RRs = S RELATIVE 1U. THE. “ . ,. . ..R,. .1... 1.. I... r. . . 2.27)....T _. NANCI . a; . . .. L . T. _ . V , . . ‘ . I l .V .: "IRRLRmDY 13:"?! a 1 RR EMF RRIS 0F PRICELEVEL ADJUSTE STATEM h. 40 Rolf? S P hesis forthe Degre T V .: .. . ,. . . . , _ .2. . . ..v .. R . . .V . , . . , ~ MICHEGAN STATE . UNWERSR l}. 4‘ E MRRERJER -’ - RAW RR _ , 5M Michigan Sta £0 Univ 85"” frat-f“! This is to certify that the thesis entitled THE RELATIVE USEFULNESS TO INVESTORS OF PRICE-LEVEL ADJUSTED FINANCIAL STATEMENTS: AN EMPIRICAL STUDY presented by Patrick Bruce McKenzie has been accepted towards fulfillment of the requirements for Ph.D Accounting degree in ’ / / 7/ l R _ ’4 , /, /, w»: _/ ;‘é ‘ "YIL/ V -/ -« 1 ~ ’ Major professor /r b I‘\(‘ .\\ L I / j ,x“ ’7’, Date 4/"L,7’/..{‘TY' // [/1 p.71 0-169 ulnar ¢;m: zhaupreq r. .;._..2::23.= as. m.._.,.:_..::.s._ofimzwn .. 3 L a . a . ammmmmammoaaammmm mmmmmmmmmmmmmmm 9...: :5: :55 I!) :3) CI, :7) :3: 1:1 :5 ca: 32 cannanaununuana-ag-u ,mnuaauuaaammwmummmmmmnm aw----------ps~p-ps-~h-h~h-~__p mmmumumuauamumauaauuamwuammwuuamumummmmmmammw anmammmmnmmanonmmnmmmrmnmmmmwmmmmmmm“annmmnom V .vew.~ew.¢w.e~.vvvv.c.~nw.v..vve~e‘ov wage..~ mnnnnnmnmnnnnnnnnnmn.nannnnmnnnnnnamm «NNN--~N----~N----~N-~N-~NNNNNNNN .—.—_______.—______—___._________—_________: m nmmnmmn Ezrzzflzmc:=8338338==3323323.3253:33:?338332223232 a:aeoeeauauanaaaaoaaaeemaaaaaeaceaeococ9999:: tr 4 i u r 2 3. 2.. .3 .m on. 2 3 T. 2 mummmwmmmm www.mamwmw ~_-rhar- mw.mmmuumm nwmmmmmmwm .....VQ.s. "mmmnnmmmn ---.NNN __________ NM 3 2 2 z E 2 cacaaaoa _r r f _ c a swim. mu, ABSTRACT THE RELATIVE USEFULNESS TO INVESTORS OF PRICE-LEVEL ADJUSTED FINANCIAL STATEMENTS: AN EMPIRICAL STUDY By Patrick Bruce McKenzie The purpose of this research was to gather evidence of an empir- ical nature to examine the proposition advanced in Accounting Researdh Study No. 6 that financial statements adjusted for changes in the general price-level are more useful to investors than are the conventional his- torical cost financial statements that are not adjusted for such changes. The concept of usefulness was defined as the predictive capacity of various types of input data to yield future data of interest to the com- mon stock investor. The research methodology of this thesis relied upon a simple forecasting model and a multiple correlation model to evaluate the rela- tive usefulness of two sets of financial data. One set of financial data was adjusted for changes in the general price—level while the other set was unadjusted. The forecasting model used two sets of past income values (adjusted and unadjusted) to predict certain future income values. The income values forecast were compared to the actual income values and the related percentage forecast error was determined. The comparison of the forecast errors produced by the two sets of input data was used to determine whether price—level adjusted input data were more useful (a better predictor) than the unadjusted input data. An input value that Patrick Bruce McKenzie resulted in a lower percentage forecast error was interpreted as being a better predictor. A multiple linear correlation model was used to evaluate the relative predictive capacity (usefulness) of two sets of financial ratios (a price-level adjusted set and an unadjusted set). The finan- cial ratios for one period were the independent variables in the corre- lation model while the dependent variable was return to the investor (stated as a percentage) in the immediately following period. The two sets of coefficients of multiple determination produced by the correla- tion model were compared to determine whether the price-level adjusted financial ratios were more useful than the unadjusted ratios in the sense of having a greater predictive capacity. Greater predictive capacity was indicated by a higher coefficient of multiple determination. The forecast model and multiple correlation model were imple- mented by price-level adjusting the financial statements of nine domestic trunk air carriers for the nine calendar years 1959 through 1967. The required financial ratios were computed for use in the correlation model, and the adjusted income values (both net and Operating) were used in the forecasting model. Return to the investor was computed for the nine calendar years 1960 through 1968. The empirical findings produced by the forecast model and multiple correlation model and their related statistical evaluation failed to support the proposition advanced in Accounting Research Study No. 6 that financial statements adjusted for changes in the general price-level are more useful to investors than unadjusted statements. Patrick Bruce McKenzie In addition, there was virtually no difference in the usefulness (as defined in this study) of the two sets of financial statements. One argument for general price—level adjustments relates to the contention that it is unrealistic in accounting to assume that changes in the value of the dollar may be ignored. In other words, the level of inflation warrants the presentation of price-level adjusted finan— cial statements. The empirical evidence gathered in this dissertation does 225 support the above position. The level of inflation in the United States in the years that affect the price-level adjustments made in this study was apparently so small that only a negligible difference in the usefulness of the two sets of financial statements was observed. Accordingly, the author concludes that price-level adjusted financial statements are not necessary during periods of inflation that are com- parable to the level encountered during this study (principally 1951 through 1967 in which the average annual rate of inflation was 2%). THE RELATIVE USEFULNESS TO INVESTORS OF PRICE-LEVEL ADJUSTED FINANCIAL STATEMENTS: AN EMPIRICAL STUDY By Patrick Bruce McKenzie A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Accounting and Financial Administration 1970 :ms 7 ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to Dr. George Mead for his constant guidance and encouragement throughout the course of this research. His important contribution as well as those of Dr. Herbert Miller and Dr. Frank Bacon are gratefully acknowledged. Dr. Miller was especially helpful in formulating the direction and pur- pose of this dissertation. In addition to needed financial support, Dr. James Don Edwards, Chairman of the Department of Accounting and Financial Administration, more importantly provided encouragement and inSpiration throughout the course of my doctoral studies. His contribution will always be remembered. A fellowship from the American Institute of Certified Public Accountants was the prime monetary support for this dissertation. The American Accounting Association and General Electric Foundation provided financial help during the early stages of my doctoral studies. The finan— cial support provided by these three groups is gratefully acknowledged. Also I extend sincere appreciation to the people at the Civil Aeronautics Board who were so helpful and cooperative during the data collection stage of this research. My sincere thanks also go to Mrs. Jo McKenzie for her most skillful and professional job in the typing of this manuscript. Finally, a special debt is owed to my wife, Sheila, for con- tributing more than her share of the effort required to complete this study. ii TABLE OF CONENTS Page ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES O O O O O O 0 O O O O O O O O O O O 0 O O O O O O O Viii Chapter I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . l 1.1 Purpose of research 1.2 Predictive capacity as a concept of usefulness 1.3 A priori arguments 1.4 Existing knowledge 1.4.1 Statement of the Accounting Principles Board No. 3 1.5 Companies and time periods studied 1.6 Summary of basic research approaches 1.6.1 A simple linear forecasting model 1.6.2 A multiple linear correlation model 1.7 Limitations and other considerations II. METHODOLOGY: GENERAL PRICE-LEVEL ADJUSTMENTS . . . . . 30 1 Introduction 2 Monetary and nonmonetary items .3 General price index used 4 Current assets 2.4.1 Marketable securities 2.4.2 Flight equipment--expendab1e parts and miscellaneous materials and supplies 2.4.3 Prepaid expenses 2.5 PrOperty and equipment, accumulated depreciation and depreciation expense 6 Investments and Special funds 7 Deferred charges 8 Current liabilities 9 Long-term liabilities 0 Deferred Federal income taxes 1 Other deferred credits 2 Capital stock and additional paid-in capital 3 Retained earnings 4 Operating revenues, expenses and income before taxes iii Chapter III. IV. Page 2.15 Net income 2.16 Accuracy of price-level adjustment methodology FORECAST MODEL: METHODOLOGY AND FINDINGS . . . . . . . 69 Overview A forecast model Forecast error measures Mean absolute deviation Ordinal rankings of means and deviations Percentage of overforecasts Comments on empirical findings Statistical evaluation of empirical findings 3.8.1 Statistical evaluation based upon ordinal comparisons of absolute percentage forecast errors 3.8.2 Statistical evaluation based upon the size of differences in absolute percentage forecast errors 3.9 Observations concerning statistical evaluation of empirical findings WNO‘UI-FUNH wwwwwuww CORRELATION MODEL: METHODOLOGY AND FINDINGS. . . . . . 