The Joint Information Content of General Purchasing Power and Current Cost Accounting
By
Ahmad Bello (PhD)*
Abstract:
The conventional accounting method stresses on using historical cost accounting as prime valuation model due to its reproducibility and dependability. This model has evolved over the years and proved to be reasonably satisfactory during period of relatively stable prices. In the absence of price stability, however, accounting reports can become extremely unsatisfactory and misleading. The paper looks into the value relevance of joint reporting of alternative inflation model with traditional historical model. Using price model, the results from the estimated two models vividly show more information content on joint report to single reporting regime. This suggests that joint reporting regime yields additional value relevant to users.
1.0 INTRODUCTION
Valuation and measurement are central theme to accounting. One of the long debated issues on the accounting valuation and measurement is whether the accounting information will really be effective in period of inflation, taking cognizance of historical cost principle (HCA) as the dominant reporting model. According to Alagiah (2009), HCA is a valuation system that has provided a framework for explaining “invested costs” where costs of all resources committed to an entity are accounted for and locked into a double entry framework with a record of the sources. Another issue of concern is on the business objective to remain as going concern, with this objective, both capital and operating capability of the business has to be maintained and improved. With accounting valuation using historical cost principle it’s questionable, if not impossible that a concern can maintain its capital as well as productive capability in period of inflation. Inflation has been and remains a measure concern of accountants and economics when it comes to financial reporting issue. Despite the recognition of dwindling effect inflation has on the quality of accounting information and the inability of the current valuation principle [Historical cost] to remedy the situation, yet a consensus is arrived on the effective method to be used in accounting for inflation. As reported by Bushman (2000), an issue that has exercise many accountants and economists over the years is whether these effect are of such nature and magnitude as to require the whole transformation of traditional accounting data before reliable inference can be drawn.
Though, the traditional accounting methods as they evolved over the years have proved to be reasonably satisfactory during period of relatively stable prices. In the absence of price stability, however, accounting reports can become extremely unsatisfactory and misleading.
The resistance to change by many standard accounting setters to the conventional inflation accounting methods is not unassociated to the hope inflation would disappear or reduce to an insignificant level. These hope seem unlikely to be fulfilled in the near future, however, as the worldwide inflation seems to be worse especially in developing countries, the need to empirically assess the effect of price level adjustments on the relevant quality of financial reports becomes inevitable.
Based on the above backdrop, the study aims in evaluating the empirical strength of joint information reporting of GPP and HCA as proposed in some countries, U.S. & UK. The rest of the paper is organized as thus: section two provides theoretical literature survey, section three explains the methodological approach used in the study; section four discusses the empirical results and section five closes with conclusion.
2.0. Theories and Literature Evidence
Debates on inflation accounting have subsided but not dead. As demonstrated by Salvary,(2004) debates on this issue is as perennial as grass; as soon as there are continues and significance increases in the level of price the debate will resume with much vigor. One important consideration of the debate is the effect of price level changes on the financial statements (Bega and Aharoni, 1977).
Thus, in periods of rising prices, financial accounting information which is based on HCA policy is been criticized on the grounds that it reflects a number of outdated values while the value is changing. The implication of this as pointed by Hoghes, et al (2004:73), is that inflation creates earning illusion as an artifact of the mismatching of expenses based on allocations of historical cost with current revenue in determining earnings. This mismatching distorts mapping of aggregates earnings and book values into equity earning value such that value relevant information is lost. Other studies offering support to such criticisms are: Mosich and larsen, (1982), Chambers, (1975), Neihens, (1978) as well as Dandago, (2001).
Recent studies conducted on the value relevance of inflation adjustments, suggests that inflation accounting provides incremental information beyond that of historical accounting (Gordon, 2001). Friday (2003) reports that accounting data adjusted for changing prices explain significant variation of cross sectional Mexican Firm Value of equity. While a similar study in Nigeria, though non empirical, by Kantudu (2003), suggested that adjusting for inflation would improve quality of reporting. Konchitchki [2009], reports that although inflation effects are not recognized in nominal financial statements, they have significant economic consequences, even during period in which inflation is relatively low. Similarly, Wettington (2010), is of the view that failure to adjust for price level could have a devastating effect on the quality of financial report. He further suggests valuation methods that would closely mark the market.
2.1 Hypothesis development and Model building.
The study maintains that joint information content of GPP and HCA increases value relevant of financial information over traditional historical cost reporting. The hypothesis is depicted as thus:
H0:AdjR2 HCC < AdjR2HCC/GPP
To test the hypothesis, modified price model was used, given as:
------(1)
Ceretis paribus market price should be equal to book value, but this is not the case because of Conservative accounting practice. Introducing inflation accounting model into equation (1)
Decomposing 2 into 3 pave way to joint model given as thus:
3.0 Research Design Sample Data & Analysis Tools
Data for the study were sourced from annual financial statement of the four cement companies. The firms are public limited companies listed on the Nigeria stock exchange. By virtue of being public and as a requirement of been listing, annual financial report has to be made available to the exchange.
The financial statements in Nigeria are reported based on Historical cost. Except in certain cases were the provision of SAS3 gives latitude of current cost reporting through revaluation. Historical cost data is available from the annual financial reports. GPP conversion was made based on the IAS [29], PSSAP7, Goldsmith framework, and Ohlson’s clean surplus accounting theory. With the aid of Microsoft excel simple programme was designed to convert the historical accounting figures to inflation adjustments. Consumer price index was used as conversion [inflation factor for GPPP. The clean surplus accounting theory used in estimating income data was tested using HCA income data for past years and run on SPSS. The output validates the theory with 67% level of precision.
Sample Size and Population.
The sample analyzed for the study consisted of four major firms engage in cement production business within the time frame of the study. A study of this nature requires highly capital intensive business. Previous researches have reported no impact of inflation adjustments in service oriented businesses [Bernad and Ruland, 1987]. Within the time frame only four cement companies were quoted on Nigerian stock market. All the four were taken as sampled population. Therefore the study employs census.
Techniques of Data Analysis
Studies that investigate accounting information signals, oftenly used regression model in making inference. In line with that, this study adopts multiple regression technique in analyzing the data. Ordinary Least Square was used to estimate the regression coefficient through SPSS.
As found in the value relevant literature three methods are oftenly used in establishing regression equations:
1. The valuation approach, which relate prices with book value by harmonizing the dependent and the independent with variable with opening book value [Gordon, 2001].
2. The levels approach, which also relates prices with book value, but in this case the two variables are harmonized with opening share prices, [Kothari, 1992].
3. The changed based approach, which relates the variation in prices with changes in book value when the dependent and independent variables are harmonized with opening prices for the year [Kothari, 1992, Bao & Bao, 2001].
The third approach was adopted for this study. All the regressors were denominated by unit of equities outstanding at the end of the accounting year. By doing this we mitigate against the effect of random walk of shares and non normality of variable [homoscedacity] This technique was used in similar studies like; Biddle, et al, 1997; Gordon, 2001; & Friday, 1997].
The multiple regression used is computed as:
Y= 0 + 1x1+ 2x2+.... n xn+Î ...... (5).
Where:
Y= share price, b=regresson intercept [constant], 1...... 6 are the coeffient of the regressors and Î= stochastic variable known as the error term or regression residual.
With the aid of SPSS the OLS is estimated and inference is made based on coeffient of determination, R2, Adjusted R2 computed as:
4. Results and Discussions:
This section presents and discusses empirical results for testing the hypothesis formulated. The first table presents value relevant of conventional reporting using HCA model. And the second table presents joint information content of HCA and GPP. The fitness of the two models were used as comparative benchmark for determining incremental relevant signal quality.(Bernand & Ruland,1987).
Table1.
