Erasmus University Rotterdam Erasmus School of Economics Master Accounting, Auditing and Control Master's Thesis Accounting, Auditing & Control Successful-Efforts

5.3.1 Regression models

The first hypothesis is as follows:

where:

ASS_it: total assets for firm i in year t, deflated by beginning economic value.

∆ASS: The change in total assets for firm i in year t-t_-1, deflated by beginning economic value.

: is a disturbance term.
For the testing of the firms that use the successful-efforts method some variables for the capitalized development expenditures are incorporated. Those capitalized development expenditures are divided in past capitalized development expenditures; the capitalized part of the development expenditures in test year; and the third part is the amortized expenditures in that year. Beside that the variable total assets is included, which is the total book value excluding the capitalized development expenditures at year end. The following cross-sectional contemporaneous regression model is constructed for the cash-expense method:

where:

ASSBCD_it: the total assets before the capitalized development expenditures at year end for firm i in year t, deflated by beginning economic value.

DCAPold_it: the development expenditures that are capitalized in previous years for firm i in year t, deflated by beginning economic value.

DCAPnew_it: the development expenditures that are capitalized for firm in year t, deflated by beginning economic value.

RDAM_it: past capitalized R&D expenditures that are written-off or impaired for firm i in year t, deflated by beginning economic value. (amortization = write-off + impairment)

∆ASSBCD: The change in total assets before the capitalized development expenditures at year end for firm i in year t-t_-1, deflated by beginning economic value.

∆DCAPold_it: The change in the development expenditures that are capitalized in previous years for firm i in year t-t_-1, deflated by beginning economic value.

∆DCAPnew_it: The change in the development expenditures that are capitalized for firm i in year t-t_-1, deflated by beginning economic value.

∆RDAM_it: The change in the past capitalized development expenditures that are written-off or impaired for firm i in year t-t_-1, deflated by beginning economic value. (amortization = write-off + impairment)

: is a disturbance term.
After determining the regression models the tests will be performed. The tests will be done as one group. The regression analysis will be done with scaled and absolute variables. The third step is to analyse the results from the test. This first analysis contains analyzing the direction coefficients for the β. The β shows how the returns react on a change of a certain dependent variable. The second analysis is to assess the P-values. The P-values show if a variable has a significant influence on the returns. The third analysis is for the adjusted R-squared to provide evidence for explanatory power of the models.
The second set of models is for hypothesis two:
H2: The earnings disclosed by recognizing R&D expenditures using the successful-efforts method is more value relevant than the cash-expense method, in the automotives industry.
This hypothesis is an earnings-returns relation and specified for the R&D expenditures disclosed in the profit and loss account. The methodology will be described per step. The first step is to develop a cross-sectional multiple regression model. The model used is from Healy et al. (2002). They described the cash-expense method and the successful-efforts method. The reason to use the models of Healy et al. (2002) is that this was a research that highly specified the parts in the financial statement. The first model is for the cash-expense method. This model is less sophisticated because the only variable in the model are the R&D expenditures in the reported year.
The cash-expense model that will be used is as follows:
RET_it =β_0t+β_1tNIBRD_it+β_2tRDEXP_Cit+β_3t∆NIBRD_it+β_4t∆RDEXP_Cit+ε_it

where:

NIBRD_it = net income before R&D expense for firm i in year t, deflated by beginning economic value.

RDEXP_it = the R&D expense for firm i in year t, deflated by beginning economic value.

∆NIBRD_it = the deflated change in net income before R&D.

∆RDEXP_it = the deflated change in R&D expense.

: is a disturbance term.
The second model that needs to be tested is the successful-efforts model. This model is more detailed. Like the cash-expense model the dependent variable is the return. The first independent variable is the net income before R&D expense. The first specified variable for R&D is the R&D expenditures that are directly expensed in the year that the R&D expenditures incurred. This is the total amount spend on R&D and the part that is capitalized deducted. The second variable specific for R&D is the amortization. The amortization contains the write down of the capitalized development expenditures and the impairment losses. In the model of Healy et al. (2002) the last variable is called the write down, for this research the word amortization is used. The amortization does not only incorporate the write down but also the impairments.
The successful-efforts model that will be used is as follows:

where:

NIBRDC_it: net income before R&D expenses and amortization for firm i in year t, deflated by beginning economic value.

RDEXPC_it: the part of the R&D expenditures that are expensed by firm i in year t, deflated by beginning economic value.

RDAM_it: past capitalized R&D expenditures that are written-off or impaired for firm i in year t, deflated by beginning economic value. (amortization = write-off + impairment)

∆NIBRDC: The change of net income before R&D expenses and amortization for firm i in year t-t_-1, deflated by beginning economic value.

∆RDEXPC_it: The change in the part of the R&D expenditures that are expensed by firm i in year t-t_-1, deflated by beginning economic value.

∆RDAM_it: The change in the past capitalized development expenditures that are written-off or impaired for firm i in year t-t_-1, deflated by beginning economic value. (amortization = write-off + impairment)

: is a disturbance term.
After determining the models the second step is to perform the tests. The regression analysis will be done with scaled and absolute variables. The second step is to perform the regression analysis. The third step is to analyse the results from the test. This first analysis contains analyzing the direction coefficients for the β. The β shows how the returns react on a change of a certain dependent variable. The second analysis is to assess the P-values. The P-values show if a variable has a significant influence on the returns. The third analysis is for the adjusted R-squared to provide evidence for explanatory power of the models.
After the regression analyses there are for each year eight adjusted R-squares, four for the cash-expense model and four for the successful-efforts model. The fourth step is to test if the second group has a significant higher R² (more value relevant) than the first group. According to Moore et al. (2003, 443) to test this hypothesis it is necessary to use the dependent t-test, if assumptions have been proved, otherwise we use the non-parametrical Mann-Whitney t-test (t-test for matched pairs). The following hypothesis will be tested:

Where:

5.3.2 Variables

In this section the variables will be further discussed. In the previous section the regression models were formed. This section provides the specific data items that are used for the variables per model.