95 4 1 Overview 4 2 Financial ratios 4 3 Return to the investor 4.4 A multiple linear correlation model 4 5 Empirical findings 4 6 Statistical evaluation of empirical findings 4.6.1 Statistical evaluation based upon ordinal comparisons of coefficients of multiple determination 4.6.2 Statistical evaluation based upon the size of differences in coefficients of multiple determination 4.7 Observations concerning statistical evaluations of empirical findings SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . 118 5.1 Summary and Conclusions 5.2 Purpose of research and summary of research design 5.3 Summary of empirical findings and related evaluation 5.4 Major limitations of this study iv Chapter Page 5.5 Conclusions and implications for contemporary accounting 5.6 Directions for future research SELECTED BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . 127 Table 10 11 12 LIST OF TABLES Marketable Securities Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 C O O O O O I O I O O O O O I O O 0 Flight Equipment—-Expendable Parts and Miscellaneous Materials and Supplies Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 . . . . . . . . . . . . . . . . Prepaid Expenses Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 . . . . . . . . . . . . . . . . Trans World Airlines Summary of Pr0perty and Equipment (000 omitted). . . . . . . . . . . . . . Gross Property and Equipment Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at Decem er 31 O O O O O O O C O O O O O I O O O 0 Net Property and Equipment Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 . . . . . . . . . . . . . . . . . . 0 Flight Equipment Deposits Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at Decem er 31 I I O O O O O O O O I O O O C I O Other Investments and Special Funds Stated as a Per Cent of Total Assets Both on a Historical Basis at December 31 . . . . . . . . . . . . . . . Total Deferred Charges Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 . . . . . . . . . . . . . . . . . . Other Deferred Charges Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 O O O O O O O O O O O O O 0 O O O Unearned Transportation Revenue Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 O O O O O I O O O O I O I O O O O O Other Deferred Credits Stated as a Per Cent of Total Assets Both on a Historical Cost Basis at December 31 O O O O O O O O O O O O O O O O O 0 vi Page 40 42 44 51 52 59 Table Page 13 Percentage by Which Annually Adjusted Operating Income Before Depreciation is (Greater) Less Than Quarterly Adjusted Operating Income Before Depreciation. . . . . . . . . . . . . . . . . . . 64 14 Mean Absolute Percentage Forecast Errors All Companies . . . . . . . . . . . . . . . . . . 75 15 Mean Absolute Deviations of the Percentage Forecast Errors. . . . . . . . . . . . . . . . . . . . . 77 16 Ordinal Rankings of Means and Deviations as Shown in Tables 14 and 15: No Limits on Percentage Forecast Errors . . . . . . . . . . . . . . . 79 17 Percentage of Overforecasts for all Companies By Period. . . . . . . . . . . . . . . . . . . . . . . . 80 18 Percentage of Forecasts Having Lower Absolute Percentage Forecast Error When Using Adj us ted Input Data 0 O O O O O O O O O O O O I I O O O O 89 19 Possible Combinations of Seven Financial Ratios used x at a Time 0 I O O O O I O O O O O O I O O O O O I 9 7 20 Percentage of Coefficients of Multiple Determination That Were Greater Based on Adjusted Financial Ratios O O O O O O O O O O O O O O O O I O I O O O O O O 103 21 Mean Coefficient of Multiple Determination by Period and for Each Group of Possible Combinations (Stated as Percentages). . . . . . . . . . . . . . . . . 105 22 Frequency Distribution for Coefficient of Multiple Determination (CMD) Comparisons for Each of the 120 Possible Combinations of Financial Ratios- . - - - . 107 23 Number of Financial Ratio Combinations Rejecting the Null Hypothesis for Each Significance Interval and Each Type of Period Based Upon Randomization Test for Matched Pairs . . . . . . . . . . . . . . . . . 116 24 Annual Levels of Inflation and Deflation in the United States as Measured by the Gross National Product Implicit Price Deflator. . . . . . . . . . . . . 123 vii LIST OF FIGURES Figure Page 1 Combinations of Input Data and Forecast Objective. . . . . 20 2 Four Income Measures . . . . . . . . . . . . . . . . . . . 7O 3 Combinations of Input Data and Forecast Objective. . . . . 71 4 Schematic of Forecast Model. . . . . . . . . . . . . . . . 72 5 Hypothetical Cases Where Actual Amounts are Near Zero or Negative. . . . . . . . . . . . . . . . . . 74 6 Outline of Four Comparison Models Used to Test Research Hypotheses . . . . . . . . . . . . . . . . 83 7 Summary of Periods Studied . . . . . . . . . . . . . . . . 98 viii CHAPTER I INTRODUCTION 1.1 Purpose of research It is the purpose of this study to gather evidence of an empiri— cal nature to test the proposition that financial statements adjusted for changes in the general price-level are more useful than the conven- tional historical cost financial statements that are not adjusted for such changes. In other words and more specifically, given the degree of inflation encountered in the past decade, are price-level adjusted finan- cial statements warranted as a replacement for or as a supplement to the conventional financial statements that have been and are currently being presented in published annual reports for investors and prospective investors? It was assumed in this study that accounting data are used in decision-making by present and potential investors. A staggering amount of accounting literature has been devoted to the subject of price-level adjustments, their nature and necessity, and the related problems of implementation. For example, in the decade of the 1950's alone, there were 136 articles covering 1048 pages in 223 Accounting Review, The Journal of Accountancy, and the NAA Bulletin on price-level problems.1 And the rapid pace did not slacken in the 1960's. 1Henry W. Sweeney, Stabilized Accounting (New York: Harper & Brothers, 1936; New York: Holt, Rinehart and Winston, Inc., 1964), p. xxx. Although the price-level controversy is far from being settled, the Accounting Principles Board of the American Institute of Certified Public Accountants issued the following statement in 1961 which provided the impetus for Accountigngesearch Study No. 6: Reporting_the Financial Effects of Price-Level Changes: the Board " . . . agreed that the assump- tion in accounting that fluctuations in the value of the dollar may be "2 ignored is unrealistic . . . This statement was strongly approved by Sprouse and Moonitz in AccountinggResearch Study No. 3: A Tentative Set of Broad Accounting Principles for Business Enterprises.3 Apparently through a priori reasoning alone, the following con- clusion was reached in Accounting_Research Study No. 6: In a nutshell, financial data adjusted for price-level effects provide a basis for a more intelligent, better informed allo- cation of resources, whether these resources are in the hands of individuals, business entities or of government. As there is a dearth of empirical evidence in support of or in opposition to the usefulness of general price-level adjustments, hopefully this study can make a significant empirical contribution to this controversial area. Empirical evidence in most areas of accounting is lacking as evidenced in 2Staff of the Accounting Research Division of the American Insti— tute of Certified Public Accountants, Accountinijesearch Study No. 6: Reportingfthe Financial Effects of Price-Level Changes (New York: American Institute of Certified Public Accountants, 1963), p. 1. 3Robert T. Sprouse and Maurice Moonitz, Accountigg Research Study No. 3: A Tentative Set of Broad Accounting_Princip1es for Business Enterprises (New York: American Institute of Certified Public Accountants, 1962), pp. 17-18. 4Staff of Accounting Research Division, Accountigg Research Study No. 6, p. 16. the following statement by Chambers: Compared with the volume of mere Opinion and dogma which fill the so-called theoretical literature of the past fifty years, the amount of actual evidence, the empirical evi- dence, is pitifully small.5 Accordingly, the purpose of this study is to gather evidence to test the position advanced in Accounting Research Study No. 6. 1.2 Predictive capacity as a concept of usefulness The relative predictive capacity of the two sets of financial data, one set adjusted for changes in the general price-level and the other unadjusted, will be the ultimate test of usefulness in this research. In other words, usefulness in this study is defined in the positive sense as meaning predictive capacity. That which is to be pre- dicted as well as the accounting data on which the prediction will be based will be described later in this chapter. This definition of usefulness is in keeping with the current trend in accounting that emphasizes the predictive value of information. The Committee to Prepare a Statement of Basic Accounting Theory (1966) of the American Accounting Association stated that the "utility of infor- mation lies in its ability to reduce uncertainty about the actual state of affairs of concern to the user."6 Sprouse stated that " . . . the primary purpose of measurement of last year's income reported to investors 5R. J. Chambers, "Prospective Adventures in Accounting Ideas," The Accounting Review, Vol. XLII, No. 2 (April, 1967), p. 251. 6Committee to Prepare a Statement of Basic Accounting Theory, A Statement of Basic Accounting_Theory (Evanston, Illinois: American Accounting Association, 1966), p. 8. is to provide a basis for predicting future year's income."7 Anthony in discussing criteria for financial accounting data defines usefulness as being " . . . useful to the external world, to the person who seeks information from the financial statements."8 The so-called "informed investor" ideally would prefer information that enables him to "predict: (1) future earnings, and (2) the safety of his principal."9 Once again, the emphasis is on the predictive capacity of information provided by the accounting system. The Statement of Basic Accounting Theory in discussing external uses of accounting information stated: Almost all external users of financial information reported by a profit-oriented firm are involved in efforts to pre- dict the earnings of the firm for some future period. Such predictions are most crucial in the case of present and prospective equity investors and their representatives . . . The past earnings of the firm are considered to be the most important single item of information relevant to the predic- tion of future earnings. The notion of predictive value can be traced to the widely advo- cated methodology of positive economics as espoused by Milton Friedman and his predecessors which emphasizes the predictive capabilities of 7Robert T. Sprouse, "The Measurement of Financial Position and Income: Purpose and Procedure," Research in Accounting Measurement, R. K. Jaedicke, Y. Ijiri, and O. Nielsen (eds.) (Evanston, Illinois: American Accounting Association, 1966), p. 106. Robert N. Anthony, "Research in Accounting Measurement," Research in Accounting Measurement, R. K. Jaedicke, Y. Ijiri, and O. Nielsen (eds.) (Evanston, Illinois: American Accounting Association, 1966), P. 259. 91bid., p. 260. 10Committee to Prepare a Statement of Basic Accounting Theory, Basic Accounting Theory, p. 23. models, theories, and information in general. As enunciated by Friedman, the "goodness" of a model is judged primarily by the correlation of pre- dicted outcomes and observed reality. The test of the theory or model is pragmatic in all situations; that is, the most relevant model is the one that most accurately predicts the observable phenomena of future economic events.11 The concept of positive economics can be applied to this study as follows: if one set of the two sets of financial data is signifi- cantly more predictive than the other set, then that set will be judged as more useful in its ability to reduce uncertainty about the future. Remember that one set of financial data will be based on conventional historical cost accounting data and the other set will be based on accounting data adjusted for changes in the general price-level. 1.3 A priori arguments No attempt was made in this study to evaluate the relative merits of the theoretical and pragmatic arguments for and against general price-level adjustments. These arguments are well-documented in accounting literature and can be summarized briefly as follows: Arguments against general price-level adjustments:12 1. The current level of inflation in the United States is so insignificant as to have a negligible effect if price-level adjustments were to be made. There- fore, the costs of implementation would far outweigh any possible benefits from adjusted data. 11Milton Friedman, "The Methodology of Positive Economics," Essays in Positive Economics (Chicago: University of Chicago Press, 1953), pp. 3-43. 12Eldon S. Hendriksen, Accounting Theory (Homewood, 111.: Richard D. Irwin, Inc., 1965), p. 182 and Staff of Accounting Research Division, Accounting Research Study No. 6, pp. 44-51. The proposed adjustments may have undesirable conse- quences as investors are used to the conventional financial statements. Public confusion could result as investors might erroneously believe that current cost information was being presented in the price- level adjusted statements. Companies are very reluctant to reduce reported net income by charges that are not recognized for tax purposes. The purpose of accounting should be to determine mone- tary income rather than economic income as an objective measure of the results of past decisions (money return on money invested) is desirable. Therefore, replace- ment costs and price-level adjusted costs are not rele- vant to the accounting process. Arguments for general-price level adjustments: 1. The income statement and balance sheet would be more meaningful if all items were measured by a stable yardstick and thus stated in terms of a common dollar. Historical costs adjusted for changes in the general price-level would be relevant whereas unadjusted his- torical costs have serious limitations. Accounting involves communication of observed economic reality between the accountant and the financial state- ment user. Price-level adjusted data more closely approximates economic reality and therefore would in- crease the quality of communication between observer and user. "Financial data adjusted for price-level changes pro- vide a basis for a more intelligent, better informed allocation of resources, whether those resources are in the hands of individuals, of business entities, or of government."13 Much of the price-level controversy contained in the foregoing arguments can be directly attributed to a broader problem that has con— fronted accountants for decades: a failure by the accounting profession 13Staff of Accounting Research Division, Accounting Research Study No. 6, p. 16. to attain a general agreement on the fundamental objective or purpose that accounting should fulfill. Many of the continuing practical and theoretical controversies in accounting can be traced to this failure. The solution of this broader problem would be an important step in the settlement of several accounting dilemmas including the price-level dispute. 1.4 Existingfiknowledge In reviewing the current and past accounting literature that would be relevant to this study, several significant research efforts were noted. Although ignored at the time and not "discovered" until the inflationary period following World War II, the comprehensive work of Henry W. Sweeney, consisting of fifteen articles published in the leading accounting journals during 1927 to 1935 and his now classic Stabilized Accounting14 published in 1936, indeed establish him as the founder of price-level accounting in the United States. His work represented the earliest comprehensive study into the theoretical problems and practical implications of assuming a stable measuring unit in times of inflation or deflation. The bulk of his conclusions and recommendations were incorporated in Accountinijesearch Study No. 6 some 28 years after first being published. In 1949, Ralph C. Jones published the first significant article on price-level accounting since Sweeney's work in 1936. His article 14Henry W. Sweeney, Stabilized Accounting (New York: Harper & Brothers, 1936; New York: Holt, Rinehart and Winston, Inc., 1964). entitled the "Effect of Inflation on Capital and Profits: The Record of Nine Steel Companies" presented empirical evidence of the difference in financial statements caused by the changing price—level.15 In 1951, the American Accounting Association in Supplementary Statement No. 2: Price Level Changes and Financial Statements concluded " . . . that knowledge of the effects of the changing value of dollar . . . may be useful information . . . "16 As a result of this directive, the American Accounting Associ- ation enlisted Ralph C. Jones to direct a price-level case study having the following objectives: 1. To develop and test techniques and methods for the preparation of supplementary financial statements expressed in constant-value units . . 2. To compare the supplementary statements expressed in uniform dollars with the conventional statements expressed in historical dollars . . 3. . . . for judging the need for and the usefulness of figures and statements in dollars of uniform pur- chasing power.17 Representing one of the early major empirical efforts in this area, the results of this milestone study were very striking. For example, one 15Ralph C. Jones, "Effect of Inflation on Capital and Profits: The Record of Nine Steel Companies," The Journal of Accountancy, LXXXVIII (January, 1949), pp. 9-27. 16Committee of Concepts and Standards Underlying Corporate Financial Statements, Supplementary Statement No. 2: Price Level Changgs and Financial Statements (American Accounting Association, 1951). P. 2. 17Ralph C. Jones, Price Level Changes and Financial Statements: Case Studies of Four Companies (American Accounting Association, 1955), p. 2. company studied for an eleven year period overstated the earning rate on average equity of stockholders by almost 100% of the rate based on the adjusted statements; also dividends as a per cent of net earnings were 562 for the published statements as compared to 89% for the adjusted statements.18 This study provided the impetus for more research, mostly of a nonempirical nature, in the area of general price-level adjustments. The empirical work that was undertaken was similar to the case studies made by Jones in that they concentrated on showing the differences in the financial statements that would result from making price—level adjustments, and largely ignored the more important problem of the rela- tive usefulness of the two sets of data. The American Institute of Certified Public Accountants formally acknowledged the price-level problem in a letter to the membership re- affirming its opinion expressed in Accounting_Research Bulletin No. 33 which was published in 1947: The committee on accounting procedure has reached the con— R clusion that no basic change in the accounting treatment of depreciation of plant and equipment is practicable or desir- able under present conditions to meet the problem created by the decline in the purchasing power of the dollar. Should inflation proceed so far that original dollar costs lose their practical significance, it might become necessary to restate all assets in the terms of the depreciated cur- rency, as has been done in some countries. But it does not seem to the committee that such action should be recommended now if financial statements are to have maximum usefulness to the greatest number of users.19 18Ibid., p. 67. 19American Institute of Certified Public Accountants, Accounting Research and Terminology Bulletins, Final Edition (New York: American Institute of Certified Public Accountants, 1961), pp. 68-69. 10 In 1963, AccountinggResearch Study NO. 6: Reporting the Finan- cial Effects of Price-Level Changes was issued under the directorship of Maurice Moonitz. As a result of this research effort, the following con- clusions were reached: 1. 4. In They found "clear evidence of the widespread concern of businessmen and accountants with the need for changes in financial reporting to reflect the effects of inflation and deflation." The "recognition of price-level changes in financial statements is practical, and not misleading or dangerous to investors." At least one reliable general price-level index is avail- able in the United States. "The effects of price-level changes should be disclosed as a supplement to conventional statements."20 1967, the Accounting Principles Board discussed internally a draft of a prOposed Opinion on price-level accounting that included the following recommendations: 1. 2. 3. In cases of severe inflation or deflation, price-level statements should preferably be presented as the basic financial statements. The degree of U. S. inflation does not require price- level data to be included for fair presentation. However, the presentation of price-level data in addi- tion to conventional data is encouraged.21 This draft was the predecessor of Statement of the Accounting Principles Board No. 3 which is discussed in the next section. 20 Staff of Accounting Research Division, Accountinijesearch Study NO. 6, p. xi. 21Accounting Principles Board, Exposure Draft: Financial State- ments Restated for Price-Level Changes, July 14, 1967, pp. 6-7. (Mdmeographed). 11 Before issuing an actual pronouncement, the Accounting Principles Board felt that more knowledge Of the implementation process and effects was needed. It therefore authorized a field test which 18 companies of widely varying sizes and types were asked to adjust their financial statements for changes in the general price—level. Two significant con- clusions resulted: 1. The participants in general agreed that . . . practical problems should not present a significant barrier to preparation of general price—level financial statements. 2. The results . . . seem to support the view of Accounting Research Study NO. 6 that presentation of supplementary general price-level financial statements would make avail- able potentially useful information that otherwise is not disclosed.22 During the last five years, three empirical studies involving the use Of questionnaires and interviews with financial analysts and other related groups attempted to evaluate the usefulness of price-level adjusted data. Morton Backer concluded that the 48 analysts be inter- viewed were almost unanimously Opposed to the adjustment of income by the use of a general price—index. In general, these analysts were against the introduction of more subjective measurements. However, they did favor a system of multiple measurements for balance sheet items.23 Estes investigated the perceived usefulness to external finan- cial statement users (financial analysts, financial executives, and bank loan officers and credit men) of price-level and current value 22Paul Rosenfield, "Accounting for InflationeA Field Test," The Journal of AOcountancy, CXXVII (June, 1969), p. 50. 23Morton Backer, "Financial Reporting and Security Investment Decisions," Financial Executive, XXXIV (December, 1966), pp. 50-60. 12 information. For the price-level adjusted information, 70% of the respondents believed that such supplemental data would be useful as Opposed to 30% who indicated that it would not be useful.24 In a research effort sponsored by the American Accounting Asso- ciation, Dyckman used questionnaires sent to financial analysts to examine experimentally the effects of price-level adjustments on deci- sion makers. The following behavioral conclusions resulted:25 1. The inclusion of statements adjusted for the changing value of the dollar or, alternatively, the reliance on adjusted statements alone, can influence relative investment evaluations. This study, however, suggests that this relationship is not a strong one. 2. The study respondents would prefer that, if price- level adjustments are to be made, they be supplied in the form of supplementary reports to conventional reports. In summary, the existing price-level literature can be classified into three categories: (1) a priori research (normative theory), (2) empirical evidence (positive theory), and (3) official positions of authoritative bodies. It was not deemed necessary in this study to eval- uate the a priori research. The empirical evidence falls into two classes: descriptions of differences between conventional and price- level adjusted financial statements of actual companies and surveys of investor Opinions about the perceived usefulness of price-level adjusted 24Ralph W. Estes, "An Assessment of the Usefulness of Current Cost and Price-Level Information by Financial Statement Users," Journal of Accounting Research, VI (Autumn, 1968), p. 200. 25T. R. Dyckman, Studies in Accountinngesearch #1: Investment Analysis and General Price-Level Adjustments (Evanston, Illinois: American Accounting Association, 1969), p. 17. 13 data. Sub-section 1.4.1 describes the most recent and comprehensive statement on price-level adjustments by an authoritative body. The major implication of the preceding research efforts for this study is that price-level adjusted financial statements may provide use- ful information. This thesis examines the prOposition that price-level adjusted information is more useful than information unadjusted for changes in the general price-level. 1.4.1 Statement of the Accounting Principles Board No. 3 In mid-1969, the Accounting Principles Board (APB) issued its Statement No. 3: Financial Statements Restated for General Price—Level Changes.26 While the Board's "statement" purposely lacked a Board "opinion", it nevertheless sets forth the conclusions and recommendations of the APB regarding general price-level adjusted financial reports. In addition to describing the basic nature Of financial statements adjusted for changes in the general price-level, this statement provides detailed guidance on how to prepare and present general price-level financial statements. The major recommendations and conclusions of Statement NO. 3 can be summarized as follows:27 1. The Board believes that general price-level financial statements or pertinent information extracted from them present useful information not available from basic historical-dollar financial statements. 2. General price-level information may be presented in addition to the basic historical-dollar financial 26Accounting Principles Board, Statement of the Accounting Prin- .Elgles Board No. 3: Financial Statements Restated For General'Price- LeVel Changes (New York: American Institute of Certified Public Accountants, June 1969). 27Ibid., pp. 12-13. 14 statements, but general price—level financial state- ments should not be presented as the basic statements. The Board believes that genral price-level information is not required at this time for fair presentation of financial position and results of operations in con- formity with generally accepted accounting principles in the United States. The Board recognizes that the degree of inflation or deflation in an economy may become so great that con- ventional statements lose much of their significance and general price-level statements clearly become more meaningful, and that some countries have experienced this degree of inflation in recent years. The Board concludes that general price-level statements reported in the local currency of those countries are in that respect in conformity with accounting principles gen- erally accepted in the United States... Statement NO. 3 provided the following general guidelines to be used in the adjustment or restatement of conventional financial state- 28 ments for changes in the general price-level: l. The same accounting principles . . . should be used in preparing general price-level financial statements except that changes in the general purchasing power are recognized . . . General price-level financial statements are an extension of and not a departure from the "historical cost" basis of accounting. An index of the general price-level, not an index of the price of a specific type of goods or services, should be used to prepare general price-level finan- cial statements. . . . the GNP Deflator is the most comprehensive indicator of the general price level in the United States. Consequently, it should normally be used . . . General price-level financial statements should be presented in terms of the general purchasing power of the dollar at the latest balance sheet date. Monetary and nonmonetary items should be distin- guished for the purpose of preparing general price- level financial statements. Monetary items are 28 Ibid., pp. 13-19. 15 stated in terms Of current general purchasing power in historical-dollar statements. On the other hand, nonmonetary items are generally stated in terms of the general purchasing power of the dollar at the time they were acquired. 5. The amounts of nonmonetary items should be restated to dollars of current general purchasing power at the end of the period. 6. Monetary assets and liabilities . . . are stated in dollars of current general purchasing power; conse- quently, they should appear in current general price— level statements at the same amounts. 7. The amounts of income statement items should be re— stated tO dollars of current general purchasing power at the end of the period. 8. General price-level gains and losses should be calcu- lated by means of the general price index and included in current net income . . . as a separate item in gen- eral price-level income statements. 9. General price-level financial statements of earlier periods should be updated to dollars of the general purchasing power at the end of each subsequent period for which they are presented as comparative information. 10. All general price-level information presented should be based on complete general price-level calculations. As this statement represents the most comprehensive and authori- tative pronouncement on the mechanics of price-level adjustments, it has been used as a guide for the price-level adjustments made in this study. Its pronouncements are very similar to the price-level theory and pro— cedures espoused by Sweeney,29 Jones,30 and Accounting Research Study 29Sweeney, Stabilized Accounting. 0Jones, Price Level Changes and Financial Statements: Case Studies of Four Companies. 16 No. 6.31 1.5 Companies and time periods studied The Air Transportation industry, more specifically the nine domestic trunk carriers publishing financial statements on a calendar year basis, was selected for analysis and implementation of this study. The carriers within this industry were suitable because the Civil Aero- nautics Board requires these air carriers to file certain detailed re- ports of Operating activity and financial position; these data happen to include information necessary in making price—level adjustments. Information of this nature is seldom if ever available in published annual reports. For this reason, it was not feasible to randomly sample from some larger pOpulation. This limitation will be discussed in greater detail in subsequent chapters. In summary, the primary criterion for selection of companies studied was the availability of detailed information that would facili- tate the price-level adjustment process. Other ancillary reasons for selecting the domestic trunk carriers include (1) the number of companies within this group represents a magnitude that was manageable in the actual implementation of the adjustments for changes in the general price~levelg (2) these air carriers are subject to a uniform system of accounts and reports which permits more meaningful intercompany comparisons; (3) this industry has a relatively heavy investment in prOperty and equipment so that price-level adjustments should produce significantly different 31Staff of Accounting Research Division, Accountinngesearch Study No. 6. 17 financial statements; (4) all carriers were subject to similar general economic and Operating conditions; and (5) though there were some excep- tions, the nine carriers followed similar accounting policies, especially in the determination of depreciation. Domestic trunk carriers include only domestic Operators primarily within the continental United States serving the larger cities and con- sist of: 1. American Airlines, Inc. (16.8%) 2. Braniff Airways, Incorporated (3.9%) 3. Continental Air Lines, Inc. (2.8%) 4. Delta Air Lines, Inc. (7.7%) 5. Eastern Air Lines, Inc. (13.0%) 6. National Airlines, Incorporated (4.1%) 7. Northeast Airlines, Inc. (1.9%) 8. Northwest Airlines, Incorporated (6.0%) 9. Trans World Airlines, Inc. (16.5%) 10. United Air Lines, Inc. (23.2%) 11. Western Air Lines, Inc. (4.1%) Relative share of the market is indicated above in brackets based on 1968 revenue passenger miles. Two of these air carriers, Delta Air Lines and National Airlines, issue their published annual reports on a June 30 fiscal year basis. Be— cause of comparability problems, these two carriers were excluded from this study. The time period studied was the ten year period 1959 through 1968 which includes the air carriers' transition into the so-called jet age. 1.6 Summary of basic research approaches The purpose of this section and the two sub-sections that follow is to outline the two research approaches that were employed in this study to examine the proposition eSpoused in Accountinijesearch Study 18 ‘Ng;_§_that price-level adjusted financial data are more useful than con- ventional financial data that have not been adjusted for changes in the general price—level. Recall that useful has been defined as predictive capacity. The first approach consisted of a simple linear regression model that was used to forecast future income values based upon two sets of past income values. One set of these past income values was adjusted for changes in the general price-level while the other set was not adjusted for these changes. Sub-section 1.6.1 presents a brief summary of this model. The second approach employed a multiple linear correlation model to evaluate the relative predictive capacity of two sets of financial ratios (one set price-level adjusted and the other unadjusted). The financial ratios were the independent variables in this correlation model while the dependent variable was investor return measured by market price appreciation (or depreciation) and cash dividends. Sub—section 1.6.2 describes this model in more detail. 1.6.1 A simple linear forecasting model As indicated in a previous section, usefulness implies to many "predictive capacity", and the item that the investor would most like to predict would be his future returns from alternate investments. In order to make this type of prediction, many investor decision models depend heavily on a prediction of future earnings. Sprouse has stated " . . the primary purpose Of measurement of last year's income reported to investors is to provide a basis for predicting future year's 19 income."32 Hayes has concluded that: . . . the valuation of most common stocks involves two prin- cipal steps or procedures. The first is the preparation of some estimate of the probable range of the earnings potential for the future . . . The second step . . . is to establish a reasonable price for the estimated earning power . . . In the majority of cases the statistical record of past earnings re- flected by the income . . . statements constitutes the start- ing point for the calculation of possible future earning power . . . The objective (of any adjustments made by the investor) is to make the past record indicative to the greatest extent possible of the economic activities which seem most likely to prevail in the future.33 In summary, past income values can be used (and are used, it seems rea- sonable to assert) by investors to forecast future income values. The forecasted future income value can then be used by the investor to facilitate his market decisions. The efficiency of this decision making process depends in part upon the reliability of the forecasted income value; the more closely this value corresponds to the income value actually reported the greater the efficiency of this process will be. So, to be useful in investor decision models as Specified above, the forecasted income value must be reasonably close to the actual income value. In order to determine the relative usefulness (that is, rela- tive predictive capacity) of adjusted financial data versus unadjusted financial data, forecasts were made using a simple linear time series regression model employing the six combinations of input data and 2Sprouse, "The Measurement of Financial Position and Income: Purpose and Procedure," Research in AccountinggMeasurement, p. 106. 33Douglas A. Hayes, Appraisal and Management of Securities (New York: The Macmillan Company, 1956), pp. 284—85. 20 forecast objective summarized in Figure 1. FIGURE 1 COMBINATIONS OF INPUT DATA AND FORECAST OBJECTIVE Input Forecast Combination Data Objective l 01 01 2 AOI A01 3 A01 01 4 NI NI 5 ANI ANI 6 ANI NI Codes used: OI Operating Income before income taxes, as reported AOI = Operating Income before income taxes, price-level adjusted NI = Net Income after income taxes, as reported ANI = Net Income after income taxes, price— level adjusted There were four different types of input data and four differ- ent forecast Objectives. Each type of input data was used to predict itself (combinations 1, 2, 4 and 5) and the two adjusted types of input data were used to predict the corresponding unadjusted forecast objec— tive (combinations 3 and 6). The selection of these six combinations of input data and forecast objective is discussed in Chapter III. The four most recent annual income values were used as inputs into a time series regression model to determine the forecasted income value for the next year and the associated forecast error stated as a percentage of the actual value. Using nine years of income values, there were five forecasts for each of nine companies and each of the 21 six forecast models. Accordingly, there were 45 forecasts for each forecast model. This methodology is similar to that employed by Frank34 and Parker35 to test the predictive capacity Of various income concepts. Frank investigated the historical cost versus current cost concepts of income argument while Parker examined the current Operating income versus all-inclusive income controversy. In summary, the following research hypothesis was tested employ— ing the statistical methodology outlined in Chapter III: the use of income data adjusted for changes in the general price-level should result in more reliable forecast values, using a simple linear forecasting model, than would be obtained using unadjusted income data. Chapter III contains a detailed description of the foregoing forecast model, a definition of "more reliable forecast value,’ and a summary of the related empirical findings. 1.6.2 A multiple linear correlation model The usefulness of financial ratios in analyzing and comparing firms is well-established in financial literature. For example, William Beaver's doctoral dissertation and subsequent published journal articles investigated the predictability of financial ratios based on conventional 34Werner Frank, "A Study of the Predictive Significance of Two Income Measures," Journal of Accounting Research, VII (Spring, 1969), pp. 123-136. 35James E. Parker, A Study of the Predictive Significance of Several Income Measures Relative to the Accounting for Extraordinary Items and Prior Period Adjustments, Unpublished Doctoral Dissertation, Michigan State University, East Lansing, Michigan, 1969. 22 data.36 More specifically, he compared the financial ratios of 70 "failed" companies over a period of five years prior to failure with the same ratios of comparable nonfailed companies. The nonfailed companies were comparable as to asset size and industry classification. As a result of his research, Beaver concluded that certain financial ratios "can be useful in the prediction of failure for at least five years prior to the event."37 These financial ratios have traditionally been based on conven- tional historical cost financial statements. Therefore, the following research hypothesis was used to test the position espoused in Accounting Research Study No. 6: financial ratios based on financial statements adjusted for changes in the general price-level are significantly more useful (predictive) than the same ratios based on conven- tional accounting financial statements. This properly places the "burden of proof" on the price-level adjustment concept as it is (1) more complex and costly to apply, (2) more difficult to understand, and (3) unconventional. Eldon Hendriksen in discussing general price—level adjustments and financial ratios stated: 6For example, see William H. Beaver, Financial Ratios as Predictors of Failure (Unpublished Ph.D. dissertation, Graduate School of Business, University of Chicago, 1965); "Alternate Accounting Measures As Predictors of Failure," AccountingiReview, Vol. XLIII (January, 1968), pp. 113-122; "Market Prices, Financial Ratios, and the Prediction of Failure," Journal of Accounting Research, Vol. VI (Autumn, 1968), pp. 179-192. 37William H. Beaver, "Financial Ratios as Predictors of Failure," Empirical Research in Accounting: Selected Studies, 1966 (Baltimore: Institute of Professional Accounting, Graduate School of Business, University of Chicago, 1967), p. 102. 23 Financial ratios computed from unadjusted balance sheet items may provide misleading information; these same ratios may be improved by stating both the numerator and denominator of the ratios in similar terms. The rate of return on investment is one Of the best examples of a ratio that can be greatly improved by using figures restated in terms of a common dollar. In the same chapter, he reached these conclusions: Useful financial ratios can be obtained from the adjusted income statement and balance sheet, and useful comparisons can be made by restating the prior year's balance sheet in terms of the current purchasing power of the dollar . . . . The adjusted rate of return ratio is useful in the appraisal Of a firm by management, stockholders, or outsiders.39 The financial ratios that were used in this model are those that are currently popular among financial analysts and investors, and are generally accepted as having predictive value. The following list is a result Of searching various textbooks, financial journals, and research efforts by Horrigan and Beaver40 to ascertain financial ratios with the characteristics mentioned above: net income to average total assets Operating income to average total assets cash flow to average total assets cash flow to average total debt total revenues to average total assets net income to total revenue operating income to total revenue \IO‘UIbWNI-J 38Eldon S. Hendriksen, Accounting Theory (Homewood, 111.: Richard D. Irwin, Inc., 1965), p. 170. 391bid., pp. 171-172. See a previous footnote for efforts by Beaver. For Horrigan, see James O. Horrigan, "The Determination of Long-term Credit Standing with Financial Rates," Empirical Research in Accounting: Selected Studies, 1966 (Baltimore: Institute of Professional Accounting, Graduate School of Bus., Univ. of Chicago, 1967), pp. 44-62; "Some Empirical Bases of Financial Ratio Analysis," The Accounting Review, Vol. XL (July, 1965), pp. 558-568; "A Short History of Financial Ratio Analysis," The Accounting Review, Vol. XLIII (April, 1968), pp. 284-294. 24 These financial ratios were computed based on two sets of finan- cial data, conventional and price—level adjusted, and represent the independent or explaining variables in a multiple linear correlation model. The ten balance sheets Of the nine air carriers studied for the calendar years ended December 31, 1958 through 1967 and the nine related statements of income and retained earnings of each carrier for the calendar years 1959 through 1967 were adjusted for changes in the gen- eral price-level. At this point, the seven financial ratios were com- puted for each time period and for both sets of data. The dependent or explained variable was "return to the investor" (stated as a percentage) in the period immediately following the finan- cial ratio period. Return to the investor in period j was defined as follows: R = (MPej J 1(ij - MPbl) + CD1 where MPej = market price per share at end of period j MPb = market price per share at beginning of period j CDj = cash dividends per share paid during period j In words, return to the investor was defined as the sum of (1) the change in market price during a period and (2) the cash dividends received during the same period both stated as a percentage Of the mar- ket price at the beginning of the period. The relative predictive capacity of the two sets of financial ratios was evaluated using the multiple linear correlation model 25 described in detail in Chapter IV. The output from this model was two sets of coefficients of multiple determination. The coefficient of multiple determination indicates the percentage of the variation in dependent variable (return to the investor) that can be explained by changes in the independent variables (financial ratios). Accordingly, the coefficients of multiple determination were used to evaluate the research proposition stated earlier in this sub-section. Chapter IV presents the empirical findings from implementation of the correlation model described in this sub-section and also outlines the statistical methodology employed to evaluate the related research hypothesis. 1.7 Limitations and other considerations In addition to the limitations and considerations already men- tioned, several items require attention at this point. As mentioned previously, the companies to be analyzed were not randomly selected from some larger population. Due to this factor, the inferences to be derived from the statistical tests as outlined in Chapters III and IV must be prefaced with a qualification. In other words, the validity of the inferences will rest primarily on the prOposition that the nine com- panies studied are representative of some larger population. The a priori arguments for and against general price-level adjustments are well-documented in accounting literature, and this study has not evaluated these arguments. As a part of the present controversy, much has been written about the merits of using specific price-level index to make financial statements more relevant to investors' needs. 26 Resolving this dispute is outside the scope of this study. Accounting Research Study No. 6 and Statement of the Accounting_Principles Board Ng;_§_both recommended that price-level adjustments be computed using a general price-level index for the rather obvious reason that the use of a specific index would introduce current costs as a replacement for his- torical costs.41 Price-level adjustments made using a general price index are still on the historical cost basis. Accordingly, this study used a general price index following the reasoning espoused above. One segment of this thesis investigated the relative capacity of financial ratios based on two sets of data to predict actual return to the common stock investor. Market price appreciation (or depreciation) is a segment of this return. To the extent that market prices are influ- enced by conventional unadjusted accounting data, an advantage in favor of this set of data will exist. Since price-level adjusted data are not generally available to directly influence market prices, there is no offsetting factor to this advantage, and it is difficult to construct a model on an a priori basis that would indicate a relationship between market prices and unavailable financial ratios based on data adjusted for changes in the general price-level. However, the philOSOphy Of positive economics does not require a logically constructed theory prior to empirical observation. This philosophy is similar to the inductive approach. Hendriksen has written the following: 41Staff of Accounting Research Division, Accounting Research Study No. 6, pp. xi-xii, and Accounting Principles Board, Statement No. 3, p. 3. 27 The advantage of the inductive approach is that it is not necessarily constrained by a preconceived model or struc- ture. The researcher is free to make any observations he may deem relevant. According to Friedman, "the only relevant test of the validity of a hypothesis is comparison of its predictions with experience."43 If a hypothesis or theory has predictive abilities, then it is judged useful. However, understanding depends primarily upon the explanation of this relationship. In all the sciences, it is common that a logically derived theory follows the empirical Observations of actual relationships. One segment Of this thesis observes the relationship between certain finan- cial data adjusted for changes in the general price-level and actual investor return. If a significant relationship is observed, then a logically derived theory could be constructed to explain and understand this phenomenon. The construction of a formally developed theory lies outside the scope of this research. However, speculation as to what such a theory would consist of is considered to be an essential part of this research effort. Predictive capacity played a major role in both approaches to evaluating the relative usefulness of two sets of financial data. In- volved in the first approach was a determination of the predictive 42Eldon S. Hendriksen, Accounting Theory (Homewood: Richard D. Irwin, Inc., 1965), p. 6. 43Milton Friedman, "The Methodology of Positive Economics," Essays in Positive Economics (Chicago: University of Chicago Press, 1953), pp. 8-90 28 capacity of different income measures. This did not imply that pre- dictive capacity should be the only criterion used in evaluating which set of financial data is more relevant. One concept of income may be preferred over another because it is more theoretically "correct" even though it is less predictive. In summary, predictive capacity is not the only criterion for determining which alternate concepts should be used, but it is probably one of the most important of many criteria. Also involved in the first approach was an evaluation of fore- cast errors. For example, in comparing a 20% overforecast of a future income value with a 20% underforecast, the overforecast might be pre- ferred by the investor due to the way that he has relied on the fore- casted amount. This study has summarized the forecast errors and their distribution so that the individual reader can evaluate the importance of over-and under-forecasts depending upon his particular decision model. In discussing the concept of predictive ability as a criterion for evaluating accounting data, Beaver, Kennelly, and Voss conclude that:44 The preference for an accounting measure may apply only within the context of a specific predictive purpose or pre- diction model. It may be impossible to generalize about the 'best' measurement alternative across different contexts. Even within a specific context, the conclusions must be considered as tentative. The inability to generalize is a possibility, but not an inevitability. We have cited only potential difficulties, whose relevance can only be assessed empirically, not by a priori speculation. What is important is to know to what extent we can generalize across purposes, and the only hope of acquiring this knowledge is to conduct the predictive 44William H. Beaver, John W. Kennelly, and William M. Voss, "Predictive Ability as a Criterion for the Evaluation of Accounting Data," Accountinijeview, XLIII (October, 1968), p. 675. 29 studies. If we discover that different measures are best for different purposes, it would be erroneous to believe that the predictive studies are any less important because Of that discovery. The inability to generalize, if it does exist, is not a flaw of the predictive ability methodology. It merely reflects the state of accounting theory, but in neither case is it an indictment of the methodology that eXposes that fact. Extension of research efforts into the predictive ability of accounting data is necessary for the fulfillment of accounting's decision-making orientation and for the mean- ingful evaluation of alternative accounting measures. Accordingly, this study has relied heavily on the criterion of predic- tive ability to empirically test the proposition that price-level adjusted financial data are more useful than financial data unadjusted for changes in the general price—level. The results of this study are of course subject to the limita- tions set forth in this thesis and depend heavily upon the applicability of the predictive capacity criterion and approach to represent a proper measure of usefulness to investors. CHAPTER II METHODOLOGY: GENERAL PRICE—LEVEL ADJUSTMENTS 2.1 Introduction The purpose of this chapter is to outline in detail the mechan- ical procedures for making general price-level adjustments. These pro— cedures are well-documented in various sources in current accounting literature.1 The American Institute of Certified Public Accountants has provided the most recent and comprehensive pronouncement on this subject in Statement of the Accounting Principles Board No. 3 (SAPB No. 3).2 This statement was used as the primary guideline for making the general price-level adjustments in this study. The discussion of the adjustment methodology contained in this chapter assumes that the reader has a fundamental understanding of the 1For example, Henry W. Sweeney, Stabilized Account}ng_(New York: Harper & Brothers, 1936; New York: Holt, Rinehart and Winston, Inc., 1964); Ralph C. Jones, Price Level Changes and Financial Statements-- Case Studies of Four Companies (American Accounting Association, 1955); Ralph C. Jones, Effects of Price Level Changes on Business Income, Capital, and Taxes (American Accounting Association, 1956); Perry Mason, Price-Level Changes and Financial Statements--Basic Concepts and Methods (American Accounting Association, 1956); Staff of the Accounting Research Division, Accounting_Research Study No. 6: Reporting the Financial Effects of Price-Level Changes (American Institute of Certified Public Accountants, 1963); and Ralph D. Kennedy and Stewart Y. McMullen, Financial Statements--Form, Analysis, and Interpretation, Chapters 18- 22 (Homewood: Richard D. Irwin, Inc., 1968). ZAccounting Principles Board, Statement of the.Accounting. Principles Board No. 3: Financial Statements Restated for General Price‘Level Changes (New York: American Institute of Certified Public Accountants, June 1969). 3O 3 '1 31 concept and mechanics of general price-level adjustments. In addition, the reader also should be acquainted with the general recommendations of SAPB No. 3. The availability of adequate data sources normally presents a very formidable if not insurmountable hurdle for the external analyst attempting to make price-level adjustments to conventional financial statements. Fortunately, this hurdle is reduced considerably in the case Of the air carriers under study due to certain reporting requirements Of the Civil Aeronautics Board (CAB) and the Federal Aviation Act of 1958. The CAB prescribes a Uniform System of Accounts and Reports (USAR) for all certificated air carriers. In accordance with the Federal Aviation Act of 1958, as amended, all air carriers are required to file CAB Form 41 which reports financial and Operating data on anywhere from a monthly to an annual basis.3 The accounting data contained in this report together with that available in individual air carrier annual reports and Moody's Transportation Manual4 was sufficient to make the required general price-level adjustments with a minimum of simplifying assumptions. Fixed assets and inventory are the two areas that normally cause the greatest problems in making price-level adjustments. These problems in price-level adjustments have two important aspects: the first is the 3Civil Aeronautics Board, Uniform System Of Accounts and Reports ifOI:Certificated Air Carriers, U. S. Government Printing Office, 1965. 4Moody's Transportation Manual (New York: Moody's Investors Service, Inc., 1945 through 1967). - pat ISL bl 32 data required to make the adjustment and the second is the materiality of the item in relation to its overall effect on the financial state— ments. Fortunately, due to the second factor, the relatively small amounts of "inventory" maintained by air carriers permitted certain assumptions to be made in place of a detailed analysis that would nor— mally be required. 2.2 Monetary and nonmonetary items In making general price-level adjustments to specific financial statement accounts, a dichotomous classification must be made. Each account must be classified as being monetary or nonmonetary in nature; this differentiation then dictates the type of adjustment necessitated. According to SAPB NO. 3, balance sheet items are monetary "if their amounts are fixed by contract or otherwise in terms of numbers of dollars regardless of changes in specific prices or in the general price-level."5 Changes in the general price-level cause holders of monetary items to either gain or lose general purchasing power. This formal computation and recognition of purchasing power gain or loss is unique to price-level accounting. For a set of general price-level adjusted financial statements to be articulated with one another, the purchasing power gain or loss must be included. A holder of cash or any other monetary asset suffers a real loss intgeneral purchasing power during a period Of inflation. The converse is true for any monetary liability. The computation of this gain or 5Accounting Principles Board, SAPB No. 3, p. 8. 33 loss can be illustrated by an example. Assume that a $1,000 monetary asset is held without change during a period in which the general price- level index rises from 100 to 110. The holder Of this asset suffers a $100 loss in general purchasing power (in terms of end Of the period dollars): Purchasing power of $1,000 monetary asset at beginning of the period (in terms of end-of—period dollars) $1,000 x 110/100 = $1,100 Purchasing power of $1,000 monetary asset at the end of the period (in terms of end-of—period dollars) $1,000 x 110/110 = 1,000 (that is, no adjustment) Loss in general purchasing power § 100 The computation of purchasing power gain or loss is somewhat more complicated in actual practice, as rarely will the net amount of monetary items remain constant throughout an entire year. The computation (as set forth in SAPB No. 3) necessitates many assumptions about the timing of additions to and reductions from all monetary accounts. In order to avoid the assumptions and computational problems of this approach, the methodology outlined in Section 2.15 was employed to determine price- level adjusted net income that by-passed the separate computation of Purchasing power gain or loss. Nonmonetary items are simply those items other than monetary as chafined above. Changes in the general price—level do not by themselves (”Ruse holders of nonmonetary items to either gain or lose general pur— c1lasing power. The application of SAPB NO. 3 concepts is described in 34 following sections. In general, though, all material nonmonetary balance sheet items, such as property and equipment, and capital stock, were aged initially as of December 31, 1958. The aging process was similar to that used to age accounts receivable in order that an appro- priate allowance for doubtful accounts be maintained. In essence, the aging process reconstructed an account balance by date of individual acquisitions, issuances, or transactions depending upon the type of account. In this way, the purchasing power of dollars spent for exist- ing fixed assets or of dollars contributed by issuances of capital stock was established and restated in terms of a common-sized dollar. Aging was of course not required for monetary balance sheet items, such as cash, receivables, and payables, because these are already by their nature stated in terms of the "current" dollar or current pur- chasing power. Remember that this study is concerned with only changes in the general price-level, and that the adjustments made will only in very rare cases reflect the "current cost" of balance sheet items. In other words, the distinction between changes in specific price-level and the general price-level is very basic, conceptually and the difference can be very material in actual application. Financial statements adjusted for changes in the general price-level are still on a historical (or invested) cost basis even though the unit of measurement has been changed. However, if historical cost were adjusted for changes in specific indexes, the current cost basis would be approximated. The purpose of general price-level adjustments is to reflect the financial statement items in terms of common or stable dollars; that is, dollars of the same general 35 purchasing power. 2.3 General price index used As recommended in Accountinijesearch Study NO. 6 and SAPB No. 3, the GNP Implicit Price Index was used to make the actual general price- level adjustments.6 This index is available on a quarterly basis since 1947 and on an annual basis since 1929. All financial statements were restated in terms of the ' 'current" dollar, which for this study was the dollar during the fourth quarter of 1967. The dollar of the fourth quar- ter of each year could have been selected, but to greatly simplify the computational manipulations involved, all financial data were restated in terms of one "current" dollar. Once this has been performed, to obtain, say, the December 31, 1964, balance sheet in terms of the 1964 fourth quarter dollar, all items in the December 31, 1964, balance sheet stated in "current" dollars, i.e. 1967 fourth quarter, would be multi- plied by a conversion factor. This conversion factor would consist of the GNP Implicit Price Index for the fourth quarter of 1964 divided by the comparable index number for the fourth quarter of 1967. It is important to note that the magnitude of the financial ratios being com- puted in this study will remain unchanged regardless of whether the two components of the ratio are 2223 stated in "current" dollars, 1964 fourth quarter dollars, or any other sized dollars. Accountinijesearch Study No. 6 proposed that a 1945 cut-off date be used in the aging of all nonmonetary balance sheet items, for 6Staff of the Accounting Research Division, Accounting Research §§udy No. 6, p. 111 and Accounting Principles Board, SAPB No. 3, p. 14. 36 the following reason: ...so many of the goods and services currently available resulted from wartime (World War II) and postwar technology, the precision of comparisons Of current price levels with those prevailing in periods prior to WOrld War II are un- reliable.7 Accordingly, fixed assets acquired or capital stock issued prior to 1945 were assigned the 1945 index in the adjustment process. The sections that follow outline in detail the methodology of general price-level adjustments for the specific financial statement accounts . 2.4 Current assets Cash and notes and accounts receivable, less the related allow- ance for doubtful notes and accounts, are clearly monetary items and were adjusted accordingly. Marketable securities were also classified as a monetary item and are discussed in the following sub-section. Flight equipment--expendable parts, miscellaneous material and supplies, and prepaid expenses are nonmonetary in nature, and sub-sections that follow outline the specific procedures used to adjust these accounts for changes in the general price-level. 2.4.1 Marketable securities SAPB No. 3 classifies marketable securities that represent invest- ments in stocks as a nonmonetary item.8 A problem arises, however, in 7Staff of the Accounting Research Division, Accounting Research Study No. 6, p. xii. 8Accounting Principles Board, SAPB No. 3, p. 26. 37 the classification of marketable securities that represent investments in bonds. If the bonds are "held for price speculation", then they should be treated as nonmonetary. On the other hand, "if the bonds are held primarily for the fixed income characteristic, they are monetary."9 According to.§§é§: there are three accounts that are classified as marketable securities; the account titles and descriptions follow:10 1. Special deposits. Record here funds or securities deposited with fiscal agents or others for payment Of current Obligations. 2. United States Government securities. Record here the cost of United States Government securities. 3. Other temporary cash investments. Record here the cost of securities and other collectible obligations acquired for the purpose of temporarily investing cash, other than those issued by the United States Government or associated companies. Marketable securities classified as special deposits, United States Government securities, and other temporary cash investments were treated as monetary items for several reasons. These investments are made primarily for their fixed income characteristics and clearly price speculation is not a dominant Objective. Other temporary cash invest- ments could conceivably include investments in common stocks which would dictate a nonmonetary classification. However, discussions with CAB field auditors and internal auditors at Trans World Airlines indicated that amounts so classified would be very small indeed. Therefore, all uMarketable securities were adjusted as monetary items. Table 1 illus- trates the relative size Of marketable securities to total assets. 91bid. 10Civil Aeronautics Board, USAR, p. 33. 38 .muuonom Hmann< uoauumo ufi< "ouusom 0.HH 0.0 m.m H.m 0.¢H ¢.HH m.ma 0.0 n.m~ H.mH 0.0 monHA Mud cumuwos m.m 0.0a H.5H w.qa 0.0 0.0 0.0 0.0 0.0 m.N H.0 mused uw< wouwns 0.0 m.0 n.~ n.q 0.5 0.m ~.0H n.n m.m 0.HH 0.n moswaufi< 0Huo3 manna m.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.0 0.0 «.0 moafiauw< unmanuuoz ¢.N 0.0 «.ma 0.m 0.0 0.0 0.0 0.0 H.N 0.0 0.0 moafiaufi< unmonuuoz q.¢ N.w 0.0 N.m H.H 0.N 0.0 0.0 0.0 m.m m.¢H omega HH< anoummm m.¢ w.m H.m ~.w m.0 H.0H m.0 0.0 0.0 0.0 0.0 mOfiHHufi< Hmuaonwuaou q.m 0.0 «.0 H.q 0.m 5.0 n.n 0.H m.m 5.5 ¢.¢H mhmauw< wmwamum 0.0a «.0H 0.mH m.NH N.MH n.0H 0.0a m.m~ 0.HH 0.mH n.0H mmnwflufim amowuma< and: Roma coma nomH coma moma NomH HomH 00mH mmmH mmmH muowuumo ufi< Hm Mmmzmumn H< mHmzH mmmfio m mammm ZOHH