1 Customer Concentration and Employment Risk in Supplier Firms Yanan Zhang Central University of Finance and Economics School of Accountancy E-mail: [email protected]Yun Ke Brock University Goodman School of Business E-mail: [email protected]Woo-Jong Lee Seoul National University Business School E-mail: [email protected]This version: January 1, 2018
54
Embed
Customer Concentration and Employment Risk in Supplier …...RSIs due to uncertainty regarding their customers’ future performance and payments, which possibly results in underinvestment
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Customer Concentration and Employment Risk in Supplier Firms
We address this question by analyzing a sample of 13,690 firm–years for U.S. suppliers from
1992 to 2015. Following prior study (e.g., Patatoukas 2012), we identify supplier firms with
major customers based on their disclosure of major customers that represent 10% or more of
their total sales. Specifically, we use the Compustat Segment Customer database to identify
customer–supplier information to construct two measures capturing the extent of a supplier’s
customer concentration.7 One measure is based on sales to major customers and the other is
based on an application of the Herfindahl–Hirschman index.
We document our main findings as follows. First, relying on a recently developed empirical
construct of employment efficiency (Jung et al. 2014), we find that the employment of suppliers
with a more concentrated customer base deviates from the optimal level to a greater extent. This
finding indicates that a customer-dependent supplier often makes employment decisions that are
not fully based on its economic needs, implying customer dependency creates suboptimal labor
investment in supplier firms. By partitioning the sample into two subgroups based on over- or
underinvestment in labor, respectively, we show that customer concentration is associated with
both types of labor employment risk. Further tests reveal that changes in employment are
positively associated with customer concentration, indicating that suppliers with a major
customer indeed tend to hire more employees to accommodate the customer. The results are
robust to alternative measure of employment risk and to an instrumental variable (IV) approach
to address potential endogeneity.
Second, we examine whether the low employment efficiency associated with customer
concentration is driven by customers’ needs for outsourcing. If a major customer exerts its of this practice to save their own inventory management costs by transferring the function to suppliers. Therefore,
trust along the supply chain is critical for the successful alignment of production functions. 7 Note that, following prior studies (Banerjee et al. 2008; Campello and Gao 2017; Patatoukas 2012), we restrict our
sample to firms with major customers because doing so ensures sufficiently asymmetric levels of bargaining power
between suppliers and customers. We thus do not contrast suppliers with and without major customers; instead, we
examine a cross-sectional variation of customer concentration among suppliers with major customers.
7
bargaining power to transfer its labor-intensive modules to supplier firms in an attempt to
improve its own operational efficiency, employment in supplier firms is then naturally affected
by the outsourcing motives of customer firms (Ramanna and Roychowdhury 2010). We find that
the positive relation between customer concentration and employment inefficiency is more
pronounced when customer outsourcing is more likely, implying that outsourcing motives
mediate the relation between customer concentration and employment risk.
Third, following Irvine et al. (2016) which maintain that selling, general, and administrative
(SG&A) expenses best reflect the RSIs of suppliers, we further find that customer concentration
is negatively associated with the elasticity of SG&A costs and the relation is mediated by
excessive employment. This finding indicates that relationship-specific employment is at least
partly a channel through which customer concentration is linked to excessive employment.
Fourth, additional analyses further reveal that the adverse impact of customer concentration
on supplier employment efficiency is contingent on certain firm characteristics. Specifically, we
find that customer concentration plays a more significant role in aggravating employment
inefficiency in supplier firms when 1) the supplier’s bargaining power is particularly weak due to
low customer switching costs, 2) the business is less complex (i.e., revenue sources are less
diversified), 3) customer-specific investment is greater, 4) operational uncertainty is greater, and
5) the information environment is poor.
Lastly, to explore the consequence of employment risk associated with customer
concentration, we investigate firm performance and provide some evidence that higher
employment risk induces suppliers with a more concentrated customer base to perform
worse—that is, with lower returns on assets (ROA)—than suppliers with a more diverse
8
customer base. This result is consistent with inefficient employment working as an operational
constraint by imposing rigid labor costs via customer concentration.