/ 1 / 2 / 3 / AdjR2 / F / F SIGNBETA / 8.293 / .371 / .400 / -.245 / .39 / 9.9 / .0001
T.Stat / 4.638 / 2.430 / 2.344 / -1.662
P / .0000 / .019 / .042 / .1046
M.R / .65
R2 / .43
DW / 2.37
N / 42
Source: Authors computation using SPSS
Table 1 presents the regression results from SPPS output and the result is used to test the value relevance of historical cost accounting information. The F statistic 9.9 at 00001 indicates a strong fitness of the model. The three regressors: book value accounting earnings and change in earnings explain 39% of the price movement as indicated by the adjusted R2 at 1% percent level of significance. The regressors also show 65% association with the regressand as indicated by M.R. The individual contribution of each the regressor is displayed in the beta row (coefficient). As expected 1 # 2 # 3 and 1 …1 >0 .This proves the independent relevance of the criterion variables on the regressand. While Book Value [BV] positively and significantly contributes 37% to the independent variable at 1% level of significance with strong and positive T statistic, like wise earnings (E) with 40% at 1% level of significant, change in earnings variable [CE] shows an insignificant negative contribution of 24%. The negative sign of C.E. was as a result of negative figures in changes in earning from the data. Likewise the sstatistical insignificant of the coefficients was due to the presence of multicolinearity. Molticollinearity diagnosis result was generated from SPSS output and shows present of intercorrelation among the independent variables. To remedy this effect Partial correlation test was run. The result shows a high positive correlation between earnings components and change in earnings. The probability of the correlation reveals `significance of the independent variable at 1%. This test led us to accept the significant value of change in earnings [Keller & Warrack, 2003]. The DW test of autocorrelation shows 2.37 which is a good indication of absence neither positive nor negative autocorrelation.
Table 2: Test of Joint Value Relevance of HCA and GPP
/ 1 / 2 / 3 / 4 / 5 / 6 / AdjR2 / F / SG.FBeta / 8.7 / .729 / .404 / -279 / -.454 / .049 / .241 / .52 / 8.2 / 0000
TSTAT / 4.3 / 3.7 / 2.6 / -1.91 / -2.5 / .376 / 1.68
P / 0001 / .0006 / .0115 / .064 / .014 / .102 / .70
M.R / .77
R2 / .59
DW / 1.75
N / 41
Source: Authors computation using SPSS
Table 2 presents SPSS output of the test of marginal information value relevance of additional disclosure of adjusted inflation, jointly report on HC and GPP model. The results indicate evidence of marginal value relevance of incremental disclosure of inflation adjusted numbers on historical accounting. There is significant increase in adjusted R2 of the regressors, which vividly support the assertion that incremental disclosure brings an additional quality of financial information to the user. Both the F value and F significance support the fitness of the model at 1% level. The regressors explain 52% behaviour of the regress. All the regressors coefficient are significant at 5% level with the exception of GPP earnings and change in earnings, which because of multicollinearity effect failed to reveal any significant contribution. The muliticolinearity diagnostics using correlation results revealed significance of the two variables. The Durbin Watson test shows 1.75. This result lies between lower dc and upper du with value of 1.05 to 1.78. At 5% significant level 44n K>5. The overall correlation of regressors on the regressand is 77% with the coefficient of determination of 59%.
From the regression output in table 2 the adjusted R2 a metric which measures overall contribution of the regress and coefficients and measure of quality of accounting information, revels a 52% variation price is explained by the regressors. As interpreted hitherto additional disclosure of GPP information has improve the fitness of the model with Fsg (000), F(8.2) and an increase of 21% from HCA disclosure which accounts for only 31%. This result supports the alternative hypothesis of incremental information content of additional disclosure by GPP variable beyond HCA variable. This finding conforms with findings in Matoclsy(1984), who documents an evidence of joint information content of historical and inflation adjusted accounting income numbers. More also, the findings of other studies: Griffin (1980), R0 (1980) and Beaver, Griffin and Lands man (1982).The F significant which 0000 make us to reject the null hypothesis.
5.0. Summary and Conclusion: