level implications of early stage venture capital investment — an empirical investigation

    J. Empir. Financ.

    (2007)
  • E.F. Fama

    Market efficiency, long-term returns, and behavioral finance

    J. Financ. Econ.

    (1998)
  • E.F. Fama et al.

    Common risk factors in the returns on stocks and bonds

    J. Financ. Econ.

    (1993)
  • J.R. Graham et al.

    The economic implications of corporate financial reporting

    J. Account. Econ.

    (2005)
  • D. Ikenberry et al.

    Market underreaction to open market repurchases

    J. Financ. Econ.

    (1995)
  • Sudha Krishnaswami et al.

    Information asymmetry, monitoring, and the placement structure of corporate debt

    J. Financ. Econ.

    (1999)
  • F.Y. Eric Lam et al.

    Limits-to-arbitrage, investment frictions, and the asset growth anomaly

    J. Financ. Econ.

    (2011)
  • B. Lev et al.

    The capitalization, amortization and value-relevance of R&D

    J. Account. Econ.

    (1996)
  • T. Loughran et al.

    Uniformly least powerful tests of market efficiency

    J. Financ. Econ.

    (2000)
  • S. Roychowdhury

    Earnings management through real activities manipulation

    J. Account. Econ.

    (2006)
  • L. Wang et al.

    Economic freedom and cross‐border venture capital performance

    J. Empir. Financ.

    (2012)
  • D. Aboody et al.

    Information asymmetry, R&D, and insider gains

    J. Financ.

    (2000)
  • K. Arrow

    Economic welfare and the allocation of resources for invention

  • J.B. Berk et al.

    Optimal investment, growth options, and security returns

    J. Financ.

    (1999)
  • J.B. Berk et al.

    Valuation and return dynamics of new ventures

    Rev. Financ. Stud.

    (2004)
  • J.I. Bernstein et al.

    Interindustry R&D spillovers, rates of return, and production in high-tech industries

    Am. Econ. Rev.

    (1988)
  • S. Birley et al.

    Growth and performance contrasts between ‘types’ of small firms

    Strateg. Manag. J.

    (1990)
  • J.S. Bracker et al.

    Planning and financial performance of small, mature firms

    Strateg. Manag. J.

    (1986)
  • J.L. Brown et al.

    Stock options, R&D, and the R&D tax credit

    Account. Rev.

    (2008)
  • J.R. Brown et al.

    Financing innovation and growth: cash flow, external equity, and the 1990s R&D boom

    J. Financ.

    (2009)
  • B.J. Bushee

    The influence of institutional investors on myopic R&D investment behavior

    Account. Rev.

    (1998)
  • M.M. Carhart

    On persistence in mutual fund performance

    J. Financ.

    (1997)
  • M. Carlson et al.

    Corporate investment and asset pricing dynamics: implications for the cross-section of returns

    J. Financ.

    (2004)
  • L.K.C. Chan et al.

    The stock market valuation of research and development expenditures

    J. Financ.

    (2001)
  • K. Chan et al.

    Earnings quality and stock returns

    J. Bus.

    (2006)
  • L. Chen et al.

    An alternative three-factor model

  • L. Cohen et al.

    Misvaluing innovation

    Rev. Financ. Stud.

    (2013)
  • M.J. Cooper et al.

    Asset growth and the cross-section of stock returns

    J. Financ.

    (2008)
  • K.M. Cremers et al.

    Governance mechanisms and equity prices

    J. Financ.

    (2005)
  • K.D. Daniel et al.

    Market reactions to tangible and intangible information

    J. Financ.

    (2006)
  • We acknowledge useful comments from Po-Hsuan Hsu, Roger Huang, Inmoo Lee, Bruce Lehman, Michael Lemmon, Chen Lin, Ji-Chai Lin, Leonardo Madureira, Kirill Novoselov, Ajai Singh, Theodore Sougiannis, Scott Weisbenner, two anonymous referees and seminar participants at Case Western Reserve University, City University of Hong Kong, the University of Hong Kong, Yuan-Ze University, and National Taiwan University. Chan acknowledges financial support from the General Research Fund (GRF 741810H) offered by the Research Grants Council of Hong Kong.

    1

    Tel.: + 886 2 29393091x81239.

    2

    Tel.: + 886 2 33661075.

    View full text