Abstract
We investigate whether asymmetric cost behavior (also termed cost stickiness) and investors’ assessment of asymmetric cost behavior are affected by firms’ long-term growth expectations. Using a sample of US firms for the period 1990–2014, we first predict and find that cost stickiness, though a short-term phenomenon, is greater when firms have high rather than low long-term growth expectations. Second, we predict that unexpected cost stickiness is negatively evaluated by investors. Investigating cumulative abnormal returns surrounding earnings announcement dates, we find support for this prediction. Third, we investigate this finding in more detail, dependent on long-term growth. We argue that the reasons for cost asymmetry differ between firms with high versus low long-term growth expectations. We expect these differences to result in differing investor reactions. In line with this prediction, our results reveal that investors react more negatively to unexpected cost stickiness when a firm has low long-term growth opportunities. This finding supports the assumption that investors perceive agency motives as more likely to explain the unexpected cost stickiness for these firms.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Because the capital market will react only to new information, we focus on unexpected cost stickiness for this hypothesis, i.e., the portion of costs of idle capacity that is not expected by the capital market.
We use US firms to ensure high data availability and comparability with prior studies.
We find that our measure for long-term growth prospects provides information about future sales growth in addition to measures suggested by prior research.
Adjustment costs are costs that a firm incurs to reduce committed resources and to replace those resources if demand is restored (Anderson et al. 2003).
While long-term growth expectations are likely to affect expected cost of idle capacity, there is no reason to believe that costs of the same adjustment (e.g., for hiring a new employee) will differ between low- and high-growth firms.
In contrast, Kama and Weiss (2013) and Dierynck et al. (2012) show that depending on motivations underlying managers’ resource adjustments, agency-driven incentives can lead to inefficiently low cost stickiness. These authors show that when sales decrease, managers reduce costs more aggressively in the presence of incentives to meet earnings targets, thereby resulting in a lower degree of cost stickiness.
We again refer to the different drivers of cost stickiness when developing H3. More precisely, we use the different drivers to explain how investors’ evaluate cost stickiness, an implication not examined in Chen et al. (2012).
For our operationalization of unexpected cost stickiness, we argue that the capital market forms its expectation of the level of cost stickiness based on the level four quarters ago. In summary, therefore, unexpected cost stickiness is operationalized as the difference between current cost stickiness and cost stickiness one year earlier.
Compared with the original method to detect cost stickiness suggested by Anderson et al. (2003), the Weiss measure can be used as an independent variable to examine the impact of cost stickiness on capital market evaluation. Many prior studies follow Anderson et al. (2003) and use a regression model to estimate cost stickiness at the industry or firm level, which results in an estimated regression function with one static regression coefficient that captures the degree of cost stickiness of the sample firms in the sample period. Thus, this measure is constant over time and/or is not firm specific. To analyze the market reaction to cost stickiness, we need a firm- and time-specific measure of cost stickiness. Thus, we use the cost stickiness measure suggested by Weiss (2010).
To minimize data loss, we use STICKYi,t-8 to calculate ∆STICKYi,t if data on cost stickiness for t-4 is missing. However, if we use only cost stickiness of t − 4 to calculate ∆STICKYi,t, our results for H2 and H3 are inferentially identical. Further, our results are unchanged if we define ∆STICKYi,t as the difference between cost stickiness in the current period and cost stickiness in the most recent of four fiscal quarters with available data.
We decided to use a firm-based instead of an industry-based measure because we expect a firm-based measure to more precisely capture growth expectations. For example, an innovative firm in the transportation industry might expect (and have) higher growth rates than well-established and mature firms in that industry. Further, we believe a firm-based measure is better able to account for multi-segment firms.
If we use industry fixed effects instead of firm fixed effects, our results for H1, H2, and H3 are unchanged.
Our results are inferentially identical if we use asset intensity and employee intensity without a logarithm.
We include a robustness test using the growth of order backlog and analysts’ sales forecasts. However, information on analysts’ forecasts is missing for about 65% of the observations we use in our main analysis, and data on the growth of order backlog is not available for about 69% of our observations.
In an additional analysis, we also use analysts’ forecast errors as a proxy for unexpected earnings.
Sales growth is calculated as the difference between the sales of firm i in quarter t and its sales in the same quarter one year prior, scaled by the sales of firm i in the same quarter one year prior.
We exclude financial firms (SIC codes 6000-6999) and utility firms (SIC codes 4900-4999) because the structure of their cash flows and financial statements differs substantially from those of all other industries.
To calculate all variables, we require data for the five preceding fiscal quarters for every observation. Thus, we also gather data for the years 1988 and 1989 to calculate these variables.
We also delete duplicate fiscal quarters (97) and 4488 observations for which a change in fiscal year occurred during the previous four quarters. These and the other criteria described in the text result in 172,113 firm-quarter observations.
These missing values occur because sales and SG&A costs do not change in the same direction in the current period or because over the last four quarters, there was either no quarter with increasing sales and increasing SG&A costs or no quarter with decreasing sales and decreasing SG&A costs. In the additional analysis section, we provide a robustness check that uses an alternative measure for STICKYi,t to limit the loss of observations and rule out the possibility that our results are affected by a selection bias.
For the analysis of H2 and H3, we must delete 20,906 firm-quarter observations due to missing values for ∆STICKYi,t and LAGSTICKYi,t.
We conduct a median split and define GDP growth above 2.9% as high and GDP growth below or equal to 2.9% as low.
We report two-tailed tests throughout.
We find a significant negative coefficient for LTGROWTHi,t although we included alternative growth measures from prior literature in the regression model as control variables into the regression model. Further, adding our main dependent variable LTGROWTHi,t to the regression model leads to an increase of adjusted R-squared from 0.1180 (in Regression Model 1 excluding LTGROWTHi,t) to 0.1185 (in Regression Model 1 including LTGROWTHi,t).
We use yearly values for order backlog because of limited data availability for quarterly values.
However, the results should be regarded with caution because, due to data limitations, we use yearly data for order backlog, which reflects the current fiscal quarter less accurately than quarterly data. Further, sample size is reduced by about 65% because of missing data for analyst forecasts. In addition, high variance inflation factors for ln∆OB and MISSING∆OB indicate that the results might be biased due to multicollinearity.
For the fundamental signals earnings quality and audit qualification, quarterly data are not available. Thus, we do not include those variables.
Because we do not have quarterly data on the number of employees, we use the number of employees at the next fiscal year end after quarter t.
References
Abarbanell JS, Bushee BJ (1997) Fundamental analysis, future earnings, and stock prices. J Account Res 35:1–24
Anderson M, Banker RD, Janakiraman SN (2003) Are selling, general, and administrative costs ‘sticky’? J Account Res 41:47–63
Anderson M, Banker RD, Rong H, Janakiraman SN (2007) Cost behavior and fundamental analysis of SG&A costs. J Account Audit Finance 22:1–28
Anthony JH, Ramesh K (1992) Association between accounting performance measures and stock prices. J Account Econ 15:203–227
Arellano M (1993) On the testing of correlated effects with panel data. J Econom 59:87–97
Balakrishnan R, Petersen MJ, Soderstrom NS (2004) Does capacity utilization affect the “stickiness” of cost? J Account Audit Finance 19:283–299
