- Email: [email protected]
R&D investment–cash flow sensitivity under managerial optimism
Accepted Manuscript R&D investment cash flow sensitivity under managerial optimism Ezzeddine Ben Mohamed, Mohammed Abdelshakour Shehata PII: DOI: Reference:
S2214-6350(17)30001-1 http://dx.doi.org/10.1016/j.jbef.2017.02.001 JBEF 98
To appear in:
Journal of Behavioral and Experimental Finance
Received date: 3 May 2016 Revised date: 17 December 2016 Accepted date: 10 February 2017 Please cite this article as: Mohamed, E.B., Shehata, M.A., R&D investment cash flow sensitivity under managerial optimism. Journal of Behavioral and Experimental Finance (2017), http://dx.doi.org/10.1016/j.jbef.2017.02.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
*Abstract
## % " ! % $ $$
"$ " $ !
! ! ! " ! "!$ " # !$ # !$$ !
A bstract The aim of this paper is to explore the effect of managerial optimism on the R&D cash-flow (hereafter, R &D ICF) sensitivity. Departing from 864 yearly observations between 108 public firms listed at the NYSE from1999-2010, we construct a measure of managerial optimism as it described by Malmendier and Tate (2005) and we use a standard Q-model of investment. Our results report that firms with optimistic CEOs apply a strong positive and significant R &D ICF sensitivity. Running estimation for sub-sample firms, we find that the sensitivity of R &D investment to cash flows is stronger for more constrained group than the less constrained group. K ey words: Managerial optimism, R &D expenditure, ICF sensitivity, financial constraints. J E L C lassification: G02, G30, G31 and G32
*Manuscript
R & D Investment C ash F low Sensitivity under M anagerial O ptimism
A bstract The aim of this paper is to explore the effect of managerial optimism on the R&D cash-flow (hereafter, R &D ICF) sensitivity. Departing from 864 yearly observations between 108 public firms listed at the NYSE from1999-2010, we construct a measure of managerial optimism as it described by Malmendier and Tate (2005) and we use a standard Q-model of investment. Our results report that firms with optimistic CEOs apply a strong positive and significant R &D ICF sensitivity. Running estimation for sub-sample firms, we find that the sensitivity of R &D investment to cash flows is stronger for more constrained group than the less constrained group. K ey words: Managerial optimism, R &D expenditure, ICF sensitivity, financial constraints. J E L C lassification: G02, G30, G31 and G32
1
Introduction Heaton (2002) argues that corporate investment distortions are led to managerial optimism. Optimistic CEOs perceive the capital market undervaluing their firms, and so they will be reluctant to issue new equity. Optimistic CEOs can reject positive NPV projects when internal funds are exhausted, and firms are constrained. While they have the tendency to invest more than non-optimistic do with the availability of internal funds. In a direct extension and test of these predictions, Malmendier and Tate (2005) empirically demonstrate that the ICF sensitivity is caused by managerial optimism bias. Similar results are also reported by Lin et al, (2005) and Campbel et al, (2009). However, the effect of managerial optimism on R &D ICF sensitivity is still unexplored. It is for interest to discuss the potential effect of this bias on spending in a specific asset such as the R &D activities. In fact, the transaction costs' theory as it pioneered by Coase (1937) and developed by Williamson (1988) evokes the concept of a specific asset. According to Williamson (1985), asset specificity refers to assets that once in place would be costly to redeploy to other activities in case of a contract breaks down because there would be a loss of productive value. According to this theory, R & D spending, as a specific asset, increases transaction costs on debt financing. For this, firms should prefer equity financing among debt; this is in order to avoid bankruptcy (Bah et al., 2001). If managers are optimistic, they will see their firms always as under valuated by the stock markets (Heaton, 2002). In terms of financial strategy, this means that external financing will be perceived as very costly compared with internal cash flow financing mode. The use of equity to finance a specific investment such as R & D spending will be seen as high costly and so they should return down to their internal source of financing. Then, the R &D ICF sensitivity should be more pronounced in presence of optimistic CEOs. The R &D ICF sensitivity will be more pronounced in the presence of financial constraints as it demonstrated by Heaton, (2002). This paper is an extension of previous works in ICF sensitivity under managerial optimism, especially that of Malmendier and Tate (2005). For this, we will reconsider a similar hypothesis to explore the effect of managerial optimism bias on R &D spending.
2
This paper is divided into four additional sections. The second section, deals with methodological details and variables' measurement. The third section includes data description. Section four, shows our results. Finally, the last section offers concluding remarks and discusses implications from our findings. 2. Methodology and empirical specification We conserve the common methodology based on Q-model of investment, which is applied by previous works focusing on investment-cash flow sensitivity under behavioral corporate finance (Malmendier and Tate, 2005; Lin et al., 2005).
(1) Where I denotes R&D expenditure, Q is the market value to its replacement value; C F stands for internal cash flows; X
is a dummy variable that is used as a proxy for managerial optimism. Following Malmendier and Tate (2005), we propose a measure based on CEOs first five stock net purchases. CEO will be classified as optimistic if he bought stock on net at least one more year than sold stock on net during his first five years, and he should also increase his ownership by at least ten percent of their stock ownership in his firm in a given year. R&D expenditure represents all direct and indirect costs related with the creation and development of new processes, techniques, applications and products with commercial possibilities. We calculate cash flow as Earnings Before Interest, Taxes and Depreciation (EBITDA). Q is the market value of assets over the book value of assets from the beginning of the year. We use the board independence (IND) and the board size (BSIZE). We construct an indicator of efficient board size as it advanced by Malmendier and Tate (2005). Board is efficient if its size is fewer than 12 members. The independence of the board is directly computed by the number of outside members. We control also for the effect of ownership structure by introducing the effect of CEOs' ownership in their firms. A standard empirical approach to study the ICF sensitivity is to estimate a fixed effect regression as it described by equation (1), and then we re-estimate our model for more and less financially constraint firms. For this, we split the full sample into sub-samples following the standard literature of financial constraints to construct sub-firms groups from more constrained to less constrained firms. We adopt two classifications to detect if a firm is 3
financially constrained or not. The first classif " . We follow Gertler and Gilchrist, (1994) to classify firms into more or less constrained. The second classification is inspired from Lin et al., (2005); we use the interest coverage to proxy for firms' financing ability. For each firm, we calculate its average interest coverage, the ratio of interest expense to the sum of the interest expense and the cash flow. We rank firms from small to large, and we define the smaller 50% as less constrained while, the larger 50% are defined as more constrained. 3. Data description Our data consist of 864 annual observations concerning 108 large public industrial American firms traded at the NYSE between 1999 and 2010. A quasi-random sampling procedure was applied. In fact, only firms with available internal transactions were selected. This is because we need such transaction in order to construct our main variable; the optimism bias. We use different information sources: (1) We use the Thomson Financial database in order to
! (2) We use the Thomson World scope in order to determinate other variables such as R&D expenditures1, cash flow and information to
" are deriving from SEC financial database and Thomson Reuters. Finally, corporate governance variables are from SEC Financial database and Thomson Reuters-Officers and Directors. 4. E mpirical results Using fixed effect panel with OLS regressions, our results show that the coefficient of R &D ICF sensitivity is positive and significant at the first-percent level. This means that optimistic CEOs will invest more in R &D activities when internal cash flows are ample. Our results corroborate previous findings by Malmendier and Tate (2005). Colum 2, 3 and 4 of table two report results of regression of our model using a fixed panel effect with a proxy of board independence (IND). We find that the effect of managerial optimism still positive, and an independent board may succeed to reduce the R &D ICF sensitivity. Our result is robust to the introduction of firm size and ownership as control variables.
1
R&D expenditure are available from Worldscope Suplementary Report, Annual Item ; F ield 01201, included on Thomson Worldscope.
4
Table three reports results from the estimation of the Q-model of investment using two subsamples constructed using the interest coverage criterion. Managerial optimism causes R &D ICF sensitivity phenomena for both groups. However, the coefficient of R & D ICF sensitivity is robust for the more constrained firms and this even with the inclusion or not of control variables. Without control variables, this relationship is intense with a value of 0.067 and significant at five-percent level (t =2.088). We obtain a similar result when we re-estimate our model for the inclusion of board independence. The coefficient between R &D expenditure and cash flow jumps to a value of 0.091 and it is significant at the five-percent level (t=2.43). In contrast, less constrained firms apply lower R &D ICF sensitivity to the presence of managerial optimism. With fixed effect and no control variable, we obtain a positive and significant coefficient, but it is less intense, when it is compared by the same coefficient at the first sub-sample. It is equal to 0.065 and strongly significant (t= 2.63). Our empirical finding concerning the effect of financial constraints on R &D ICF sensitivity still robust: the coefficient of sensitivity in less constrained firms is lower than that for more constrained firms. It is 0.071 and significant at the first percent level (t=2.69). Table four generates results of regression of our model using firm size as a criterion to construct sub-samples of more to less constrained firms. The more constrained
&D investment is sensitive to their internal cash flow. The R &D ICF sensitivity coefficient for more constrained firms varies from 0.034 which is significant at the five-percent level to 0.065 with a signification at the five-percent level. We find also that the board independence exerts a negative effect on the considered relationship. We can detect also that more constrained firms have the most sensitive R &D to cash flow. In contrast, large firms are less exposed to sensitivity because they are assumed to be less financially constrained. The coefficient of R &D ICF sensitivity is weak, . Conclusion Beyond the traditional theories that have reduced R &D ICF sensitivity primarily to agency costs and asymmetric information problems, we find that the optimism bias increases R&D ICF sensitivity. Our result is robust when introducing a multitude of control variables. An interesting empirical finding in this paper is that this sensitivity is greater for firms with financial constraints. Optimistic CEOs align their investment strategies with the internal finance availability level.
5
Even with the specificity of investment in R&D, which is a specific investment, with managerial optimism, CEOs will always prefer to finance such investment by internal funds. Finally, firms must pay a special attention to their internal corporate governance system. So they can overcome possible distortions of their investments in R&D, which is caused by the irrationality of their principal agents. References Bah, R., Dumontier, P., 2001. R&D intensity and corporate financial policy: some international evidence. Journal of Business Finance and Accounting, 28 (5;6), 671-692. Broussard, J.P., Buchenroth, S.A., Pilotte, E.A., 2004. CEO incentives, cash flow, and investment. Financial Management 33(2), 51-70. Campbell, C., Johnson, S., Rutherford, J., Stanley, B. (2011). CEO confidence and forced turnover. Journal of Financial Economics, 101(3), 695712. Charles J. H., 1998. Ownership, liquidity, and investment. RAND Journal of Economics, 29(3), 487-508. Coase, R. H., 1937. The nature of the firm. Economica, 4(16), 386-405. Gertler, M., Gilchrist, S., 1994. Monetary Policy Business Cycles, and the Behavior of Small Manufacturing Firms. Quarterly Journal of Economics, 109(2), 309-340. Glaser, M., Weber, M., 2007. Overconfidence and trading volume. Geneva Risk Insurance Review, 32(1), 1-36.
and
Heaton, J.B., 2002. Managerial optimism and corporate finance. Financial Management, 31(2), 33-45. Lin, Y., Hu, S., Chen, M., 2005. Managerial optimism and corporate investment: Some empirical evidence from Taiwan. Pacific-Basin Finance Journal 13(5), 523 546. Malmendier, U., Tate, G., 2005. Does overconfidence affect corporate investment? CEO overconfidence measures revisited. European Financial Management, 11(5), 649-59. Puri, M., Robinson, D.T., 2007. Optimism and economic choice. Journal of Financial Economics , 86 (1), 71-99. Williamson, O.E, 1988. The logic of economic organization. Journal of Law, Economics, & Organization, 4(1), 65-93.
