The market value of R&D : a study of the effects on high-tech companies

(1)

Faculty of Economics and Business

The market value of R&D

A study of the effects on high-tech companies

Bachelor thesis Student: Anneline Mendes de Leon Student Number: 10418997 Supervisor: dr. R. Perez Ribas University of Amsterdam, Faculty of Economics and Business Bsc. in Economics and Finance June 2016

(2)

Statement of Originality

This document is written by Anneline Mendes de Leon who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

This paper investigates whether high-tech firms should engage in R&D investment or not, by investigating how the market reacts to R&D. This is tested for four industries from 1998 until 2015. The industries observed are Chemicals and Allied Products, Industrial Machinery and Equipment and Electronics and Other Electric Equipment. Also, a wider range of high-tech firms is examined. Two hypotheses are tested in particular. The first states that firms should engage in R&D investment because R&D positively influences their market value, up to a certain point. I test this by estimating the elasticity of market value and R&D. The second is that growth in R&D should follow market sensitivity to this type of investment. The main findings in this thesis are that market responses to R&D tend to change over time. However, these changes are not always reflected in the decisions of firms to increase or decrease their R&D efforts.

(4)

1. Introduction

Research and Development (R&D) expenditure plays a crucial role for the high-technology industry, because R&D efforts are a critical driving force behind successful innovation. Innovation is necessary for companies in this industry because they face a competitive and dynamic environment (Wang et al., 2013). For high-technology companies R&D is one of the biggest sources of competitive

advantage (Bloch, 2005). For managers in technology-based firms the valuation of their R&D efforts is important for their decision-making. Since investment in R&D is expensive and risky, they want to have positive returns. It is also important for investors, because they try to measure the

fundamental value of firms as a guide to investment (Hall, 2000). It is challenging to interpret what the “correct” amount of R&D intensity should be. If firms are not spending enough on R&D, an increase in their R&D intensity is likely to increase their market values, whereas an increase for firms spending too much on R&D may result in lower market values (Bracker and Ramaya, 2011).

This thesis investigates to what extent firms should invest in R&D. I want to know what the attitudes of shareholders and investors towards R&D should be. I do this by testing how an increase in R&D expenditure translates in terms of market value. I estimate the elasticity of R&D and market value and I make a comparison between four industry ranges. Since R&D efforts trigger innovation, I expect them to positively influence market value. However, since R&D expenses are coupled with risk this may not always be true (Hall et al., 2010). Furthermore, the existence of time lags between the development and commercialization of R&D may indicate negative effects on market values in the short run (Pakes and Schankerman, 1984). My main hypothesis is as follows: firms engage in R&D investment up to a certain optimal point, since the R&D intensity of high-technology firms positively influences the change in their market values.

To test my hypothesis, I use data of North American companies, based on their Standard Industrial Classification (SIC) codes. The industries for which I test my hypothesis are Chemicals and Allied Products, Electronics and Other Electrics and Industrial Machinery and Equipment. Also, I use a wider range of high-technology industries in my comparison. These sectors are chosen because according to Hirschey et al. (2012) their R&D expenditure over sales ratio from 2006 to 2010 is relatively high compared to the average of all industry groups: 10.7, 3.6 and 7.3 per cent

respectively, compared to 0.6 per cent on average. Furthermore, these industries are in the top four of industries with the most firms. They consist of 437, 385 and 229 firms respectively and sufficient data is available. They can be plausibly compared to each other because they are all highly

(6)

2 To test whether the relationship of R&D to market value is changing over time a period of seventeen years is used: 1998 until 2015. This period is divided into three phases. The first one is from1998 until and including 2001, the second phase is from 2002 until and including 2008. The last one is from 2009 until and including 2014, which is the last year for which enough data on R&D is available. Both the early 2000’s recession as the Great Recession happen in this period. In the first period the early 2000’s recession takes place, in the second period the Great Recession takes place, which is still going when the third period starts. Since all periods have years of recession, they are better comparable.

The consensus of prior research is that R&D intensity positively affects firm performance and results in greater valuation in financial markets. Different kinds of indicators are used to evaluate the effectiveness of R&D expenditure. This study contributes to prior research because it uses a different measure of R&D intensity: R&D expenditure over total operating expenditure. This ratio is chosen because accounting standards in the United States treat R&D expenses as operating expenses and by using this ratio the importance of R&D expenditure relative to other expenses can be observed. Other studies mostly measure R&D intensity as R&D expenses over total sales. Furthermore, this study examines whether there is an optimal amount and whether the relationship differs among these four comparable industries. I find that the industries differ in their impacts of R&D.

Considering the whole period of 1998 until 2014, all industries show positive relationships to their market values. However, if separate periods are considered, there are negative elasticities of R&D intensity to market value, which means firms should not increase their R&D intensity. My findings indicate that there is an optimal amount of R&D.

Based on my main hypothesis, my second order hypothesis is that the R&D intensity increases every period, because I expect R&D to have positive effects on market value. In other words, I expect R&D to follow market sensitivity of this type of investment. I find that this is not the case for all periods and industries I used. This indicates that it holds for certain periods that firms spend less on R&D expenditure relative to other operating expenses even though R&D positively affects market value.

The main conclusion in literature is that there is a positive effect of R&D expenditure on firm performance. Thus, the negative effects found contradict existing literature. Also, this study shows a decrease in R&D intensity over time, whereas prior research conclude that there is an increase (Bracker and Ramaya, 2011).

(7)

3

2. Literature Review

Previous studies namely focus on two relationships: the relationship of R&D expenses and the productivity of firms and the relationship of R&D expenses and the market value of firms.

According to the US General Accepted Accounting Principle (GAAP) all R&D costs must be expensed. The major difference between capital and expenses is that expenses are not expected to generate future profits (Deng and Lev, 2006). This indicates that according to GAAP, R&D expenses do not necessarily result in future benefits.

However, the main conclusion found in literature on the effect of R&D expenditure on firm performance is different. Griliches (1981) finds a significant relationship between the market value of firms and their R&D expenditures. This is confirmed by Hirschey (1982) and Hirschey and Weygandt (1985). Hirschey (1982) empirically finds that advertising and R&D expenditures have positive and significant market value effects. Hirschey and Weygandt (1985) suggest that because of the systematic influences of R&D expenditures on market value, they should be considered as forms of intangible capital investment by the accounting standards, not as expenses. Sougiannis (1994) finds that a one dollar increase in R&D expenses results in a five dollar increase in market value on average. Hu (2001) also comes to this conclusion for Chinese high technology sectors. Ettlie (1998) finds significant positive influence of R&D investment on firm performance by surveying 600 durable goods firms. Bae et al. (2008) also find this positive relationship. According to Huang and Diewert (2011) successful R&D investments add to knowledge assets, and these assets provide a flow of services over time and not just in one period. Because they suggest capitalization of R&D

investments, they make an estimation of depreciation rates of R&D investment. Deng and Lev (2006) also conclude that R&D expenditures should be capitalized because they are significantly associated with future cash flows of at least three years.

There also exists research in which a significant relationship between R&D expenditure and market value is not found (e.g. Johnson (1967), Newman (1968) and Milburn (1971)). Demirel and Mazzucato (2012) observe that for large pharmaceutical firms R&D expenditure may have negative impact on growth. This could be caused by the low productivity of their R&D since the 1990’s. Furthermore, they conclude that small firms are more likely to grow as a result of their R&D efforts and large firms may have a negative effect on growth performance. Lin et al. (2006) did not find a significant relationship between R&D intensity and market value for US technology companies from 1969 until 1999.

(8)

4 Lev and Sougiannis’ (1999) research is similar to Fama and French (1993), who analyze the effect of specific factors on returns. Fama and French (1993) use the firms’ size, the book to market ratio, leverage, CAPM beta and earnings to price ratio. They find that the size and the book to market ratio both have impact on stock returns, with size having a negative effect and book-to-market having a positive effect. Lev and Sougiannis (1999) add the ratio of R&D capital to the book-to-market value of equity. They find that this factor has significant and positive effect on returns, which may reflect greater risk, due to the uncertainty of R&D investments. Hall and Lerner (2010) conclude that this uncertainty is especially high at early stages of R&D projects. Coad and Rao (2010) state that there is a long time lag required for a commercially valuable discovery to materialize. The length of this lag is important for managers.

Chan et al. (2001) examines whether stock prices fully value firms’ intangible assets and specifically R&D. They find that there is no direct link between R&D spending and future stock returns. This conclusion is consistent with the efficient market hypothesis by Fama (1970), who states that stock prices include all available information. This means that the value of R&D is already reflected in the stock prices, thus there is no link between R&D and future stock returns. Hsieh et al. (2003) conclude that R&D investments earn an above average return and provide a firm a

competitive advantage. Bloch (2003) studies whether stock prices of Danish R&D intensive firms fully reflect the value of R&D. When he uses the ratio of R&D to market value of the firm, he does find positive effects on future returns. According to Ho et al. (2005) intensive investment in R&D positively affects one-year stock performance for manufacturing firms, but this is not the case for non-manufacturing firms. Their results are inconclusive for three-year stock market performance. Gollotto and Kim (2003) also find positive relationships between R&D and stock values for dot.com companies with high R&D expenditures.

Franzen et al. (2007) find that the R&D intensity of US firms increase significantly from 1980 to 2004. One finding of this thesis is that the R&D intensity of the Chemicals and Allied products is relatively high compared to the other industries tested. According to Chen and Chang (2010) it is the industry with the highest R&D intensity among all industries in the United States. High levels of R&D intensity signal about the strategic importance of innovation to the firm (Lin et al., 2006).

(9)

5

3. Data and Empirical Model

3.1 Data

The data come from COMPUSTAT, consisting of 324,829 observations. After the drop out due to missing data and after the creation of new variables the final sample consists of 56,998 observations. These observations are then first filtered for a wide range of high technology industries, based on the sic codes 2800-5000, second for the Chemicals and Allied Products industry (sic codes 2810-2890), then for the Industrial Machinery and Equipment industry (sic codes 3510-3590) and finally for the Electronic and Other Electric industry (sic codes 3610-3690). These three are part of the medium to high R&D intensity industrial sectors, according to the EU R&D Scoreboard 2014.

Most variables are retrieved directly from the database, whilst other variables are

constructed. The variables used from the database are the sic codes, total assets, total debt, capital expenditures, cash flows of financing activities, investing activities and operating activities, research and development expenses, operating expenses and the market value, defined as the number of shares outstanding times the year end share price. There are four computed variables. They can be found in Table 1.

As explained before, the R&D intensity ratio is computed by dividing R&D by operating expenditure. The other variables computed are leverage, defined as total debt divided by total assets, total cash flows, defined as the sum of net cash flows of financing, investing and operating activities, divided by total assets and the log of the market value/operating expenses - ratio.

Because of the missing data before 1998 and after 2014, the time frame chosen is from 1998 until and not including 2015, which is still large enough to see trends over time.

Variable Description R&D xrd/xopr LnMarket ln(mkvalt/xopr)

Leverage dt/at

Cash Flows (fincf+ivncf+oancf)/at Size ln(at)

(10)

6

3.2 Descriptive Statistics

In Table 2 the mean and standard deviation of the five constructed variables can be found. As can be seen in the table, the mean of R&D intensity decreases in each period in the four industries. Despite an increase in total R&D expenditure, the intensity decreases because of a larger increase in

operating expenses. This is not in line with my expectation that the intensity increases in every period. The decrease in R&D intensity may be the result of a negative market respond to R&D, which is investigated in the next section. It could also be due to the fact that other operating expenses (e.g. are rents, marketing costs and inventory costs) become relatively more important.

3.3 Empirical Model

To run a regression on these data, panel data is used. According to Hsiao (2003) the advantages of panel data sets over cross-sectional or time series data sets are that panel data give more accurate inference of model parameters. They give a large number of data points, which increases the

degrees of freedom and reduces the collinearity among explanatory variables. Thus, the efficiency of the estimates improves. Furthermore, panel data allow to control for omitted variables.

Fixed effects are added to the xtreg regression, because the variables are non-random. Fixed effects investigate the relationship between predictor and outcome variables within entities (Torres-Reyna, 2007). The standard errors are clustered by firm.

Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.

0.193 0.262 0.194 0.574 1.376 0.559 0.473 11.930 17.886 408.994 4.504 2.670 4.179 0.376 0.371 0.397 1.144 1.296 1.278 0.524 11.749 0.222 17.899 412.303 4.005 2.651 3.844 0.106 0.115 0.117 0.295 1.227 0.209 0.204 1.590 0.202 22.752 335.329 4.952 2.656 4.389 0.150 0.147 0.154 0.381 1.291 0.532 0.269 2.521 0.270 14.773 249.131 4.620 2.389 4.332 Size R&D LnMarket Leverage Cash Flows Size 0.301 0.194 0.259 0.192 0.236 1.553 0.527 1.346 0.650 1.282 0.642 17.183 0.393 8.919 0.467 11.100 18.173 249.934 16.833 400.942 19.136 498.463 2.484 4.512 2.587 4.717 2.875 0.487 0.383 0.365 0.357 0.301 1.521 1.029 1.273 1.214 1.181 0.515 0.376 2.372 0.863 19.290 15.688 173.037 14.720 351.422 22.961 547.898 2.483 4.072 2.567 4.008 2.828 0.128 0.104 0.113 0.098 0.104 1.474 0.274 1.141 0.408 1.084 0.592 0.167 0.429 0.262 2.853 13.857 176.802 18.863 360.075 36.991 402.702 2.389 4.929 2.601 5.516 2.856 0.162 0.153 0.146 0.144 0.138 1.460 0.392 1.247 0.256 1.206 2.651 0.262 2.772 0.278 1.993 24.838 206.400 9.507 211.082 15.008 318.213 2.213 4.580 2.311 4.887 2.588

Electronic and Other Electric Equipment R&D

LnMarket Leverage Cash Flows Size

Industrial Machinery and Equipment R&D

LnMarket Leverage Cash Flows Size

Chemicals and Allied Products R&D

LnMarket Leverage Cash Flows

Wide Range High-Tech Industries

Table 2: Descriptive Statistics

(11)

7 In order to explain the trend of R&D expenditure over time, I use factors to estimate the financial return of R&D expenditure. I add Leverage, Cash Flows and Size as control variables and I also include a dummy for the fiscal year. Furthermore, the squared R&D intensity is included in order to test whether the R&D intensity is linearly related to market value (Ho et al., 2005). Through estimating how the R&D expenses translate in terms of market value, I investigate how shareholders or investors should see R&D investments.

To research whether changes in R&D explain changes in market value, I do the following regression:

𝐿𝑛𝑀𝑎𝑟𝑘𝑒𝑡 = 𝛽0+ 𝛽1𝑅&𝐷 + 𝛽2𝑅&𝐷2+ 𝛽3𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 + 𝛽4𝐶𝑎𝑠ℎ 𝐹𝑙𝑜𝑤𝑠 + 𝛽5𝑆𝑖𝑧𝑒 + 𝛽6𝐷𝑌𝐸𝐴𝑅+ 𝑓𝑒 + 𝑢 Where “LnMarket” stands for the log of market value divided by total operating expenses.

My hypothesis is that R&D has a positive effect on LnMarket, so: H0: β1 = 0

(12)

8

4. Results and Analysis

As can be seen in Table 3, the R&D coefficient for the wide range of high-tech industries is positive and significant for the period as a whole. This is in line with my hypothesis that the market values R&D investment. However, for the three separate periods this is not always the case. The coefficient increases from the first to the second period, which means the contribution of the R&D intensity of the firms in this industry increases, but after the second period the coefficient decreases to a negative and significant value. This means that an increase in R&D negatively affects market value. This contradicts prior research, where a positive effect is found (e.g. Chauvin and Hirschey, 1993).

An explanation for the negative coefficient could be the fact that the market is slow in fully incorporating the value of R&D investments (Eberhart et al., 2004). Another reason for this could be the big amounts of R&D spending that are unsuccessful. There are risks associated with R&D

spending and furthermore, as mentioned before, it may take relatively much time before R&D investments translate in benefits (DiMasi et al., 2003). Another factor that could be of importance is the recession that started in 2008. The period in which the coefficient is negative is the period after the Great Recession. This may cause the inverse relationship between R&D spending and market value, but further investigation would be necessary.

The coefficients of R&D2, though not significant in every period, imply that there is an optimal amount of R&D spending. After a certain point, it does not pay off for firms in these industries to increase their R&D intensity. The negative relationship implies linearity. This non-linear relationship between R&D and market value is confirmed by Ehie and Olibe (2010).

The coefficients of leverage are insignificant and small. This means leverage does not affect market value changes much. Cash flows show positive and significant coefficients, which means the larger the cash flows, the more the increase in market value. This positive relationship is expected because if a company reports that it has more cash available, the market should value its price higher. The size of a firm also shows a positive relationship with market value, which makes sense since the larger the firm, the higher the market value.

Apart from the last period, it holds that for firms in these industries it is beneficial, up to a certain point, to invest in R&D. The hypothesis can be accepted for the period as a whole, but not for every period separately.

(13)

9 Table 3: Regression Outcomes for the Wide Range of High-Tech Industries

1998-2015 1998-2001 2002-2008 2009-2014 Regression Coefficients R&D 0.688*** 0.158 0.764*** -0.307*** (0.050) (0.122) (0.108) (0.119) R&D2 -0.058*** -0.007 -0.179*** 0.210*** (0.008) (0.014) (0.047) (0.042) Leverage 0.001 0.001 0.001 0.001 (0.001) (0.001) (0.003) (0.001) Cash Flows 0.062*** 0.202*** 0.025*** 0.067*** (0.006) (0.017) (0.007) (0.008) Size 0.159*** 0.183*** 0.208*** 0.166*** (0.007) (0.022) (0.012) (0.013) Constant -0.161*** -0.375*** -0.745*** -0.078 (0.048) (0.100) (0.057) (0.064)

Fixed Effects Yes Yes Yes Yes

Year Dummy Yes Yes Yes Yes

No. Of Obs. 31,764 7,145 14,244 10,375

R-Squared 0.147 0.115 0.263 0.083

***, **, * represent statistical significant at the 1%, 5% and 10% levels, respectively. Standard errors are clustered by firm.

For Chemicals and Allied Products firms in Table 4, the same pattern exists. As in the range of industries discussed before, the coefficient is positive and significant for the period as a whole, positive and insignificant for 1998 until 2001, positive and significant for 2002 until 2008 and negative and significant for 2009 until 2014. The same explanations apply for this industry as for the Wide Range of High-Tech Industries, discussed above.

Also, the R&D2 coefficient shows that there is an optimal R&D intensity, as discussed above, but not all of the coefficients are significant.

For the coefficients of Leverage, Cash Flows and Size the same can be concluded as for the Wide Range of Industries, except for the Leverage coefficients that show significance in the last two periods. They are positive, but small. This indicates that during these periods risk had a positive effect on market value.

Firms in this industry should increase their R&D intensity up to a certain point, because it translates into benefits for their market values. However, during the most recent period this is contradicted by the negative coefficient, implying that increases in R&D intensities are not profitable. This could be due to one of the reasons mentioned for the prior industry. Thus, my hypothesis can be accepted for the period as a whole, but not for every period separately.

(14)

10 Table 4: Regression Outcomes for Chemicals and Allied Products

1998-2015 1998-2001 2002-2008 2009-2014 Regression Coefficients R&D 0.487*** 0.198 0.528*** -0.631*** (0.065) (0.149) (0.148) (0.180) R&D2 -0.038*** -0.011 -0.090 0.200** (0.009) (0.016) (0.058) (0.085) Leverage 0.001 -0.038 0.013** 0.001* (0.001) (0.068) (0.006) (0.001) Cash Flows 0.031*** 0.159*** 0.000 0.054*** (0.006) (0.034) (0.009) (0.009) Size 0.171*** 0.234*** 0.284*** 0.209*** (0.012) (0.040) (0.022) (0.020) Constant 0.428*** 0.150 -0.457*** 0.615*** (0.093) (0.173) (0.098) (0.094)

No. Of Obs. 9,272 1,755 4,142 3,375

R-Squared 0.159 0.101 0.275 0.113

The R&D coefficients in the Industrial Machinery and Equipment industry in Table 5 follow a similar pattern. As in the two cases mentioned above, they increase in the second period compared to the first and they decrease in the third period compared to the second. However, they start as negative coefficients in the first period, increase to positive coefficients in the second period and they decrease but stay positive in the last period. This is in contrast with the first two cases, where the coefficient was positive (but insignificant) in the first period and negative in the last period. With the exception of 2002 until 2008, all coefficients are insignificant.

The coefficient of 2002 until 2008 is significant and relatively large. In line with this finding, I expect firms in this industry to increase their R&D intensity, but as can be seen in Table 2, this is not the case. The reason for this could be that more importance is given to other expenses.

In contrast to other industries, this industry does not show a negative coefficient of R&D2 for the period as a whole. This means, there is no curvilinear relationship, rather the relationship is linear and no optimal point exists, so firms should keep on increasing their R&D intensity. This is not in line with my expectations.

(15)

11 For the coefficients of Cash Flows and Size the same applies as for the other industries, but the coefficients of Leverage are different. They are negative in every period, implying that the higher leverage or risk of a firm, the larger the decrease in market value.

Overall, this industry does not show significant results except for the period of 2002 until 2008. Though this means that a higher R&D intensity would result in an even higher change in market value, firms do not increase this ratio during this period, as can be seen in Table 2. My hypothesis can be accepted for the period as a whole, but not for the last two periods. However, there is no optimal amount of R&D in this industry, which is not in line with my hypothesis.

Table 5: Regression Outcomes for Industrial Machinery and Equipment

1998-2015 1998-2001 2002-2008 2009-2014 Regression Coefficients R&D 0.482 -0.858 1.154** 0.765 (0.398) (0.900) (0.556) (0.854) R&D2 0.002 0.764 -1.567** -2.336* (0.52) (1.101) (0.727) (1.235) Leverage -0.053*** -0.075 -0.144*** -0.048*** (0.011) (0.074) (0.051) (0.007) Cash Flows 0.139*** 0.774*** 0.111*** -0.003 (0.027) (0.091) (0.030) (0.036) Size 0.180*** 0.105* 0.124*** 0.289*** (0.02) (0.061) (0.038) (0.036) Constant -0.679*** -0.356 -0.633*** -1.248*** (0.131) (0.275) (0.193) (0.197)

No. Of Obs. 4,444 1,204 1,955 1,285

R-Squared 0.155 0.198 0.256 0.163

The coefficients of the R&D intensity in the Electronic and Other Electric Equipment industry can be seen in Table 6. The pattern of the coefficient is similar to that of the Industrial Machinery and Equipment Industry.

The only significant coefficients are the ones of the period as a whole and of the period of 2002 until 2008. The coefficient of 2002 until 2008 is remarkably large. It is the largest compared to the coefficients in all other periods and industries. However, firms in this industry decrease their R&D intensity.

(16)

12 There is no reason to conclude that there is an optimal intensity of R&D for every period separately, since the coefficients of R&D2 are insignificant. However, for the period as a whole, this can be concluded, since it is negative and significant.

When considering the period as a whole, it is beneficial for firms within the Electronic and Other Electric Equipment industry to increase their R&D intensity up to a certain point, since it will increase their market values and there is an optimal amount. My hypothesis can be accepted for this industry.

Table 6: Regression Outcomes for Electronic and Other Electric Equipment

1998-2015 1998-2001 2002-2008 2009-2014 Regression Coefficients R&D 1.244*** -0.874 1.547*** 0.658 (0.233) (0.541) (0.544) (0.637) R&D^2 -0.431* 0.617 -1.074 0.411 (0.226) (0.388) (0.775) (0.824) Leverage 0.003 0.023* -0.012** 0.045*** (0.005) (0.013) (0.006) (0.012) Cash Flows 0.195*** 0.144*** 0.122*** 0.351*** (0.018) (0.024) (0.036) (0.061) Size 0.139*** 0.237*** 0.254*** 0.105*** (0.015) (0.051) (0.027) (0.032) Constant -0.351*** -0.598*** -1.211*** -0.312* (0.098) (0.220) (0.139) (0.167)

No. Of Obs. 6,874 1,579 3,118 2,177

R-Squared 0.232 0.218 0.334 0.118

(17)

13

5. Discussion

My findings are that benefits for firms from R&D intensity differ among the industries studied, as can be seen in Figure 2. The negative elasticities found in certain periods show that market responses negatively to R&D. However, for the period as a whole, the results show that for firms in every industry studied, their efforts in R&D translate in increases in their market values, which is in line with my hypothesis.

