Nội dung

Journal of Economics and Development, Vol.18, No.2, August 2016, pp. 59-70 ISSN 1859 0020 Risk Attitude and Corporate Investment under Output Market Uncertainty: Evidence from The Mekong River Delta, Vietnam Le Khuong Ninh Can Tho University, Vietnam Email: lekhuongninh@gmail.com Le Tan Nghiem Can Tho University, Vietnam Email: letannghiem@gmail.com Huynh Huu Tho Can Tho University, Vietnam Email: huynhhuuthoct@gmail.com Abstract This paper aims to detect the impact of firm managers’ risk attitude on the relationship between the degree of output market uncertainty and firm investment. The findings show that there is a negative relationship between these two aspects for risk-averse managers while there is a positive relationship for risk-loving ones, since they have different utility functions. Based on the findings, this paper proposes recommendations for firm managers to take into account when making investment decisions and long-term business strategies as well. Keywords: Competition; corruption; investment; market uncertainty; risk attitude. Journal of Economics and Development 59 Vol. 18, No.2, August 2016 1. Introduction structured as follows. Section 1 introduces the paper. Section 2 gives a review of the related literature. Section 3 defines the empirical model out of the literature reviewed. Section 4 discusses the findings using a set of primary data on 667 non-state firms in the MRD. Section 6 concludes the paper and renders recommendations. Investment is crucial to the development of firms since it helps enhance product quality and increase market share. Thus, good investment decisions will raise firms’ efficiency and then trigger economic growth (Maki et al., 2005). However, making right investment decisions is basically difficult, due to the output market uncertainty facing firm managers, among others (e.g., competition and financing constraints). 2. Literature review When making investment decisions, firm managers do face output market uncertainty. To put it differently, they do not know the exact future sales. Thus, they tend to postpone investment in order to fetch more relevant information and determine the right time to invest (Berk, 1999). According to Nishihara and Shibata (2014), unless firms have to invest to preempt competitors, most investment projects can be postponed, since for most of the time, investment opportunities remain for a certain period prior to absolutely expiring. Indeed, having an investment opportunity (a real option) is analogous to owning a European call option to buy a stock. Then, the owner can exercise it right away, or later, to get a financial asset with a certain value (e.g., a stock). When possessing a real option (i.e., an investment opportunity), the firm can decide to invest right away or at any future point of time to obtain a real asset with a certain value (e.g., a factory). Like call options, the value of real options stems from the managerial flexibility in making use of the uncertainty about the future value of the real asset (Luehrman, 1998). Due to uncertainty, firm managers tend to wait for more information that helps to avoid failure. As well perceived, investment decisions are much dependent on firm managers’ risk attitude toward output market uncertainty (Bo and Sterken, 2007; Femminis, 2008). Being skeptical about the loss that may result from poor investment decisions, risk-averse managers tend to postpone investment intents so as to acquire more relevant information. In this situation, they possibly forgo good investment opportunities. On the other hand, risk-loving managers who are normally over-optimistic about their own competence and market prospect will proceed with investment opportunities, irrespective of their uncertain outcomes. This tendency is accentuated by successes in the past. Such over-optimistic behaviour may be problematic if output markets would somehow turn worse. Thus, investment decisions of both risk-averse and risk-loving managers seem to have drawbacks that should be avoided. The aim of this paper is to examine the impact of managers’ risk attitude on investment by non-state firms in the Mekong River Delta (MRD) under output market uncertainty. Findings of this paper will lay down a credible ground for recommendations that enable firms to make better investment decisions and proper long-term business strategies. This paper is Journal of Economics and Development Thus, researchers have tried to examine the impact of output market uncertainty on firm 60 Vol. 18, No.2, August 2016 ies pay attention to full-fledged investment decisions, thanks to the inevitable assertion that output market uncertainty affects investment via the channel of managers’ risk attitude (Nakamura, 1999; Bo and Sterken, 2007; Femminis, 2008; Chronopoulos et al., 2011; Whalley, 2011; Aistov and Kuzmicheva, 2012). According to them, risk-loving managers tend to accelerate investment as the degree of uncertainty goes up because of self-confidence, ambition to get over challenges and sanguineness about the future. For those managers, the satisfaction resulting from a success surely dominates the disappointment of failing. Thus, a higher degree of uncertainty will induce them to invest more. investment. Most of the empirical studies on this topic (Guiso and Parigi, 1999; Ghosal and Loungani, 2000; Le, Hermes and Lanjouw, 2004) came up with evidence of negative relationships between the uncertainty and firm investment. According to them, the higher the degree of output market uncertainty, the lower the level of investment, due to the fact that uncertainty may increase the user cost of investment. A higher degree of uncertainty also makes firm managers cautious in taking up investment projects because, in that case, it is hard to control and mitigate the adverse impact of market gyrations. As a result, investment will decline. However, these studies have ignored firm managers’ risk attitude or implicitly assumed that their risk attitude is virtually identical. On the other hand, risk-averse managers who do prefer certain values to uncertain ones will opt for investment projects with more certain profits. In terms of utility, risk-averse managers feel worse off if losing more, than better off if winning. Being skeptical about losing, they need time to acquire more relevant information before making investment decisions so as to minimize the possibility of failure and regret. Thus, investment will drop as the degree of output market uncertainty picks up. In other words, the relationship between output market uncertainty and investment depends on firm managers’ risk attitude. As a matter of fact, firm managers would belong to either risk-averse or risk-loving group of people, due to differences in utility and motivation, among others (Block et al., 2015). Thus, researchers started to examine the relationship between managers’ risk attitude and firm investment. For them, investment decisions of managers are normally aimed at maximizing expected profits rather than actual ones. Then, the utility function U(π) of a risk-averse manager is a concave curve of profit π, because of the law of diminishing marginal utility. On the other hand, the utility function U(π) of a risk-loving manager is a convex curve of profit π, due to the law of increasing marginal utility. As a result, investment decisions by those groups of managers largely diverge. 3. Empirical model Given the argument previously presented, the empirical model used to detect the impact of managers’ risk attitude on the relationship between the degree of output market uncertainty and firm investment is specified as follows: Different from those studies that just focus on single aspects of relevant issues (such as output market uncertainty, risk attitude, competition or financing constraints), recent studJournal of Economics and Development INVi = β0 + β1UNCERi + β2UNCERi x RISKi­ + β3RISKi­ + εi (1) 61 Vol. 18, No.2, August 2016 In Model (1), INVi is the ratio of planned investment in machinery, land and buildings to total fixed assets of firm i. UNCERi is the degree of output market uncertainty, measured by the coefficient of variation of expected sales of firm i (Guiso and Parigi, 1999; Le, Hermes and Lanjouw, 2004).1 Coefficient β1 is expected to be negative since the theory postulates that output market uncertainty may have a negative impact on firm investment. tainty and investment for risk-averse managers. For those managers, since RISKi = 0 then ∂INVi / ∂UNCERi = β1 . Thus, it is expected that β1<0 . For risk-loving managers, there is a positive impact. For those, since RISKi = 1 then ∂INVi / ∂UNCERi = β1 + β 2 . Therefore, β2 is supposed to be positive and |β2|>|β1|. εi is an error term. To be complete, the empirical model should include the determinants of investment identified by other studies (Bo and Lensink, 2005; Guiso and Parigi, 1999; Polder and Veldhuizen, 2012), such as retained profit, growth rate of sales, degree of competition, etc. Given these factors, the empirical model of this paper then becomes: RISKi is used to proxy