SciELO - Scientific Electronic Library Online

 
vol.44 número1Matching espacial para georreferenciar datos de encuestas de hogarPrecios de activos y desigualdad de la riqueza en un modelo simple con shocks idiosincráticos índice de autoresíndice de assuntospesquisa de artigos
Home Pagelista alfabética de periódicos  

Serviços Personalizados

Journal

Artigo

Indicadores

Links relacionados

  • Em processo de indexaçãoCitado por Google
  • Não possue artigos similaresSimilares em SciELO
  • Em processo de indexaçãoSimilares em Google

Compartilhar


Estudios de economía

versão On-line ISSN 0718-5286

Estudios de Economía vol.44 no.1 Santiago jun. 2017

http://dx.doi.org/10.4067/S0718-52862017000100081 

Article

The influence of the trade-off between profitability and future increases in sales on cost stickiness

La influencia del dilema entre rentabilidad inmediata y crecimiento futuro de las ventas en la histéresis de los costes

Josep Mª Argilés-Bosch*  , Josep García-Blandón**  , Diego Ravenda***  , Maika M. Valencia-Silva****  , Antonio D. Somoza* 

*Universitat de Barcelona (Departament d’Empresa).Barcelona, Spain.

**IQS, (Faculty of Economics). Barcelona, Spain.

***Toulouse Business School (Department of Management Control, Accounting and Auditing). Toulouse, France.

****ESADE Business School. Barcelona, Spain.

Abstract:

This study analyses cost stickiness under the dilemma between current profitability and future sales increase. When activity decreases firms are faced to keep profitability adjusting resources, while they should also consider long term consequences and keep slack resources which allow building firms capacities to adapt to external challenges and take advantage of future opportunities. We find empirical evidence that changes in current firm profitability and one year ahead sales increase significantly influence resource adjustment in periods when sales decrease. We find a significant moderating effect of changes in profitability, as well as a significant stressing effect of one year ahead sales increase, on cost stickiness.

Keywords: Cost stickiness; profitability; increases in sales; slack; resource adjustment.

Resumen:

Este estudio analiza la histéresis de los costes bajo el dilema de conseguir una rentabilidad inmediata o futuros incrementos de ventas. Cuando la actividad disminuye las empresas se enfrentan al dilema de mantener la rentabilidad recortando gastos, pero tienen que considerar también las consecuencias a largo plazo de mantener capacidad sobrante o recursos de holgura que les permitan construir capacidades para adaptarse a los desafíos que les plantea el entorno, así como afrontar con ventaja las oportunidades futuras. En este trabajo encontramos evidencia empírica de que los cambios en la rentabilidad presente, así como los aumentos futuros en las ventas influencian significativamente el ajuste de recursos de las empresas en los períodos en los que disminuyen las ventas. Encontramos un efecto moderador de los cambios en la rentabilidad presente, así como un efecto acentuador de los aumentos futuros en las ventas, en la histéresis de los costes.

Palabras clave: Histéresis de los costes; rentabilidad económica; aumento de las ventas; recursos sobrantes; disminución de costes.

Introduction

Cost stickiness has attracted considerable academic attention during recent years. Cooper and Kaplan (1998, p. 246-7) alleged that two steps are required for costs decrease when activity decreases: reducing demands for resources and lowering the spending. According to this, costs’ response to an increase in activity is significantly stronger in comparison to their response to a decrease in activity. This type of cost behaviour is labelled sticky, and stickiness the correspondent effect. Therefore, sticky costs occur due to asymmetric adjustments of resources when activity increases and decreases.

Anderson et al. (2007) identified three factors causing cost stickiness: fixity of costs, management failure to control costs, and management decisions to maintain resources during a downturn. In the same vein, Yasukata and Kajiwara (2011) distinguished two different causes of cost stickiness: as a result of managers being unable to adjust firms’ capacity, and as the deliberate decision of managers. On the one hand cost downward adjustment is more difficult than upward adjustment because firms face difficulties in removing committed resources. Some costs, as for example maintenance or general services, are difficult to avoid. The adjustment of resources entails also increases in severance pay for employees, returns of materials and services, organizational adjustments, etc. In this respect, some prior studies found empirical evidence on the inability of managers to adjust resources, as well as on the incidence of objective factors and firm characteristics on cost stickiness (e.g. Anderson et al., 2003; Calleja et al., 2006; Argilés-Bosch and García-Blandón, 2011; Dalla and Perego, 2014; Balakrisnan et al., 2014). On the other hand, other studies stress on the deliberate decisions of managers, depending on expectations (Cooper and Kaplan, 1992; Yasukata and Kawijara, 2011; Banker et al., 2014), past sales growth (Anderson et al., 2013), the type of adjusted resources (Balakrishnan and Gruca, 2008), manager’s building incentives (Chen et al., 2012), CEO compensation and expectations for value creation of input resource expenditures (Banker et al,. 2011a; Banker et al., 2011b), incentives to meet earnings targets (Kama and Weiss, 2013), changes in regulation (Balakrishnan et al., 2004; Holzhacker et al., 2015), capacity utilization (Cannon, 2014), demand uncertainty (Banker et al., 2013), etc. However no previous study analysed cost stickiness under the scope of attaining current profitability and future sales growth. Banker et al. (2011b) analysed value creation by selling, general and administrative expenses (SGA), but they neither focused on profitability nor analysed the specific case of sales decreases. Kama and Weiss (2013) did not focus on profitability, they studied cost behaviour of firms with incentives of attaining specific targets of avoiding losses and decrease in annual earnings. They used dummy variables distinguishing between firms suspicious of attaining these targets and the remaining firms, but they did not analyse the whole range of profitability or changes in profitability influencing resource adjustments. Moreover, no previous study considered the trade-off between profitability and sales growth.

The trade-off between efficiency and flexibility is an enduring postulate in management literature (e.g.Adler et al., 1999; Tan and Wang, 2010). Thomson (1967) described it as the important paradox of administration. It has been extensively analysed by management research, but it has not deserved similar attention in accounting research. More specifically, no previous empirical research on cost stickiness focuses on this dilemma. Banker et al. (2011a) support the premise that cost behaviour is driven by deliberate resource commitment decisions made by forward-looking managers. Some studies found association between firm’s costs and future earnings per share (e.g.Anderson et al., 2007, Baumgarten et al., 2010), future profitability (Banker et al., 2006; Banker and Chen, 2006) and a ratio of increases in sales to market capitalization (Baumgarten et al., 2010). Banker et al. (2011b) found that the potential of future value-creation influences resource expenditure. Weiss (2010) analysed the incidence of cost behaviour on analysts’ forecasts. The incidence of expectations on cost stickiness has also been analysed, as previously mentioned. This study contributes analyzing cost stickiness under this scope: the management of firm resources and the dilemma between efficiency and flexibility. In this vein, we link the extant research on cost accounting with management research. We use one year ahead sales increase and change in current profitability as indicators of flexibility and efficiency respectively.

We use a sample of American industrial firms with at least 20 years of data in COMPUSTAT. We find empirical evidence that managers, in decisions on resource adjustment when activity decreases, consider changes in current profitability, as well as future increases in sales. When enduring a decrease in activity managers are faced to the dilemma of cutting costs to uphold firm profitability, versus keeping enough resources to ensure firm’s future advantageous position and to allow firm’s sales recovery. They consider that resource adjustment hinders the ability of firms to build capabilities that enable them advantages to attain future increases in sales, while on the other hand they also consider cutting costs as a measure to attain targets of profitability. We find empirical evidence that changes in current firm profitability and one year ahead sales increase significantly influence resource adjustment decisions in periods of sales decrease. More precisely, we find a significant moderating effect of change in profitability, as well as a significant stressing effect of one year ahead sales increase, on resource adjustment when activity decreases. However, firms with high profitability are less urged to adjust resources when activity decreases, while firms with low profitability are more urged to cut costs in such circumstances. We find empirical evidence of such behaviour with firm-year observations.

Our results provide useful insights for the management and cost accounting literature, enhancing our understanding of resource adjustment. They are important for various users of financial information. They are useful in assessing financial statements, and more specifically costs changes. External analysts of financial information should be aware that resource adjustments in periods of sales decrease may entail immediate gains in current profitability, or may prevent decreases in current profitability, but it may also restrict future sales recovery and/ or sales increase. The maintenance of slack resources when activity decreases, even for firms facing gloomy sales expectations, may be a good indicator of a proactive strategic orientation towards building firm capabilities. Our results have significant implications for the management accounting literature, providing new knowledge on management decisions on resource adjustment, with respect to the balance between short and long term consequences.

The remainder of the paper is organized as follows. The next section discusses theoretical issues and raise hypotheses, we then discuss methodological issues, describe the sample used in the study, explain results and raise concluding remarks.

Theory and hypotheses development

Stigler (1939) was the first to suggest the existence of a trade-off between flexibility and efficiency. Thomson (1967) described it as the important paradox of administration. He mentioned two opposed extremes of the organization. On the one hand, the technical core of the organization is concerned on scoring well on assessments of technical rationality. On the other hand, the upper reaches of the administration are concerned in providing the capacities and resources which allow the organization to make demands on its environment and to take advantage of opportunities afforded by that environment. Fuss and McFadden (1978) argued that the improvement in efficiency is usually a loss in flexibility. They argue that there is a competition over firm’s scarce resources between flexibility and efficiency. There are two important strategic choices, efficiency and flexibility, that explain management decisions.

According to Cyert and March (1956) firms build flexibility through investing in resources. They introduced the notion of slack as a pool of excess resources that helps firms to adjust to unexpected fluctuations. Slack is potentially utilizable resources that can be diverted or redeployed for the achievement of organizational goals (George, 2005). A certain level of excess resources provides flexibility to experiment, take risks and make proactive choices. Firms use this slack to build capabilities that make firms competitive, allow them to take strategic choices, act as buffers in periods of economic difficulties, facilitate the adjustment to unanticipated contingencies, etc. Resources act as inducements to experiment, take risks, make proactive choices, build capabilities, as well as to take advantage of market circumstances and opportunities. The presence or absence of excess resources determines to what extent firms are able to take profit of available advantages. Surplus resources make it easier to pursue firms’ projects. Thompson (1967) argued that slack resources provide a reactive protection against threats and facilitate proactive initiatives. According to this author, the presence of excess resources provides the flexibility for a firm to decide on a course of action when trying to adapt to its environment. Slack resources allow firms to adapt to complex competitive landscapes, as well as to redeploy organizational means for the achievement of long term goals. They act as buffers that help to pursue and build new capabilities. Damanpour (1987) and Greve (2003) argued that slack resources have a positive effect on innovation. Firms with no surplus resources are more constrained to deal with operational issues. Managers and employees direct their efforts to urgent matters and rush orders in such cases. Coordination, planning, process improvement, innovation, market research, etc. are set aside, or may not be properly dealt with, because any available resource is committed to daily or imminent operational activities, when there are no slack resources. Any unanticipated event is an important disruption that may hardly be managed in these cases. In this respect, firms with slack resources are in better position to make proactive choices that can be translated into future sales increases. They have quicker response to such circumstances. They also have more potential ability to respond to competitor strategies, to cater special needs, to deliver new orders from customers, to shift resources into or away from their existing markets, etc. (George, 2005). Indeed, they are in better position to meet increases in demand. Firms with no slack resources are more likely to lose opportunities for expansion of sales. They will be unable to mobilize the necessary resources quickly enough when opportunities for sales present themselves. Weiss (2001) refer to tactical flexibility as the ability of a firm to adjust overall output to exogenous shocks. In this study we focus on tactical flexibility and approach it as the achievement of future increases in sales. Mishina et al. (2004) and Bradley et al. (2011) found empirical evidence on the influence of slack on future sales growth. In this respect, when sales decrease managers may decide to cut resources proportionately, or almost proportionately, with respect to sales increases, or they may decide to cut resources to a lesser extent, because managers’ expectations for future sales are optimistic, but also because managers may give priority to allow futures sales recovery. The resource-based view (Penrose, 1959) and the behavioural theory (Cyert and March, 1963) of the firm argue on the beneficial effects of resources on sales growth. They provide both, a reactive bulwark against threats and the implementation of proactive strategic initiatives. According to it, managers need slack to fuel innovation and increase sales. Figure 1 depicts scenarios A and B, with lower and greater cost stickiness allowing lower and greater future sales increase respectively.

We thus formulate the following hypothesis:

H1. The attainment of greater (lower) future sales growth drives greater (lower) cost stickiness when activity decreases.

The adjustment of resources when activity decreases produces an immediate effect of improving firm income, with respect to firms that do not adjust resources, as well as of reducing the amount of investment used. The subsequent effect is an improvement in firm profitability, by increasing income and by reducing the investment in resources (Starr and MacMillan, 1990). The adjustment of resources has an additional effect of improving efficiency, because it reveals more clearly the scarcity of resources and stimulates their efficient use. In this vein, Baker and Nelson (2005) argue that firms with fewer resources are likely to leverage them more efficiently. According to these authors, resource constraints alter the behaviour by which resources are expended, forcing managers to apply the appropriate level and type of resource that any specific challenge demands. Therefore, firms more focused on the specific concern of improving figures of profitability will be more prone to adjust resources when they face decreases in activity.

