The moderating effect of a CEO’s financial expertise on the relationship between managerial entrenchment and asymmetric cost behavior

Bubeck, Stephan Klaus; Lunardi, Micheli Aparecida

doi:10.15728/bbr.2024.1858.en

Tabla de contenido Anterior Actual Siguiente

Tabla de contenido

ARTICLE • BBR, Braz. Bus. Rev. 23 • 2026 • https://doi.org/10.15728/bbr.2024.1858.en linkcopy

The moderating effect of a CEO’s financial expertise on the relationship between managerial entrenchment and asymmetric cost behavior

Authorship SCIMAGO INSTITUTIONS RANKINGS

This study analyzed the moderating effect of a CEO's financial expertise on the relationship between managerial entrenchment and the asymmetric cost behavior of Brazilian public companies, analyzing 178 companies in the period between 2012 and2021. Descriptive statistics, a correlation matrix, and regressions using the GMM technique were used. The results showed that entrenched managers with financial expertise increase COGS and TC more quickly in periods of sales growth and reduce them more slowly when sales decline. From the perspective of Agency Theory, these results suggest signs of empire building, which could be harmful to the company in the long term. This study contributes by bringing a new perspective on the risks related to managers' characteristics, suggesting that financial expertise can be used as a mechanism for entrenched managers to strengthen their control. In this way, shareholders can demand mechanisms to promote greater vigilance over the influence of managers on the company's cost structure.

Keywords:
Asymmetric Cost Behavior; Managerial Entrenchment; CEO’s Financial Expertise; Agency Theory

Panel A - Composition of the companies in the sample
Companies	AF	RF
(+) Companies listed on B3	480	100,0 %
(-) Companies in the financial sector	79	16,5 %
(=) Subtotal	401	83,5 %
(-) Companies with periods of no movement for NSR, COGS or SG&A	191	39,8 %
(-) Companies with periods with negative values for NSR, COGS or SG&A	4	0,8 %
(-) Companies with periods with variation above 100 % in NSR	28	5,8 %
(=) Total	178	37,1 %
Panel B - Composition of companies by sector
Sectors	AF	RF
Health care	5	2,8 %
Real estate	17	9,6 %
Cyclical consumption	39	21,9 %
Non-cyclical consumption	20	11,2 %
Energy	4	2,2 %
Industry	29	16,3 %
Basic materials	24	13,5 %
Educational and academic services	3	1,7 %
Utilities	31	17,4 %
Technology	6	3,4 %
(=) Total	178	100,0 %

Variable	Description	ES	Author(s)
EXSH	Variable calculated by the number of shares held by the members of the executive board.	+	Adapted from: Hu & Kumar (2004); Florackis & Ozkan (2009); Lin et al. (2014)
BIND	Variable calculated by the number of independent members of the board of directors in relation to the total numbers of board members.	-	Hu & Kumar (2004); Florackis & Ozkan (2009); Salehi et al. (2018, 2020, 2021)
BCOM	Variable based on the natural logarithm of the board of directors' compensation.	-	Salehi et al. (2018, 2020, 2021)
DCEO	Dichotomous variable, where 1 if the CEO is also a member of the board of directors, and 0 (zero), if not.	+	Hu & Kumar (2004); Florackis & Ozkan (2009); Lin et al. (2014); Salehi et al. (2018, 2020, 2021)
ECEO	Variable calculated by the number of years the CEO held the position in the company continuously.	+	Adapted from: Salehi et al. (2018, 2020, 2021)

Dimension	Variable	Item measurement
Academic	BF	Dichotomous variable, where 1 is considered if the CEO has academic background (undergraduate and/or graduate) related to finance (courses in Administration, Accounting, Economics and Finance), and 0 (zero), otherwise.
Professional	EFS	Dichotomous variable, where 1 is considered if the CEO has professional experience in companies in the financial sector, and 0 (zero), otherwise.
	EFD	Dichotomous variable, where 1 is considered if the CEO has professional experience as a financial director, and 0 (zero), otherwise.
	EC	Continuous variable, which refers to the number of years the CEO has experience in this position.

Panel A: Dependent variables
Variables	Mean	SD	Minimum	Maximum	P25	P50	P75
logCOGS	0.0263	0.1271	-1.4925	1.4122	-0.0094	0.0314	0.0737
logSG&A	0.0212	0.1163	-1.3039	1.3636	-0.0172	0.0261	0.0673
logTC	0.0261	0.1060	-1.3025	1.0893	-0.0074	0.0307	0.0696
Panel B: Independent variables
Variables	Mean	SD	Minimum	Maximum	P25	P50	P75
logNSR	0.0249	0.1136	-1.1804	0.2991	-0.0082	0.0338	0.0721
EN	0.0000	1.2976	-2.4871	8.4539	-0.8225	-0.2161	0.5056
FE	0.0000	1.1794	-1.9220	3.0122	-0.8070	-0.0217	0.2628
Panel C: Control variables
Variables	Mean	SD	Minimum	Maximum	P25	P50	P75
TAM	21.8413	1.8656	16.5333	27.6184	20.5469	21.8990	23.0625
IDA	47.3596	28.4359	2.0000	149.0000	21.0000	48.0000	67.0000
ALA	1.9854	26.6463	-774.1732	412.7411	0.6327	1.2580	2.4199
Panel D: P&L accounts
Accounts	Total (R$ Million)		Mean (R$ Million)		% TC		% NSR
COGS	11,636,104.49		6,537.14		86.53 %		70.30 %
SG&A	1,811,776.64		1,017.85		13.47 %		10.95 %
TC	13,447,881.13		7,554.99		100.00 %		81.25 %
NSR	16,551,637.65		9,298.67		-		100.00 %
Observations: 1,780

Variables	logCOGS	logSG&A	logTC	logNSR	EN	FE	SIZ	AGE	LEV
logCOGS	1	0.42***	0.94***	0.84***	-0.04*	0.03	0.14***	-0.05**	0.07***
logSG&A	0.35***	1	0.60***	0.50***	-0.04*	0.03	0.13***	-0.07***	0.06***
logTC	0.95***	0.52***	1	0.86***	-0.05**	0.03	0.17***	-0.06**	0.08***
logNSR	0.79***	0.44***	0.83***	1	-0.05**	0.04*	0.14***	-0.03	0.08***
EN	-0.03	-0.03	-0.03	-0.03	1	-0.30***	-0.37***	0.13***	-0.14***
FE	0.00	0.02	0.02	0.00	-0.31***	1	0.14***	-0.15***	0.10***
SIZ	0.11***	0.07***	0.12***	0.12***	-0.38***	0.14***	1	-0.23***	0.44***
AGE	-0.04	-0.06***	-0.05*	-0.03	0.17***	-0.16***	-0.20***	1	-0.09***
LEV	0.00	0.01	0.00	-0.01	0.00	0.00	0.03	-0.01	1

Variables	Dependent = logCOGS		Dependent = logSG&A		Dependent = logTC
Variables	Coefficient	z	Coefficient	z	Coefficient	z
Constant	0.0009	0.22	0.0137***	2.75	0.0027	0.86
logNSR	0.9213***	17.06	0.3556***	5.23	0.8120***	19.81
d-logNSR	-1.1608**	-2.44	-1.9291***	-3.20	-1.5798***	-4.36
d-EN	0.0789**	2.08	0.0403	0.83	0.0527*	1.81
d-SIZ	0.0519**	2.33	0.0840***	2.98	0.0657***	3.87
d-AGE	-0.0007	-0.45	0.0037*	1.83	0.0019	1.53
d-LEV	-0.0031**	-2.28	-0.0004	-0.24	-0.0016	-1.57
Wald chi2	1,042.47		184.74		1,311.72
Prob > chi2	0.0000		0.0000		0.0000

Dependent variables	1 % Variation on NSR			Asymmetric Behavior	Conclusion Hypothesis 1
	Increase	Decrease
	logNSR (i)	d-EN (ii)	Total (i + ii)
logCOGS	0.9213***	0.0789**	1.0002**	Anti-sticky	Accepted
logSG&A	0.3556***	0.0403	0.3959	Not significant	Rejected
logTC	0.8120***	0.0527*	0.8647*	Anti-sticky	Accepted

Variables	Dependent = logCOGS		Dependent = logSG&A		Dependent = logTC
Variables	Coefficient	z	Coefficient	z	Coefficient	z
Constant	0.0011	0.28	0,0140***	2.79	0.0032	1.01
logNSR	0.9200***	17.00	0,3538***	5.21	0.8078***	19.69
d-logNSR	-1.0768**	-2.25	-1,8858***	-3.11	-1.5013***	-4.12
d-(EN*FE)	-0,0602**	-2.14	-0,0284	-0.79	-0.0515**	-2.42
d-SIZ	0.0464**	2.07	0,0812***	2.86	0.0610***	3.57
d-AGE	-0.0005	-0.33	0,0039*	1.91	0.0019	1.62
d-LEV	-0.0028**	-2.01	-0,0003	-0.15	-0.0013	-1.28
Wald chi2	1,040.08		184.81		1,314.77
Prob > chi2	0.0000		0.0000		0.0000

Dependent variables	1 % Variation in NSR			Asymmetric Behavior	Conclusion Hypothesis 2
	Increase	Decrease
	logNSR (i)	d-(EN*FE) (ii)	Total (i + ii)
logCOGS	0,9200***	-0,0602**	0,8598***	Sticky	Accepted
logSG&A	0,3538***	-0,0284	0,3254	Not significant	Rejected
logTC	0,8078***	-0,0515**	0,7563**	Sticky	Accepted

Panel A: Continuous variables
Variables	Mean	SD	Minimum	Maximum	P25	P50	P75
EXSH	1.0825	5.2870	0.0000	0.6241	0	0	0.0859
BIND	22.9779	24.2202	0.0000	100.0000	0	19.5408	40.0000
BCOM	13.2449	3.2920	0.0000	18.0890	12.9152	13.9523	14.7876
ECEO	7.4472	9.3149	0	66	2	4	9
Observations: 1,780
Panel B: Dichotomous variables
Variables	Category		Observations			Frequency ( %)
DCEO	1		836			46.97 %
DCEO	0		944			53.03 %
Total		1,780		100.00 %

Variables	EXSH	BIND	BCOM	DCEO	ECEO
EXSH	1	0.18***	0.23***	-0.07***	0.15***
BIND	-0.08***	1	0.40***	-0.08***	-0.12***
BCOM	-0.18***	0.30***	1	-0.28***	-0.10***
DCEO	0.10***	-0.08***	-0.15***	1	0.26***
ECEO	0.25***	-0.16***	-0.10***	0.30***	1

Bartlett Sphericity Test
Chi-square	577,798		p-value	0,000
Kaiser-Meyer-Olkin (KMO)	0,5681
Factor analysis for principal component factor - Number of extracted components (eigenvalues)
Factor	Eigenvalue	Difference	Proportion	Cumulative
Factor 1	1,68365	0,623633	0,3367	0,3367
Factor 2	1,06002	0,152459	0,2120	0,5487
Factor 3	0,907562	0,139252	0,1815	0,7302
Factor 4	0,768309	0,187855	0,1537	0,8839
Factor 5	0,580455	-	0,1161	1,0000

Panel A: Continuous variable - transformed into nominal
Variables	Mean	SD	Minimum	Maximum	P25	P50	P75
EC	1.3753	0.7142	0	3	1	1	2
Observations: 1,780
Panel B: Dichotomous variables
Variables	Category		Observations				Frequency ( %)
BF	1		1,218				68.43 %
BF	0		562				31.57 %
Total			1,780				100.00 %
EFS	1		273				15.34 %
EFS	0		1,507				84.66 %
Total			1,780				100.00 %
EFD	1		344				19.33 %
EFD	0		1,436				80.67 %
Total			1,780				100.00 %

Variables	BF	EFS	EFD	EC
BF	1	0.18***	0.13***	-0.05**
EFS	0.18***	1	0.22***	-0.05**
EFD	0.13***	0.22***	1	-0.04*
EC	-0.08***	-0.06**	-0.07***	1

Bartlett Sphericity Test
Chi-square	181,651		p-value	0,000
Kaiser-Meyer-Olkin (KMO)	0,5852
Factor analysis for principal component factor - Number of extracted components (eigenvalues)
Factor	Eigenvalue	Difference	Proportion	Cumulative
Factor 1	1,3911	0,4206	0,3478	0,3478
Factor 2	0,9705	0,1026	0,2426	0,5904
Factor 3	0,8679	0,0974	0,2170	0,8074
Factor 4	0,7705	-	0,1926	1,0000

location_on

Fucape Business School Av. Fernando Ferrari, 1358, Boa Vista, 29075-505, Vitória, Espírito Santo, Brasil, (27) 4009-4423 - Vitória - ES - Brazil
E-mail: bbronline@bbronline.com.br

rss_feed Acompañe los números de esta revista en su lector de RSS

[TFN1111] Legend: AF: Absolute Frequency; RF: Relative Frequency; B3: Brasil, Bolsa, Balcão; NSR: Net Sales Revenue; COGS: Cost of Goods Sold; SG&A: Selling, General and Administrative Expenses.

[TFN2222] Source: Research data (2023).