for risk attitude of the top manager of firm i. To construct this variable, the manager was asked to choose between two hypothetical cases: (i) investing a certain amount of money to earn 10% profit for sure or (ii) investing the same amount of money to earn 20% profit with a probability of 50% or nothing with the remaining probability of 50%. RISKi takes a value of 0 (risk-averse) for the manager who chooses case (i) and 1 (risk-loving) for the one choosing case (ii). The previous empirical studies proved that risk-loving managers tend to invest more as the degree of output market uncertainty increases (Antonides and Van der Sar, 1990; Driver and Whelan, 2001; Andrade and Stafford, 2004; Akdoğu and Mackay, 2008). Therefore, coefficient β3 is supposed to be positive. INVi = β0 + β1UNCERi + β2UNCERi x RISKi­ + β3RISKi­ + β4PROi + β5IRRi + β6DSALi + β7COMPi + β8COMPi2 + β9FAGEi + β10BRIi + β11BRIi2 + β12FSIZEi + β13MANUi + β14SERVi + εi (2) PROi is the ratio of after-tax profits to total assets of firm i. Bo and Lensink (2005) and Bayraktar (2014) argue that, in the case of credit rationing due to information asymmetry, transaction cost and limited liability, firm investment is largely related to internal finance (mainly retained profits) because of difficulties in getting access to external finance (e.g., bank credit). Therefore, coefficient β4 is thought to be positive. UNCERi x RISKi is an interaction of UNCERi and RISKi. This interactive term is used to detect the impact of managers’ risk attitude on the relationship between output market uncertainty and investment of firm i. Studies (Bo and Sterken, 2007; Femminis, 2008; Chronopoulos et al., 2011; Whaley, 2011; Aistov and Kuzmicheva, 2012) argue that there is a negative impact of managers’ risk attitude on the relationship between the degree of uncerJournal of Economics and Development IRRi is a proxy for the irreversibility of used assets of firm i. Managers of the surveyed firms were asked to evaluate the possibility to resell used assets in order to construct variable IRR1i, which takes a value of 1 if the answer is “easy” and 0 if the answer is “not easy”. We also use the information about the expected resell value 62 Vol. 18, No.2, August 2016 of used assets to construct variable IRR2i (i.e., the ratio of the expected resell value of used assets to their replacement cost). Since the irreversibility of used assets depends on both IRR1i and IRR2i (Guiso and Parigi, 1999), we utilize the principal component technique to combine these two variables to create IRRi = w1IRR1i + w2IRR2i , with w1 and w2 being component parameters. The higher the value of IRRi , the higher the possibility for firms to resell used assets. Since investment decisions are normally hard to reverse (either partially or totally), a higher possibility to resell used assets induces firms to invest more, other things being equal. Coefficient β5 is then expected to be positive. not plausible to add up the quantity of different goods (Polder and Velhuizen, 2012). In sum, PEi can be written as follows: PEi = As just explained, fierce competition may squeeze PEi. Therefore, in order to make it easier to grasp the impact of the degree of competition on investment, we use COMPi = |PEi|. A higher value of COMPi means a higher degree of competition facing firm i. COMPi2 is also used to reveal the presence of an inverted-U shaped relationship between the degree of competition and investment by the firm. Nielsen (2002), Aghion et al. (2005), Moretto (2008), Akdoğu and Mackay (2008) and Polder and Veldhuizen (2012) assert that firms operating in a less severely competitive environment often have high costs due to moral hazard that results in inefficiency. As competition pressure strengthens, firms are forced to raise investment to mitigate costs, enhance efficiency and preempt competitors so as to tackle the risk of squeezed market share. Yet, if competition pressure goes beyond a certain point, it becomes too fierce, market niches evaporate and benefits from investing are no longer present, firms will then scale down investment. Thus, coefficient β7 is expected to be positive and β8 negative. FAGEi is the number of years in operation (age) of firm i. Since young firms are more eager to invest to grow and expand market share so as to avoid failing. Thus, β9 is supposed to be negative (Hansen, 1992; Moohammad et al., 2014). DSALi is the annual growth rate of sales by firm i (%). A fast growth of sales means a better prospect for firms. Therefore, firms will embark on more investment to make use of good available opportunities (Guiso and Parigi, 1999; Bo and Sterken, 2007). As a result, coefficient β6 is supposed to be positive. COMPi is the degree of competition facing firm i, measured by its profit elasticity (PEi). PE was coined by Boone (2000) and further developed by Boone (2001, 2008), Polder and Veldhuizen (2012), etc. According to those studies, the degree of competition can be identified by the ratio of percentage change of profit (π) to percentage change of marginal cost (MC), which means: Since it is often difficult to measure MC, researchers replace it by average cost (AC). In addition, the average cost of firms that operate in different sectors will be the ratio of total cost (TC) to total revenue (TR), because it is Journal of Economics and Development ∆π i / π i (%) < 0, π i = π i / TRi ∆ACi / ACi (%) BRIi is the ratio of bribes that firm i paid to public officials to its total assets. BRIi2 is included to detect the non-monotonic relation63 Vol. 18, No.2, August 2016 Table 1: Summary of the signs of the coefficient of independent variable Variables Definitions/Measures UNCERi Degree of output market uncertainty, measured by the coefficient of variation of expected sales of firm i Being 0 for risk-averse managers and 1 for risk-loving ones RISKi Signs of the coefficients of independent variables Negative Positive UNCERi x RISKi Interaction of UNCERi and RISKi Positive PROi Ratio of after-tax profits to total assets of firm i. Positive IRRi A proxy for the irreversibility of used assets of firm i Positive DSALi Annual growth rate of sales by firm i (%) Positive COMPi Degree of competition facing firm i, measured by its profit elasticity (PEi) Positive COMPi2 Square of COMPi Negative FAGEi Number of years in operation of firm i Negative BRIi Ratio of bribes that firm i paid to public officials to its total assets Positive BRIi2 Square of BRIi Negative FSIZEi Logarithm of total assets of firm i Positive/Negative MANUi Being 1 for manufacturing firms and 0 otherwise Positive/Negative SERVi Being 1 for service firms and 0 otherwise Positive/Negative 2008). As a result, β10 is expected to be positive and β11 to be negative. ship between bribes and investment by the firm. If bribed, bureaucratic officials are ‘greased’ to provide better services to firms, enabling them to take up available investment opportunities. However, despite bribes, some corrupt officials deliberately stay intact so as to force firms to bribe more. If forced to bribe too much, expected profits from investment projects will go down and firms will reduce investment accordingly. Therefore, there exists an inverted-U shaped relationship between bribes and firm investment (Svensson, 2005; Le Khuong Ninh, Journal of Economics and Development FSIZEi is the size of firm i, measured by the logarithm of the firm’s total asset value. In fact, large non-state firms tend to be more conservative about making big investments since it is difficult to find good opportunities. Thus, β12 is thought to be negative (Hansen, 1992; Le Khuong Ninh et al., 2007; Akdoğu and MacKay, 2008). MANUi and SERVi are used to test for the possible discrepancy of investment among firms in 64 Vol. 18, No.2, August 2016 different sectors (i.e., manufacturing, trade, and services). MANUi takes a value of 1 for manufacturing firms and 0 otherwise. SERVi takes a value of 1 for service firms and 0 otherwise. Coefficients β13 and β14 can be either positive or negative, depending on the environments in which the firms operate. performance, actual and planned investment by the firms, among others. To give a full picture of the characteristics of the surveyed firms, we use descriptive statistics. Then, we utilize Tobit model to estimate the impact of managers’ risk attitude on the relationship between the degree of output market uncertainty and investment by the surveyed firms. 4. Data and estimation method The primary data used in this paper were directly collected from 667 non-state firms in the MRD (Vietnam), using a questionnaire prepared in advance that had been corrected after several pilot surveys. Due to some reasons (such as being unable to contact the top manager, wrong address, missing information, etc.), we were able to get information from 667 non-state firms. The sample includes 42 firms in An Giang province (accounting for 6.3% of the total number of the surveyed firms), 24 in Bac Lieu (3.6%), 22 in Ben Tre (3.3%), 44 in Ca Mau (6.6%), 194 in Can Tho (29.1%), 43 in Dong Thap (6.5%), 53 in Hau Giang (7.9%), 43 in Kien Giang (6.5%), 52 in Long An (7.8%), 44 in Soc Trang (6.6%), 24 in Tien Giang (3.6%), 25 in Tra Vinh (3.7%) and 57 in Vinh Long (8.5%). The data collected include the information about general characteristics, 5. Findings 5.1. Characteristics of the surveyed firms According to the survey, the average age of the firms is 10 years and their average asset value is 146,913 VND million (Table 2). There are 231 liability-limited firms (accounting for 34.6% of the total number of firms surveyed), 193 joint-stock ones (28.9%), and 180 sole proprietorship ones (27%). There are 154 firms exporting part of or total output (accounting for 23.1% of total number of firms surveyed). Average sales of the surveyed firms in 2013 is 210,402 VND million (increasing by 17.4% compared to that in 2012). Average profits of those firms are 16,761 VND million (increasing by 6.8% compared to that in 2012). However, their average costs went up markedly (by 18.4% compared to that in 2012). Return on Table 2: General information about the surveyed firms (2013) Indicators Mean Standard deviation Min Max Age (year) 10 9 2 52 Total assets (VND million) 146,913 492,392 130 6,750,400 Sales (VND million) 210,402 539,048 50 5,450,131 Profit (VND million) 16,761 77,904 –705,087 1,200,000 Investment (VND million) 14,402 60,835 0 793,000 Source: Authors’ survey in 2014. Journal of Economics and Development 65 Vol. 18, No.2, August 2016 Table 3: Investment by the firms Investment in 2013 Financing sources Equity Loans from joint-stock banks Loans from state banks Loans from foreign-owned banks Loans from government projects Others Total investment Source: Authors’ survey in 2014. Amount (VND million) 9,472.03 2,976.26 1,432.91 221.11 30.58 269.51 14,402.41 % of total 65.77 20.66 9.95 1.54 0.21 1.87 100.00 Planned investment in 2015 Amount % of (VND total million) 5,142.61 58.57 2,169.25 24.71 1,022.82 11.65 90.67 1.03 19.34 0.22 335.13 3.82 8,779.81 100.00 Change in 2015 compared to 2013 (%) –45.71 –27.11 –28.62 –58.99 –36.76 24.35 –39.04 sales (ROS) of those firms was 8%. All this implies that the firms had reasonable growth rates but did not well utilize resources, so the costs are high. vestment decisions, firm managers were also concerned with output market uncertainty. The coefficient of variation of the future sales of the firms is 37.7%. About 46.3% of the surveyed firms paid bribes and the average bribe per firm is 192.2 VND million per year. Bribing seems to be pervasive as 45.6% of the firms did it on purpose to get things done faster and 48.5% saw it as an implicit norm. The firms bribed by giving gifts (accounting for 56.0% of total number of firms), travel (54.3%) or in cash (52.8%). Le, Hermes and Lanjouw (2004) also estimated the coefficient of variation of expected sales for firms in the MRD in 2000 and came up with a figure of 17.9%. This result implies that the degree of output market uncertainty facing firms in the region has gone up substantially. The reason for that is the economic downturn during the time our data were collected. 5.2. Estimation results Average investment by the firms in 2013 is 14,402.4 VND million. Due to economic downturn and suppressed market demand, planned investment of the firms in 2015 is just 8,779.8 VND million (decreasing by 39.04% compared to that in 2013). Financing sources for investment by the firms are equity (mainly retained profits) and bank loans. According to the survey, equity is an important financing source of investment by the firms, which accounts for as much as 65.77% of total investment outlays of the firms in the sample. When making inJournal of Economics and Development This section aims to examine the impact of managers’ risk attitude on the relationship between the degree of output market uncertainty and investment by the surveyed firms. Before doing that, we carefully check the data for hypotheses on multicollinearity and heteroskedasticity. All coefficients between independent variables (rij) are smaller than 0,8 (0,0002 ≤ |rij| ≤ 0,532), proving that there is no multicollinearity effect. In addition, we have used the Robust estimation option in Stata to correct the 66 Vol. 18, No.2, August 2016 Table 4: Estimation results Dependent variable: INV – planned investment in 2015 Variables C UNCERi UNCERi×RISKi Model 2a Model 2b Model 2c –0.038 – –0.115* (–0.045) –0.035 – –0.126* (–0.048) –0.024 – –0.151** (–0.059) 0.215** (0.084) RISKi 0.085* (0.035) 0.252*** 0.246*** 0.251*** PROi (0.098) (0.094) (0.098) 0.045*** 0.047*** 0.048*** IRRi (0.018) (0.018) (0.019) 0.002*** 0.002*** 0.002*** DSALi (0.001) (0.001) (0.001) 0.005*** 0.005*** 0.005*** COMPi (0.002) (0.002) (0.002) –0.000** –0.000** –0.000** COMPi2 (–0.000) (–0.000) (–0.000) –0.002 –0.002 –0.002 FAGEi (–0.001) (–0.001) (–0.001) 7.375*** 7.223*** 6.972*** BRIi (2.856) (2.763) (2.722) –58.675* –57.327* –55.203* BRIi2 (–22.722) (–21.929) (–21.557) –0.002 –0.002 –0.002 FSIZEi (–0.001) (–0.001) (–0.001) –0.006 –0.011 –0.011 MANUi (–0.002) (–0.004) (–0.004) –0.046 –0.050 –0.051 SERVi (–0.017) (–0.018) (–0.019) Observations (n) 667 667 667 F value 5.050 4.550 4.650 Significance 0.000 0.000 0.000 Log likelihood –334.551 –332.607 –331.879 Notes: In the first line is coefficient βi. In the brackets is ∂INV / ∂X i . ***: 1% significance level; **: 5% significance level; and *: 10% significance level. Source: Authors’ survey in 2014. problem of heteroskedasticity. in Table 4. The estimate shows that coefficient The impact of output market uncertainty on investment by the firms with managers’ risk attitude being excluded (Model 2a) is presented β1 of UNCERi has a value of -0.115 at a signif- Journal of Economics and Development icance level of 10%, implying that the degree of output market uncertainty has a negative 67 Vol. 18, No.2, August 2016 6. Conclusion and recommendations impact on investment by the surveyed firms. RISKi is added to Model 2b (Table 4) to estimate the impact of managers’ risk attitude on investment, regardless of output market uncertainty. The coefficient of RISKi is 0.085 at a significance level of 10%. This would mean that risk-loving managers tend to investment more than risk-averse ones do, others being equal. The findings show that the impact of output market uncertainty on investment is negative if managers’ risk attitude is not considered. This relationship becomes stronger for risk-averse managers since they need time to acquire more relevant information before making investment decisions so as to minimize the possibility of failure and regret. Yet, there exists a positive relationship between output market uncertainty and investment of risk-loving managers since they tend to accelerate investment as the degree of uncertainty goes up due to self-confidence, ambition to get over challenges and sanguineness about the future. In addition, the impacts of the degree of competition and bribes on firm investment both have the shape of an inverted-U. In addition, the higher the reversibility, the higher the investment by the firms. However, recent studies argue that there is an interaction between the degree of output market uncertainty and managers’ risk attitude to influence firm investment. Therefore, Model 2c (Table 4) aims to find evidence for this argument. Indeed, coefficient β2 of the interactive term UNCERi×RISKi has a value of 0.215 at a significance level of 5%. Obviously, risk-loving managers (RISKi = 1) tend to invest more as the degree of output market uncertainty goes up, since ∂INVi / ∂UNCERi = −0.059 + 0.084 = 0.025 Yet, risk-averse managers (RISKi = 0) tend to scale down investment as the degree of output market uncertainty picks up, since ∂INVi / ∂UNCERi = −0.059 . As argued, investment decisions of both groups of managers (i.e., risk-averse and risk-loving) may bring about bad outcomes if the perceptions about the degree of output market uncertainty are not precise. To make good predictions of the future, firms should have an own unit that is in charge of forecasting market tendency that will help firm managers make better investment decisions. In addition, to a certain extent, firms should consider diversifying operations to mitigate risks resulting from market gyrations and using risk hedging instruments (such as forwards, futures and swaps). The Government can also consider establishing an agency specializing in providing market information to firms. Coefficient β4 of PROi has a positive value at a significance level of 1%, implying that retained profits have a positive impact on investment by the firms, since they are usually credit rationed by commercial banks. In addition, coefficient β5 of IRRi also has a positive value at a significance level of 1%, meaning that the easier it is to resell used assets, the higher the level of investment. Similarly, coefficient β6 of DSALi also has a positive value at a significance level of 1%. In addition, most coefficients of other variables have expected signs, except for those of FAGEi, FSIZEi, MANUi, and SERVi. Journal of Economics and Development 68 Vol. 18, No.2, August 2016