6
Table 1 Full sample: summary of descriptive statistics Panel A: CEO's data Years as CEO CEO and President and Chairman CEO ownership Panel B: Firms' data R&D expenditure($M) Q Cash flow($M) Panel C:classifications' criteria Interest coverage Firm size Panel D: Managerial optimism Managerial optimism Optimistic CEOs (%)
Observations 778 832 630
Mean 13.303 0.673 0.010
Median 12 1 0.002
864 864 864
0.175 1.173 1.104
0.018 1.120 0.152
744 797
0.102 1.328
0.097 1.280
864
0.791
1 0.76
7
Standard deviation 7.300 0.469 0.025
Minimum 8 0 0.000
Maximum 59 1 0.206
0.572 0.567 4.904
0 0.003 -3.512
6.506 4.878 60.908
0.161 74.601
-0.5308 0.928
0.9130 696.504
0
1
0.406
Table 2 OLS regression of investment on cash flow and optimism measure
Intercept × 100 Cash flow Q × 100 Optimism measure × 100 Cash flow × Q Cash flow × Optimism measure IND. IND × Cash flow BSIZE BZIS E × Cash flow
Optimism Fixed effect No control 6.91 (3.68)*** -0.003 (-0.597) 1.58 (8.52)*** -0.80 (-4.314)*** -0.037 (-4.66)*** 0.089 (12.29)***
Optimism Fixed effect Control: IND 0.2 (0.367) 0.021 (0.964) 1.6 (8.72)*** -0.8 (-4.74)*** -0.039 (-5.11)*** 0.091 (14.16)*** 0.05 (1.011) -0.0025 (-1.055)
Optimism Fixed effect Control: TCA 0.02 (0.042) -0.05 (-1.37) 1.3 (7.54)*** -0.7 (-3.64)*** -0.032 (-3.63)*** 0.072 (8.77)***
Optimism Fixed effect Control: OWN 3.83 (1.167) 0.034 (1.89)*** 1.8 (7.84)*** -0.5 (-4.03)*** -0.044 (-4.52)*** 0.066 (8.98)***
0.05 (1.42) -0.32 (-1.52) -0.03 (-2.36)** -0.55 (-4.18)*** 0.13 709
OWN OWN × Cash flow Adjusted R-squared 0.088 0.087 0.13 Observations 784 758 758 *** ** * , and denote that results are significant at the 1%, 5% and 10% levels
8
Table 3: Sub-samples using interest coverage
More constrained firms Fixed effect No control 1.50 Intercept × 100 (3.10)*** -0.042 Cash flow (-1.43) 1.08 Q × 100 (3.58)*** -0.86 Optimism measure × 100 (-1.89)* -0.022 Cash flow × Q (-2.97)*** Cash flow × Optimism 0.067 (2.088)** measure IND. IND × Cash flow Adjusted R2 observations
0.04 389
Less constrained firms Fixed effect No control 0.05 (0.078) 0.005 (0.211) 0.63 (2.25)** 1.03 (1.84)* -0.01 (-1.05) 0.065 (2.63)***
Fixed effect With control -0.93 (-1.04) 0.086 (1.48) 1.16 (3.75)*** -1.11 (-2.23)** -0.031 (-3.62)*** 0.091 (2.43)** 0.0029 (3.15)*** -0.015 (-2.31)** 0.07 380
0.17 355
*** **
, and * denote that results are significant at the 1%, 5% and 10% levels
9
Fixed effect With control -0.03 (-0.037) -0.006 (-0.13) 0.55 (1.87)* 0.71 (1.12) -0.031 (-4.17)*** 0.071 (2.69)*** 0.0006 (0.84) 0.0005 (0.14) 0.17 343
More constrained firms
Intercept × 100
Less constrained firms
Fixed effect No control
Fixed effect With control
Fixed effect No control
Fixed effect With control
-0.85 (-1.83)*
-1.8 (-2.86)***
1.40 (3.83)***
0.89 (0.88)
Table 4:Sub-sample using firms size criteria
10
0.27 (1.26) 1.91 Q × 100 (6.68)*** 1.05 Optimism measure × 100 (2.67)*** -0.04 Cash flow × Q (-4.75)***
0.077 (2.45)** 1.91 (6.49)*** 0.67 (1.53) -0.04 (-4.95)***
-0.009 (-0.69) 0.43 (1.30) -0.71 (-1.47) 0.06 (3.01)***
0.02 (0.24) 0.52 (1.41) -0.71 (-1.42) 0.06 (2.38)**
Cash flow × Optimism measure
0.065 (2.41)**
-0.009 (-0.27)
-0.009 (-0.27)
Cash flow
0.034 (1.73)*
IND × Cash flow Adjusted R2 observations
0.0004 (0.48)
0.0017 (2.21)** -0.0091 (-2.022)**
IND.
0.18 408
-0.002 (-0.36)
0.12 389
0.19 405
*** **
, and * denote that results are significant at the 1%, 5% and 10% levels
11
0.12 368
S2214-6350(17)30001-1 http://dx.doi.org/10.1016/j.jbef.2017.02.001 JBEF 98
To appear in:
Journal of Behavioral and Experimental Finance
Received date: 3 May 2016 Revised date: 17 December 2016 Accepted date: 10 February 2017 Please cite this article as: Mohamed, E.B., Shehata, M.A., R&D investment cash flow sensitivity under managerial optimism. Journal of Behavioral and Experimental Finance (2017), http://dx.doi.org/10.1016/j.jbef.2017.02.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
*Abstract
## % " ! % $ $$
"$ " $ !
! ! ! " ! "!$ " # !$ # !$$ !
A bstract The aim of this paper is to explore the effect of managerial optimism on the R&D cash-flow (hereafter, R &D ICF) sensitivity. Departing from 864 yearly observations between 108 public firms listed at the NYSE from1999-2010, we construct a measure of managerial optimism as it described by Malmendier and Tate (2005) and we use a standard Q-model of investment. Our results report that firms with optimistic CEOs apply a strong positive and significant R &D ICF sensitivity. Running estimation for sub-sample firms, we find that the sensitivity of R &D investment to cash flows is stronger for more constrained group than the less constrained group. K ey words: Managerial optimism, R &D expenditure, ICF sensitivity, financial constraints. J E L C lassification: G02, G30, G31 and G32
*Manuscript
R & D Investment C ash F low Sensitivity under M anagerial O ptimism
A bstract The aim of this paper is to explore the effect of managerial optimism on the R&D cash-flow (hereafter, R &D ICF) sensitivity. Departing from 864 yearly observations between 108 public firms listed at the NYSE from1999-2010, we construct a measure of managerial optimism as it described by Malmendier and Tate (2005) and we use a standard Q-model of investment. Our results report that firms with optimistic CEOs apply a strong positive and significant R &D ICF sensitivity. Running estimation for sub-sample firms, we find that the sensitivity of R &D investment to cash flows is stronger for more constrained group than the less constrained group. K ey words: Managerial optimism, R &D expenditure, ICF sensitivity, financial constraints. J E L C lassification: G02, G30, G31 and G32
1
Introduction Heaton (2002) argues that corporate investment distortions are led to managerial optimism. Optimistic CEOs perceive the capital market undervaluing their firms, and so they will be reluctant to issue new equity. Optimistic CEOs can reject positive NPV projects when internal funds are exhausted, and firms are constrained. While they have the tendency to invest more than non-optimistic do with the availability of internal funds. In a direct extension and test of these predictions, Malmendier and Tate (2005) empirically demonstrate that the ICF sensitivity is caused by managerial optimism bias. Similar results are also reported by Lin et al, (2005) and Campbel et al, (2009). However, the effect of managerial optimism on R &D ICF sensitivity is still unexplored. It is for interest to discuss the potential effect of this bias on spending in a specific asset such as the R &D activities. In fact, the transaction costs' theory as it pioneered by Coase (1937) and developed by Williamson (1988) evokes the concept of a specific asset. According to Williamson (1985), asset specificity refers to assets that once in place would be costly to redeploy to other activities in case of a contract breaks down because there would be a loss of productive value. According to this theory, R & D spending, as a specific asset, increases transaction costs on debt financing. For this, firms should prefer equity financing among debt; this is in order to avoid bankruptcy (Bah et al., 2001). If managers are optimistic, they will see their firms always as under valuated by the stock markets (Heaton, 2002). In terms of financial strategy, this means that external financing will be perceived as very costly compared with internal cash flow financing mode. The use of equity to finance a specific investment such as R & D spending will be seen as high costly and so they should return down to their internal source of financing. Then, the R &D ICF sensitivity should be more pronounced in presence of optimistic CEOs. The R &D ICF sensitivity will be more pronounced in the presence of financial constraints as it demonstrated by Heaton, (2002). This paper is an extension of previous works in ICF sensitivity under managerial optimism, especially that of Malmendier and Tate (2005). For this, we will reconsider a similar hypothesis to explore the effect of managerial optimism bias on R &D spending.
2
This paper is divided into four additional sections. The second section, deals with methodological details and variables' measurement. The third section includes data description. Section four, shows our results. Finally, the last section offers concluding remarks and discusses implications from our findings. 2. Methodology and empirical specification We conserve the common methodology based on Q-model of investment, which is applied by previous works focusing on investment-cash flow sensitivity under behavioral corporate finance (Malmendier and Tate, 2005; Lin et al., 2005).
(1) Where I denotes R&D expenditure, Q is the market value to its replacement value; C F stands for internal cash flows; X
is a dummy variable that is used as a proxy for managerial optimism. Following Malmendier and Tate (2005), we propose a measure based on CEOs first five stock net purchases. CEO will be classified as optimistic if he bought stock on net at least one more year than sold stock on net during his first five years, and he should also increase his ownership by at least ten percent of their stock ownership in his firm in a given year. R&D expenditure represents all direct and indirect costs related with the creation and development of new processes, techniques, applications and products with commercial possibilities. We calculate cash flow as Earnings Before Interest, Taxes and Depreciation (EBITDA). Q is the market value of assets over the book value of assets from the beginning of the year. We use the board independence (IND) and the board size (BSIZE). We construct an indicator of efficient board size as it advanced by Malmendier and Tate (2005). Board is efficient if its size is fewer than 12 members. The independence of the board is directly computed by the number of outside members. We control also for the effect of ownership structure by introducing the effect of CEOs' ownership in their firms. A standard empirical approach to study the ICF sensitivity is to estimate a fixed effect regression as it described by equation (1), and then we re-estimate our model for more and less financially constraint firms. For this, we split the full sample into sub-samples following the standard literature of financial constraints to construct sub-firms groups from more constrained to less constrained firms. We adopt two classifications to detect if a firm is 3
financially constrained or not. The first classif " . We follow Gertler and Gilchrist, (1994) to classify firms into more or less constrained. The second classification is inspired from Lin et al., (2005); we use the interest coverage to proxy for firms' financing ability. For each firm, we calculate its average interest coverage, the ratio of interest expense to the sum of the interest expense and the cash flow. We rank firms from small to large, and we define the smaller 50% as less constrained while, the larger 50% are defined as more constrained. 3. Data description Our data consist of 864 annual observations concerning 108 large public industrial American firms traded at the NYSE between 1999 and 2010. A quasi-random sampling procedure was applied. In fact, only firms with available internal transactions were selected. This is because we need such transaction in order to construct our main variable; the optimism bias. We use different information sources: (1) We use the Thomson Financial database in order to
! (2) We use the Thomson World scope in order to determinate other variables such as R&D expenditures1, cash flow and information to
" are deriving from SEC financial database and Thomson Reuters. Finally, corporate governance variables are from SEC Financial database and Thomson Reuters-Officers and Directors. 4. E mpirical results Using fixed effect panel with OLS regressions, our results show that the coefficient of R &D ICF sensitivity is positive and significant at the first-percent level. This means that optimistic CEOs will invest more in R &D activities when internal cash flows are ample. Our results corroborate previous findings by Malmendier and Tate (2005). Colum 2, 3 and 4 of table two report results of regression of our model using a fixed panel effect with a proxy of board independence (IND). We find that the effect of managerial optimism still positive, and an independent board may succeed to reduce the R &D ICF sensitivity. Our result is robust to the introduction of firm size and ownership as control variables.
1
R&D expenditure are available from Worldscope Suplementary Report, Annual Item ; F ield 01201, included on Thomson Worldscope.
4
Table three reports results from the estimation of the Q-model of investment using two subsamples constructed using the interest coverage criterion. Managerial optimism causes R &D ICF sensitivity phenomena for both groups. However, the coefficient of R & D ICF sensitivity is robust for the more constrained firms and this even with the inclusion or not of control variables. Without control variables, this relationship is intense with a value of 0.067 and significant at five-percent level (t =2.088). We obtain a similar result when we re-estimate our model for the inclusion of board independence. The coefficient between R &D expenditure and cash flow jumps to a value of 0.091 and it is significant at the five-percent level (t=2.43). In contrast, less constrained firms apply lower R &D ICF sensitivity to the presence of managerial optimism. With fixed effect and no control variable, we obtain a positive and significant coefficient, but it is less intense, when it is compared by the same coefficient at the first sub-sample. It is equal to 0.065 and strongly significant (t= 2.63). Our empirical finding concerning the effect of financial constraints on R &D ICF sensitivity still robust: the coefficient of sensitivity in less constrained firms is lower than that for more constrained firms. It is 0.071 and significant at the first percent level (t=2.69). Table four generates results of regression of our model using firm size as a criterion to construct sub-samples of more to less constrained firms. The more constrained
&D investment is sensitive to their internal cash flow. The R &D ICF sensitivity coefficient for more constrained firms varies from 0.034 which is significant at the five-percent level to 0.065 with a signification at the five-percent level. We find also that the board independence exerts a negative effect on the considered relationship. We can detect also that more constrained firms have the most sensitive R &D to cash flow. In contrast, large firms are less exposed to sensitivity because they are assumed to be less financially constrained. The coefficient of R &D ICF sensitivity is weak, . Conclusion Beyond the traditional theories that have reduced R &D ICF sensitivity primarily to agency costs and asymmetric information problems, we find that the optimism bias increases R&D ICF sensitivity. Our result is robust when introducing a multitude of control variables. An interesting empirical finding in this paper is that this sensitivity is greater for firms with financial constraints. Optimistic CEOs align their investment strategies with the internal finance availability level.
5
Even with the specificity of investment in R&D, which is a specific investment, with managerial optimism, CEOs will always prefer to finance such investment by internal funds. Finally, firms must pay a special attention to their internal corporate governance system. So they can overcome possible distortions of their investments in R&D, which is caused by the irrationality of their principal agents. References Bah, R., Dumontier, P., 2001. R&D intensity and corporate financial policy: some international evidence. Journal of Business Finance and Accounting, 28 (5;6), 671-692. Broussard, J.P., Buchenroth, S.A., Pilotte, E.A., 2004. CEO incentives, cash flow, and investment. Financial Management 33(2), 51-70. Campbell, C., Johnson, S., Rutherford, J., Stanley, B. (2011). CEO confidence and forced turnover. Journal of Financial Economics, 101(3), 695712. Charles J. H., 1998. Ownership, liquidity, and investment. RAND Journal of Economics, 29(3), 487-508. Coase, R. H., 1937. The nature of the firm. Economica, 4(16), 386-405. Gertler, M., Gilchrist, S., 1994. Monetary Policy Business Cycles, and the Behavior of Small Manufacturing Firms. Quarterly Journal of Economics, 109(2), 309-340. Glaser, M., Weber, M., 2007. Overconfidence and trading volume. Geneva Risk Insurance Review, 32(1), 1-36.
and
Heaton, J.B., 2002. Managerial optimism and corporate finance. Financial Management, 31(2), 33-45. Lin, Y., Hu, S., Chen, M., 2005. Managerial optimism and corporate investment: Some empirical evidence from Taiwan. Pacific-Basin Finance Journal 13(5), 523 546. Malmendier, U., Tate, G., 2005. Does overconfidence affect corporate investment? CEO overconfidence measures revisited. European Financial Management, 11(5), 649-59. Puri, M., Robinson, D.T., 2007. Optimism and economic choice. Journal of Financial Economics , 86 (1), 71-99. Williamson, O.E, 1988. The logic of economic organization. Journal of Law, Economics, & Organization, 4(1), 65-93.
6
Table 1 Full sample: summary of descriptive statistics Panel A: CEO's data Years as CEO CEO and President and Chairman CEO ownership Panel B: Firms' data R&D expenditure($M) Q Cash flow($M) Panel C:classifications' criteria Interest coverage Firm size Panel D: Managerial optimism Managerial optimism Optimistic CEOs (%)
Observations 778 832 630
Mean 13.303 0.673 0.010
Median 12 1 0.002
864 864 864
0.175 1.173 1.104
0.018 1.120 0.152
744 797
0.102 1.328
0.097 1.280
864
0.791
1 0.76
7
Standard deviation 7.300 0.469 0.025
Minimum 8 0 0.000
Maximum 59 1 0.206
0.572 0.567 4.904
0 0.003 -3.512
6.506 4.878 60.908
0.161 74.601
-0.5308 0.928
0.9130 696.504
0
1
0.406
Table 2 OLS regression of investment on cash flow and optimism measure
Intercept × 100 Cash flow Q × 100 Optimism measure × 100 Cash flow × Q Cash flow × Optimism measure IND. IND × Cash flow BSIZE BZIS E × Cash flow
Optimism Fixed effect No control 6.91 (3.68)*** -0.003 (-0.597) 1.58 (8.52)*** -0.80 (-4.314)*** -0.037 (-4.66)*** 0.089 (12.29)***
Optimism Fixed effect Control: IND 0.2 (0.367) 0.021 (0.964) 1.6 (8.72)*** -0.8 (-4.74)*** -0.039 (-5.11)*** 0.091 (14.16)*** 0.05 (1.011) -0.0025 (-1.055)
Optimism Fixed effect Control: TCA 0.02 (0.042) -0.05 (-1.37) 1.3 (7.54)*** -0.7 (-3.64)*** -0.032 (-3.63)*** 0.072 (8.77)***
Optimism Fixed effect Control: OWN 3.83 (1.167) 0.034 (1.89)*** 1.8 (7.84)*** -0.5 (-4.03)*** -0.044 (-4.52)*** 0.066 (8.98)***
0.05 (1.42) -0.32 (-1.52) -0.03 (-2.36)** -0.55 (-4.18)*** 0.13 709
OWN OWN × Cash flow Adjusted R-squared 0.088 0.087 0.13 Observations 784 758 758 *** ** * , and denote that results are significant at the 1%, 5% and 10% levels
8
Table 3: Sub-samples using interest coverage
More constrained firms Fixed effect No control 1.50 Intercept × 100 (3.10)*** -0.042 Cash flow (-1.43) 1.08 Q × 100 (3.58)*** -0.86 Optimism measure × 100 (-1.89)* -0.022 Cash flow × Q (-2.97)*** Cash flow × Optimism 0.067 (2.088)** measure IND. IND × Cash flow Adjusted R2 observations
0.04 389
Less constrained firms Fixed effect No control 0.05 (0.078) 0.005 (0.211) 0.63 (2.25)** 1.03 (1.84)* -0.01 (-1.05) 0.065 (2.63)***
Fixed effect With control -0.93 (-1.04) 0.086 (1.48) 1.16 (3.75)*** -1.11 (-2.23)** -0.031 (-3.62)*** 0.091 (2.43)** 0.0029 (3.15)*** -0.015 (-2.31)** 0.07 380
0.17 355
*** **
, and * denote that results are significant at the 1%, 5% and 10% levels
9
Fixed effect With control -0.03 (-0.037) -0.006 (-0.13) 0.55 (1.87)* 0.71 (1.12) -0.031 (-4.17)*** 0.071 (2.69)*** 0.0006 (0.84) 0.0005 (0.14) 0.17 343
More constrained firms
Intercept × 100
Less constrained firms
Fixed effect No control
Fixed effect With control
Fixed effect No control
Fixed effect With control
-0.85 (-1.83)*
-1.8 (-2.86)***
1.40 (3.83)***
0.89 (0.88)
Table 4:Sub-sample using firms size criteria
10
0.27 (1.26) 1.91 Q × 100 (6.68)*** 1.05 Optimism measure × 100 (2.67)*** -0.04 Cash flow × Q (-4.75)***
0.077 (2.45)** 1.91 (6.49)*** 0.67 (1.53) -0.04 (-4.95)***
-0.009 (-0.69) 0.43 (1.30) -0.71 (-1.47) 0.06 (3.01)***
0.02 (0.24) 0.52 (1.41) -0.71 (-1.42) 0.06 (2.38)**
Cash flow × Optimism measure
0.065 (2.41)**
-0.009 (-0.27)
-0.009 (-0.27)
Cash flow
0.034 (1.73)*
IND × Cash flow Adjusted R2 observations
0.0004 (0.48)
0.0017 (2.21)** -0.0091 (-2.022)**
IND.
0.18 408
-0.002 (-0.36)
0.12 389
0.19 405
*** **
, and * denote that results are significant at the 1%, 5% and 10% levels
11
0.12 368