Ball R, Brown PR (1968) An empirical evaluation of accounting income numbers. J Accout Res 6:159–178
Banker RD, Byzalov D (2014) Asymmetric cost behavior. J Manag Account Res 26:43–79
Banker RD, Chen JYS, Kama I, Weiss D (2013) Cost behavior, earnings expectations and stock prices. SSRN. http://ssrn.com/abstract=2312550. Accessed 13 Jan 2019
Banker RD, Byzalov D, Ciftci M, Mashruwala R (2014) The moderating effect of prior sales changes on asymmetric cost behavior. J Manag Account Res 26:221–242
Beaver WH (1968) The information content of annual earnings announcements. J Account Res 6:67–92
Bertrand M, Mullainathan S (2003) Enjoying the quiet life? Corporate governance and managerial preferences. J Polit Econ 111:1043–1075
Carhart MM (1997) On persistence in mutual fund performance. J Finance 52:57–82
Chen CX, Lu H, Sougiannis T (2012) The agency problem, corporate governance, and the asymmetrical behavior of selling, general, and administrative costs. Contemp Account Res 29:252–282
Chen CX, Gores T, Nasev J (2013) Managerial overconfidence and cost stickiness. SSRN. http://ssrn.com/abstract=2208622. Accessed 13 Jan 2019
Churchill NC, Lewis VL (1983) The five stages of small business growth. Harv Bus Rev 61:30–50
Ciftci M, Mashruwala R, Weiss D (2016) Implications of cost behavior for analysts’ earnings forecasts. J Manag Account Res 28:57–80
Dickinson V (2011) Cash flow patterns as a proxy for firm life cycle. Account Rev 86:1964–1994
Dierynck B, Landsman WR, Renders A (2012) Do managerial incentives drive cost behavior? Evidence about the role of the zero earnings benchmark for labor cost behavior in private Belgian firms. Account Rev 87:1219–1246
Fama EF, French KR (1993) Common risk factors in the returns on stocks and bonds. J Financ Econ 33:3–56
Flamholtz E, Randle Y (2007) Growing pains—transitioning from an entrepreneurship to a professionally managed firm, 4th edn. Wiley, San Francisco
Gaure S (2011) OLS with multiple high dimensional category dummies. https://ideas.repec.org/p/hhs/osloec/2010_014.html. Accessed 30 Sept 2018
Guimarães P, Portugal P (2010) A simple feasible procedure to fit models with high-dimensional fixed effects. Stata J 10:628–649
Hausman JA (1978) Specification tests in econometrics. Econom 46:1251–1271
Jegadeesh N, Livnat J (2006) Revenue surprises and stock returns. J Account Econ 41:147–171
Jovanovic B (1982) Selection and the evolution of industry. Econometrica 50:649–670
Kama I, Weiss D (2013) Do earnings targets and managerial incentives affect sticky costs? J Account Res 51:201–224
Lev B, Thiagarajan SR (1993) Fundamental information analysis. J Account Res 31:190–215
Miller D, Friesen PH (1984) A longitudinal study of the corporate life cycle. Manag Sci 30:1161–1183
Mueller DC (1969) A theory of conglomerate mergers. Quart J Econ 83:643–659
Myers SC (1984) The capital structure puzzle. J Finance 39:575–592
Quinn RE, Cameron K (1983) Organizational life cycles and shifting criteria of effectiveness: some preliminary evidence. Manag Sci 29:33–51
Rajgopal S, Shevlin T, Venkatachalam M (2003) Does the stock market fully appreciate the implications of leading indicators for future earnings? Evidence from order backlog. Rev Account Stud 8:461–492
Schaffer M, Stillman S (2016) XTOVERID: Stata module to calculate tests of overidentifying restrictions after xtreg, xtivreg, xtivreg2, xthtaylor. http://EconPapers.repec.org/RePEc:boc:bocode:s456779. Accessed: 13 Jan 2019
Sergio C (2014) REGHDFE: Stata module to perform linear or instrumental-variable regression absorbing any number of high-dimensional fixed effects. Boston College Department of Economics. https://ideas.repec.org/c/boc/bocode/s457874.html. Accessed 05 Apr 2018
Weiss D (2010) Cost behavior and analysts’ earnings forecasts. Account Rev 85:1441–1471
Wernerfelt B (1985) The dynamics of prices and market shares over the product life cycle. Manag Sci 31:928–939
Williamson OE (1963) Managerial discretion and business behavior. Am Econ Rev 53:1032–1057
Acknowledgements
We thank Ralf Ewert (Editor), two anonymous reviewers, Jan Diebecker, Thorsten Knauer, Christian Rose, Friedrich Sommer, participants at the 39th European Accounting Association Annual Congress, and participants at the American Accounting Association Annual Meeting 2016 for their helpful comments and suggestions.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Variable | Description |
|---|---|
CAR[− 1;1]i,t | Cumulative abnormal returns for an event period of 3 days surrounding the announcement day of quarterly earnings of firm i and quarter t, calculated using the market model with an estimation period of 250 days with a minimum number of 100 returns and a 90-day gap between the estimation and event window. The market return is calculated as the value-weighted return on all NYSE, AMEX, and NASDAQ stocks from CRSP |
BHAR[− 1;60]i,t | Buy-and-hold abnormal returns for an event period of 1 day before until 60 days after the announcement day of the quarterly earnings of firm i and quarter t, calculated with an estimation period of 250 days with a minimum number of 100 returns and a 90-day gap between the estimation and event window. The market return is calculated as the value-weighted return on all NYSE, AMEX, and NASDAQ stocks from CRSP |
STICKY i,t | \( ln\left( {\frac{\Delta SG\& A}{\Delta SALES}} \right)_{{i,\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\tau } }} - ln\left( {\frac{\Delta SG\& A}{\Delta SALES}} \right)_{{i,\bar{\tau }}} \) where \( \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\tau } \) is the most recent of four fiscal quarters of firm i with a decrease in sales and \( \bar{\tau } \) is the most recent of four fiscal quarters with an increase in sales |
LAGSTICKY i,t | Cost stickiness for the same quarter one year ago (STICKYi,t−4) or if STICKYi,t−4 is missing for two years ago (STICKYi,t−8) |
∆STICKY i,t | \( STICKY_{i,t} {-} LAGSTICKY_{i,t} \) where LAGSTICKYi,t is equal to STICKYi,t−4 (or if STICKYi,t−4 is missing equal to STICKYi,t−8) |
∆SG&A i,t | Difference between the SG&A costs (Compustat item XSGAQ) of firm i in quarter t and its SG&A costs in quarter t–1 |
∆SALES i,t | Difference between the sales (Compustat item SALEQ) of firm i in quarter t and its sales in quarter t–1 |
LTGROWTH i,t | Dummy variable equal to 1 if the current life cycle stage of firm i in quarter t is the growth stage and zero otherwise. The current firm life cycle stage of firm i in quarter t equals the growth stage if the accumulated cash flow data of firm i over the last four fiscal quarters (t − 3 to t) corresponds to the following pattern: Cash flow from operating activities (\( \mathop \sum \nolimits_{n = 0}^{3} OANCFQ_{i,t - n} \)) is positive, cash flow from investing activities (\( \mathop \sum \nolimits_{n = 0}^{3} IVNCFQ_{i,t - n} \)) is negative, and cash flow from financing activities (\( \mathop \sum \nolimits_{n = 0}^{3} FINCFQ_{i,t - n} \)) is positive |
OANCFQ i,t | For the first fiscal quarter, the quarterly cash flow from operating activities OANCFQi,t equals the year-to-date net cash flow from operating activities (Compustat item OANCFY). For the second, third and fourth fiscal quarters, OANCFQ equals OANCFYi,t − OANCFYi,t−1 |
IVNCFQ i,t | For the first fiscal quarter, the quarterly cash flow from investing activities IVNCFQi,t equals the year-to-date net cash flow from investing activities (Compustat item IVNCFY). For the second, third and fourth fiscal quarters, IVNCFQ equals IVNCFYi,t − IVNCFYi,t−1 |
FINCFQ i,t | For the first fiscal quarter, the quarterly cash flow from financing activities FINCFQi,t equals the year-to-date net cash flow from financing activities (Compustat item FINCFY). For the second, third and fourth fiscal quarters, FINCFQ equals FINCFYi,t − FINCFYi,t−1 |
lnAINT i,t | Logarithm of the ratio of total assets (Compustat item ATQ) to sales (Compustat item SALEQ) of firm i in quarter t |
lnEINT i,t | Logarithm of the ratio of number of employees (Compustat item EMP) at the next fiscal year end after quarter t to sales (Compustat item SALEQ) of firm i in quarter t |
SUCCDEC i,t | Dummy variable that equals 1 if sales (Compustat item SALEQ) declined not only in the fiscal quarter, which was used to calculate STICKYi,t, but also in the previous fiscal quarter, and zero otherwise |
SUCCINC i,t | Dummy variable that equals 1 if sales (Compustat item SALEQ) increased not only in the fiscal quarter, which was used to calculate STICKYi,t, but also in the previous fiscal quarter, and zero otherwise |
∆GDP i,t | GDP growth (as percentage) in the current fiscal quarter. Source: http://www.bea.gov/briefrm/gdp.htm. |
∆IB i,t | Difference between net income before extraordinary items (Compustat item IBQ) of firm i in quarter t and net income before extraordinary items of firm i in quarter t − 4, scaled by the market value of equity (Compustat items PRCCQ · CSHOQ) in quarter t–4 |
∆CAPEX i,t | Difference between capital expenditures of firm i in quarter t (CAPEXi,t) and capital expenditures of firm i in quarter t − 4, scaled by the market value of equity (Compustat items PRCCQ · CSHOQ) in quarter t − 4 |
CAPEX i,t | Quarterly capital expenditure for firm i in quarter t is calculated as the year-to-date capital expenditure (Compustat item CAXY) for the first fiscal quarter and CAXYi,t− CAXYi,t−1 for the second, third and fourth fiscal quarters |
∆SG i,t | Difference between the sales growth of firm i in quarter t (SGi,t) and its sales growth in quarter t − 1 |
SG i,t | Sales growth is calculated as the difference between the sales (Compustat item SALEQ) of firm i in quarter t and its sales in quarter t − 4, scaled by the sales of firm i in quarter t − 4 |
∆RISK i,t | Difference between the systematic risk of firm i in quarter t (RISKi,t) and the systematic risk in quarter t − 1 |
RISK i,t | Systematic risk estimated from a regression of the monthly stock returns of firm i on a value-weighted market index over a 60-month estimation period prior to the end of a fiscal quarter and a minimum of ten returns |
∆ASSETS i,t | Difference between total assets (Compustat item ATQ) of firm i in quarter t and total assets of firm i in quarter t − 4, scaled by the market value of equity (Compustat items PRCCQ·CSHOQ) in quarter t − 4 |
∆LEV i,t | Difference between the leverage ratio (LEVi,t) of firm i in quarter t and the leverage ratio of firm i in quarter t − 1 |
LEV i,t | Leverage ratio is defined as long-term debt (Compustat item DLTTQ) divided by the book value of common equity (Compustat item CEQQ) |
CAPEXS i,t | Quarterly capital expenditure (CAPEXi,t) scaled by sales for firm i in quarter t |
NOAG i,t | Growth of net operating assets, calculated as the difference between the net operating assets (Compustat item NOAQ) of firm i in quarter t and its net operating assets in quarter t − 4, scaled by the net operating assets of firm i in quarter t − 4 |
RD i,t | Research and development expenses (Compustat item XRDQ) of firm i in quarter t, scaled by its sales (Compustat item SALEQ) in quarter t |
EPS i,t | EPSi,t (Compustat item EPSPXQ) is the quarterly earnings per share |
ROA i,t | Average return on assets over the last four quarters, calculated as the sum of net income for quarter t − 3 to quarter t for firm i, scaled by the average total assets of firm i for quarter t and quarter t − 4 |
PM i,t | Profit margin calculated as the net income (Compustat item NIQ) of firm i in quarter t, scaled by the sales (Compustat item SALEQ) of firm i in quarter t |
STICKY2 i,t | \( \left( {\frac{\Delta SG\& A}{\Delta SALES}} \right)_{{i,\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\tau } }} - \left( {\frac{\Delta SG\& A}{\Delta SALES}} \right)_{{i,\bar{\tau }}} \) where \( \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\tau } \) is the most recent of four fiscal quarters of firm i with a decrease in sales and \( \bar{\tau } \) is the most recent of four fiscal quarters with an increase in sales |
LAGSTICKY2 i,t | Cost stickiness for the same quarter one year ago (STICKY2i,t−4) or if STICKY2i,t−4 is missing for two years ago (STICKY2i,t–8). |
∆STICKY2 i,t | \( STICKY2_{i,t} {-} LAGSTICKY2_{i,t} \) where LAGSTICKY2i,t is equal to STICKY2i,t−4 (or if STICKY2i,t−4 is missing equal to STICKY2i,t−8) |
FINSLACK i,t | Financial slack, calculated as the ratio of cash and short-term investments (Compustat item CHEQ) to total assets (Compustat item ATQ) |
SUE i,t | \( \frac{{EPS_{i,t} - E\left( {EPS_{i,t} } \right)}}{{\sigma_{i,t} }} \) Standardized unexpected earnings for firm i in quarter t, where EPSi,t (Compustat item EPSPXQ) is the quarterly earnings per share, E(EPSi,t) is the expected quarterly earnings per share prior to the earnings announcement, and σi,t is the standard error of quarterly earnings growth: \( E\left( {EPS_{i,t} } \right) = EPS_{i,t - 4} + \frac{1}{8}\mathop \sum \limits_{j = 1}^{8} \left( {EPS_{i,t - j} - EPS_{i,t - j - 4} } \right) \) \( \sigma_{i,t} = \frac{1}{7}\sqrt {\mathop \sum \limits_{j = 1}^{8} \left( {EPS_{i,t - j} - E\left( {EPS} \right)_{i,t - j} } \right)^{2} } \). |
lnBM i,t | Logarithm of the book-to-market ratio of firm i in quarter t, which is calculated as the book value of common equity (Compustat item CEQQ) of firm i in quarter t divided by the market value of common equity (Compustat items PRCCQ · CSHOQ) of firm i at the end of quarter t. When the book value of common equity is negative, we replace the lnBM variable with the lowest number of the distribution of the logarithm of the book-to-market ratio. |
lnSIZE i,t | Logarithm of the market value of equity (Compustat items PRCCQ · CSHOQ) of firm i at the end of quarter t |
lnVAR i,t | \( ln\left( {\frac{\Delta SG\& A}{\Delta SALES}} \right)_{i,t} \) lnVARi,t captures the proportion of the SG&A cost response to a change in sales. A greater value of lnVAR indicates greater cost variability |
CHGEPS i,t | \( \frac{{EPS_{i,t} - EPS_{i,t - 4} }}{{P_{i,t - 4} }} \) Difference between the earnings per share (Compustat item EPSPXQ) of firm i in quarter t and the earnings per share of firm i in quarter t − 4, scaled by the stock price (Compustat item PRCCQ) of firm i at the end of quarter t − 4 |
∆INVT i,t | \( \frac{{Inventories_{i,t} }}{{SALES_{i,t} }} - \frac{{Inventories_{i,t - 4} }}{{SALES_{i,t - 4} }} \) Difference between total inventories (Compustat item INVTQ) scaled by the sales (Compustat item SALEQ) of firm i in quarter t and total inventories scaled by the sales of firm i in quarter t − 4 |
∆RECT i,t | \( \frac{{Accounts receivable_{i,t} }}{{SALES_{i,t} }} - \frac{{Accounts receivable_{i,t - 4} }}{{SALES_{i,t - 4} }} \) Difference between receivables (Compustat item RECTQ) scaled by the sales of firm i in quarter t and receivables scaled by the sales (Compustat item SALEQ) of firm i in quarter t − 4 |
∆CAPEX2 i,t | \( \frac{{Firm Capital Expenditures_{i,t} }}{{Industry Capital Expenditure_{i,t} }} - \frac{{Firm Capital Expenditures_{i,t - 4} }}{{Industry Capital Expenditure_{i,t - 4} }} \) Difference between firm capital expenditure scaled by the industry capital expenditure of firm i in quarter t and firm capital expenditure scaled by the industry (defined by four-digit SIC codes) capital expenditure of firm i in quarter t − 4. Quarterly capital expenditure for firm i in quarter t is calculated as the year-to-date capital expenditure (Compustat item CAXY) for the first fiscal quarter and CAXYi,t− CAXYi,t−1 for the second, third and fourth fiscal quarters. |
∆GM i,t | \( \frac{{Gross Margin_{i,t - 4} }}{{SALES_{i,t - 4} }} - \frac{{Gross Margin_{i,t} }}{{SALES_{i,t} }} \) Difference between the gross margin scaled by the sales (Compustat item SALEQ) of firm i in quarter t − 4 and the gross margin scaled by the sales of firm i in quarter t. Gross margin of firm i in quarter t is calculated as difference between the sales (Compustat item SALEQ) and cost of goods sold (Compustat item COGSQ) of firm i in quarter t. |
∆TAX i,t | \( \left( {\left( {\frac{1}{12}\mathop \sum \limits_{j = 1}^{12} \frac{{TXT_{i,t - j} }}{{PI_{i,t - j} }}} \right) - \frac{{TXT_{i,t} }}{{PI_{i,t} }}} \right) \cdot \frac{{EPS_{i,t} - EPS_{i,t - 4} }}{{P_{i,t} }} \) where TXTi,t is total income taxes (Compustat item TXTQ), PIi,t is pretax income (Compustat item PIQ), EPSi,t is the quarterly earnings per share (Compustat item EPSPXQ) and Pi,t is the closing price (Compustat item PRCCQ) of firm i in quarter t |
∆LF i,t | \( \frac{{\frac{{SALES_{i,t - 4} }}{{Employees_{i,t - 4} }} - \frac{{SALES_{i,t} }}{{Employees_{i,t} }}}}{{\frac{{SALES_{i,t - 4} }}{{Employees_{i,t - 4} }}}} \) where SALESi,t is the sales (Compustat item SALEQ) of firm i in quarter t and Employeesi,t is the number of employees (Compustat item EMP) of firm i at the next fiscal year end after quarter t |
∆LEV2 i,t | \( \frac{{Long - term\;Debt_{i,t} }}{{Equity_{i,t} }} - \frac{{Long - term \;Debt_{i,t - 4} }}{{Equity_{i,t - 4} }} \) Difference between the leverage ratio (LEVi,t) of firm i in quarter t and the leverage ratio of firm i in quarter t − 4 |
ln∆AF i,t | Logarithm of the ratio of mean sales forecast (I/B/E/S item: SALI1MN) for quarter t + 1 to actual sales (I/B/E/S item: I0SAL) in quarter t |
ln∆OB i,t | Logarithm of ∆OBi,t. ∆OBi,t is order backlog (Compustat item OB) at the next fiscal year end after quarter t divided by order backlog at the previous fiscal year end before quarter t. We manually set the value for order backlog growth to one (0% growth) if data for the current and previous year were missing (and also if both are zero), thus assuming that data on order backlog is missing because a firm has no order backlog. |
MISSING∆OB i,t | Dummy variable that is one if the ∆OB is manually set to one |
FE i,t | \( \frac{{Actual EPS_{i,t} - analyst consensus forecast_{i,t} }}{{Price_{i,t - 1} }} \) Actual EPSi,t (I/B/E/S item: I0EPS) is actual earnings per share for firm i in quarter t. Analyst consensus forecasti,t (I/B/E/S item: EPSI1MN) is analysts’ consensus forecast for firm i in quarter t. Pricei,t is the stock price of firm i at the end of quarter t (Compustat item: PRCCQ) |
Rights and permissions
About this article
Cite this article
Silge, L., Wöhrmann, A. Market reaction to asymmetric cost behavior: the impact of long-term growth expectations. Rev Manag Sci 15, 309–347 (2021). https://doi.org/10.1007/s11846-019-00341-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11846-019-00341-8