The R&D intensity is relatively large for Chemicals and Allied Products. This fact is confirmed by other studies that find that this industry is the most active with respect to R&D activities (Lager, 2002, Reichstein and Salter, 2006). The ratio is decreasing over time, which is not in accordance with my expectations. Despite the fact that the elasticities increase in the second period compared to the first, the R&D intensity decreases when comparing these periods. In other words, even though the market values R&D intensity more, firms decrease their intensity. The negative elasticity in 2009 until 2014 comes together with a relatively large decline in R&D intensity. This can be logically explained, since according to these numbers, it is not profitable to increase R&D intensities. Although R&D expenses alone increase, the intensity decreases. This means operating expenses in total increases by a larger amount than R&D expenses. Thus, the importance of R&D relative to other expenses decreases.

The industry with the second largest R&D intensity is the wide range of high-tech industries. Since all three industries are in this range, I do not discuss this industry separately. The R&D intensity of the Electronic and Other Electric Equipment industry shows a slight decrease over time, with the same reasoning as the Chemicals and Allied Products industry mentioned above. But the changes are relatively small, which also holds for the Industrial Machinery and Equipment industry. The latter has lower ratio values, which is also observed by Coa and Road (2008), who measure the ratio as R&D expenditure over total sales. The increase in the elasticities of the first compared to the second period would imply the intensity to increase, since the increase would translate in a larger increase in market value. However as can be seen in Figure 1, this did not happen. The same reasoning as before applies.

The values of the elasticities vary per industry, but they show the same pattern: the elasticities increase for the period of 2002 until 2008 compared to 1998 until 2001 and decrease when comparing 2002 until 2008 to 2009 until 2014. Reason for the decrease between the last two periods could be that the R&D process does not pay off immediately. Rather, it is a process that takes a long time, often a decade or more (DiMasi et al., 2003). Their benefits are likely to materialize much later. Because of the time lag, the recent R&D expenditure may not have

(18)

14 translated in an increase in market value yet. Another explanation for the negative elasticity could be that there is a downward trend in the success rate of R&D processes for more recent periods (DiMasi et al., 2016). Since the success rates are lower, the chance that R&D expenditure pays off is lower.

Although the patterns of the relationships are similar, their values differ. Electronic and Other Electric Equipment and Industrial Machinery and Equipment sectors show similar values, but they differ from the Chemicals and Allied Products and the wider range of industries. This can be logically explained because the characteristics of the first two sectors are more similar to each other than to the Chemicals and Allied Products.

When looking at the elasticities over the whole period, the Electronic and Other Electric Equipment shows the highest value. This means that firms in this industry benefit the most from increases in their R&D efforts, comparing to the other industries. Reason for this could be that the firms in this industry are faster in translating R&D expenditure to firm performance. The speed of this transformation should be further investigated.

Regarding the non-linearity of the impact of R&D intensity on market value, there is enough evidence to conclude that there is an optimal amount of R&D intensity. The negative coefficient of the squared R&D intensity means that firms could overspend on R&D. R&D spending is designed to generate benefits in the future (Bracker & Ramaya, 2011). If a firm does not spend enough on R&D, it is passing up positive Net Present Value opportunities. If a firm spends too much on it, then it undertakes Net Present Value opportunities. Both too much as too little spending on R&D can be harmful.

Figure 1: Mean of R&D-Ratio per Industry

0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 1998-2001 2002-2008 2009-2014 1998-2014

Wide Range High-Tech Industries

Chemicals and Allied Products Industrial Machinery and Equipment

Electronic and Other Electric Equipment

(19)

15

Figure 2: Elasticities of R&D to Market Value per Industry

-1 -0.5 0 0.5 1 1.5 2 1998-2001 2002-2008 2009-2014 1998-2014

Wide Range of High-Tech Industries

Chemicals and Allied Products Industrial Machinery and Equipment

Electronic and Other Electric Equipment

(20)

16

6. Conclusion

The purpose of this paper is to investigate to what extent firms should engage in R&D investment. This thesis concludes that when the period as a whole is considered, firms should increase their R&D efforts because they have positive effects on their market values. Based on the results of this thesis I also conclude that there exists a non-linear relationship between the market values and R&D

intensities, implying that there is an optimal amount of R&D spending. Thus, firms should increase their R&D intensity, but up to a certain point. This holds for every industry except the Industrial Machinery and Equipment industry, which shows insignificant results.