This study contributes to the literature in three important aspects. First, to our knowledge,
we are the first to empirically document employment risk transferred between suppliers and
customers. Whereas the supply chain risk literature has evaluated the overall performance impact
of supply chain risk, we consider employment risk as a specific channel through which the
operational risk of suppliers with major customers is binding. Second, we consider employment
as reflecting RSIs that suppliers with a major customer may strategically pursue. That is,
although the labor investment of supplier firms is far from optimal, it can be viewed as a way to
retain major customers. Our empirical findings on the negative association between customer
concentration and employment efficiency put an emphasis on that suppliers should make balance
between potential benefits and related costs of having a major customer. Third, we present direct
evidence that suppliers’ employment decisions are affected by their customers. This finding
therefore complements the findings of Campello and Gao (2017) and Dhaliwal et al. (2016), who
commonly report the negative consequences of concentrated customer bases in terms of
financing ease (i.e., the cost of raising equity and debt capital). Our study differs from these
studies by presenting the adverse effects from an operational view (i.e., employment).
The remainder of the paper is organized as follows. In Section 2, we review the literature
and develop our hypothesis. Section 3 presents the sample selection and the research
methodology, followed by our empirical results in Section 4. We conduct several additional
analyses in Section 5. Finally, Section 6 concludes the study with implications for practitioners
and academics and future research avenues.
9
2. Literature review and hypothesis development
Supply chain operational risk
Customers and suppliers establish and maintain economic links via various implicit and explicit
arrangements, such as long-term contracts, strategic alliances, and RSIs (Hui et al. 2012). The
supply chain’s contractual nature imposes a certain risk to both parties, known as supply chain
risk. In operations or management research, supply chain risk has typically been categorized in
many different ways and from many perspectives (Christopher and Peck 2004). For example,
Thun and Hoenig (2011) distinguish between internal and external supply chain risks that
encompass purchasing, demand, and environmental issues, whereas Guertler and Spinler (2015)
subdivide supply chain risk into supply, demand, product, and process risks. Recently,
operational risk has gained growing attention (Guertler and Spinler 2015; Hora and Klassen 2013;
Heckmann et al. 2015; Mitra et al. 2015). Because operational risk reflects the complexity,
uncertainty, and diversity of risk sources that are valid for supply networks; operational risk is
considered a better conceptual basis for the notion of supply chain risk compared to financial risk,
which is understood as market, credit, currency, and liquidity risk (Heckmann et al. 2015).
Therefore, the conventional approach in operations management research analyzes the benefits
and costs of management of reliable suppliers, with the customer firm at the center.
In contrast, accounting and finance studies focus instead on supplier firms and assess the
potential impact of their major customers (e.g., Ak and Patatoukas 2016; Campello and Gao
2017; Dhaliwal et al. 2016; Patatoukas 2012).8 In particular, the requirement to disclose major
customers allows related studies in accounting and finance to apprehend supply chain risk of
suppliers in the context of operational constraints. This stream of literature examines the impact
8 The difference is partly attributable to the popular use of disclosure requirements in accounting and finance. We
refer to SFAS No. 131 and SEC regulation S-K, which require a supplier with an individual customer that comprises
10% or more of firm sales to report on the major customer.
10
of the customer–supplier relationship on firm performance and the cost of equity (Dhaliwal et al.
2016; Patatoukas 2012), capital structure (Banerjee et al. 2008; Titman and Wessel 1988), bank
loan contracting (Cen et al. 2016; Kim et al. 2011), earnings management and accounting
conservatism (Hui et al. 2012; Raman and Shahrur 2008), analysts’ forecasts (Guan et al. 2015),
and tax avoidance (Cen et al. 2017). However, research on the influence of major customers on
the investment strategy of supplier firms is relatively scarce. Our study aims to fill part of this
gap by documenting the impact of customer–supplier relationships on corporate employment.
Sources of operational risk for supplier firms – demand uncertainty and outsourcing
Drawing from operations management and other management literature, we specifically consider
two sources of operational risk for supplier firms with major customers: demand uncertainty and
the lack of outsourcing opportunities. First, demand uncertainty can put suppliers in supply
chains in a disadvantageous position. Lee et al. (1997) and Lee and Whang (2002) analyze the
bullwhip effect, both empirically and theoretically, and show that demand information distortion
being exaggerated towards the suppliers can cause longer lead times and lower their supply chain
efficiency. 9 Particularly, Garavelli (2003) suggests that the bullwhip effect translates into
business uncertainty, cost rigidity, and operational risk suppliers of the supply chain.
Furthermore, the literature relates outsourcing opportunities to operational risk. Various
advantages of outsourcing in terms of lower manufacturing costs, reduced investment in plant
and equipment, capacity flexibility, enhanced focus on core competencies, and the promotion of
suppliers’ competition have long been widely documented across multiple areas of management
(Beach et al. 2000; D’Aveni and Ravenscraft 1994; Gilley and Rasheed 2000; Lei and Hitt; 1995;
Prahalad and Hamel 1990). Therefore, a stylized factor in the operations management literature
9 The bullwhip effect prevails for a majority of firms along the supply chain (Bray and Mendelson 2013) and can
arise in various relationships, such as between a retailer and a manufacturer or between a manufacturer and its
supplier (Lee et al. 1997).
11
is that an outsourcing strategy benefits customers with tighter supply chain integration, while
some recent studies introduce into this relation several contingency factors such as product
complexity, the complexity of business environment, uncertainty, and other country-level
variables (Kim 2009; van der Vaart et al. 2012; Wiengarten et al. 2016).
By contrast, suppliers are characterized by limited outsourcing options, which is one of the
most common policies for reducing operational risk (Chen and Xiao 2015; Holzhacker et al.
2015; Wiengarten et al. 2016). Suppliers consist mostly of manufacturers of parts and interim
goods, as well as natural raw material providers. In such cases, supplier firms must invest in
manufacturing facilities and have few outsourcing opportunities due to the inherent nature of
their operations, whereas customer firms can easily adopt a flexible multiple-sourcing strategy to
mitigate risk (Chen and Xiao 2015). We therefore anticipate that, when a major customer
attempts to outsource to a supplier and thus transfer its own operational constraints to the
supplier, the supplier will end up retaining excessive capacity and related labor that would have
been curtailed otherwise. This is, in fact, the spirit of the relationship-specific investment, whose
costs in terms of labor have not yet been studied in the literature.
Employment risk as part of supply chain risk
Rigidity in employment works as an operating constraint, given that labor is a key production
element. In addressing this issue, prior studies have primarily focused on labor unions (e.g.,
Bronars and Deere 1991; 1993; Hirsch 1992) or country-level employee protection laws (Van
Long and Siebert 1983). In general, prior studies suggest that the constraints of employment
decisions, such as firm-level labor unions or employment protection legislation, provide
managers with nontrivial adjustment costs in employment and thus prompt inefficient labor
investment decisions.
12
We consider customer concentration as another determinant of employment efficiency. As
documented above, suppliers with major customers suffer from significant operational
constraints induced by supply chain risk. We therefore hypothesize that such suppliers are also
exposed to employment risk to a great extent. In the presence of operational constraints,
suppliers’ employment can deviate from optimal levels and remain off-equilibrium due to the
high adjustment costs. On the one hand, over-hiring (i.e., employment levels above optimum) is
likely for suppliers with major customers, because they unavoidably retain excessive labor that
would not be otherwise needed. The sticky nature of employment due to its greater downward
than upward adjustment costs will exacerbate labor over-investment among such suppliers. On
the other hand, under-hiring (i.e., employment levels below optimum) can also arise, because
resources for new employment will be limited due to excessive capacity exclusively devoted to
major customers (Jensen and Meckling 1976; Montgomery 1989). Such a “hold-up” problem
deters managers from making sufficient commitment in employment. We therefore expect that
employment efficiency will be lower as customer base is more concentrated in supplier firms.
In contrast, the alignment theory indicates some merits of supply chains for suppliers. Major
customers can help suppliers streamline their business activities and thus achieve greater
operating efficiency through collaboration and information sharing along the supply chain (e.g.,
Kinney and Wempe 2001). Consistent with the idea, Ak and Patatoukas (2016) document a
positive association between customer-base concentration and inventory management efficiency.
Such efficiency advantages in supply-chain practices via collaboration and information sharing
ultimately enable dependent suppliers to better cope with demand uncertainty and thus manage
their resources more efficiently. Accordingly, employment efficiency does not have to be always
low for suppliers with a concentrated customer base. In sum, whether customer concentration
13
and employment risk are positively or negatively related is an empirical question. We therefore
set up our hypothesis in a null form.
HYPOTHESIS: Customer concentration is not associated with the employment risk of
supplier firms.
3. Sample, research design, and descriptive statistics
Sample
We construct our sample by intersecting a number of data sources. Since this study focuses on
supplier firms with major customers, we follow prior literature (e.g., Ak and Patatoukas 2016;
Irvine et al. 2016; Patatoukas 2012) and obtain customer–supplier data from the Compustat
Segment database, which provides the name of each major customer along with the sales
amounts. We then retrieve financial statement data from Compustat’s annual data sets, stock
return and stock price data from the Center for Research in Security Prices monthly stock files,
institutional ownership data from the Thomson Reuters Institutional Holdings database, and
industry unionization data from the Union Membership and Coverage Database.10 Our sample
period is from 1992 to 2015. We begin in 1992 because five years of operating cash flow data are
required to calculate cash flow volatility and the cash flow data first became available in 1988.