The lowest line depicts the initial increase in costs when sales increase (from t-2 to t-1). A symmetric behavior would entail that costs would decrease along this line when sales decrease (from t-1 to t). However, an asymmetric behavior currently takes place, which may be different according to deliberate firms’ decisions. In scenario A (thicker line) managers cut more resources than in scenario B (thinner line), given a decrease in sales in period t with respect to period t-1. Therefore, firm in scenario B is in an advantageous position to allow sales recovery with respect to firm in scenario A. Distances A and B between costs increase and decrease indicate cost stickiness for scenario A and B respectively. The greater slope lines depict cost curve when sales increase.

FIGURE 1 COST STICKINESS AND FUTURE SALES INCREASE 

The agency theory describes the conflict between shareholders and managers. The later do not always act in the best interest of shareholders and choose actions to maximize their own utility (Jensen and Meckling, 1976). Some authors (e.g. Jensen, 1986; Phan and Hill, 1995) argue that surplus resources make it easier for managers to pursue self-serving behaviours, which are not usually reconciled with firm performance. On the contrary, surplus resources often jeopardize performance. Hope and Thomas (2008) found empirical evidence of this management behaviour. Accordingly, firms more focused on profitability will give priority to capacity adjustment, the elimination of slack resources and the avoidance of the opportunistic behaviour of managers, thus preventing cost stickiness.

On the other hand, the agency theory argues that incentives to meet earnings targets promote self interested management behaviour. There is a bulk of empirical research finding that managers’ concern for meeting earnings benchmarks induces cost reduction (e.g.Roychowdhury, 2006; Cohen et al., 2008). More specifically, Kama and Weiss (2013) found that managers faced to meet earnings targets diminish cost stickiness. In this vein, managers aiming at preserving firm profitability will cut slack resources excessively when sales decrease, even if it jeopardizes futures increases in sales, as depicted by scenarios A and B in Figure 1.

We thus formulate the following hypothesis:

H2. The attainment of favourable (unfavourable) changes in profitability drives lower (greater) cost stickiness when activity decreases.

Methodology

The purpose of our analysis is to test the influence of the attainment of future sales increases and current profitability on resource adjustment when activity decreases. We rely on the model used in the seminal study of Anderson et al. (2003), as well as in most of the subsequent research on cost stickiness (e.g. Dalla and Perego, 2013; Banker et al., 2011a), and formulate the following equation:

(1)

where SGA indicates selling general and administrative expenses for a firm i at year t, S sales, ROA return on assets, CHROA change in ROA (measured as the difference between current and previous year ROA), SEXP sales expectations, and ASSTS and DEBTS the ratios of total assets (ASS) and long term debt to sales respectively. D and DNEXT are dummy variables indicating respectively, with value 1, that sales decrease with respect to previous year and that sales increase in the following year with respect to current year respectively (and 0 otherwise). Y are also dummy variables, indicating that an observation belongs to a given year, with value 1 and 0 otherwise. They control for specific seasonal or temporary effects. All variables used in this study are explained in the annex.

Changes in costs depend on changes in sales. The logarithmic specification of the model entails that the coefficient β1 measures the percentage increase in costs for a 1% increase in sales. If the traditional fixedand variable-cost model is valid, upward and downward changes in costs, given changes in output, will be equal and consequently β2 = 0. Because of D takes the value of 1 when sales decrease between periods t-1 and t, the sum of coefficients β1 + β2 measures the percent decrease in costs that follows a 1% decrease in output. If costs are sticky, the changes in costs following output increases should be greater than under output decreases, thus being β2<0, conditional on β1>0.

We use SGA in our dependent variable. More precisely, the dependent variable is the logarithm of SGAt / SGAt-1. They are a sort of slack resources. They are more linked to strategic activities than most other operating costs. They include a bulk of resources related to managing, planning and building firm capacities.

They include a considerable share of fixed costs. They behave more as fixed costs than most other operating costs, such as for example raw materials or direct labour. They include resources used in strategic firm activities that allow firms to be in an advantageous position to face competition. These resources allow firms to take actions to adapt to environmental circumstances, as well as to be in an advantageous position to attain future sales increases. SGA are an interesting approach for resources committed to coordination, planning, process improvement, innovation, marketing research, building capabilities, etc., which enable firms to be in advantageous position and allow sales recovery. They are usually used in analyzing cost stickiness (e.g.Anderson et al., 2003; Chen et al., 2012; Banker and Byzalov, 2014; Anderson et al., 2013). Moreover, they have been persistently used in empirical research on slack versus financial performance, as a measure of slack (e.g.Singh, 1986; Bergh and Lawless, 1998; Reuer and Leiblein, 2000). Anderson et al. (2007) found future positive returns related to increases in these costs in revenue declining periods. Therefore, we focus on SGA in our analysis on cost stickiness.

We are interested in analyzing how firms’ attainment of future increases in sales deters managers from adjusting resources in periods when activity decreases. According to our hypotheses stickiness is more pronounced for those firms being able to get greater future increases in sales. We use the interaction variable with the logarithm of one year ahead to current year sales (St+1/St) as a proxy for this effect. We assume that firms giving priority to future increases in sales are more able to get higher sales increases, with respect to those that do not give priority to supply firms’ customers or clients demand. Therefore, this variable offers a plausible proxy measuring comparative firms’ orientation to future increases in sales. Given that we also control for the effect of sales expectations on cost stickiness, the interaction variable with St+1/St captures the net effect of surplus resources devoted to meet unexpected demand fluctuations and/or to make proactive choices trying to increase sales above expectations. We are also interested in testing that the focus on firm profitability influences a lower degree of cost stickiness. As mentioned, studies linking flexibility or slack to efficiency usually use financial performance, and more precisely ROA, as an approach for efficiency, or merely as a measure of financial performance (e.g.Tan and Peng, 2003; Eben and Johnson, 2005; Tan and Wang, 2010; Bradley et al., 2011). Firms giving priority to current profitability will be more prone to apply measures that render higher profitability in the short run, and thus a more favourable change in profitability, with respect to those that do not have such priority. We use the interaction variable with CHROA as a proxy for this effect. Therefore, β3 and β4 allow tests for our hypotheses H1 and H2 respectively, being negative and positive their hypothesized signs respectively.

Managers’ current expectations for future sales are also a factor recognised as influencing resource adjustment when business activity decreases. Managers take into account that it will take time to reacquire resources if they have been reduced, and to bring those resources’ performance up to earlier levels. In their seminal study, Anderson et al. (2003) suggested that managers are deterred from adjusting resources, or delay this adjustment, when they believe that the decrease in sales is temporary. In this vein, they found that the degree of stickiness decreases with longer periods of time. Indeed, Yasukata and Kajiwara (2011) found a significant lower resource adjustment in periods of sales decreases matched with managers’ optimism about future sales. As we do not have information on managers’ expectations with respect to sales in the following year, we use mechanical predictive models based on current year sales and recent changes in sales. Following previous research using similar parsimonious predictive models for earnings and cash flow prediction (e.g.Carnes et al., 2003; Kim and Kross, 2005), we estimate the following equation:

(2)

where ∆S is sales changes with respect to previous year, measured as St-St-1. We also include dummy variables for years, controlling for specific yearly effects. We perform fixed effects estimations of Equation (2) and calculate SEXP with the corresponding predicted values, which are used in the corresponding interaction variable in Equation 1. A negative sign is expected for β5. Given the usefulness of mechanical or statistical procedures -as compared to expert opinion- in medical research (e.g.: Grove et al., 2000; Ægisdóttir et al., 2006), as well as in audit research (e.g.: Koh and Killough, 1990; Simnett, 1996, Gadenne and Iselin, 2000; Bhimani et al., 2009) , and the recognised human bias in business prediction and judgement (e.g.: Furnham and Boo, 2011; Kim and Prather-Kinsey, 2010; Louis et al., 2013), our variable is a plausible proxy for manager’s expectations for future sales.

We include the interaction variable with ROA as additional control variable, considering that more (less) profitable firms are less (more) urged to adjust resources when activity decreases. Moreover, less (more) profitable firms have likely higher (lower) SGA to sales ratio, thus being in less (more) need to get slack. Therefore a negative sign is expected for β6.

Following previous studies we also include assets and debt intensity in Equation 1 (ASSTS and DEBTS respectively), as additional interaction control variables. Firms will find significant difficulties and rigidities in depleting resources, especially for firms requiring more assets to support their sales. Asset-depletion is in most cases unrelated to circumstantial changes in sales, and consequently we expect an effect of asset intensity on cost stickiness. On the other hand, indebtedness moderates the sticky behaviour of firms. The fact that indebted firms have to meet interest payments, and may be subject to more scrutiny from creditors, encourages managers to apply efforts to adjust resources. Previous studies found empirical evidence of negative and positive coefficients for these variables, respectively (e.g.: Chen et al., 2012; Calleja et al., 2006; Dalla and Perego, 2013). We consequently expect these signs for β7 and β8 respectively. There are no expectations for the dummy variables for the years included in the model.

Sample

We use COMPUSTAT data for all industrial American firms (SIC codes 2 and 3) with data from 1979 to 2009. We do not consider firms in the service industry because of their different cost structure and business approach. Table 1 displays sample and data screening. We start with 37,730 year-data observations from 1,453 different firms. We remove 4,065 observations, and 32 firms, with missing data in sales and SGA for the current and previous years. The remaining 1,421 firms and 33,665 observations are the available sample for our study. Sales decrease with respect to previous year in around 30% of observations (10,236), while they increase in around 70% (23,427). Almost all firms experience both, increases and decreases in sales, throughout all years. In an attempt to clean their sample from the effects of mergers, acquisitions and other special operations Dalla and Perego (2013) refused observations with more than 50% sales changes with respect to previous year. We begin performing our study with the full sample, and look for robust results applying a lower restrictive removal criterion. Excluding observations with more than 70% sales change (i.e.: when the ratio of sales at t to sales at t-1 is lower than 0.3 and higher than 1.7) provides a final sample of 32,663 firm-year observations. The screening procedure leads to different samples depending on the availability of data for variables included in the different models and sample selection criteria. Given that Equation 1 includes lagged and forward variables, lower numbers of observations are available in the subsequent estimations. Some authors (e.g.Weiss, 2010; Anderson and Lanen, 2007; Chen et al., 2012) excluded observations where sales and costs move in opposite directions. However, Banker and Byzalov (2014) found that this sample selection criterion is fundamentally flawed. These authors also criticized the criterion of excluding observations for which costs exceed sales, as ambiguous and potentially source of bias. We also refuse this criterion, because it excludes observations with negative income.

Table 2 displays descriptive statistics for our sample. Sales and total assets are converted into values of 2009 applying the US inflation ratio. They are lower in periods when sales decrease with respect to periods when sales increase. The average sales increase across all periods is 21.08%, while the average decrease is 13.16% (i.e.: 1-0,8684), with a 10.67% mean increase for the overall sample. The asymmetric cost behaviour can be observed for SGA, with this univariate analysis: on average they increase 17.41% in periods of sales increasing 21%,

TABLE 1 SAMPLE AND DATA SCREENING 

TABLE 2 DESCRIPTIVE STATISTICS: MEAN VALUES 

Whole sample Sales increases Sales decreases
S* 4,114.676 4,293.625 3,705.918
ASS* 4,064.307 4,118.683 3,940.643
St/St-1 1.1067 1.2108 0.8684
SGAt/SGAt-1 1.1090 1.1741 0.9602
SGATS 0.2426 0.2369 0.2555
ROA 0.0818 0.1100 0.0174
ASSTS 0.9664 0.9340 1.0405
DEBTS 0.1765 0.1688 0.1941
CHROA -0.0039 0.0137 -0.0443

* Constant values: 000 $.

S is sales, ASS total assets, SGA selling general and administrative expenses, SGATS the ratio of SGA to sales, ROA return on assets, ASSTS the ratio of assets to sales, DEBTS the ratio of long term debt to sales and CHROA the change in ROA with respect to previous year. All data refer to year t, with the exception of variables with t-1 subscript.

while they only decrease 3.98% in periods of sales decreasing 13.16%. The average SGA to sales ratio (SGATS) is 24.26% for the whole sample, being slightly lower (higher) in periods when sales increase (decrease). The table displays an improvement in most ratios when sales increase, with respect to periods when they decrease: ROA, CHROA, ASSTS and DEBTS. Profitability improves in periods of increasing sales, while it declines when sales decrease. The slight decrease in profitability for the overall sample can be explained in terms of the increasing competitive pressure.

Results

Regression results

Table 3 displays estimations for Equation 1. We perform panel data regressions correcting for autocorrelation disturbances. The Hausman test rejects the null hypothesis of no correlation between individual effects and explanatory variables: χ2 values of 308.68 (with 31 d.f.), 363.79 (with 36 d.f), 377.11 (with 30 d.f.), 242.21 (with 36 d.f.) and 222.71 (with 36 d.f.) for the estimations in columns A, B, C, D and E respectively. As individual effects are correlated with the regressors in all estimations, the random effects estimator is inconsistent, while the fixed effects estimator is consistent and efficient. We therefore perform panel data estimations with fixed effects.

All models estimated in this table present a significant goodness of fit, with overall R-squares ranging from 40 to 53%. The coefficients for the dummies for years are not displayed, because of simplicity, and because they have no interest for our analysis. Columns A and B display estimations with the full sample, while estimations in columns C to E are performed excluding observations when sales change more than 70% with respect to previous year. Columns A and C display results for the basic model for cost stickiness, while columns B, D and E present results for the full model formulated in Equation 1. We explain results referring to β coefficients with its corresponding subscripts.