Market responses to R&D change over time. When comparing the four industries studied, it can be seen that they all have the same pattern of changes in elasticities. They all increase in the second period compared to the first and they all decrease in the third relative to the second period. Apart from the positive relationship between R&D intensity and market value over the period as a whole, there are also negative relationships found in certain periods. This means that an increase in R&D intensity causes a decrease in market value. Possible explanations for this are that R&D spending is coupled with risk. Investments in R&D can be unsuccessful. Also, there exist time lags between the R&D process and the commercialization of R&D.

In my expectation these changes in market responses are reflected in R&D intensities. During some periods, R&D negatively affects market value, which explains the decrease of R&D intensity. But since the elasticities over the whole period are positive, I expect the intensity to increase for this period. However, I find that this is not the case. The intensities decrease in every period, even when there is a positive market response to R&D. In other words, for firms it is beneficial to increase their R&D investment, but instead of an increase in intensity, there is a decrease in every period for every industry. The reason for this is that operating expenses in total increase by more than R&D expenses. A reason for the decrease in R&D intensity could therefore be that other operating expenditures, for example marketing expenditure, become relatively more important, but further investigation should be pursued on the change of relative importance of other costs.

(21)

17

References

Bloch, Carter (2003). The Effect of R&D Expenditures on Stock Market Returns for Danish Firms. The Danish Centre for Studies in Research and Research Policy, retrieved from

http://www.afsk.au.dk.

Bloch, Carter (2005). R&D Investment and Internal Finance: the Cash Flow Effect. Economics of Innovation and New Technology, Vol. 14, No.3, pp. 213-223.

Bracker, Kevin and Krishnan Ramaya (2011). Examining the Impact of Research and Development Expenditures on Tobin’s Q. Academy of Strategic Management Journal, Vol. 10, pp. 63-79. Chan, Louis K.C., Josef Lakonishok, and Theodore Sougiannis (2001). The stock market valuation of

research and development expenditures. The Journal of Finance Vol. 56, No. 6, pp. 2431-2456.

Chauvin, Keith W. and Mark Hirschey (1993). Advertising, R&D Expenditures and the Market Value of the Firm. Financial Management, Vol. 22, No. 4, pp. 128-140.

Chen, Yu-Shan and Ke-Chiun Chang (2010). The Relationship Between a Firm’s Patent Quality and its Market Value – The Case of US Pharmaceutical Industry. Technological Forecasting & Social Change, No. 77, pp. 20-33.

Coad, Alex and Rekha Rao (2010). Firm Growth and R&D Expenditure. Economics of Innovation and New Technology, Vol. 19, No. 2, pp. 127-145.

Demirel, Pelin and Mariana Mazzucato (2012). Innovation and Firm Growth: Is R&D Worth it? Industry and Innovation, Vol. 19, No. 1, pp. 45-62.

Deng, Zhen and Baruch Lev (2006). In-Process R&D: To Capitalize or Expense? Journal of Engineering and Technology Management, Vol. 23, No. 1, pp. 18-32.

DiMasi, Joseph A., Henry G. Grabowski and Ronald W. Hansen (2016). Innovation in the

Pharmaceutical Industry: New Estimates of R&D costs. Journal of Health Economics, Vol. 47, pp. 20-33.

DiMasi, Joseph A., Ronald W. Hansen, Henry G. Grabowski (2003). The Price of Innovation: New Estimates of Drug Development Costs. Journal of Health Economics, Vol. 22, No. 2, pp. 151-185.

Doukas, J. and L.N. Switzer (1992). The Stock Market's View of R&D Spending and Market Concentration. Journal of Economics and Business, pp. 95-114.

(22)

18 Eberhart, Allan C., William F. Maxwell and Akhtar R. Siddique (2004). An Examination of Long-Term

Abnormal Stock Returns and Operating Performance Following R&D Increases. The Journal of Finance, Vol. 59, No. 2, pp. 623-650.

Ehie, Ike C. and Kingsley Olibe (2010). The Effect of R&D Investment on Firm Value: An Examination of US Manufacturing and Service Industries. International Journal of Production Economics, Vol. 128,

pp. 127-135.

Ettlie, John E (1998). R&D and Global Manufacturing Performance. Management Science, Vol. 44, No. 1, pp. 1-11.

Fama, Eugene F (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance, Vol. 25, No. 2, pp. 383-417.

Fama, Eugene F. and Kenneth R. French (1993). Common Risk Factors in Returns on Stocks and Bonds. Journal of Financial Economics, Vol. 33, No. 1, pp. 3-56.

Franzen, Laurel, Kimberly J. Rodgers and Timothy T. Simin (2007). The Journal of Finance Vol. 62, No. 6, pp. 2931-2967.

Gollotto, Jamie C. and Sungsoo Kim (2003). Market valuation of dot.com companies: R&D versus hype. Managerial Finance, Vol. 29, No. 1, pp. 61-72.

Griliches, Zvi (1986). Productivity, R&D, and Basic Research at the Firm Level in the 1970's. The American Economic Review, Vol. 76, pp. 141-54.

Griliches, Zvi (1981). Market value, R&D, and patents. Economic letters, Vol. 7, No. 2, pp. 183-187. Hall, Bronwyn H. (1993). The Stock Market’s Valuation of R&D Investment during the 1980’s. The

American Economic Review, Vol. 83, No. 2, pp. 259-264.