We exclude utilities (Standard Industrial Classification, or SIC, codes 4900–4999) and financial
firms (SIC codes 6000–6999) and require firm–year observations with non-missing data in
estimating our test variables. These procedures yield a final sample of 13,690 supplier–year
observations.
10 The Union Membership and Coverage Database is available at www.unionstats.com.
This table presents the Pearson correlation coefficients for the variables used in the main tests. The sample period is from 1992 to 2015. We exclude firms missing
the financial data required to compute the variables used in our regression analyses. To alleviate potential problems associated with extreme outliers, we winsorize
all continuous variables at the first and 99th percentiles. All the variables are defined in Appendix A. Here, *, **, and *** indicate, respectively, the 10%, 5%, and
1% significance levels (two tailed).
40
TABLE 3
Effect of customer concentration on labor employment risk
This table reports the main results of the impact of customer concentration on abnormal net hiring based on regression model (2). The sample period is from
1992 to 2015. Columns (1) and (2) show the full-sample results. Columns (3) and (4) present the results for the overinvestment subsample, where actual net
hiring is above expected levels. Columns (5) and (6) demonstrate the results of the underinvestment subsample, where actual net hiring is under expected
levels. In both subsamples, the dependent variable is ABS_AB_NET_HIRE. All the variables are defined in Appendix A. Here, *, **, and *** indicate,
respectively, 10%, 5%, and 1% significance (two tailed). The t-values in parentheses are based on standard errors clustered by firm.
This table reports the main results of the impact of customer concentration on labor employment risk using an alternative proxy (ABS_EMP_CHG). We
measure this variable as the absolute value of the percentage change in the employment numbers from years t to t - 1, adjusted by the two-digit SIC industry
median value. The sample period is from 1992 to 2015. Columns (1) and (2) show the full-sample results. Columns (3) and (4) present the results of a positive
change in the employment numbers in the subsample where EMP_CHG > 0. Columns (5) and (6) present the results of a negative change in the employment
numbers in the subsample where EMP_CHG < 0. In both subsamples, the dependent variable is ABS_EMP_CHG. All the variables are defined in Appendix
A. Here, *, **, and *** indicate, respectively, 10%, 5%, and 1% significance (two tailed). The t-values, in parentheses, are based on standard errors clustered
by firm.
44
TABLE 5
Channel test: Outsourcing
Dep. Var. = ABS_AB_NET_HIREt
(1) (2)
CC_SALESt-1 -0.034
(-1.586)
CC_SALESt-1 × OUTSOURCINGt-1 0.011***
(3.186)
CC_HHIt-1
-0.059
(-1.254)
CC_HHIt-1 × OUTSOURCINGt-1
0.030***
(2.865)
OUTSOURCINGt-1 -0.005*** -0.004**
(-2.938) (-2.562)
AQt-1 -0.005*** -0.004***
(-3.529) (-3.486)
MTBt-1 0.002** 0.002*
(1.999) (1.888)
SIZEt-1 -0.000 0.000
(-0.118) (0.059)
QUICKt-1 0.005* 0.004*
(1.955) (1.699)
LEVt-1 0.010 0.012
(0.410) (0.500)
DIVDUMt-1 -0.010 -0.010
(-1.372) (-1.426)
STD_CFOt-1 0.073 0.059
(1.065) (0.879)
STD_SALESt-1 0.005 0.007
(0.311) (0.459)
TANGIBLEt-1 -0.054** -0.058**
(-2.282) (-2.444)
LOSSt-1 0.006 0.005
(0.867) (0.746)
INSTIt-1 -0.031*** -0.028**
(-2.701) (-2.443)
STD_NET_HIREt-1 0.023*** 0.022***
(3.300) (3.170)
LABOR_INTENSITYt-1 -0.087 -0.044
(-0.253) (-0.131)
UNIONt-1 -0.009 -0.017
(-0.124) (-0.225)
ABS_AB_INVEST_OTHERt 0.016** 0.016**
(2.383) (2.490)
Constant 0.199*** 0.187***
(3.285) (3.028)
Industry fixed effects YES YES
45
Year fixed effects YES YES
Observations 2,576 2,576
Adjusted R-squared 0.098 0.107
This table reports the channel effect of customers’ outsourcing activities on the relation between suppliers’
customer concentration and labor investment efficiency. In a similar vein as that of Ramanna et al. (2010), we
measure outsourcing activity (OUTSOURCING) as the negative value of major customers’ weighted average
abnormal net hiring over sales for each supplier–year, ranked into deciles. A higher value of OUTSOURCING
indicates a higher level of outsourcing activities. All the other variables are defined in Appendix A. Here, *, **, and
*** indicate, respectively, 10%, 5%, and 1% significance (two tailed). The t-values in parentheses are based on
standard errors clustered by firm.
46
TABLE 6
Channel test: RSI
Dep. Var. = SGA_ELASTICITYt
(1) (2)
CC_SALES_Rt -0.356***
(-3.262)
CC_SALES_Rt × AB_NET_HIRE_Rt 0.434***
(3.386)
CC_HHI_Rt
-0.345***
(-3.153)
CC_HHI_Rt × AB_NET_HIRE_Rt
0.444***
(3.447)
MVEt 0.000** 0.000**
(2.055) (2.066)
AGEt 0.003 0.003
(1.484) (1.517)
GROWTHt 0.201*** 0.200***
(6.062) (6.053)
CONGLOt -0.023 -0.021
(-0.389) (-0.367)
FLEVt -0.004 -0.004
(-0.299) (-0.303)
Constant 2.009*** 1.995***
(3.400) (3.374)
Industry fixed effects YES YES
Year fixed effects YES YES
Observations 26,372 26,372
Adjusted R-squared 0.003 0.003
This table investigates the channel effect of RSI. In detail, the results demonstrate the impact of abnormal net hiring
on the relation between customer concentration and SG&A elasticity. The variable CC_SALES_R is the decile rank
of CC_SALES, CC_HHI_R is the decile rank of CC_HHI, and AB_NET_HIRE_R is the decile rank of the raw value
of abnormal net hiring. Following Irvine et al. (2016), we calculate the elasticity of SG&A expenses with respect to
sales (SGA_ELASTICITY) as the change in the logarithmic value of SG&A expenses from years t to t - 1, divided by
the change in the logarithmic value of sales from years t to t - 1. Several control variables are included. The variable
MVE is the logarithm of the market value of equity, AGE is the logarithm of firm age, GROWTH is the annual sales
growth rate, CONGLO is a dummy variable that equals one if the firm reports more than one business segment, and
FLEV is calculated as the book value of assets divided by the book value of equity. All the other variables are
defined in Appendix A. Here, *, **, and *** indicate, respectively, 10%, 5%, and 1% significance (two tailed). The
t-values in parentheses are based on standard errors clustered by firm.
47
TABLE 7
Cross-sectional analysis
Panel A: Effect of supplier bargaining power (measured by customers’ switching costs)
Dep. Var. = ABS_AB_NET_HIREt
(1) (2)
CC_SALESt-1 0.079***
(5.820)
CC_SALESt-1 × SWITCH_COSTt-1 -0.009***
(-3.983)
CC_HHIt-1
0.148***
(5.071)
CC_HHIt-1 × SWITCH_COSTt-1
-0.018***
(-3.723)
SWITCH_COSTt-1 -0.004*** -0.005***
(-2.817) (-3.580)
Control variables YES YES
Industry fixed effects YES YES
Year fixed effects YES YES
Observations 13,613 13,613
Adjusted R-squared 0.114 0.114
Panel B: Effect of supplier revenue diversification
Dep. Var. = ABS_AB_NET_HIREt
(1) (2)
CC_SALESt-1 0.026**
(2.334)
CC_SALESt-1 × BUS_SEGt-1 0.024*
(1.881)
CC_HHIt-1 0.036
(1.424)
CC_HHIt-1 × BUS_SEGt-1 0.063**
(2.103)
BUS_SEGt-1 -0.013*** -0.011***
(-3.327) (-3.083)
Control variables YES YES
Industry fixed effects YES YES
Year fixed effects YES YES
Observations 12,801 12,801
Adjusted R-squared 0.112 0.111
48
Panel C: Effect of customer-specific investment
Dep. Var. = ABS_AB_NET_HIREt
(1) (2)
CC_SALESt-1 -0.044
(-1.353)
CC_SALESt-1 × CSPECIFICITYt-1 0.014***
(2.604)
CC_HHIt-1 -0.110
(-1.639)
CC_HHIt-1 × CSPECIFICITYt-1 0.035***
(3.478)
CSPECIFICITYt-1 -0.004* -0.003*
(-1.721) (-1.791)
Control variables YES YES
Industry fixed effects YES YES
Year fixed effects YES YES
Observations 1,750 1,750
Adjusted R-squared 0.106 0.115
Panel D: Effect of supplier operational uncertainty
Dep. Var. = ABS_AB_NET_HIREt
(1) (2)
CC_SALESt-1 0.001
(0.072)
CC_SALESt-1 × STD_EARNt-1 0.006***
(2.850)
CC_HHIt-1
0.017
(0.464)
CC_HHIt-1 × STD_EARNt-1
0.009*
(1.647)
STD_EARNt-1 0.000 0.001*
(0.497) (1.840)
Control variables YES YES
Industry fixed effects YES YES
Year fixed effects YES YES
Observations 13,141 13,141
Adjusted R-squared 0.110 0.110
49
Panel E: Effect of the supplier’s information environment
Dep. Var. = ABS_AB_NET_HIREt
(1) (2)
CC_SALESt-1 0.070***
(5.490)
CC_SALESt-1 × AQt-1 -0.006***
(-3.039)
CC_HHIt-1 0.129***
(4.600)
CC_HHIt-1 × AQt-1 -0.011**
(-2.255)
AQt-1 -0.001 -0.002**
(-1.017) (-2.464)
Control variables YES YES
Industry fixed effects YES YES
Year fixed effects YES YES
Observations 13,690 13,690
Adjusted R-squared 0.111 0.111
This table reports the cross-sectional results of the impact of customer concentration on labor employment risk.
Panel A presents the effects of customer’s switching costs, SWITCH_COST, measured as the supplier’s sales
divided by two-digit SIC industry sales, ranked into deciles by year. Panel B presents the effect of supplier
diversification with the proxy of BUS_SEG, an indicator variable equal to one if the supplier firm has only one
business segment and zero otherwise. Panel C reports the effect of investor-specific investment as proxied by the
customer’s input specificity (CSPECIFICITY). Following Nunn (2007), we measure CSPECIFICITY using
customers’ weighted average level of non-homogeneous inputs over sales. Panel D presents the effect of suppliers
operational uncertainty as measured by sales volatility, STD_SALES, where STD_SALES is the standard deviation
of sales from years t - 5 to t - 1, ranked into deciles by year. Panel E presents the effect of the supplier’s information
environment as proxied by AQ, the accounting quality measure, ranked into deciles by year. All the other variables
are defined in Appendix A. Here, *, **, and *** indicate, respectively, 10%, 5%, and 1% significance (two-tailed).
The t-values in parentheses are based on standard errors clustered by firm.
50
TABLE 8
Instrumental variables regressions
Panel A: First-stage results
Industry concentration measures as IVs
Dep. Var. = CC_SALESt-1 CC_HHIt-1
(1) (2)
IND_CC_SALESt-2 0.675***
(66.975)
IND_CC_SALESt-3 0.166***
(16.490)
IND_CC_HHIt-2 1.096***
(48.824)
IND_CC_HHIt-3 0.469***
(21.068)
AQt-1 0.000 -0.001*
(0.068) (-1.889)
MTBt-1 -0.000 -0.001***
(-0.368) (-3.287)
SIZEt-1 0.001** 0.002***
(2.153) (3.829)
QUICKt-1 0.002*** 0.002***
(3.210) (2.592)
LEVt-1 -0.008 -0.019***
(-1.478) (-2.754)
DIVDUMt-1 0.000 0.000
(0.041) (0.027)
STD_CFOt-1 0.051** 0.067***
(2.487) (2.677)
STD_SALESt-1 -0.001 0.002
(-0.360) (0.377)
TANGIBLEt-1 0.030*** 0.041***
(6.650) (7.341)
LOSSt-1 0.002 0.005**
(1.075) (2.164)
INSTIt-1 -0.010*** -0.021***
(-3.499) (-5.990)
STD_NET_HIREt-1 0.006** 0.011***
(2.139) (3.508)
LABOR_INTENSITYt-1 -0.130 -0.331***
(-1.409) (-2.923)
UNIONt-1 0.023** 0.102***
(2.309) (8.300)
ABS_AB_INVEST_OTHERt 0.007*** 0.008***
(3.518) (3.339)
51
Constant 0.025*** 0.116***
(5.331) (21.018)
Observations 9,412 9,412
Adjusted R-squared 0.683 0.522
Test of endogeneity, weak instruments, and overidentification