Results in column A suggest the existence of the typical sticky behaviour for SGA for the full sample: β1 and β2 are positive and negative respectively, and significant at the 1% confidence level. SGA increased 0.68% per 1% sales increase, while they only decreased 0.58% (the combined value of β1+β2) per 1% sales decrease. These results are in accordance with most previous research on cost stickiness.

As mentioned, column B displays results for the whole sample and the full model formulated in Equation 1. With the exception of the interaction variable for asset intensity, all control variables present the expected signs, and are significant with p<0.05. The coefficient β5 (corresponding to the interaction variable for the ratio of the expected sales at t+1 to sales at t) is significantly negative with p<0.01, indicating the existence of the expected stressing effect of expectations on cost stickiness. According to this result, resource adjustment in SGA when activity decreases is lower for higher future expected sales, and vice-versa. The coefficient β6 (corresponding to the interaction variable for ROA) is significantly negative (at p<0.01), according to expectations, indicating that the urgency to adjust SGA grows with lower profitability in periods of decreasing sales, and vice-versa. The significant positive sign of β8 (at p<0.05) reveals a moderating effect of indebtedness on cost stickiness. The more firms are indebted, the more they are urged to meet interest payments, and consequently to adjust SGA, when sales decrease.

With respect to our variables of interest, the coefficient β3 (corresponding to the interaction variable for one year ahead increase in sales) is significantly negative with p<0.01, indicating a stressing effect on cost stickiness and providing support for our hypothesis H1. This finding suggests that managers focusing on flexibility, and therefore giving priority to future increases in sales, prefer to adjust less SGA when activity decreases. Managers consider that the more

TABLE 3 FIXED EFFECTS ESTIMATIONS FOR THE INCIDENCE OF ONE YEAR AHEAD SALES INCREASE AND CHANGE IN PROFITABILITY ON COST STICKINESS. DEPENDENT VARIABLE: log (SGAt/SGAt-1) 

*, ** and *** indicate significance at the 10 per cent, 5 per cent and 1 per cent levels based on two-tailed tests, respectively.

SGA is selling general and administrative expenses, S is sales, D a dummy variable indicating that sales decrease with respect to previous year, DNEXT a dummy variable indicating that sales increase one year ahead with respect to current year, ROA return on assets, CHROA the change in ROA with respect to previous year, SEXP sales expectation, ASSTS the ratio of total assets to sales and DEBTS the ratio of long term debt to sales.

they reduce resources, the less they will be able to achieve sales increases in the following year. On the contrary, the coefficient β4 (corresponding to the interaction variable for profitability change) is significantly positive at p<0.01. This result suggests that upholding current profitability influences decisions on adjustment in SGA when activity decreases, and provides support for our hypothesis H2. Managers apply greater adjustments in SGA in order to avoid greater decreases in current profitability when sales decrease. They must balance the decision with the attainment of future increases in sales.

Columns C, D and E display results with the restricted sample. As mentioned before, these estimations are performed excluding cases with more than 70% sales change with respect to previous year. Indeed, column E offers estimations with standardized ROA values, using industry-year means and standard deviations (in the interaction variables with ROA and CHROA). Results are similar to those included in columns A and B. The coefficient for the interaction term with asset intensity is now negative and significant at p<0.01 (see β7 in columns D and E). As expected, asset intensity is a source of difficulties and rigidities when firms are faced to cut down SGA when sales decrease. We also repeat estimations excluding cases with more than 50% sales change, similarly to Dalla and Perego (2013), and results (not displayed) are substantially the same with respect to our variables of interest. Therefore, we find strong support for our hypotheses H1 and H2.

Assuming that results in columns B, D and E in Table 3 may be biased, because some of the covariates may be endogenous, we perform an additional analysis in order to provide robust results. One of our interaction variables in Equation 1 is built with one year ahead sales increase, which in its turn depends on additional factors, as for instance size (usually bigger firms grow less than smaller), SGA slack and increase in capacity, measured with investment in property plant and equipments (PPE). We regress log(Si,t+1/Si,t) depending on logASSi,t,, SGATSi,t and log(PPEi,t+1/PPEi,t). All variables present the expected significant signs (negative the former and positive the remaining variables). Given that the interaction variable with the corresponding residuals is significant when it is included in Equation 1, the suspected variable is endogenous, suggesting that it is plausible that managers consider future sales growth in decisions on resource adjustment, taking into account firm slack, size and investment. We consequently include additional variables in our model as instruments for this variable. More precisely, we use:

(3)

(4)

(5)

as instruments for:

(6)

We use the two-stages least-squares within estimator, available in STATA for the fixed effects option. Results with the full and restricted samples, displayed in columns A to C in Table 4, are essentially the same as those in Table 3. They are robust to a reverse casualty problem. More precisely, β3 and β4 are significantly negative and positive (with p<0.01), respectively in all cases. Therefore, our results are not likely biased by endogeneity, and they provide robust support for our hypotheses H1 and H2. As additional robustness tests, we repeat estimations with instrumental variables and excluding observations where sales and SGA move in opposite directions, following Anderson et al. (2003) sample selection criterion. Results (not displayed) provide additional strong support for our hypotheses H1 and H2, given that β4 is significantly positive with p<0.01 in all estimations, and β3 is significantly negative with p<0.05 (with p<0.1 in estimations with standardized ROA).

While firms frequently reverse the sign of the change in ROA (an approximately 50% average for all years), the full reversion is substantially lower: only 27.7% and 21.9% of increases and decreases in ROA are fully reverted in the next year respectively. Moreover, on average only around 19% of firms move above and below the median ROA across the periods analyzed in our sample. Therefore, a substantial part of changes in ROA can be considered a change in efficiency, and this variable is a plausible appropriate proxy for testing our hypothesis on the efficiency driver.

We repeat our analysis for total operating expenses, and results (not displayed) are essentially the same.

Additional analysis: cost behavior when sales increase

We complete our study analyzing the trade-off between profitability and future sales increases when sales increase. We formulate the following equation, similar to Equation 1, where the dummy variable IN indicates that sales increase in a given period with respect to previous year:

(7)

TABLE 4 FIXED EFFECTS ESTIMATIONS, ADJUSTING FOR ENDOGENEITY, FOR THE INCIDENCE OF ONE YEAR AHEAD SALES INCREASE AND CHANGE IN PROFITABILITY ON COST STICKINESS. DEPENDENT VARIABLE: log(SGAt/SGAt-1

*, ** and *** indicate significance at the 10 per cent, 5 per cent and 1 per cent levels based on two-tailed tests, respectively.

SGA is selling general and administrative expenses, S is sales, D a dummy variable indicating that sales decrease with respect to previous year, DNEXT a dummy variable indicating that sales increase one year ahead with respect to current year, ROA return on assets, CHROA the change in ROA with respect to previous year, SEXP sales expectation, ASSTS the ratio of total assets to sales and DEBTS the ratio of long term debt to sales.

The Hausman test rejects the null hypothesis of no correlation between individual effects and the explanatory variable, therefore indicating that the fixed effects estimator is consistent and efficient. We again find that log(Si,t+1/Si,t) is endogenous and consequently we use the two-stages least-squares within estimator using:

(8)

(9)

(10)

as instruments for:

(11)

Table 5 shows estimations. All models present significant goodness of fit in all estimations. The significant positive coefficient θ2 indicates how much costs increase when sales increase is larger than how much they decrease when sales decrease. As compared with our previous findings when analyzing decreases of sales, this result constitutes a complementary piece of evidence regarding the asymmetric behaviour of costs. Asset intensity is related with firms’ rigidities to adjust their organizations to sales fluctuations, consequently moderating this greater cost increase. The significant (at p<0.01) negative sign of θ7 is in accordance with this expectation, while it is not for the restricted sample. The significant (at p<0.05) positive sign of θ8 with the restricted sample indicates a stressing effect of indebtedness on cost increase when sales increase, suggesting that indebted firms are not urged to meet interest payments when activity increases. A likely explanation for this unexpected sign is that debt is a source of financial resources to endeavour expansion plans, with the subsequent achievement of growth through debt. Creditors’ requirements are less demanding in periods of economic growth. They loosen conditions for indebtedness in favourable economic settings. However, the coefficient for this variable is not significant at p<0.1 with the full sample. The significant positive sign of the interaction variable with ROA (θ6) reveals that firms with lower previous profitability are more prone to save resources with respect to more profitable firms, even in periods of economic growth, and vice versa for firms with an advantageous profitability background. The sign of θ5 is negative, significant at p<0.01 and contrary to expectations. A likely explanation for this surprising sign is that expectations for future sales are mainly based on past performance. Managers tend to have positive expectations for future sales when they have experienced favourable past performance. Given that SGA are mainly fixed, the increase in these costs is lower than the increase in sales when they soar.

TABLE 5 FIXED EFFECTS ESTIMATIONS, ADJUSTING FOR ENDOGENEITY, FOR THE INCIDENCE OF ONE YEAR AHEAD SALES INCREASE AND CHANGE IN PROFITABILITY ON COSTS WHEN SALES INCREASE. DEPENDENT VARIABLE: log(SGAt/SGAt-1) 

*, ** and *** indicate significance at the 10 per cent, 5 per cent and 1 per cent levels based on two-tailed tests, respectively.

SGA is selling general and administrative expenses, S is sales, IN a dummy variable indicating that sales increase with respect to previous year, DNEXT a dummy variable indicating that sales increase one year ahead with respect to current year, ROA return on assets, CHROA the change in ROA with respect to previous year, SEXP sales expectation, ASSTS the ratio of total assets to sales and DEBTS the ratio of long term debt to sales.

As for our variables of interest, the coefficient of θ3 is positive and significant at p<0.01 in estimations with the full and restricted samples, indicating the existence of a stressing effect of the interaction variable for future sales increase. There is a positive relationship between future sales growth and increases in SGA in periods of increasing sales. This result suggests that firms focusing on flexibility keep slack in SGA as a means to build capabilities that will fuel future growth. On the contrary, the sign of θ4 is negative and significant at p<0.01, indicating that orientation to profitability moderates the higher increase in SGA when activity increases. Unfavorable changes in profitability are associated with greater SGA, while positive changes in profitability are associated with lower expenditures in these expenses, which means that firms aiming at improving immediate profitability focus on monitoring SGA. These results provide support for the hypothesis on the flexibility orientation, as well as on the profitability strategic orientation, of firms when activity increases. They suggest that the trade-off between profitability and sales growth exists not only when sales decrease, but also when sales increase.

Results with fixed effects without instruments for the suspected endogeneous variable are essentially the same. We do not display these results for simplicity reasons.

Conclusions

In this study we analyze cost stickiness under the scope of the dilemma between current profitability and future sales growth. We use a sample of industrial firms with at least 20 years of data in COMPUSTAT. We focus on SGA considering that they are a sort of slack resources. They are substantially related to resources used in managing and planning, as well as in building firm capabilities. These resources allow firms to be in an advantageous position to face competition. We find evidence that the change in profitability and one year ahead sales increase significantly influence resource adjustment when firm sales decrease. According to our results, the higher (lower) the increase in one year ahead sales, the higher (lower) the asymmetric cost behaviour, considering that we control for sales expectations and other variables. Given that resource adjustment hinders the ability of firms to meet future increases in demand, managers consider that they will lose chances to increase future sales if they cut costs when activity decreases. We also find a significant positive moderating effect of change in profitability on cost stickiness. In periods of decreasing sales managers apply greater resource adjustment when they want to avoid greater decrease in profitability. They avoid maintaining slack resources despite it may curtail to take advantage of future opportunities. Our results suggest that managers balance advantages and drawbacks of both options, future sales increase and change in current profitability, when they must consider adjusting SGA when activity decreases. We also find that firms with high profitability are less urged adjusting resources than those with low profitability when activity decreases, and vice versa for firms with low profitability. All control variables included in our models present the expected significant signs. Results are robust to different estimation methods, sample selection criteria and alternative variables used.

We also find consistent results with our hypotheses when activity increases. SGA increase more when activity increases. In this setting future sales growth stresses the effect on SGA increases, while change in profitability moderates the effect on the increase in these resources. Our results suggest that the trade-off between profitability and sales growth exists not only when sales decrease, but also when sales increase.

Cost stickiness is not merely the result of firm failure to adjust resources when activity decreases. It is also a managers’ choice influenced by their expectations and strategies. More precisely, strategies of profitability and sales growth are important factors explaining resource adjustments when firm sales decrease.

Overall, our results enhance our understanding of firms’ costs behaviour and provide useful insights for the management and financial accounting literature. Depending on their strategic choices, when activity decreases managers may adjust resources to meet short term targets in terms of profitability, but they may also keep slack resources, despite gloomy sales expectations, to meet targets in terms of flexibility. One of the limitations of our study is that it does not use data on managers’ expectations. We use mere mechanical indicators of them. An additional limitation is that we do not use true indicators of firms’ strategies on flexibility and efficiency. We have analyzed mere specific issues related to flexibility and efficiency. More research is needed with more accurate measures of both, managers’ expectations and firms’ strategies.

Acknowledgments

We are grateful to the editor of Estudios de Economía, and to an anonymous reviewer whose comments contributed to improve substantially the paper. We also acknowledge the valuable comments of Martha Rodríguez and Klender Cortez for their valuable comments.

References

Adler, P.S., B. Goldoftas y D.I. Levine (1999). “Flexibility versus efficiency? A case study of model changeovers in the Toyota production system”, Organization Science, Vol. 10 (1): 43-67. [ Links ]

Ægisdóttir, S.A., M.J. White, P.M. Spengler, A.S. Magherman, L.A. Anderson, R.S. Cook, C.N. Nichols, G.K. Lampropoulos, B.S. Walker, G. Cohen y J.D. Rush (2006). “The Meta-Analysis of Clinical Judgment Project: Fifty-Six Years of Accumulated Research on Clinical Versus Statistical Prediction”, The Counseling Psychologist, Vol. 34(3): 341-382. [ Links ]

Anderson, M., O. Asdemir y A. Tripathy (2013). “Use of precedent and antecedent information in strategic cost management”, Journal of Business Research, Vol. 66: 643-650. [ Links ]

Anderson, M.C., R.D. Banker, R. Huang, y S.N. Janakiraman (2007). “Cost behavior and fundamental analysis of SG&A costs”, Journal of Accounting, Auditing & Finance, Vol. 22 (1): 1-28. [ Links ]

Anderson, M.C., R.D. Banker y S. Janakiraman (2003). “Are selling, general and administrative costs sticky?”, Journal of Accounting Research, Vol. 41 (1): 47-63. [ Links ]

Anderson, S.W. y W.N. Lanen (2007). “Understanding cost management: what can we learn from the evidence on ‘sticky costs’?”, “Understanding cost management: what can we learn from the evidence on ‘sticky costs’?”, https://research.mbs.ac.uk/accounting-finance/Portals/0/docs/2008/UnderstandingCostManagementWhatCanWeLearnfromtheEmpiricalEvidenceonStickyCosts.pdf . Accessed 18 February 2014 [ Links ]

Argilés-Bosch, J.M. y J. García-Blandón (2011). “The influence of size on cost behaviour associated with tactical and operational flexibility”, Estudios de Economía, Vol. 38 (2): 419-455. [ Links ]

Balakrishnan, R. y T.S. Gruca (2008). “Cost stickiness and core competency: a note”, Contemporary Accounting Research, Vol. 25 (4): 993-1006. [ Links ]

Balakrishnan R., E. Labro y N. Soderstrom (2014). “Cost structure and sticky costs”, Journal of Management Accounting Research, Vol. 26 (2): 91-116. [ Links ]

Balakrishnan, R., M.J. Petersen y N.S. Soderstrom (2004). “Does capacity utilization affect the stickiness of cost?”, Journal of Accounting Auditing & Finance, Vol. 19 (3): 283-299. [ Links ]

Baker, T. y R.E. Nelson (2005). “Creating something from nothing: resource construction through entrepreneurial bricolage”, Administrative Science Quarterly, Vol. 50: 329-366. [ Links ]

Banker R.D. y D. Byzalov (2014). “Asymmetric cost behaviour”, Journal of Management Accounting Research, Vol. 26 (2): 43-79. [ Links ]

Banker, R.D., D. Byzalov, M. Ciftci y R. Mashruwala (2014). “The moderating effect of prior sales changes on asymmetric cost behaviour”, Journal of Management Accounting Research, Vol. 26 (2): 221-242. [ Links ]

Banker, R.D., D. Byzalov y J.M. Plehn-Dujowich (2011a). “Sticky cost behavior: theory and evidence”. http://astro.temple.edu/~dbyzalov/sticky.pdf . Accessed 20 September 2012. [ Links ]

Banker, R.D., D. Byzalov y J.M. Plehn-Dujowich (2013). “Demand uncertainty and cost behaviour”, The Acounting Review, Vol. 89 (3): 839-865. [ Links ]

Banker, R.D. y T. Chen (2006). “Predicting earnings using a model based on cost variability and cost stickiness”, Accounting Review, Vol. 81 (2): 285-307. [ Links ]

Banker, R.D., R. Huang y R. Natarajan (2006). “Does SG&A expenditure create a long-lived asset?”, “Does SG&A expenditure create a long-lived asset?”, https://aaahq.org/mas/MASPAPERS2007/cost_behavior/Banker,%20Huang%20and%20Natarajan.pdf . Accessed 22 April 2014. [ Links ]

Banker, R.D., R. Huang y R. Natarajan (2011b). “Equity incentives and longterm value created by SG&A expenditure”, Contemporary Accounting Research, Vol. 28 (3): 794-830. [ Links ]

Baumgarten, D., U. Bonenkamp y C. Homburg (2010). “The information content of the SG&A ratio”, Journal of Management Accounting Research, Vol. 22; 1-22. [ Links ]

Bergh, D. y M. Lawless (1998). “Portfolio restructuring and limits to hierarchical governance: the effects of environmental uncertainty and diversification strategy”, Organization Science, Vol. 9; 87-102. [ Links ]

Bhimani, A., M.A. Gulamhussen y S. Lopes (2009). “The effectiveness of the auditor’s going-concern evaluation as an external governance mechanism: Evidence from loan defaults”, The International Journal of Accounting, Vol. 44; 239-255. [ Links ]

Bradley, S.W., J. Wiklund y D.A. Shepherd (2011). “The importance of slack for new organizations facing ‘Tough’ environments”, Journal of Management Studies, Vol. 48 (5): 1071-1097. [ Links ]

Calleja, K., M. Stelarios y D.C. Thomas (2006). “A note on cost stickiness: some international comparisons”, Management Accounting Research, Vol. 17: 127-140. [ Links ]

Cannon, J.N. (2014). “Determinants of ‘sticky costs’: an analysis of cost behaviour using United States air transportation industry data”, The Accounting Review, Vol. 89 (5): 1645-1672. [ Links ]

Carnes, T.A., J.P. Jones, T.B. Biggart y K.J. Barker (2003). “Just-in-time inventory systems innovation and the predictability of earnings”, International Journal of Forecasting, Vol. 19 (4): 743-749. [ Links ]

Chen, C.X., H. Lu y T. Sougiannis (2012). “The agency problem, corporate governance and the asymmetrical behavior of selling, general, and administrative costs”, Contemporary Accounting Research, Vol. 29 (1): 252-282. [ Links ]

Cohen, D.A., A. Dey y Z. Lys (2008). “Real and Accrual-Based Earnings Management in the Pre- and Post-Sarbanes-Oxley Periods”, The Accounting Review, Vol. 83: 757-787. [ Links ]

Cooper, R. y R.S. Kaplan (1992). “Activity based cost systems: measuring the cost of resource usage”, Accounting Horizons, Vol. 6 (September): 1-13. [ Links ]

Cooper, R. y R.S. Kaplan (1998). The Design of Cost Management Systems Text, Cases and Readings. Prentice Hall, New Jersey. [ Links ]

Cyert, R. y J. March (1956). “Organizational factors in the theory of oligopoly”, Quarterly Journal of Economics, Vol. 70 (1): 44-64. [ Links ]

Cyert, R. y J. March (1963). A behavioural theory of the firm. Prentice Hall, New Jersey. [ Links ]

Dalla Via, N. y P. Perego (2014). “Sticky cost behaviour: evidence from small and medium sized companies”, Accounting & Finance, Vol. 54 (3): 753-778. [ Links ]

Damanpour, F. (1987). “The adoption of technological, administrative, and ancillary innovations: Impacts of organizational factors”, Journal of Management, Vol. 13: 675-688. [ Links ]

Furnham, A. y H.C. Boo (2011). “A literature review of the anchoring effect”, The Journal of Socio-Economics, Vol. 40: 35-42. [ Links ]

Fuss, M. y D.L. McFadden (1978). “Flexibility versus efficiency in ex ante plan design”. In: Fuss M, McFadden DL (ed.). Production Economics: a Dual Approach to Theory and Applications, Vol I: The theory of Production. North-Holland, Amsterdam; 312-364. [ Links ]

Gadenne, D. y E.R. Iselin (2000). “Properties of accounting and finance information and their effects on the performance of bankers and models in predicting company failure”, Journal of Business Finance & Accounting, Vol. 27 (1): 155-193. [ Links ]

George, G. (2005). “Slack resources and the performance of privately held firms”, Academy of Management Journal, Vol. 48 (4): 661-676. [ Links ]

Greve, H.R. (2003). Organizational Learning from Performance Feedback: a Behavioral Perspective on Innovation and Change. Cambridge University Press, Cambridge. [ Links ]

Grove, W. M., D.H. Zald, B.S. Lebow, B.E. Snitz y Ch. Nelson (2000). “The Meta-Analysis of Clinical Judgment Project: Fifty-Six Years of Accumulated Research on Clinical Versus Statistical Prediction”, Psychological Assessment, Vol. 12 (1): 19-30. [ Links ]

Holzhacker, M., R. Krishnan y M.D. Mahlendorf (2015). “The impact of changes in regulation on cost behaviour”, Contemporary Accounting Research, Vol. 32 (2): 534-566. [ Links ]

Hope, O.K. y W.B. Thomas (2008). “Managerial empire building and firm disclosure”, Journal of Accounting Research, Vol. 46 (3): 591-626. [ Links ]

Jensen, M.C. (1986). “Agency costs of free cash flow, corporate finance, and takeovers”, American Economic Review, Vol. (76) 2: 323-329. [ Links ]

Jensen, M.C. y W.H. Meckling (1976). “Theory of the firm: managerial behavior, agency costs and ownership structure”, Journal of Financial Economics, Vol. 3 (4): 305-60. [ Links ]

Kama, I. y D., Weiss (2013). “Do earnings targets and managerial incentives affect sticky costs?”, Journal of Accounting Research, Vol. 51 (1): 201-224. [ Links ]

Kim, M. y W. Kross (2005). “The ability of earnings to predict future operating cash flows has been increasing - not decreasing”, Journal of Accounting Research, Vol. 43 (5): 753-780. [ Links ]

Kim, M. y J. Prather-Kinsey (2010). “An additional source of financial analysts’ earnings forecast errors: imperfect adjustments for cost behavior”, Journal of Accounting, Auditing & Finance, Vol. 25 (1): 27-51. [ Links ]

Koh, H.C. y L.N. Killough (1990). “The use of multiple discriminant analysis in the assessment of the going-concern status of an audit client”, Journal of Business Finance & Accounting, Vol. 17 (2): 179-192. [ Links ]

Louis, H., A.X. Sun y O. Urcan (2013). “Do Analysts Sacrifice Forecast Accuracy for Informativeness?”, Management Science, Vol. 59 (7): 1688-1708. [ Links ]

Martinelli, C. (1997). “Small firms, borrowing constraints, and reputation”, Journal of Economic Behavior & Organization, Vol. 33: 91-105. [ Links ]

McKee, T.E. (2003). “Rough sets bankruptcy prediction models versus auditor signalling rates”, Journal of Forecasting, Vol. 25: 569-586. [ Links ]

Mishina, Y., T.G. Pollock y J.F. Porac (2004). “Are more resources always better for growth? Resource stickiness in market and product expansion”, Strategic Management Journal, Vol. 25: 1179-1197. [ Links ]

Penrose, E. (1959). Theory of the growth of the firm. Willey, New York. [ Links ]

Phan, P. y C. Hill (1995). “Organizational restructuring and economic performance in leveraged buyouts: an ex-post study”, Academy of Management Journal, Vol. 38 (3): 704-739. [ Links ]

Reuer, J. y M. Leiblein (2000). “Downside risk implications of multinationality and international joint ventures”, Academy of Management Journal, Vol. 43: 203-214. [ Links ]

Roychowdhury, S. (2006). “Earnings Management Through Real Activities Manipulation”, Journal of Accounting and Economics, Vol. 42: 335-70. [ Links ]

Simnett, R. (1996). “The effect of information selection, information processing and task complexity on predictive accuracy of auditors”, Accounting Organizations and Society, Vol. 21 (7/8): 699-719. [ Links ]

Singh, J. (1986). “Performance slack and risk taking in organizational decision making”, Academy of Management Journal, Vol. 29: 562-585. [ Links ]

Starr, J.E. y I.C. MacMillan (1990). “Resource cooptation via social contracting: resource acquisition strategies for new ventures”, Strategic Management Journal, Vol. 11 (Special issue): 79-92. [ Links ]

Stigler, G. (1939). “Production and distribution in the short run”, The Journal of Political Economy, Vol. 47(3): 305-327. [ Links ]

Tan, J. y M.W. Peng (2003). “Organizational slack and firm performance during economic transitions: two studies from an emerging economy”, Strategic Management Journal, Vol. 24 (13): 1249-1263. [ Links ]

Tan, J. y L. Wang (2010). “Flexibility-efficiency tradeoff and performance implications among Chinese SOEs”, Journal of Business Research, Vol. 63: 356-362. [ Links ]

Thompson, J.D. (1967). Organizations in Action. McGraw-Hill, New York. [ Links ]

Weiss, C. (2001). “On flexibility”, Journal of Economic Behaviour & Organization, Vol. 46 (3): 347-356. [ Links ]

Weiss, D. (2010). “Cost behavior and analysts’ earnings forecasts”. Accounting Review, Vol. 85 (4): 1441-1471. [ Links ]

Yasukata, K. y T. Kajiwara (2011). “Are sticky costs the result of deliberate decision of managers?”, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1444746.Accessed 21 June 2012. [ Links ]

1 McKee (2003) found that auditor signalling rate for bankrupt companies, when other opinion modifications related to going concern issues were also considered, was similar to models’ prediction accuracies.

Annex: Variable definition

Received: December 2015; Accepted: June 2016

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License