Hall, Bronwyn H. (2000). Innovation and Market value. Cambridge University Press.

Hall, Bronwyn H. and J. Lerner (2010). The Financing of R&D and Innovation. Handbook of the Economics of Innovation, Vol. 1, pp. 609-630.

Hall, Bronwyn H., Jacques Mairesse and Pierre Mohnen (2010). Measuring the Returns to R&D. Handbook of the Economics of Innovation, Vol. 2, pp. 1033-1082.

(23)

19 Haneda, S. and H. Odagiri (1998). Appropriation of Returns From Technological Assets and the

Values of Patents and R&D in Japanese High-Tech Firms. Economics of Innovation and New Technology, Vol. 7, No. 4, pp. 303-321.

Hernández, H., A. Tübke, F. Hervás, A. Vezzani, M. Dosso, S. Amoroso, N. Grassano (2014). The 2014 EU Industrial R&D Investment Scoreboard. Publications Office of the European Union. Hirschey, Mark (1982). Intangible Capital Aspects of advertising and R&D Expenditures. The Journal

of Industrial Economics, Vol. 30, No. 4, pp. 375-390.

Hirschey, Mark and Jerry J. Weygandt (1985). Amortization Policy for Advertising and Research and Development Expenditures. Journal of Accounting Research, Vol. 23, No. 1, pp. 326-335. Hirschey, Mark, Hilla Skiba and M. Babajide Wintoki (2012). The Size, Concentration and Evolution of

Corporate R&D Spending in U.S. Firms from 1976 to 2010: Evidence and Implications. Journal of Corporate Finance, Vol. 18, No. 3, pp. 496-518.

Ho, Yew K., Zhenyu Xu and Chee M. Yap (2004). R&D Investment and Systematic Risk. Accounting and Finance, Vol. 44, pp. 393-418.

Ho, Y.K., Keh, H.T., Ong, J.M. (2005). The effect of R&D and advertising on firm value: an examination of manufacturing and nonmanufacturing firms. IEEE Transactions on Engineering

Management Vol. 52, pp. 3–14.

Hsiao, Cheng (1985). Benefits and Limitations of Panel Data. Econometric Reviews, Vol. 4, No. 1, pp. 121-174.

Hsieh, Ping-Hung (2003). The Return on R&D Versus Capital Expenditures in Pharmaceutical and Chemical Industries. IEEE Transactions on Engineering Management, Vol. 50, No. 2, pp. 141-150.

Hu, Albert G (2001). Ownership, Government R&D, Private R&D, and Productivity in Chinese Industry. Journal of Comparative Economics, Vol. 29, No.1, pp. 136-157.

Huang, Ning and Erwin Diewert (2011). Estimation of R&D Depreciation Rates: A Suggested

Methodology and Preliminary Application. Canadian Journal of Economics, Vol. 44, No. 2, pp. 387-412.

Johnson, Orace (1967). A Consequential Approach to Accounting for R&D. Journal of Accounting Research, Vol. 5, No. 2, pp. 164-172.

(24)

20 Lager, Thomas (2002). Product and Process Development Intensity in Process Industry: A Conceptual

and Empirical Analysis of the Allocation of Company Resources for the Development of Process Technology. International Journal of Innovation Management, Vol. 6, No. 2, pp. 105-139.

Lev, Baruch, and Theodore Sougannis (1999). Penetrating the Book-to-Market Black Box: The R&D Effect. Journal of Business Finance & Accounting, Vol. 26, No. 3.

Lin B. W., Y. Lee and S. C. Huang (2006). R&D intensity and commercialization orientation effects on financial performance. Journal of Business Research Vol. 59, No. 6, pp. 679-685.

Milburn, A. (1971). An Empirical Study of the Relationship of Research and Development Expenditures to Subsequent Benefits. Department of Accountancy, University of Illinois. Pakes, Ariel and Mark Schankerman (1984). The Rate of Obsolescence of Patents, Research

Gestation Lags, and the Private Return to Research Resources. University of Chicago Press, pp. 209-23.2

Reichstein, Toke and Ammon Salter (2006). Investigating the Sources of Process Innovation Among UK Manufacturing Firms. Industrial and Corporate Change, Vol. 15, No. 4, pp. 653-682. Sougiannis, Theodore (1994). The Accounting Based Valuation of Corporate R&D. The Accounting

Review, Vol. 69, No. 1, pp. 44-68.

Torres-Reyna, O (2007). Panel Data Analysis: Fixed and Random Effects Using STATA, Princeton Univesity, retrieved from: http://dss.princeton.edu/training.

Wang, C.H., Y. H. Lu, C.W. Huang and J.Y. Lee (2013). R&D, productivity, and market value: An empirical study from high-technology firms. Omega, Vol. 41, No. 1, pp. 143-155.

The market value of R&D : a study of the effects on high-tech companies

Faculty of Economics and Business