6 EMPIRICAL RESULTS
6.2 Determinants of corridor method users and compliance
6.2.1 Determinants of using the corridor method
Before the IAS 19R, there were three methods for recognizing actuarial gains and losses (IAS 19R: para. BC 66): 1) the equity method (i.e. to recognize actuarial gains and losses in other comprehensive income and transferred to retained earnings);
2) immediately recognize actuarial gains and losses in profit or losses (which is also the rule to recognize the actuarial gains and losses under IAS 19R); and 3) the corridor method. The corridor method allows the “within corridor”39 actuarial gains and losses not to be recognized in other comprehensive income or profit or
39 The corridor is the greater of 10% of the defined benefit obligation and 10% of the fair value of any plan assets.
losses. Hence, firms with certain characteristics have an incentive to use the corridor method. Moreover, based on chapter 4, the expected determinants of using the corridor method are leverage and ownership concentration. This section will employ 200 observations before the adopting of IAS 19R and the model 3 (in section 5.4.1) to examine the determinants of using the corridor method.
6.2.1.1 Univariate results
Table 23 shows the distribution of corridor method users among different countries. It can be seen from the table that except for Swedish firms, firms in the sample (i.e. German firms, French firms and Italian firms) prefer to be non-corridor method users. Thus, country affects the choice of using the non-corridor method.
Table 23. The distribution of corridor method users among different countries
Non-CM users CM users Total
Germany 27 54% 23 46% 50 100%
France 39 78% 11 22% 50 100%
Italy 27 54% 23 46% 50 100%
Sweden 16 32% 34 68% 50 100%
Total 109 54.5% 91 45.5% 200 100%
Pearson chi2(3) = 42.7059 Pr = 0.00
Note: Chi-square is used to test whether there is an association between different countries and corridor method users. ***Significant at the 0.01 level (2-tailed),
**Significant at the 0.05 level (2-tailed), *Significant at 0.1 level (2-tailed).
Table 24 shows that except for firms in the field of Machinery and Other Primary and Secondary industries, firms from other fields are less likely to use the corridor method. Moreover, the non-corridor method users account for 72.73% in the field of other service. Thus, industry affects firms’ choices to be corridor method users.
Table 24. The distribution of corridor method users among different industries
Non-CM users CM users Total
Chemical 15 60% 10 40% 25 100%
Machinery 30 46.15% 35 53.85% 65 100%
Other Tertiary industry 23 54.76% 19 45.24% 42 100%
Other service 24 72.73% 9 27.27% 33 100%
Other primary &
secondary industries 17 48.57% 18 51.43% 35 100%
Total 109 54.5% 91 45.5% 200 100%
Pearson chi2(4) = 14.0988 Pr = 0.007
Note: Chi-square is used to examine whether there is a relationship between industry and corridor method users. ***Significant at the 0.01 level (2-tailed), **Significant at the 0.05 level (2-tailed), *Significant at 0.1 level (2-tailed).
Even though few articles investigate the effects of culture on the use of the corridor method, the country effects of employing the corridor method have been identified (e.g. Fasshauer et al. 2008). Thus, it is expected that there will be an association between culture and the use of the corridor method.
Table 25 presents the relationship between culture and the use of the corridor method. More specifically, Table 25 shows that the mean values of UA, PD and secrecy for non-corridor method users are significantly different from those of corridor method users.
Table 25. The descriptive statistics of cultural effects on the use of the corridor method
Variables Non-CM Mean CM Mean Mean Diff
UA 109 69.706 91 56.615 13.091***
PD 109 49.936 91 41.286 8.650***
IND 109 71.248 91 71.253 -0.005
Secrecy 109 48.394 91 26.648 21.746***
Note: The observation includes 200 companies with 50 firms each from Sweden, Italy, Germany and France. CM means the corridor method users; Non-CM stands for the non-corridor method users; UA examines the uncertainty avoidance of culture; PD examines the power distance of culture; IND investigates the individual level of culture; and Secrecy examines the secrecy level of culture.
Table 26 shows the descriptive statistics for test variables of Hypotheses 6 and 7:
leverage (i.e. in Panel A) and ownership concentration (i.e. in Panel B).
In order to analyze the effects of leverage/ownership concentration on corridor method users, the observations have been divided into firms with lower leverage/ownership concentration (i.e. firms’ leverage/ownership concentration that is lower than the median of all firms’ leverage/ownership concentration) and firms with higher leverage/ownership concentration (i.e. firms’
leverage/ownership concentration that is equal to or higher than the median of all firms’ leverage/ownership concentration).
Moreover, the mean value of firms with lower leverage/ownership concentration has been compared with the mean value of firms that have higher leverage/ownership concentration, and the results show that there is no significant difference between the mean value of firms with lower leverage/ownership concentration and the mean value of firms with higher leverage/ownership concentration.
Thus, the results in Table 26 reject Hypothesis 6 as well as Hypothesis 7 and reveal that neither the leverage nor the ownership concentration affects the use of corridor method.
Table 26. Descriptive statistics Panel A. Comparison for leverage
Variables Firms with lower
LEV Mean Firms with higher
LEV Mean Mean Diff
CM 100 0.463 100 0.448 0.016
Panel B. Comparison for ownership concentration Variables Firms with higher
OC Mean Firms with lower
OC Mean Mean Diff
CM 100 0.47 100 0.44 0.03
Note: The observation includes 200 companies with 50 firms each from Sweden, Italy, Germany and France. The LEV means the leverage, which is measured as the total debts divided by total assets and has been winsorized at 1%; the OC means the ownership concentration; the CM means the corridor method user, it equals 1 if a firm is a corridor method user, otherwise, it equals 0; the firms with lower LEV / OC are the firms have the under median value LEV / OC, otherwise they are considered to be firms with higher LEV / OC.
6.2.1.2 Regression results
Table 27 shows the logit regression results regarding the determinants of employing the corridor method.
Hypothesis 6 predicts a positive relationship between leverage and the use of the corridor method. However, the results show that the coefficient of LEV is not significant. Thus, the leverage does not have significant effects on the use of the corridor method. Hence, the results do not support Hypothesis 6. This result is in contrast to the findings of Morais (2010), who finds firms with higher leverage tend to use the corridor method to mitigate the effects of debts. This could be due to the different research periods with different samples.
Hypothesis 7 suggests a negative effect of ownership concentration on the use of the corridor method. Nevertheless, the coefficient of both ownership concentration (OC) and large ownership (Large) is not significant, thus rejecting Hypothesis 7. It is believed that firms with less ownership concentration tend to use an income-increasing accounting method (Astami and Tower: 2006), and the use of the corridor method may lead to unrecognized actuarial gains and losses which affect profitability. However, the magnitude of the effects of unrecognized actuarial gains and losses on profitability decides the use of the corridor method.
Thus, a possible reason for the lack of relationship between ownership concentration and the use of the corridor method could be that the sample included in this dissertation is 200 firms with different firm size, which leads to different amount of unrecognized actuarial gains and losses.
Furthermore, it can be seen from Table 27 that the coefficient of AGL_R is significant and positive, which suggests that firms with more actuarial gains and losses (i.e. the difference between actuarial gains and actuarial losses) are more likely to use the corridor method. In addition, compared with German firms, French firms are less likely to use the corridor method.
Table 27. Determinants of using the corridor method m1
VARIABLES CM
LEV -1.723
(-1.122)
OC -0.00818
(-0.613)
Large 0.401
(0.611)
AGL_ R 31.54**
(2.452)
ROE 0.00408
(0.542)
PM 0.0185
(1.075)
TA_ 0.0452
(0.482)
O. Germany -
France -1.100**
(-2.150)
Italy 0.00767
(0.0159)
Sweden 0.832
(1.595)
O. Chemical -
Other service -0.649
(-0.983)
Machinery 0.503
(0.858)
Other_Pssector 0.549
(0.844)
Other_Tertiary sectors 0.340
(0.555)
Constant -0.337
(-0.359)
Observations 187
Pseudo R2 0.1487
Robust z-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
This table shows the regression results examining the determinants of using the corridor method. The 187 observations are obtained from the annual reports in the year before the 200 firms first adopted the IAS 19R. The CM examines whether a firm is a corridor method user, it equals 1 if the firm is a corridor method user, otherwise it is not; the LEV means the leverage which is calculated as the total debts divided by total assets and has been winsorized at 1%; the OC means the ownership concentration; Large means the large ownership, it equals 1 if an entity’s ownership concentration is more than average of all entities’ ownership concentration, it equals 0 if otherwise; AGL_R means the difference between actuarial gains and actuarial losses divided by the total assets, and its currency is million Euros; ROE means the return of shareholders’ earnings which has been winsorized at 1%; PM means the profit margin which has been winsorized at 1%; TA_
means the logarithm of total assets, and its currency is million Euros, which has been winsorized at 1%; Other service examines whether the firm is in the field of other service;
Machinery examines whether the firm is in the field of Machinery; Other_Pssector examines whether the firm is in the field of Other industries in Primary and Secondary sectors; Other_Tertiary sectors examines whether the firm is in the field of Other industries in Tertiary sectors.
Table 28 presents the robustness check for determinants of using the corridor method, which uses operating revenue to replace profit margin, employing ROEN instead of ROE. Moreover, the logarithm of total assets has been replaced by the logarithm of number of employees.
It can be seen from Table 28 that neither the ownership concentration nor the leverage has an effect on the use of the corridor method. Thus, Hypothesis 6 and Hypothesis 7 are rejected. Moreover, the robustness check suggests significant and positive effects of AGL_R on the use of the corridor method. Furthermore, French firms are still less likely to use the corridor method. However, the Swedish firms begin to show positive effects on the use of the corridor method compared to German firms. In addition, firm size and operating revenue begin to show positive effects on the use of the corridor method in the robustness check.
Table 28. Robustness check for determinants of using the corridor method m1
VARIABLES CM
LEV -0.132
(-0.0832)
OC -0.00436
(-0.350)
Large 0.372
(0.559)
AGL_R 27.19**
(2.381)
ROEN -0.00116
(-0.735)
OR_A 1.108***
(2.885)
NE_ 0.177*
(1.686)
O. Germany -
France -1.059*
(-1.954)
Italy 0.300
(0.615)
Sweden 1.129**
(2.128)
O. chemical -
Other service -1.040
(-1.603)
Machinery 0.155
(0.275)
Other_Pssector 0.0812
(0.131)
Other_Tertiary sectors -0.227
(-0.392)
Constant -2.504**
(-2.000)
Observations 190
Pseudo R2 0.1847
Robust z-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
This table shows the regression results of the robust check examining the determinants of using the corridor method. The ROE has been replaced by ROEN, the PM has been replaced by the OR_A, the TA_ has been replaced by NE_. The CM examines whether a firm is a corridor method user, it equals 1 if the firm is a corridor method user, otherwise it is not; the LEV means the leverage which is calculated as the total debts divided by total assets and has been winsorized at 1%; the OC means the ownership concentration; Large means the large ownership, it equals 1 if an entity’s ownership concentration is more than average of all entities’ ownership concentration, it equals 0 if otherwise; the AGL_R means the difference between actuarial gains and actuarial losses divided by the total assets, and its currency is million Euros; the ROEN means the ROE using net income and it has been winsorized at 1%; the OR_A means the operating revenue which has been divided by the total assets, and its currency is million Euros and it has been winsorized at 1%; the NE_ means the logarithm of number of employees; Chemical examines whether the firm is in the field of Chemistry; Other service examines whether the firm is in the field of other service; the Machinery examines whether the firm is in the field of Machinery; Other_Pssector examines whether the firm is in the field of Other industries in Primary and Secondary sectors; Other_Tertiary sectors examines whether the firm is in the field of Other industries in Tertiary sectors.
6.2.2 Determinants of compliance disclosure about the corridor