The first model is the cash-expense model for the first hypothesis that focuses on earnings. The description per variable will be provided with the specific calculation:
RET_it = the economic return for firm i in year t, computed as the change in economic value plus the cash dividend for year i, deflated by beginning economic value.

NIBRD_it = net income before R&D expense for firm i in year t, deflated by beginning economic value.

RDEXP_it = the R&D expense for firm i in year t, deflated by beginning economic value.

∆NIBRD_it = the deflated change in net income before R&D.

∆RDEXP_it = the deflated change in R&D expense.

The next model is also for the first hypothesis. This model incorporates all the variables for the successful-efforts. The returns are calculated in the same manner, so they will not be explained again.

NIBRDC_it: net income before R&D expenses and amortization for firm i in year t, deflated by beginning economic value.

RDEXPC_it: the part of the R&D expenditures that are expensed by firm i in year t, deflated by beginning economic value.

RDAM_it: past capitalized R&D expenditures that are written-off or impaired for firm i in year t, deflated by beginning economic value. (amortization = write-off + impairment)

∆NIBRDC: The change of net income before R&D expenses and amortization for firm i in year t-t_-1, deflated by beginning economic value.

∆RDEXPC_it: The change in the part of the R&D expenditures that are expensed by firm i in year t-t_-1, deflated by beginning economic value.

∆RDAM_it: The change in the past capitalized development expenditures that are written-off or impaired for firm i in year t-t_-1, deflated by beginning economic value. (amortization = write-off + impairment)

The third model that is described in the previous section is the cash-expense regression model for the book values. The variable return is already discussed. The variables are calculated by the following data items:
ASS_it: total assets for firm i in year t, deflated by beginning economic value.

∆ASS: The change in total assets for firm i in year t-t_-1, deflated by beginning economic value.

The fourth model that is described in the previous section is the successful-efforts regression model for the book values. The variables return and RDAM and ∆RDAM are already discussed. The variables are calculated by the following data items:
ASSBCD_it: the total assets before the capitalized development expenditures at year end for firm i in year t, deflated by beginning economic value.

DCAPold_it: the development expenditures that are capitalized in previous years for firm i in year t, deflated by beginning economic value.

DCAPnew_it: the development expenditures that are capitalized for firm in year t, deflated by beginning economic value.

∆ASSBCD: The change in total assets before the capitalized development expenditures at year end for firm i in year t-t_-1, deflated by beginning economic value.

∆DCAPold_it: The change in the development expenditures that are capitalized in previous years for firm i in year t-t_-1, deflated by beginning economic value.

∆DCAPnew_it: The change in the development expenditures that are capitalized for firm i in year t-t_-1, deflated by beginning economic value.

5.3.3 Control variables

The models described in the first section of this paragraph have to be controlled with control variables if the relation before controlling will be the same as after controlling. This section describes which variables will be used. First the variables that are marked as perhaps usable in chapter three and four will be discussed, if they are applicable. Afterwards, the remaining variables that can be used will be discussed. Besides the discussion the calculation of the variables is provided.

In chapter three “Institutional setting” the macro economic variables and the control variables from Ali and Hwang (2007) are presented as not usable. Those variables aren’t usable, because the firms in the automotives industry are multinational firms. The economic events have impact on all firms, so controlling for these events isn’t necessary. The five factors from Ali and Hwang (2007) are: market-orientated or bank-orientated, government influence on standard setting, continental versus British-American model (financing), influence of TAX on standard setting, and spending on audits. Those factors are able to control for regional differences. The spending on audits isn’t relevant, because all firms are audited by one of the big four accounting firms. The factor continental model versus British-American model and the factor bank- versus market-orientated aren’t applicable too, because the firms are at this point in time heavily involved in both bank loans and capital from exchange markets. Those large automotive firms are not typically market orientated or bank orientated.
The variable government involvement is based on two law systems. The first system is the common law. Those rules come from court. For each element there is a rule. This done by the FASB. The second system is the code law were there are more guidelines instead of rules. This is principle based. Those principles are codified in law. This distinction is good to take into account. The systems common law and code law will be used as control variables.
The last characteristic discussed in chapter four is the R&D intensity. According to Chan et al. (1990) and Chan et al. (2001) firms with R&D intensity above industry average receive a higher share price. The variable R&D intensity will be used as a control variable. The R&D intensity is calculated as follows:


The control models that will be used are the distinction between code law/common law and the R&D intensity. In the absolute models the variables R&D intensity is replaced by the absolute value Revenue.

4. 
   
   Directory: pub
   pub -> Project Scoring Template Pre-Construction Phase New Core/Branch Campus Projects
   pub -> Acm word Template for sig site
   pub -> The german unification, 1815-1870
   pub -> Baltic Olympiads in Informatics: Challenges for Training Together
   pub ->  Preparation of Papers for ieee transactions on medical imaging
   pub -> Forthcoming meetings and conferences
   pub -> Asilomar Conference on Circuits, Systems & Computers, Pacific Grove, ca, November 1978, pp. 55 58
   pub -> Forthcoming meetings and conferences
   pub -> Harmonised compatibility and sharing conditions for video pmse in the 7 9 ghz frequency band, taking into account radar use
   
   
   
   Download 0.83 Mb.
   
   Share with your friends: