PUBLISHED Q: IIIEIR. ADEQU , * DECISION MAKING I IV... . I .I I. 44 u'.’ I II II "I'fi I HI”... “I. ... is. for the Degree of Ph D. CY FOR INVESTMENT ”III III-I ..,..'I1' N... , . .II —-. I n "-h -. “Imp. r: ‘7 ,I... ,~ II III}I...,", -«~.,-... I. I ’4 I . -, ,2... mvaIII I I I“ I..- n .- t ... . II I: 0 W - w wan-34mm: I .I.;..I.,. W. I "w. I. -r .,.. . f "r ....-..VIII.'....:.I , ,-. my::,..!,.. “‘1 was". This is to certifg that the thesis entitled Published Quarterly Financial Data: Their Adequacy for Investment Decision Making presented by Gale E. Newell has been accepted towards fulfillment of the requirements for iidegree in Accounting Major professor Dam November 14, 1968 0-169 IL 5 BRAR Y Mick igan State University r H 551! rH £51! ABSTRACT PUBLISHED QUARTERLY FINANCIAL DATA: THEIR ADEQUACY FOR INVESTMENT DECISION MAKING BY Gale Earl Newell Firms listed with either the New York Stock Exchange or the American Stock Exchange are required by the respective exchange to make quarterly statements available to the public unless such data would be impractical to accumulate or would be misleading to the users° This study provided additional evidence to substantiate the belief that such quarterly data are used in investment decision making and therefore affect the market price of a firm's securitieso The primary purpose of this thesis is to evaluate the adequacy, for investment decision making purposes, of the published quarterly financial statements of firms listed on the American Stock Exchangeo This study reveals that firms report comparatively stable performance during the first three quarters of the year° However, the reported fourth quarter results, which represent the difference between audited annual figures and unaudited nine month figures, tend to vary significantly from the reported results of the other three quarters of the same year, This variation is especially evident in the quarterly Gale Earl Newell net income to net sales ratio and holds even when the dollar amount of sales is comparable in all four quarters of the year. In numerous instances originally reported quarterly net income was subsequently revised and this indicates that the originally reported quarterly data were recognized as being inaccurate. The deviation in the fourth quarter re- ported results from those of the first three quarters, as well as the numerous revisions of quarterly net income, implies that the first three quarter's data were misstated, and perhaps mis- leading, and indicates that reliance on quarterly net income figures in determining whether to buy or sell securities is precarious. The market price of a firm's securities depends to some extent on the firm's reported net income and it is gen— erally conceded that fluctuating earnings are not valued as highly as stable earnings, given the same expected future earnings. As management has more of an Opportunity to "manage” quarterly data than annual data this study investi- gated quarterly financial statements for evidence of "manage- ment" of reported quarterly net income. Although the lack of detail reported in quarterly statements places some limita- tions on this type of investigation there was no evidence to support the hypothesis that the tax rate applied is varied Gale Earl Newell between quarters for purposes of reporting "desirable" net in— come after taxes. The current quarter's income statement gen- erally presents the net income for the same quarter of the prior year to enable the reader to evaluate the current quar— ter's performance by comparing the two net income figures. This leaves Open the possibility that the quarterly net in— come figures are revised with the objective of improving the reader's evaluation of the current quarter's performance. This study provided no evidence that quarterly net income figures were revised with this objective in mind. While the accuracy of derived annual net income is limited due to the numerous acceptable accounting methods and the estimates necessary in preparing accounting statements, the quarterly net income is subjected to even greater limita— tions. These limitations result in part from the shortness of the time period being reported on, the lack of a quarterly audit, and the limited effort expended intheir preparation. The evidence discloses that reported quarterly net income has been unreliable in the past and there is no indication that it will be more reliable in the near future. Security analysts should be aware of the limitations inherent in re— ported quarterly net income and should consider these limita- tions when using quarterly data in making investment decisions. PUBLISHED QUARTERLY FINANCIAL DATA: THEIR ADEQUACY FOR INVESTMENT DECISION MAKING BY Gale E. Newell 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 ét 52¢ éhfi xi .9 . " ..' R44} \«L ‘~’ “ ‘ M 1968 ‘My' ‘5 ‘. ”xiii 77C0pyright by GALE EARL NEWELL 1969 ACKNOWLEDGMENTS I wish to express my sincere appreciation to Dr. Herbert E. Miller for his conscientious encouragement and guidance during this research project. His effort went far beyond the call of duty. I also wish to thank Dr. Alden C. Olson and Dr. Bernhard C. Lemke for their contributions to the successful completion of this study. The financial support and encouragement provided by Dr. James Don Edwards, Chairman of the Department of Accounting and Financial Administration, during my tenure at Michigan State University is gratefully acknowledged. The Haskins and Sells public accounting firm and the Michigan Accountancy Foundation have also provided financial assistance during this study and it is greatly appreciated. My thanks also is extended to the many faculty mem— bers and graduate students who added something to this thesis. My wife gathleen, and children Stephen and Daniel, have been whim very patient during this difficult period and I am in their debt. Kathy, who typed the numerous rough drafts, continued to be encouraging even through the more trying periods of this program. I am indebted to all the individuals and groups mentioned above for their help and advice. ii TABLE OF CONTENTS ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . LIST OF TABLES . . . . . . . . . . . . . . . . . . LIST OF FIGURES. . . . . . . . . . . . . . . . . . LIST OF APPENDICES . . . . . . . . . . . . . . . . Chapter I. II. IIIo IV. INTRODUCTION . . . . . . . . . . . . . . . Introduction. . . . . . . . . . . . . . The Problem . . . . . . . . . . . . . . Objectives of the Study . . . . . . . . Approach to be Used . . . . . . . . . . Limitations of this Study . . . . . . . Plan of the Thesis. . . . . . . . . . . THE ROLE OF QUARTERLY DATA IN INVESTMENT DECISION MAKING. . . . . . . . . . . . . Limitations of Quarterly Data . . . . . Reliability of Quarterly Data . . . . . Are Quarterly Reports Useful? . . . . . Are Quarterly Data Used?. . . . . . . . THE FINANCIAL ANALYSTS AND QUARTERLY DATA. How are Quarterly Data Used?. . . . . . The Purpose of the Questionnaire. . . . Summary . . . . . . . . . . . . . . . . FLUCTUATIONS IN QUARTERLY DATA . . . . . . Introduction. . . . . . . . . . . . . . How can Adequacy be Measured. . . . . . Allied Control Company, Inc. . . . . . iii Page ii vi Xi [—4 10 10 l3 l4 l6 l6 19 23 25 35 35 41 56 59 59 59 61 V. VI. H & B American Corporation. . . . . . . Roxbury Carpet Company. . . . . . . . . Huyck Corporation . . . . . . . . . . . Cubic Corporation . . . . . . . . . . . Great Basin Petroleum . . . . . . . . . Reasons for Fluctuations in Quarterly Data. . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . EMPIRICAL ANALYSIS . . . . . . . . . . . . Introduction. . . . . . . . . . . . . . Selection of Firms Studied and Source of Data. . . . . . . . . . . . Method Used to Investigate Quarterly Results . . . . . . . . . . . . . . . Results of Inquiry Using the Firms' Fiscal Years. . . . . . . . . . . . . Results of Inquiry Splitting the Firms' Fiscal Years. . . . . . . . . . . . . Results Adjusting for Seasonality . . . Summary . . . . . . . . . . . . . . . . Managing Quarterly Data . . . . . . . . Motives for Managing Net Income . . . . Federal Income Taxes and "Managing" Net Income. . . . . . . . . . . . . . Results of Test . . . . . . . . . . . . Analysis of First—Quarter Taxes . . . . Summary . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . THE REVISION OF QUARTERLY DATA . . . . . . Revising Quarterly Data . . . . . . . . The Cerro Corporation . . . . . . . . . Varo Incorporated . . . . . . . . . . . Are Quarterly Data Frequently Revised?. Predicting Annual Earnings. . . . . . . Revised Data and Prediction . . . . . . Results of Inquiry Regarding Revisions . . . . . . . . . . . . . Direction of Revisions. . . . . . . . . Conclusions . . . . . . . . . . . . . . iv Page 64 66 66 74 78 78 85 86 86 87 89 91 96 101 103 104 104 108 115 119 121 122 124 126 128 131 132 134 136 138 140 141 Page VII. CONCLUSIONS AND RECOMMENDATIONS. . . . . . . 143 Quarterly Data: Are they Used? . . . . . 143 Quarterly Data: Are They Useful? . . . . 145 The Effect of Unreliable Quarterly Data on Investment Decision Making. . . 149 Recommendations . . . . . . . . . . . . . 150 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . 174 Table 3-4 LIST OF TABLES Page Price Changes in Selected Stocks Following Releases of Quarterly Reports . . . . . . 27 What do you Believe the Primary Purpose of the Published Quarterly Income Statement Should be?. . . . . . . . . . . . . . . . 43 How Useful do you consider Published Quar- terly Reports to be in Your Analysis of the Investment Quality of a Firm's Securities? . . . . . . . . . . . . . . . 45 How Does the fact that Quarterly Reports Cover a Shorter Time Period than do Annual Reports affect their usefulness in Security Analysis. . . . . . . . . . . . . . . . . 46 How Does the fact that Quarterly Data are not audited affect Their Usefulness in Security Analysis. . . . . . . . . . . . . . . . . 47 How much Confidence do you place in the Accuracy of Quarterly Income Data published by firms that engage in a Substantial Amount of Merger Activity as compared with data Published by Firms not so classified? . . 48 What would be the effect on the Usefulness of Quarterly Reports for use in Your Analysis if Quarterly Balance Sheets, along with Income Statements, were made available on a Quarterly Basis? . . . . . 49 If you use published Quarterly Data in Your Security Analysis, do you have a system— atic Method for Analyzing These Data? . . 51 What is your Opinion Concerning the Usefulness (for use in your security analysis) of the vi following Reported Quarterly Information: (A) Depreciation EXpense; (B) Other In- come (Expense); (C) Provision for Federal Income Tax? . . . . . . . . . . . . . . . Allied Control Company, Inc.: Sum of Deviations of Quarterly N.I./N.S. Ratio from Line of Best Fit . . . . . . . . . . H & B American Corporation: Sum of Devia- tions of Quarterly N.I./N.S. Ratio from Line of Best Fit. . . . . . . . . . . . . Roxbury Carpet Company: Net Income/Net Sales Ratios. . . . . . . . . . . . . . . . . . Huyck Corporation: Net Income/Net Sales Ratios. . . . . . . . . . . . . . . . . . Huyck Corporation: Federal Income Tax/Net Income Before Taxes . . . . . . . . . Huyck Corporation: Sum of Deviations of Quarterly Tax/NIBT Ratios from Line of Best Fit . . . . . . . . . . . . . . . Huyck Corporation: Sum of Deviations of Quarterly "Other Income" from Line of Best Fit . . . . . . . . . . . . . . . Cubic Corporation: Net Income as a Percent of Net Sales. . . . . . . . . . . . . . . Cubic Corporation: Depreciation. . . . . . . Number of Firms in which the Various Quarters Experience the Greatest Deviation from "Normal" Using the Firm's Fiscal Year for calculating "Normal". . . . . . . . . Greatest Deviation from Normal . . . . . . . Number of Firms in which the Various Quar— ters EXperience the Greatest Deviation vii Page 54 62 64 66 70 70 72 74 76 76 92 94 from "Normal" Splitting the Firm's Fiscal Year when calculating "Normal". . . . . . Greatest Deviation from Normal . . . . . . . Number of Firms in which the Various Quar— ters Experience the Greatest Deviation from ”Normal" Splitting the Firms' Fiscal Year when calculating "Normal" and Eliminating Seasonal Data . . . . . . . . Greatest Deviation from Normal . . . . . . . Quarters in which the Net Income would have been Greater than for the same Quarter of the Prior year if the Annual Tax Rate had been Applied. . . . . . . . . . . . . Quarters in which the Net Income would have been less than for the same Quarter of the Prior Year if the Annual Tax Rate had been Applied. . . . . . . . . . . . . First Quarters in which the Net Income would have been Greater than for the same Quarter of the Prior Year if the Annual Tax Rate had been Applied . . . . . . . First Quarters in which the Net Income would have been less than for the same Quarter of the Prior Year if the Annual Tax Rate had been Applied. . . . . . . . . . . . . Cerro Corporation: Revisions of Published Quarterly Net Income. . . . . . . . . . . Varo Incorporated: Revisions of Published Quarterly Net Income. . . . . . . . . . . Data Leading to the Best Prediction of Annual Net Income. . . . . . . . . . . . . . . . Direction of Revision when the Current Quarter's Net Income is Less than the viii Page 98 99 101 102 116 117 119 120 129 132 137 6-5 6-6 Originally Reported Net Income for the same Quarter of the Prior Year. Net Income for the First Three Quarters which were subjected to Revision. Direction of Revision. ix 0 O Page 139 140 141 LIST OF FIGURES Quarterly N.I./N.S. Ratios: Allied Control. . . . . . . . . . Quarterly N.I./N.S. Ratios: H & B American Corporation. . . . Quarterly N.I./N.S. Ratios: Roxbury Carpet Corporation. . . . Quarterly N.I./N.S. Ratios: Huyck Corporation . . . . . . . . Quarterly Tax/NIBT Ratios: Huyck Corporation . . . . . . . . Quarterly ”Other Income”: Huyck Corporation . . . . . . . . Quarterly N.I./N.S. Ratios: Cubic Corporation . . . . . . . . Quarterly Depreciation EXpense: Cubic Corporation . . . . . . . . Splitting the Fiscal Year. . . . . . Criteria for Determining if Federal Income Tax Improves or Reduces Net Income After Taxes. . . . . . Page 63 65 67 69 71 73 75 77 97 113 Appendix A LIST OF APPENDICES Randomly Selected Firms. . . . . . . . . . Questionnaire. . . . . . . . . . . . . . . Calculations for Determining the Line of Best Fit — Least Squares Method — Allied Control Company, Inc. (N.I./N.S.) . . . Calculations for Determining the Line of Best Fit - Least Squares Method - H & B American Corporation (N.I./N.S.). . . . Calculations for Determining the Line of Best Fit — Least Squares Method - Roxbury Carpet Company (N.I./N.S.). . . . . . . Calculations for Determining the Line of Best Fit — Least Squares Method - Huyck Corporation (N.I./N.S.) . . . . . . . . Calculations for Determining the Line of Best Fit - Least Squares Method - Huyck Corporation (Tax/NIBT). . . . . . . . . Calculations for Determining the Line of Best Fit — Least Squares Method - Huyck Corporation (Other Income). . . . . . . Calculations for Determining the Line of Best Fit — Least Squares Method - Cubic corporation (N.IO/NOS.) O O O O O O O O Huyck Corporation and Subsidiary Companies: CONSOLIDATED STATEMENT OF EARNINGS. . . xi Page 159 163 166 167 168 169 170 171 172 173 CHAPTER I INTRODUCTION Introduction One of the major objectives of financial accounting is to prepare an income statement which reports the results of Operations for a past period of time. For numerous prac— tical reasons the computation of corporate profits cannot await the termination of the life of the organization. The shareholders and the managers of the business require periodic reports of the firm's performance and of course period mea- sures of income are needed for taxation purposes. The year has become accepted as the accounting period although it is realized, conceptually, that all measures of income for periods less than the life of the firm can only be approxi~ mations. Parties that are interested in financial statements include: stockholders, employees, customers, management, suppliers, tax authorities, regulatory agencies, creditors, and the general public. The published reports should provide the most useful information possible to these groups. In an article concerning postulates, principles, and research in accounting Gordon stated: . . . The primary role of the financial statements, the role peculiar to them, is to test the soundness of the general policies the owner has been follow— ing. If the statements indicate that his firm's performance has been good and its position is strong, the owner is reinforced in his policies. If the statements indicate that he is doing poorly, the owner questions his policies, searches for new ones, and the result is that his decisions and the future receipts of the firm are different than they otherwise would have been.1 The published financial reports are used in investment decision making and do affect the allocation of resources within our economy. Dyckman, when reviewing the investment decision, recognized the importance of published accounting information on the allocation of resources. He states: . . .Financial accounting, then must generate quanti— tative data which will assist in making investment decisions. And, further, these decisions should facilitate the orderly and intelligent Operation of the economic system including the maximization of the contribution of economic resources to society. But does financial reporting, as it is presently constituted, fulfill this objective? If, for example, a firm's securities are selling at a higher price than would be warranted by its economic situation due to financial re- porting techniques, new capital may be raised at lMyron J. Gordon, "Postulates, Principles, and Research in Accounting," The Accountinngeview, XXXIX, No. 2 (April, 1964) 257. a relatively low price, and through misinforma— tion resources are misallocated.l This suggests that one of the major functions of the financial report is to provide information which will enable investors to allocate their resources to the most desirable alternative(s). As investors use information other than that provided in the financial statements in making investment decisions, it is difficult to determine the precise effect that accounting information has on these decisions. Most decisions to purchase securities are based upon expectations of future performance and the soundness of such decisions are evaluated based on future performance. ”The future, unfortu— nately, is uncertain and cannot be analyzed: future events can only be anticipated or predicted as a result of an analy— sis performed on the past." The interim statements, of concern in this study, re- port on the activities of the firm for periods less than a year; they attempt to fill the financial information void which would exist if only annual data were published. A 1Thomas R. Dyckman, "On the Investment Decision,” The Accounting Review, XXXIX, No. 2 (April, 1964) 285. 2Robert Eugene Jensen, ”A Study of Effects of Alter— nate Accounting Systems on Security Analysis and Portfolio Selection Decisions” (unpublished Ph.D. dissertation, Stanford University, 1966), p. l. primary purpose of published quarterly data is to provide information about the activities of a firm for a particular period. The published quarterly report, as well as the annual report, is used by the investor, the potential investor, or their representatives in making investment decisions and, therefore, these quarterly statements do affect the alloca- tion of resources. Interim reports have been prepared for many years although their reliability has been questioned. For example, Gilman wrote, in 1939, ”It is of course, true that most busi- ness organizations do prepare monthly or at least quarterly reports, but these are generally regarded as Operating reports for the guidance of Operating men and are seldom considered sufficiently reliable in certain respects for general pub- lication."2 Despite the questionable reliability of interim re- ports, the New York Stock Exchange has been advocating, since 1910, the publication of interim data for firms listed on the exchange. By 1939 approximately 90 per cent of the firms 1The effect that published quarterly data can have on the price of securities is illustrated by L. J. Seidler and W. D. Benjes in "The Credibility Gap in Interim Financial Statements,” Financial Analysts Journal, Vol. 23, No. 5 (September—October, 1967), 109-15. 2Stephen Gilman, Accounting Concepts of Profit (New York: Ronald Press Company, 1939), p. 77. listed on the New York Stock Exchange were issuing quarterly reports; by 1967 this figure had reached 98 per cent.1 The practice of issuing quarterly statements became so well accepted that when, in 1962, the American Stock Exchange re- vised its listing requirements to include such statements, 60 per cent of its listed firms were already publishing them. Neither the New York Stock Exchange nor the American Stock Exchange require that the interim statements be sent to the stockholders. They do require, however, that the interim statements be published in the public press through newspapers of general circulation in New York and to the news-wire ser— vices including Dow Jones & Company, Inc. They also require that interim statements be filed with the stock exchange with which they are listed.2 The requirements of the New York Stock Exchange, and the American Stock Exchange, are adhered to by most of the firms listed on the exchanges. Listed firms are exempted from reporting quarterly data only if the publication of such data would be impractical or misleading. 1Letter from Morton B. Solomon, Executive Assistant, New York Stock Exchange, December 28, 1967. 2See American Stock Exchange's, "Listing Form L" (revised September 15, 1966) p. 6; and New York Stock Exchange's, New York Stock Exchange—Company Manual," p. A70— A71. The Securities and Exchange Commission (S.E.C.) also requires that interim statements be filed for firms under its jurisdiction. In 1946 the S.E.C. began requiring quarterly sales data for firms subject to S.E.C. requirements but, due to Opposition, it rescinded the requirement on October 9, 1953. Robert E. Taylor's study of this change in the S.E.C.'s requirements indicates that organized pressure was applied by members of the Financial Analysts Federation in Opposition to the rescinding of the requirement. The analysts contended that the publication of financial data on an annual basis was too infrequent for them in their investment decision making. As a result of this pressure the S.E.C. reinstated regulations requiring the filing of interim reports, and on June 23, 1955 they began requiring semiannual, unaudited in— come statements on form 9-K.2 Form 9—K provides for the reporting of the following data: 1. Gross sales less discounts, returns and allowances 2. Operating revenues 3. Total of 1 & 2 1Robert E. Taylor, "An Examination of the Evolution, Content, Utility, and Problems of Published Interim Reports," (unpublished Ph.D. dissertation, University of Chicago, 1963) pp. 31—38. 2Securities Exchange Act of 1934, Release NO. 5189, June 23, 1955, p. 1. 4. Extraordinary items. 5. Net income or loss before taxes on income. 6. Provisions for taxes on income. 7. Net income or loss 8. Special items. 9. Earned surplus items. The Problem: The view is widely held that reported interim earnings . . . 2 do affect investment deCiSions, however, the usefulness, or reliability, of these data for investment decision making is questioned by both analysts and accountants.3 Problem areas lLouis H. Rappaport, SEC: Accounting Practice and Procedure (New York: Ronald Press Company, 1963), p. 13-18. 2For example see: Albert J. Bows, "Standards for 1Consistency in Interim Reports," Financial Executive, XXXIV, NO. 11 (November, 1966), 24; Robert Doane Neubig, "A Study of the Effect of Periodicity on Certain Interim Corporate Re- porting Practices," (unpublished Ph.D. dissertation, The Ohio State University, 1961), p. 24; Seidler and Benjes, ”Credibility Gap," p. 109. 3For views of selected critics of interim reports see: Cohen, Manuel F., ”The SEC and Accountants: Co-Operative Efforts to Improve Financial Reporting," from an address given before the 79th annual meeting of the American Institute of Certified Public Accountants, The Journal of Accountancy, Vol. 122, No. 6 (December, 1966) 56—60; Neubig, "Effect of Periodicity” p. 10; Welsch, "Discussion of the Predictive Power" pp. 41-43; Carman G. Blough (editor) "Some of the Dangers Inherent in Quarterly Financial Statements," The Journal of Accountancy, Vol. 95, No. 2 (February, 1953) 222. encountered in deriving quarterly income include: the allo- cation of annual fixed costs to the quarters; valuing the inventory at the end of the quarter; the allocation of wind- falls or miscellaneous revenues to the quarters; and how to account for needed adjustments which are discovered only at year end. The seasonal nature of some firm's activities has also been cited as a factor which limits the usefulness of the quarterly statements of these firms, although the possible ". . .misleading effect of such fluctuations can Often be overcome by providing adequate comparison data for prior periods."l As quarterly statements are unaudited, there is Opportunity for the firm to present data which were not de- rived as a result of following principles acceptable to an independent public accountant. Because they are unaudited there is more of an Opportunity for management to ”manage" reported quarterly income than there is for them to "manage" reported annual income. If firms take advantage of the absence of an independent audit it may be that reported quarterly income is "managed” by the use of questionable accounting procedures. Manipulating the cut-off dates of 1Frank S. Capon, "The Need for Interim Statements,” The Canadian Chartered Accountant, Vol. 67, No. 4 (October, 1955) 279. revenues and/or expenses is one method of "managing" reported performance. There is little doubt that if such "management" occurs, and if these statements were audited some, although not all of these discrepancies could be reduced. The cost, however, seems to prohibit a quarterly audit as a prescriptive device and there appears to be no movement afoot advocating quar— terly audits. There is disagreement concerning whether publicly re- ported quarterly data can be useful in making investment deci— sions. Although some hold the view that quarterly data are reliable, others contend that the derivation of quarterly in— come is subject to such a high degree of uncertainty due to the limited time period covered, the lack of effort expended in their preparation, the allocation procedure and the fact that they are unaudited, that they are of limited usefulness and may even be misleading. The Opinion is widely held that quarterly data are relied upon in making investment decisions, even though the reliability of such data is Openly questioned and the problems inherent in the preparation of such data are well known. If interim data are not reliable and if such data 1 See pages 25-29 in Chapter II and pages 37-38 in Chapter III. 10 are used in making investment decisions, individuals may make ill-advised investments and their resources may be mis— appropriated. Objectives Of the Study The objective of this study is to evaluate published quarterly statements as one of the principal financial reports issued by a firm's management. The more specific objectives are (1) to investigate the quarterly information needs of the analysts in their security analysis in order to evaluate cur- rent quarterly reports; (2) to investigate the quality of quarterly data as reported to the public and to review the quality of disclosure in such data in order to evaluate their adequacy for investment decision making; (3) to investigate the revision of reported quarterly data in order to evaluate the reliability of such data as initially reported and to isolate the areas that are prone to such revision; and (4) to investigate the problems encountered in preparing quarterly data for publication in order that more meaningful suggestions can be made for improving such data. Approach to be Used This study attempts to determine how financial analysts use quarterly data and to discover their needs concerning 11 quarterly information. Research also includes an examination of published quarterly reports for a sample of firms for the period 1960 to 1968. This examination will be used for evalu- ating the adequacy of quarterly reports as well as for mea— suring the frequency and extent of the revision of data in- cluded in these reports. The approach used will be: 1. To inquire into the interim information needs of security analysts; and to review their methods or techniques for using quarterly data. A questionnaire, sent to randomly selected Chartered Financial Analysts, is used to accomplish this. The questionnaire is employed in an attempt (1) to determine how important the analysts consider the quarterly report to be in their security analysis; (2) to gain some insight into the needs of analysts concerning such reports; (3) to specify how they use such reports; and (4) to Obtain suggestions for increasing the utility of quarterly data to the analysts. 2. To investigate the characteristics of selected publicly reported quarterly data to gain some basis for the evaluation of quarterly reports. This investigation has been undertaken to determine (1) if the fourth quarter Often pre— sents a significantly different outlook than do the other three quarters of the year; and (2) if the relationship 12 between selected income statement items show substantial, unexpected, variations over the quarters of the year. In resolving the above, the relationship between re- ported net income and sales has been reviewed for the various quarters to determine if there is a substantial fluctuation between quarters and if a pattern exists. In cases where substantial fluctuations do exist an attempt has been made to determine the causes of such fluctuations. Data collected in the investigation are used to test if the reported Federal Income Tax/Net Income Before Taxes ratio is affected by discrepancies between "desired" and actual results. The "sign test" is applied to the data for this purpose. 3. To investigate the practice of revising quarterly data. This investigation has been undertaken (1) to determine if quarterly reports, as initially reported, are frequently revised; (2) to determine if initially reported quarterly data are generally overstated or understated; and (3) to pro- vide data for testing if the direction of the revision of prior quarters reported income is affected by the performance of the firm for the quarter under report. 13 Limitations of this Study: This study will be limited in that no attempt has been made to determine the precise effect of quarterly information or of certain quarterly reporting practices on investment decisions. It is assumed that reported quarterly data do affect investment decisions and this study attempts to evalu- ate the adequacy of these data. This study cannot be considered to be an all-inclusive treatise concerning the adequacy of published quarterly finan- cial data for investment decision making. Many of the spe— cific items that affect quarterly data are buried in aggregate classifications and are not of sufficient detail to permit adequate analysis. For this reason the researcher has been restricted to studying those items that are reported in some detail in the quarterly statements even though they may not be the only problem areas. This study indicates the effect on quarterly data that certain reporting techniques have had, as well as, in certain cases, the reasons for, or causes of, unexpected results. In isolating the causes of, or reasons for, unexpected relation— ships, etc., in quarterly reports, direct correspondence with the management of the firm under consideration was required. However, not all firms have responded to such inquiries and 14 because of this unwillingness to discuss the reasons or causes for such unexpected relationships the inquiry is not all- inclusive. In this study the revision of quarterly data was used as one indicator of the overstatement or understatement of initially reported quarterly net income. However, if firms failed to revise quarterly data, when subsequent evidence indicated that the initially reported net income was inaccurate, there was no way of determining if the originally reported quarterly data were misstated. Plan of the Thesis: Chapter II will provide background information con- cerning the role that quarterly data play in investment deci— sion making. It will present the limitations of quarterly data, Opinions of the critics and advocates concerning the usefulness of quarterly information, and will provide examples indicating the effect that reported quarterly earnings has had on the prices of specific firm's securities. In Chapter III the manner in which quarterly reports are used will be presented. It will present a summary of the responses to a questionnaire for purposes of establishing the needs and views of Chartered Financial Analysts concerning published quarterly information. 15 Chapters IV and V will present the results of an investigation concerning the reliability of published quar- terly data and Chapter VI will investigate the revision of such data. Specific cases revealed by the investigation will be presented as will the results Of tests used to determine if quarterly data are ”managed.” Chapter VII will summarize the findings of the research and will give recommendations concerning the reporting and the using of quarterly data. CHAPTER II THE ROLE OF QUARTERLY DATA IN INVESTMENT DECISION MAKING Limitations of Quarterly Data: Quarterly reports contain financial data as well as other information concerning the activities and performance of the reporting firm for a given period. These data provide the management of the firm with an Opportunity to inform interested parties of the progress, or lack of it, that has been made over the period in question and to provide them with other information considered relevant to their needs. The reader uses such data as his needs dictate, subject to his Opinion concerning their reliability and usefulness to him. The members of the accounting profession, and it is hOped the users of accounting data, have long recognized that any reporting of corporate profits before the termination of the life of the organization is an approximation.1 However, 1For example see: John B. Canning, The Economics of Accounting (New York: Ronald Press Company, 1929), p. l; Gilman, Accounting Concepts of Profits; American Institute of Certified Public Accountants, Accounting Research Bulletin No. 43 (New York, 1929) p. 59; Morton Backer (editor), Modern Accounting Theory, "The Measurement of Business Income Part I — The Matching Concept," (Englewood Cliffs, New Jersey, Prentice- Hall, Inc., 1966), pp. 41—71. 16 17 even with this limitation on the accuracy of annual data, the annual report has continued to be widely used by investors, potential investors, and their representatives. Quarterly data attempt to report the firm's performance for a much shorter period of time and therefore are subject to even greater limitations. Although a physical inventory is taken on an annual basis, management may consider it impractical to do so quar- terly. Another major problem encountered is that of allo- cating annual fixed costs (which may or may not be known for the year at the time the interim statement is prepared) to the quarter in question. Included among the costs which can be reasonably approximated for the year but which should be allocated to quarters include: advertising costs, product development costs, charitable contributions, and prOperty taxes. Other annual costs which are more difficult to esti— mate at the time the quarterly statements are prepared but should be allocated to the quarters for a prOper matching of costs with revenues include: bonuses, maintenance and repair expenses, pension accruals, and legal fees. These require that the accountant estimate not only the annual costs but also allocate them to quarters on an apprOpriate basis. If the yearly estimate is in error then the costs allocated to 18 the interim periods may result in misleading information being reported. Some of the problems may be due to the lack of atten— tion that has been given to the preparation and presentation of quarterly data by the accounting profession. Seidler and Benjes state, in commenting on the problems in quarterly data and the lack of precedence in judging practices according to some criteria, ". . . We do not really know what is or is not acceptable alternatives for the simple reason that organized accounting bodies have given little or no thought to their (quarterly statements) problems. There are no official pro— 1 nouncements, virtually no literature and hence no guidelines. . . The quarterly data are presented to the public without being subjected to an independent audit. The firm may, in preparing the quarterly data for publication, be unaware of, or disregard, adjustments that an independent audit would reveal. Such needed adjustments may go unrecorded until re- quired by the auditor performing the annual audit. If these adjustments are applicable to the Operations of the first three quarters, it would indicate that inaccurate quarterly data have been released to the public. Thus, making these adjustments only at year—end is insufficient for reliable 1 Seidler and Benjes, "Credibility Gap,” p. 112. 19 quarterly reporting. Goldstein states, "picking up these adjustments at the close of the calendar or fiscal year does not eliminate the damage that may have been done during the interim period. Omitting them may result in considerable embarrassment to the accountant."l Adjustments, which are inapprOpriately postponed until year—end, could affect the decisions of many investors and potential investors. Not only those who purchased the stock based on the analysis of quar— terly data but also those who sold the stock based on the analysis of such data. Reliability of Quarterly Data Norr states, ". . . Analysts are not accountants; They assume that the data included in financial statements are valid. It has been . . . a shock for many of them to discover that the mere choice of accounting methods makes for signifi- 2 The reliability of cant variations in reported earnings.” selected firm's quarterly data has been questioned, however, by financial analysts. Hal Chefitz, Vice President of Hayden l . . Joseph Goldstein, "Interim Statements on a True Accrual Basis," The New York Certified Public Accountant, XXXIV, No. 7 (July, 1953) 429. 2David Norr, "Investment Analyst's views of Financial Reporting," Financial Executive, Vol. XXXIV, No. 12 (December, 1966) 22. 20 Stone, made the following comment concerning the Syntex Corporation: I'm concerned, however, about earnings. Syntex may have borrowed from this years' first quarter to make last years' $2.04. In fiscal 1966, they showed $2.01 after saying all year they would do $2. I think they made their $2.01 by borrowing from the first quarter. If you re- call, last years' first quarter was very poor. This year, I think they may have borrowed from the first quarter to show their $2.04. Year— to-year comparisons may be favorable, because they borrowed in both years. . .1 While he was not specific concerning how they had borrowed from other quarters, he thought that they had bor- rowed, recognized that it could be done, and therefore ques— tioned the reliability of the data. The tendency for some firms to report imposing fourth quarter adjustments has not been unnoticed. The following was reported in Barrons, July 14, 1968: Frankly, we don't know what to make of Sperry Rand. Judging by the recent action of the stock, a lot of other peOple don't either. The Company it might be recalled ran into heavy selling in early April, when rumors circulated that it would have to take a big write-off in its fourth fiscal quarter (ending March) which would drOp full-year earnings well below street estimates. What made holders particularly nervous at the time was that Sperry, too Often in the past, seemed to be rolling along toward a good profit showing, only to run 1Interview with Hal Chefitz, Vice President, Hayden Stone, "No Wonder Drugs," Barrons, November 20, 1967, p. 10. 21 into some massive year-end adjustments. Well, as it turned out, there were no charges and the firm made good its estimate. However, earlier this month, the company painfully disclosed that the reports were pre- mature rather than unfounded; specifically that it would have to take a $6.5 million special write-off against earnings for the first three months of the new fiscal year (representing term- ination of a navigational system contract with Pan American). . .1 This is evidence of analysts' skepticism regarding the reliability of the quarterly data of this firm. It indicates the effect that adjustments can have on both quarterly and annual data. The effect of such year—end adjustments on fourth quarter data has been so extreme in certain cases that it has resulted in quarterly profits being greater than quarterly sales. The restatement of prior quarterly data has also been noticed by interested parties. Barrons, on August 14, 1967 reported the following concerning comparison data: Super Sleuth vs. Seagrave Corp., Chapter II. Two weeks ago, we probed the mystery of product X via the eyes and ears of Peter DuBois. Well, last week Seagrave reported second quarter net income - and at a first glance, anyway, it was sharply higher than a year ago. At second glance however, it wasn't. As Peter points out, Operating earnings actually dropped, to $527,577 from $770,892. How— ever, by restating the June, 1966 quarter to include lAlan Abelson,”Up and Down Wall Street," Barrons, July 24, 1967, p. 21. 22 for the first time a nonrecurring charge, Seagrave was able to show a favorable year—to-year comparison. The fact that quarterly data are not subject to audit has not escaped those using such data nor has the possible effect of this on the quality of the quarterly data. This is evidenced by a quote in a recent article by Harris concerning the fact that normally the C.P.A. has no responsibility toward published interim reports. He writes: ”. . . It is generally taken for granted that an interim audit sufficient to express an Opinion is not practical because of time and costs re— straints. This means that corporate financial officers, almost without restraint, can fashion earnings patterns during a 2 year." Several writers have commented on the use of account- ing techniques for smoothing annual income.3 They believe that this is advantageous to the firm and is possible on an annual basis. Because of the greater flexibility that is accepted in reporting quarterly data the opportunity for lAlan Abelson, "Up and Down Wall Street," Barrons, August 14, 1967, p. 23. 2Nelson G. Harris, "Accounting Practices — Impact on Preparation of Annual Reports," Financial Executive, XXXIV, No. 9 (April, 1966) 66, quoted from a comment by the finan— cial editor of a Philadelphia newspaper. 3For example see: Gordon, "Postulates, Principles and Research in Accounting,”; and Samuel R. Hepworth, "Smoothing Periodic Income," The Accounting Review, XXVII, No. l (January,l953). 23 smoothing is much greater in the case of reported quarterly income than in reported annual income. Holton states: ". . . I think the facts are that quarterly earnings are stabilized by accounting techniques much more than annual earnings. . ."1 If quarterly earnings are, in fact, stabilized the user should be made aware of this in order that he can conduct his analysis accordingly. Are Qparterly Reports Useful? Some writers hold the view that with today's sophis- ticated accounting and communication systems, reliable quar— terly data can be made available to investors to enable them to make sound investment decisions.2 Others question the usefulness of such data in making investment decisions. Carman Blough, in referring to the uncertainties of quarterly data, states: "Accountants seriously question whether state— ments that are subject to all of these inherent weaknesses would be very useful to anyone attempting to make investment judgments upon the basis of them. They are much more likely 1Thomas L. Holton, Discussion of the Predictive Power of First—Quarter Earnings Reports: A Replication," Empirical Research in Accounting: Selected Studies 1966, p. 38. 2C. B. Matheson, "Interim Reporting," The Canadian Chartered Accountant, Vol. 87, No. 5 (November, 1965), 421. 24 1 to be misleading in many circumstances. . . Seidler and Benjes conclude: . . . with no control by the auditors, it becomes easier to justify a 'quick and dirty' job on the interim statements and to defer the work on knotty adjustments to the year end. . . . . . certainly there is substantial differ— ence between the quality of information in the annual report and in interim statements.2 Morton Backer, who has conducted research in the area, states: . . . I found many corporate executives admitting that there is a tendency to 'flavor' quarterly earnings reports which, of course, are unaudited. I also found many security analysts who are dis— turbed about last—quarter adjustments that tend to be abnormal. . . If the market reacts to quar— terly-reports as it does, and management has com— plete discretion without independent audit to 'flavor' the results, a disturbing situation obviously exists. The aforementioned indicates that many users question the usefulness of quarterly data in investment decision making. They hold that they are of questionable reliability and may be misleading. l Carman G. Blough (editor) "Some of the Dangers Inherent in Quarterly Financial Statements," The Journal of Accountancy, Vol. 95, No. 2 (February, 1953) 222. 2Seidler and Benjes, "Credibility Gap," pp. 113—114. 3Letter from Morton Backer, Professor of Accounting, University of Massachusetts, dated February 2, 1968. 25 AreyQuarterly Data Used? If quarterly data are subject to the high degree of uncertainty, as implied by the prior discussion, one might question whether such data are used in investment decision making. If their limitations, due to the shortness of the time period covered, the lack of effort expended in their preparation, the vagueness of the allocation procedure and the fact that they are unaudited, are as severe as indicated it is apprOpriate to question whether quarterly data play a significant role in investors' decisions. Welsch, questioned whether quarterly reports are not just a publicity gimmick to be mailed with a proxy statement or a somewhat useless document which is issued only to satisfy certain pressures or to avoid certain criticisms.l It does not appear that users consider the quarterly report as such. The View is widely held that reports of interim earnings do affect investment decisions. Bows commented on the importance of quarterly reports as follows: One of the most significant figures released by a corporate financial executive is the interim report of quarterly earnings. It is used first of all to determine the company's basic earnings l . . . . Glenn A. Welsch, "Discuss10n of the Predictive Power of First-Quarter Earnings Reports," p. 41. 26 trend. It is then frequently summarized and compiled with similar reports from other com- panies in the same industry and used by in— vestors to determine whether there are any significant profit trends in the industry which might change their investment policy. Finally, many of these quarterly earnings re— ports are carefully studied by analysts, econ— omists, and government officials to determine the vigor of our economy. The wide interest in earnings indicates a significant amount of securities traded is based on the reports of quarterly earnings. . . Finally, quarterly earnings reports should be regarded as important as annual reports. It is quite clear that the volume of security trading which takes place based on quarterly reports may be as great as that which results from annual reports. It may even be desirable at some later date for the independent public accountant not only to certify the annual figures but to review the reporting of the quarterly earnings as a check on the fairness of the reported results. Neubig holds a similar view: This interim information is Often given as the reason for fluctuations in the stock market. Specific issues of stock rise or fall on the stock market as the result of news that quarterly earn- ings just reported were higher or lower than had been expected. Current economic predictions are Often confirmed or denied by the interim reports of corporations.2 Table 2—1 indicates that reported quarterly earnings have, in the past, affected the market price of certain firm's lBows, "Standards for Consistency in Interim Reports,” p. 24. 2Neubig, "Effect of Periodicity on Certain Interim Corporate Reporting Practices," p. 9. 27 .ONN .m ..nmo muannnaemno= .mmflamm eat nmNeNmm "mousomr .musflom m mo mono nonpocm How HHH um pmmoHo paw mHH um oocomo Moonm onu .mmonm may wfi> OODMGNEOmmNO soon on: wzoc onu OOSB hop axon mnu GO .mmw mo mmoH o How mHH um pmmoHo paw mwa um pmcomo Mooum OSD .co>flm mm3 nomomm onu SUHSB so how OSD .hm HOQEODmmm GO .Mamu we: m>mm unopflmmnm on“ £0fl£3 so moo map SDHB msflccflmmn OOHHOQ wmpIOBD m wOOSHUGH cofluom Mooum mo omsmn OSB .mum%HMQ¢ moansoom Mo %DONoOm MHON 3oz onu mo mcfluooe m on MHOHODOS mo usopflmonm >9 co>flm mmeflumo mumsfiaflaonm N .hnmmfioo mo pampflmmum kn Co>flm oumfiflumo hnmsflfifiaouma Ase.mv N\N e..N\H NoH:N\N «NH mN. m> ea. 66. o eaN ee\N\m xmnasm Ass.me «\m N I a\N eeuma we. m> Ne. 66. o ecN ee\mN\e manmmm Lae.ev «\m H : m\m mNum\H NN No.4 m> no. we. a cum ee\eN\m acmaanmm mmmmmz Ase.me m\e N . N\N omuNm we. m> No. me. o ecN me\e\m munmsmm Asm.NNV Hm- HHNuNeH ON.H m> oa.a 66. 0 sum Nee\eN\m mNonouoz Axe.NHV N\H m . e\m NNum\m NN so. m> em. 66. 0 arm Nee\HN\a wpoaeoua mam>m xa.ea «\m NN msanmea mm. m> mN.Hm 66. a ran me\a\Na xmnesm ODHm> pox mom mmoHOIcomo om¢ Moo? m> poflnom HMOHSOO mammEOU nan: mo K msofl>onm mmsflsumm mcflpnomom .Dm Hang “mum mmoq Eoum mmoq pouuomom OH unomom no same no CHMO mo ovum mBmOmmm MQmMBmfibo m0 mmmflmqmm OZHBOQAOM mMUOBm QmBUmqmm ZH mm02¢m0 MOHmm «HIN OHQMB 28 securities. A more recent incident indicating the effect that quarterly data have on stock prices involves Litton Industries. When it was announced that the second quarter results would be substantially lower than expected the price of the stock fell significantly. The effect of the substan— tially lower than expected earnings and the resulting effect on the price of the stock was noted in Barrons by the following: . . . Response to that bombshell was a flood of selling which totaling 304,400 shares, made it the most active stock Tuesday. That day it lost 1 nearly $6 and on Thursday it drOpped $5 more to 72. . . Assuming that quarterly reports are used, do the users make allowances for the limitations inherent in them? This is not an easy question to answer. Biggs believes that users fail to make adjustments to reported net income and states: ”. . . unfortunately, investors and most analysts work almost exclusively with earnings as they are reported, making little distinction as to the differences in the quality of these earnings. The obsession is with reported earnings, not what lies behind them. . .2 Norr, as reported earlier in this chapter, also holds this view as does Williams. Williams noted that reported net 1H. J. Nelson, "The Trader," Barrons, January 29, 1968, p. 4. 2Barton M. Biggs, ”Numbers Game," Barrons, July 24, 1967, p. 3. 29 income is the "universal yardstick" for appraising the value of a firm's securities and that it is trusted by most in— vestors as being an accurate measure of a company's perform- ance. He stated: . . . A large segment of the investing public, particularly during periods of rising stock prices, is seldom aware of, or interested in, the accounting policies which produce the earnings figures carried in the financial press. Accordingly, in cases where earnings are overstated, for whatever reason, there is a potential for investor loss. The above indicates that quarterly data are used in investment decision making and that the reported earnings figure is very significant in the analysis of such data. If quarterly data are used in security analysis, in what way are they used? Investors, potential investors or their representa- tives are not generally interested in reported earnings be— cause of their accurate reporting of what has happened for a prior period of time. They, instead, are interested in using such data as a basis for projecting future performance. In regard to the use made of published financial data the Com— mittee to Prepare a Statement of Basic Accounting Theory of the American Accounting Association states: Almost all external uses of financial infor- mation reported by a profit—oriented firm are 1Williams, ”A Look Behind Reported Earnings," p. 40. IIII-----------_______________________________________________nllllllllllll1L‘J 30 involved in efforts to predict the earnings of the firm for some future period. Such predic— tions are most crucial in the case of present and prospective equity investors and their repre— sentatives - considered by many to be the most important of the user groups. . . The past earn- ings of the firm are considered to be the most important single item of information relevant to the prediction of future earnings. . . . . . It is important to emphasize that ac- countants (with good justification) have avoided the role of forecasters in connection with reports to external users. The committee suggests that accounting information for external users should reflect their needs by reporting measurements and formulations thought to be relevant in the making of forecasts without implying that the information is wholly adequate for such prediction.1 The View that quarterly data are used for purposes of predicting future results is commonly accepted.2 It also seems clear from the nature of the intended reader that finan- cial statements are intended to provide information that is helpful in making rational investment decisions. If quarterly data are used for the above purposes what should the data in- clude or wnat should be their objective(s)? l . . . Committee to Prepare a Statement of BaSlC Accounting Theory, A Statement of Basic Accountinngheory, Evanston, Illinois, American Accounting Association, 1966, pp. 23—24. 2For example see: Gordon Shillinglaw, "Concepts Underlying Interim Financial Statements," The Accounting Review, XXXVI, NO. 2 (April, 1961) 222; Howard I. Ross, "The Current Crisis In Financial Reporting,” The Canadian Chartered Accountant, Vol. 67, No. 4 (October, 1955) 66; Benjamin Graham, David L. Dodd, and Sidney Cottle, Security Analysis: Principles and Techniques (New York: McGraw Hill Book Company, 1963) 106-107. 31 If financial statements are used by investors as a basis for predicting future performances it appears that the interim report should aid in predicting annual performance. Green and Segall,faculty members at the University of Chicago, attempted to evaluate the ability of quarterly data to improve predictions of annual performance. More specifically, they attempted to research the predictive power of first quarter earnings reports. They employed several different mOGels, which utilize quarterly data, in attempting to predict the annual earnings per share. They concluded, as a result of the study, "that first quarter reports as presently prepared, are of little help in forecasting annual earnings per share."1 In effect they said that the availability of first-quarter data, as well as prior quarterly data, did not improve the accuracy of the user's prediction of annual income over what it was if only annual data were available. It is essential that the adequacy of quarterly data be evaluated based on the needs of the users of the informa- tion. It is very difficult, if not impossible, to determine the needs of all the users of quarterly data. Therefore, it is necessary that such data attempt to fill as many needs as lDavid Green, Jr., and Joel Segall,"The Predictive Power of First—Quarter Earnings Reports," The Journal of Business, Vol. 40, No. 1 (January, 1967) 55. 32 is practical. Stone believes that "the most effective and efficient report will be achieved when the dominant group is singled out as the focal point of the report.”1 He also says, "specifically, the purpose Of the published financial state— ments is to assist the investor in making his own qualitative judgements about the firm."2 It does not seem unreasonable to assume that, if the purpose of the reports is to help the investor in his decision making, the user must be competent to understand the data presented. Therefore, we can assume that such statements should be prepared for the investor who has a working knowledge of business methods and terminology.3 An inquiry into individual investor needs would be impractical if not impossible and therefore an alternative must be used. As the security analyst Often acts as the repre— sentative of the investor, it seems that if quarterly reports can satisfy the analysts' needs most of the needs of the in— vestors will also be satisfied. In this paper the attempt 1Donald E. Stone, "The Objective of Financial Re— porting in the Annual Report," The Accounting Review, XLII, No. 2 (April, 1967) 333. 2Stone, ”Objectives of Financial Reporting,” p. 335. 3Committee on Concepts and Standards Underlying Corporate Financial Statements, "Standards of Disclosure for Published Financial Reports," The Accounting Review, XXX, NO. 3 (July, 1955) 400. 33 will be to evaluate quarterly data using analyst's needs as a guide. The overall Objective of financial information should be to direct resources into the most profitable firms. It should aid in allocating resources to those firms which will be most beneficial to society. Paton says: . . . Accounting, by making price data avail- able in a systematic and intelligent manner, is perhaps the principal instrument by which the com- plex data of the market, as they attach to the par- ticipating business, are translated into effective managerial criteria. It is the function of ac- counting to record values, classify values, and to organize and present value data in such a fashion that the owners and their representatives may utilize wisely the capital at their disposal. The accountant has not published forecasts of ex— pected performance even though some claim that the accountant is in the very best position to adequately forecast future performance.2 For example, Shillinglaw says, ”Relaxing the present constraints against the reflection of anticipation in financial statements may increase the danger of manipula— tion but this should not be overemphasized.”3 It is ques- tionable, however, if the analysts want this and questionable lWilliam Andrew Paton, Accounting Theory (Chicago: Accounting Studies Press, Ltd., 1962) 7. 2Ross, I'Current Crisis," p. 68. 3Shillinglaw, "Concepts Underlying Interim Financial Statements," p. 225. 34 if the accounting profession is prepared to assume such a responsibility. This paper Will use the needs of the analysts as a basis for evaluating quarterly data. The following chapter will consider how the analysts use such data and will attempt to give some insight into their needs. L. - CHAPTER III THE FINANCIAL ANALYSTS AND QUARTERLY DATA How Are Quarterly Data Used? The users of published corporate financial data may represent quite different interests. In addition to inves- tors, these users include, among others: creditors (both current and potential), suppliers, and regulatory agencies. In this study, quarterly data are evaluated based on their adequacy for use in investment decision making under the assumption that the investor is one of the primary users of such data. There appears to be no significant conflict be- tween the needs Of different investors or between the needs of investors and other users Of financial data. For purposes of this study, it is assumed that the needs of the investor are characteristic of the needs of all users. The Opinion that the analyst forms as a result of his analysis of the available information may affect many peOple. If the data lead him to erroneously recommend the stock as a good buy there is potential for stockholder loss. If he rec— commends that holders of the issue sell, based on his analysis 35 36 of misleading data, the effects may be equally as bad. It is not the intent of this study to explain specifi— cally the various stock valuation models or to study the com— ponents of such models. It is, however, restated here that: "the standard method of valuation of individual enterprises consists of capitalizing the expected future earnings and/or dividends at an apprOpriate rate of return."1 Hayes, in a manner very similar to the above, stated that one of the ob- jectives of the investment analysis of a reported earnings stream is, ". . . to portray the results on a basis which seems most probable to resemble the future Operation of the company. . ."2 These statements imply that analysts use re- ported corporate information to project future performance. These projected earnings are then used to value the firm's securities. If the valuation placed on the stock is differ— ent from the current market price of the stock the analysts may recommend the stock as a good buy, or he may recommend it be sold, depending if the market price appears too high, or too low in comparison to calculated value. lGraham, Dodd, and Cottle, Security Analysis, p. 435. 2Douglas A. Hayes, Investment Analysis and Manage- ment (New York: Macmillan Company, 1961) 196. 37 This implies that the adequacy of published corporate information for use by the investor depends upon its useful- ness in helping him predict the firm's future performance. The annual data published by the firm provides much of the information for this type of projection, however, available information of any type or from any source may be equally (or more) important. The quarterly report is a source of financial information which supplements annual data and it is used by analysts in their analyses. Deskin's interviews with financial analysts of mutual funds revealed that quarterly reports do provide information for this purpose. He states that: . . . Even though accounting reports are but a small proportion of the information sources, the importance of the information they contain. . . makes them disproportionately important as a source of information. . . The source of quan- titative information . . . is . . . annual or quarterly reports as Opposed to SEC registra- tion statements or Form 10K. Albrecht states: The financial analyst is concerned not so much with the ultimate consumption of capital stock (a valid concern of the CPA and ultimately, of course of the analyst too) as he is with the 1James Wesley Deskins, "The Uses of Externally Reported Financial Data by Mutual Funds and Their Implica— tions Concerning Financial Accounting Theory," (unpublished Ph.D. dissertation, The University of Texas, 1965), pp. 82-83. 38 factors which go to make stock prices. These factors include earnings, dividends, and the creation of future earnings and dividend power. . . There is, of course, much criticism levied against accounting information and its usefulness to the analyst for projecting future earnings. These criticisms include (1) con- cern over the multitude of accounting methods or principles which the accountant can use in the preparation Of the state- ments; and (2) concern over the lack of detailed information (or disclosure) included in published reports.2 The reasons for such criticisms are understandable, ". . . it is somewhat startling to hear the president of a large corporation say that he has no idea what his company's last year's earnings were — that his accountants have told him they can be figured six different ways, all having official sanction."3 Morton Backer, who is currently completing research for the National Association of Accountants, also noted this 1Philip E. Albrecht, "Analyst Views Financial Reporting Problems," Financial Executive, XXXIV, No. 9 (September, 1966) 14. 2Richard D. Bradish, "Corporate Reporting and the Financial Analyst," The Accounting Review, XL, No. 4 (October, 1965) 757—766. 3Howard C. Greer, "The Corporation Stockholder - Accounting's Forgotten Man," The Accounting Review, XXXIX, No. 1 (January, 1964) p. 24 quoting Downing Jenks, President of the Missouri Pacific Railroad, in a newspaper interview. 39 discontentment and stated: During recent years, criticism of external financial reports have mounted in intensity. In a sense these attacks are an attestation of the importance of the accounting function. However, they also reflect growing dissatisfaction with financial statements as a basis for informed in— vestment and credit decisions.1 Even though analysts use published financial data for projecting future earnings there is evidence that they do not want the accountant to do this projecting for them. Backer states the following concerning the findings of his study: All security investment decisions are con- cerned with the future. Yet 70% of the analysts interviewed (in NAA study) were Opposed to the inclusion by management of formal but condensed budgets in corporate annual reports. These analysts believe that if this were done they would have to explain and justify differences between their expectations for a company and its published budgets or that they might become overly dependent on management's forecasts and neglect verification through independent sources. They also contend that market prices of securities would be influenced by such budgets, which might induce some managements to issue overly Optimistic or even deliberately misleading reports. It was also noted that budgeting capability varies among companies. Although the usefulness of annual reports is criti— cized by those who use them they, nevertheless, are used. lMorton Backer, "Financial Reporting and Security Investment Decisions," Financial Executive, XXXIV, No. 12 (December, 1966) 50. 2Backer, "Security Investment Decisions,” p. 54. 40 Even though different net income figures can be derived by acceptable, alternative methods, net income of a corporation continues to be ". . . the universal yardstick used in ap— praising the value of its shares. Most investors place implicit trust in these figures as an accurate measure of a 1 company's performance.” As was previously mentioned, quarterly reports are subject to the same, and more, limitations than are the annual reports. "The comparability of interim reports with annual reports often leaves much to be desired. The differences, if you want to classify them as differences, result from the . . . . 2 short period of time covered by interim reports.” However, even with their inherent limitations they are used in invest— ment decisions. Seidler and Benjes state: . . . During the past several years, an increasing number of individual and institutional investors have striven to maximize investment results in a short period of time. To this class of investors the latest twelve months reported earnings and quarterly earning comparisons to be reported in the near future often assume more significance than earnings reported in an annual report.3 1Williams, "A Look Behind Reported Earnings," p. 38. 2Ralph A. Martin, ”Industry Viewpoint on Reporting Problems," Financial Executive, XXXV, No. 4 (April, 1967) 28. 3Seidler and Benjes, "Credibility Gap,” p. 109. 41 Albrecht states: . . . Financial analysts would find helpful a relative degree of uniformity in these quarterly statements. For instance, in each statement the analyst would like to see sales, depreciation, interest charges, estimated tax liability, and net income. . . Yet by no means do all companies present such information quarterly, eVen though it is obviously available. A questionnaire was used in an attempt to gain some insight into how quarterly data are used and to determine how adequate the analyst views quarterly data for investment decision making purposes. The Purpose of the Qgestionnaire: The questionnaire was used to elicit responses from Chartered Financial Analysts in order (1) to determine what analysts believe the primary objective of quarterly reports should be; (2) to determine how useful quarterly reports are in their security analysis; (3) to determine what they con— sider to be weaknesses in quarterly reports; (4) to determine which quarterly data they use in their analysis; and (5) to obtain suggestions for improving the usefulness of quarterly reports to the analysts. The questionnaire was sent to a randomly selected sample of Chartered Financial Analysts who were listed in the 1Philip E. Albrecht, "Analyst Views Financial Re— porting Problems,” p. 16. 42 1967 directory of The Financial Analysts Federation. The questionnaire was sent to 150 analysts and was completed and returned by seventy—seven of them. The remainder of this chapter will present the questions and the responses to the questions. It will also discuss their relevance to this study. It is obvious from the response to question 1 (shown on the following page) that the analysts responding believe that quarterly reported income should be meaningful to users and not just a means of keeping contact with the shareholders. There is no consensus, however, as to whether the primary purpose should be to report the actual income for the quarter as if it were a distinct accounting period in itself or whether the primary purpose should be to allow the user to better predict the annual performance of the firm. Of course, even if the eXpressed purpose is to report the actual income for the quarter the analyst can, and will, use such informa— tion to predict future performance. If quarterly reports are prepared with the expressed purpose of allowing one to better predict the annual income, the quarterly statement could, as an extreme, present pre- dicted or projected net income. This would require that the accountant, with the help of management, be placed in the posi- tion of predicting future performance. It is questionable 43 Table 3—1 WHAT DO YOU BELIEVE THE PRIMARY PURPOSE OF THE PUBLISHED QUARTERLY INCOME STATEMENT SHOULD BE? Response Frequency % a. To report the actual income for the quarter as if it were a distinct accounting period in itself. 34 42 b. To present data with the expressed purpose of allowing the user to better predict the annual performance of the firm. 44 54 c. To serve mainly a public relations function with little significance attached to the use of such data for decision making by outsiders. 0 0 d. Other 3 4 whether accountants should assume such responsibility or if their position and training qualify them to adequately perform the task. While it is true that accounting principles used for the reporting of actual results of prior Operations are flexible and allow the reported income amount to vary depend- ing upon the methods used, there are pp standards or prin— ciples for predicting the future. It would give management more of an Opportunity to manipulate the price of their stock by their prediction of future performance. However, improving 44 the user's ability to predict the future performance of the firm must be a primary objective of published quarterly statements. The past earnings of the firm are considered to be the most important single item of information in predicting future earnings1 and it appears that it should definitely be included in the quarterly reports. However, this does not mean that it is the only information which is important in predicting future performance. Other data, to supplement re— ported quarterly income, can be very beneficial in improving the user's ability to predict future performance. Comments given by respondents to this question in— cluded two statements to the effect that the year-to—date figures should be included in quarterly statements and one comment that the data for the latest 12 months should be re— ported rather than data for a particular quarter. Several respondents noted that a combination of responses a and b would be desirable. Some respondents thought that expected future develOpments should be commented on in the text of the report. 1Committee to Prepare a Statement of Basic Accounting Theory, A Statement of Basic Accounting Theory, Evanston, Illinois: American Accounting Association, 1966, p. 24. 45 Table 3-2 HOW USEFUL DO YOU CONSIDER PUBLISHED QUARTERLY REPORTS TO BE IN YOUR ANALYSIS OF THE INVESTMENT QUALITY OF A FIRM'S SECURITIES? Response Frequency % a. Very useful 35 45 b. Quite useful 28 36 c. Of limited usefulness l4 19 d. Of no usefulness 0 0 The response to this question indicates that analysts do consider quarterly reports to be useful in their analysis (81 percent of the respondents considered the quarterly re— port to be either very useful or quite useful). None of the respondents considered the quarterly reports to be of no use— fulness while only 19 percent considered them to be of limited usefulness. While this response indicates that quarterly reports are useful, it apparently does not mean that they are as use— ful as they could, or should, be. This is brought out by their responses to other questions. Replies by five respond— ents included comments to the effect that their usefulness varies widely depending on the company doing the reporting. 46 Table 3—3 HOW DOES THE FACT THAT QUARTERLY REPORTS COVER A SHORTER TIME PERIOD THAN DO ANNUAL REPORTS AFFECT THEIR USEFULNESS IN SECURITY ANALYSIS? Response Frequency % a. It has no significant effect on their usefulness. 28 37 b. It makes them less useful. 36 47 c. It makes them more useful. 12 16 d. It makes the data insignificant for use in security analysis. 0 0 Less than half (47 percent) of those responding to the above question feel that the short time period under re— port in quarterly statements make them less useful. It is worthwhile to note that no one felt that it made the data insignificant for use in security analysis. Three out of the twelve who think it makes the data more useful, commented that this was because it was more recent information. This may be what others thought when checking their response but even so they apparently were of the Opinion that any limitations due to the shortness of the time period are more than offset by the availability of more current information. The response to this question indicates 47 Table 3-4 HOW DOES THE FACT THAT QUARTERLY DATA ARE NOT AUDITED AFFECT THEIR USEFULNESS IN SECURITY ANALYSIS? Response Frequency % a. It has no significant effect on their usefulness. 45 59 b. It makes them less useful. 31 41 c. It makes them more useful. 0 0 d. It makes the data insignificant for use in security analysis. 0 0 that the financial analysts are divided concerning whether the short time period does affect the usefulness of quarterly data. The financial analysts who responded to the question— naire were divided as to the importance (or lack of impor- tance) of the audit concerning the usefulness of quarterly data for security analysis purposes. Forty-five (59 percent) of those responding felt that the fact that quarterly data were not audited had no significant effect on their useful— ness in security analysis. Thirty—one (41 percent) felt that this made the data less useful but no one thought it made the data insignificant for use in their analysis. 48 Four respondents noted that the effect of this de- pended on the company reporting. They commented that the audit would be much more useful in the case of the smaller firm with their less SOphisticated accounting system then in the case of a larger firm with a qualified accounting staff. Table 3-5 HOW MUCH CONFIDENCE DO YOU PLACE IN THE ACCURACY OF QUARTERLY INCOME DATA PUBLISHED BY FIRMS THAT ENGAGE IN A SUBSTANTIAL AMOUNT OF MERGER ACTIVITY AS COMPARED WITH DATA PUBLISHED BY FIRMS NOT SO CLASSIFIED? Response Frequency % a. More confidence 3 4 b. Less confidence 57 76 c. Same degree of confidence 15 20 Firms that engage in a substantial amount of merger activity have a greater incentive to keep the price of their stock high than do firms that are not active in merger activ- ity. In most cases the terms of the merger depend, to some extent, on the market price of the stock of the involved firms. For this reason it is possible that the quarterly reports of firms that engage in a great deal of merger activity 49 may present somewhat biased data in the hOpes of maintaining a high price for their securities. The responses to this question indicates that most analysts place less confidence in the quarterly reports of merger oriented firms. Of those responding 76 per cent indi- cated that they have less confidence in the quarterly reports of firms that engage in a substantial amount of merger activ— ity as compared to firms not so classified. Table 3-6 WHAT WOULD BE THE EFFECT ON THE USEFULNESS OF QUARTERLY REPORTS FOR USE IN YOUR ANALYSIS IF BALANCE SHEETS, ALONG WITH INCOME STATEMENTS, WERE MADE AVAILABLE ON A QUARTERLY BASIS? Response Frequency % a. Greatly increase their usefulness 34 45 b. Slightly increase their usefulness 38 50 c. Have no effect on their usefulness 4 5 The response to this question indicates that financial analysts believe that a balance sheet presented quarterly, along with the income statement, would improve the usefulness of quarterly reports. While 95 per cent felt that balance “ sheets would be helpful to-some degree only 5 per cent felt 50 that their inclusion would have no effect on their usefulness. The potential usefulness of quarterly balance sheets has also been noted in the literature. Seidler and Benjes state: The tendency towards taking short cuts in the preparation of interim reports is reinforced by the lack of any requirement, on the part of the SEC or the major stock exchanges, for the presentation of balance sheets at interim dates. Only condensed income statements are required. Though analysts often treat the balance sheet with some degree of disdain, its limitations should not be permitted to Obscure its value. Aside from its other functions, the balance sheet Operates as a control and dis- cipline over the income statement. For example, a company selling a product with the major part of its sales in the winter should normally show an ex— tremely low inventory balance at the end of the first quarter. If the inventory balance is high one might suSpect that the management has failed to write-off now obsolete merchandise or had lower sales than had been expected. In either case the analyst would be interested. If a balance sheet were presented with the quarterly report, there would be some pressure on the company to explain or write-off the high inventory balance. In the absence of a balance sheet, the problem can be con— veniently neglected. Although quarterly balance sheets are considered to be useful by analysts they are not generally included in pub- lished quarterly reports. In Taylor's study he found only ninety—six reports (out of 480 analyzed) that had any form of l Seidler and Benjes, "Credibility Gap,” pp. 113-114. 51 1 balance sheet included in the quarterly report. A conversa- tion with the Director of the Corporation Room at the Univer— sity of Maine revealed even more alarming statistics. He stated that in a 1967 survey of the statements of approxi— mately 1,000 firms, which included at least 80 per cent of Fortune's 500, fewer than twenty provided balance sheet infor- mation adequate for funds flow analysis on a quarterly basis. He also stated that only five provided balance sheet data adequate for quarterly cash flow analysis. Table 3—7 IF YOU USE PUBLISHED QUARTERLY DATA IN YOUR SECURITY ANALYSIS, DO YOU HAVE A SYSTEMATIC METHOD FOR ANALYZING THESE DATA? Response Frequency % a. Yes 36 51 b. No 34 49 The respondents, who indicated that they did use quar- terly data in their analysis, were asked to include working papers or give specific methods for using quarterly data. 1 Taylor, "Evolution, Content, Utility, and Problems of Published Interim Reports," p. 69. 52 Twenty—seven of those who use a systematic method in analyzing quarterly data commented on the methods used and/or included working papers that they used for such purposes. The intent of this question is not to discover the specific methods or techniques that are used and present them, but rather to discover the specific items of information that are used by the financial analysts when analyzing quarterly information. The working papers received in response to this request, as well as comments made concerning it, indicate that there are various methods employed, as well as various items of information used, when analyzing quarterly data. Sales, earnings before taxes, taxes, and net income are items of% information that were used in almost all cases. It was also frequently stated that the change in the amount of such items was considered important. Items which were mentioned very infrequently included: depreciation, selling and general administration expenses, and research and develOpment expendi— tures. These items may have been mentioned infrequently only because they are seldom available in quarterly reports and it may not mean that they are considered unimportant. Most of the working papers received, provided space for the recording of quarterly data for the previous several years. In one case there were forms to be filled in with quarterly data for 53 the last twelve years. Several respondents stated that they depended to a great extent, on financial services to do this type of analy— sis. Three of them specifically mentioned Financial Dynamics as the service they used. Although a letter was written to Financial Dynamics inquirying how they use quarterly data in their analysis no response was received. The response to the question given on the following page indicates that most analysts feel that knowledge of deprecia- tion expenses; other income (expenses); and the provision for Federal Income Tax, on a quarterly basis, are useful to them in their security analysis. Another question asked was: Which, if any industries (or firms) do you believe publish the most useful quarterly data for your analysis and why? While the respondents indicated many individual firms that they believed published useful quarterly reports, there was no consensus concerning any particular firm. In fact, the duPont Company was the only firm that was mentioned more than once and it was named by only two respondents. However, there was more unanimity concerning indus— tries that publish the most useful quarterly reports. Regu— lated industries, and specifically public utilities, were 54 Table 3-8 WHAT IS YOUR OPINION CONCERNING THE USEFULNESS (FOR USE IN YOUR SECURITY ANALYSIS) OF THE FOLLOWING REPORTED QUARTERLY INFORMATION: (A) DEPRECIATION EXPENSE; (B) OTHER INCOME (EXPENSES); (C) PROVISION FOR FEDERAL INCOME TAX? Provision for Depreciation Other Income Federal Response Expense (Expenses) Income Tax Frequency % Frequency %» Frequency % a. Very useful 28 37 24 32 41 54 b. Quite useful 28 37 33 44 27 36 c. Of limited usefulness 18 24 18 24 7 9 d. 0f no usefulness 2 2 0 -0 1 l mentioned frequently. Also, Oils were mentioned several times although not as Often as public utilities. Finance companies, autos, and steels were also mentioned more than once. Certain quarterly reports of public utilities and oils will be re— viewed in Chapter VII and the findings will be considered in making suggestions concerning the improvement of quarterly reports. In hOpes of gaining some suggestions for improving quarterly reports the question was asked: If you consider the quality of published annual data to be superior to 55 published quarterly data for use in ypp£_analysis, what recommendations do you have for improving quarterly data? The response to this question was extensive and, in most cases, the recommendations could be implemented with limited cost, either in time or money. The recommendation most often made was that more detail was needed in the quar- terly income statement. Often mentioned was the need for reporting depreciation expenses and other income or expenses on a quarterly basis. It was also mentioned that the report- ing for taxes on a quarterly basis was inadequate in many cases. The need for a quarterly balance sheet was also em- phasized and coincided with the response to question 6 con- cerning how useful quarterly balance sheets would be. The above recommendations, as well as others made by the respondents, would not be difficult to implement. In fact they are included in almost all annual reports and in some quarterly reports. It seems that the analysts are not recommending a radical change in quarterly reporting tech- niques as some firms are providing the quarterly information that the respondents requested. If certain firms or indus- tries can give the analysts what they need, is there any reason why they all can not? Another recommendation, which was often made by the 56 respondents, was that they should be more comments or dis— cussion of the data in text form. These comments should ex- plain changes, unusual occurences, and trends. They also believed that a discussion Of the general outlook for the firm would be helpful. Again these practices are followed by many firms in their annual reports. Other recommendations included: breaking down sales by product or area; more information on seasonal and nonrecur- ring factors; set forth basis for allocation to quarters such major items as depreciation, investment tax credit, etc. Other specific comments were: "It would be useful if the corporation would publish additional financial information if it were greater than x_% of the comparable quarter of the prior year;" "Too short a period for too much emphasis anyway;" "No public relations jazz;” "Think different standards should be used for interim reports than for annual reports;” "Com- panies should follow exact accounting procedures used in annual reports — a number of firms do not;” "In the final analysis it depends mainly on how COOperative tOp management wishes to be with security analysts." Summary The price which buyers are willing to pay for a share Of a company's stock depends, to a great extent, on the 57 estimates of future income for the firm. The analyst uses published financial statements as well as other information in making these estimates. This does not mean that the re- ports themselves should contain specific estimates of future performance, but it does mean that management should be aware Of how the statements are used and should, if practical, pro- vide the user with the additional detail he desires. This is especially true concerning information which is not con- fidential, is not harmful to the competitive position of the firm, and which is available with little inconvenience to the accountant preparing the statements. The;seventy—seven analysts (51 percent of the sample) responding to the questionnaire indicated that the published quarterly reports are useful to them in their analysis and the majority of them believe the primary purpose of such re— ports should be to present data with the expressed purpose of allowing the user to better predict the annual performance of the firm. The majority of the respondents do not feel that the lack of an audit affects the usefulness of quarterly data, but they tend to place less confidence in the quarterly data of firms that engage in a substantial amount of merger activ- ity as compared with data published by firms not so classified. The feeling that the usefulness of quarterly reports 58 would be increased if balance sheets were presented along with income statements was held by 95 per cent. The majority also felt that the quarterly reporting of depreciation ex- pense, other income (expenses), and the provision for federal income tax would be useful to them. The respondents stated that more detailed income statements would be helpful in their analysis, as would more comments or textual material concerning the reported figures. These suggestions do not appear to be unreasonable or diffi— cult to comply with. Indeed, more detail is available on a quarterly basis to internal management and much of it could be made available to the public with no threat to the com— petitive position of the firm. The same is true of other information that could be presented in text form along with the financial statements. The inclusion of a balance sheet, likewise, could be accomplished with limited effort or expense. CHAPTER IV FLUCTUATIONS IN QUARTERLY DATA Introduction The quarterly statements of selected firms indicate that quarterly data fluctuate widely between quarters, within the same firm. If expectations of the firm's future perform— ance are influenced by quarterly data, this expectation would be significantly affected by the quarter(s) being reviewed. This chapter will provide illustrations of the variations that have occurred in specific firm's quarterly data. They illus— trate why there is concern, in specific cases, over the quality of quarterly data and indicate the need for further study con- cerning their adequacy. How can Adequacy be Measured? One problem in evaluating the adequacy of published quarterly data for use in investment decision making is that the external user Of such data has little, if any, knowledge of the actual events which lead to the information in the statements. There is usually very little detail included in the quarterly statements and the textual material is limited. 59 60 One of the criticisms made frequently in the litera- ture concerning quarterly data is that there are often mas- sive year—end adjustments which are reported in, and there- fore affect the results of, the fourth quarter of the year. These adjustments can affect either revenue or expense items, however, there is no indication that the net sales figure is subject to such year-end adjustments. In fact, concerning the sales figure in quarterly statements, Carman Blough stated, "usually the only figure which has any validity is the amount of the sales. . ."1 Material adjustments which are made at year-end, and therefore affect expected results, should lead the user to question the reliability of the prev- iously reported quarterly data. The Struthers Wells Corporation noted that the 1966 year—end adjustments were so substantial that prior reported quarterly data were unreliable. In their third quarter, 1967, report they stated, ”the figures for the nine months of fis- cal 1966 as then reported are not deemed comparable because of year—end adjustments which were material in amount.” If the year-end adjustments do have a significant effect on reported fourth quarter data, the fourth quarter 1Letter from Carman G. Blough dated April 29, 1968 in response to my inquiry concerning his current feelings about the adequacy of quarterly reports for investment decision making. 61 results should vary from those of the first three quarters. Assuming that net sales figures are not adjusted, but that other revenue and expense items are, the reported net income will be affected (unless such adjustments are offsetting) and therefore, any material adjustments would be reflected in the net income/net sales (N.I./N.S.) ratio. If these adjustments do not exist, are immaterial, or are offsetting, we would ex- pect the N.I./N.S. ratio for a firm whose activity is not seasonal in nature and which adequately allocates fixed costs, to be approximately the same for all four quarters of the year. That is, there would be no reason to expect the fourth quarter to deviate significantly from that of the other three quarters. The following illustrations provide examples of such fourth quarter deviations, as well as other variations in quarterly data, for selected firms. The data from these firms were selected because their fluctuations were substan— tial and clearly indicate why the quality of quarterly data is sometimes questioned. Allied Control Company, Inc. Allied Control provides a good illustration of the variation that occurs in quarterly income statements. The quarterly N.I./N.S. ratio of Allied Control fluctuates widely during the period examined. In particular, the fourth quarter 62 ISJI./N.S. ratio for each of the seven years (1961-1967) Studied was greater than the N.I./N.S. ratio for any of the Other three quarters of that specific year. This is true (even though only in 1965 was the fourth quarter net sales :figure higher than that of any of the other three quarters <3f the year. This deviation is indicated by the scatter diagram and line of best fit as illustrated in Figure 4-1 (applicable computations for the figures illustrated in this chapter are presented in the appendix).1 The sum of the deviations from the line of best fit are presented in Table 4—1. These figures indicate that the fourth quarter deviates much more from the line than do the other three quarters of the year. Table 4—1 ALLIED CONTROL COMPANY, INC. SUM OF DEVIATIONS OF QUARTERLY N.I./N.S. RATIOS FROM LINE OF BEST FIT Sum of Deviations from Quarter Line of Best Fit First .0518 Second .0464 Third .1600 Fourth .2314 l The line of best fit was constructed using the method of least squares as presented in John R. Riggleman and‘ Ira N. Frisbee, Business Statistics (McGraw—Hill Book Com- pany, Inc., New York, 1951), pp. 302—303. 63 Figure 4—1 QUARTERLY N.I./N.S. RATIOS ALLIED CONTROL .11 .10 .09 .08 1961 .03 2° 1. .02 :01 1. (.OD (.02) 3. l 2 3 4 1 2 3 4 1 2 3 4 12 344:1;23 4 1 2 3.41.2 3 47 61 62 63 64 65 66 67 Time 64 The fact that the first three quarter's N.I./N.S. ratios lie in most cases very close to, or below, the trend line while the fourth quarter's N.I./N.S. ratios lie signifi- cantly above the trend line may indicate that Allied Control tends to be conservative in its reporting during the year. H & B American Corporation The quarterly results of H & B American, as in Allied Control, vary substantially between quarters with the fourth quarter N.I./N.S. ratios deviating significantly from those of the other three quarters. The scatter diagram and line of best fit presented in Figure 4—2 show this variation. Table 4-2 presents the sum of the various quarters' deviation from the trend line and indicates that the fourth quarter fluctu- ates much more than that of the other three quarters. Table 4—2 H & B AMERICAN CORPORATION. SUM OF DEVIATIONS OF QUARTERLY N.I./N.S. RATIOS FROM LINE OF BEST FIT Sum of Deviations from Quarter Line of Best Fit First .0885 Second .1030 Third .1303 Fourth .2129 .24 .23 .22 .21 .01 65 Figure 4—2 QUARTERLY N.I./N.S. RATIOS H & B AMERICAN CORPORATION 1964 1966 1967 1 2 I3 4 l 2 63 3 64 41 l 2 3 4 l 2 65 Time 3 66 4- 1 2 3 4 67 _] 66 Roxbupy Carpet Company The Roxbury Carpet Company, which was incorporated in 1859, manufactures woven and tufted carpets. Its quarterly data also indicate the wide fluctuations in the N.I./N.S. ratio between the quarters of the year as is indicated in Table 4—3 and Figure 4-3. Note the variations between the Table 4-3 ROXBURY CARPET COMPANY NET INCOME/NET SALES RATIOS Quarter 1961 1962 1963 1964 1965 1966 1967 First .0001 .0147 .0123 £0296) £0354) £0340) £0463) Second .0135 .0303 £0020) £0689) £0744) £0273) £0159) Third .0246 .0322 £0143) £0712) £0962) £0654) £0472) Fourth £0289) £0076) .0355) .0044 £0273) .0298 .0347 ( ) Denotes Net Loss/Net Sales various quarters and especially the fact that the greatest deviation is always between the third and fourth quarter of each of the years under review. Huyck Corporation HUka Corporation issues one of the more detailed quarterly income statements. A typical example of its quar— terly statement is presented in Figure C-1 of the appendix. 67 Figure 4—3 QUARTERLY N.I./N.S. RATIOS ROXBURY CARPET CORPORATION 1963 1967 1962 (.06) (.07) 3. (.08) (.09) 12:341234.1234:12341234112342u234 61 62 63 64 65 66 67 Time 68 It also presents more textual material in its quarterly state- ments than do most other firms although it does not publish quarterly balance sheets. Because or this detail it is pos- sible to perform some analysis on these statements that is not possible to perform on other firm's quarterly statements. The Huyck Corporation is not characterized by sea- sonality in its sales volume. i.e., there is no case in which the quarterly sales are greater than 30 per cent (or less than 20 per cent) of annual sales. The Net Income/Net Sales Relationship The net income reported for the fourth quarter is generally greater (5 out of 6 years) than that for any of the other quarters during the year. This is also reflected in the N.I./N.S. ratios in which 4 of the 6 years the fourth quarter had the highest N.I./N.S. ratio and is especially significant in 1965, 1966, and 1967. This relationship is indicated in Table 4—4 and Figure 4-4. (Taxes in the Quarterly Statements In addition to the N.I./N.S. ratio, the provision for federal income taxes, as a percent of earnings before taxes, 1This is the criterion used by David Green Jr. for detecting seasonality in "interim Reporting: Direct Costing and.Seasonalizing," p. 7. 69 Figure 4-4 QUARTERLY N.I./N.S. RATIOS HUYCK CORPORATION .090 1966 .085 V\\\S 1967 .080 .075 4 .070 .065 .060 1965 .055 2. .050 . 1045 1962 .040 .035 1964 .030 .025 1 2:: 5 .010 .005 (.005) 3. 1 2 3 4 l 2 3411.2 3 4 1 2 3¢4IL2 3 4-1 2134 62 63 64 . 65 66 67 Time 70 Table 4-4 HUYCK CORPORATION NET INCOME/NET SALES RATIOS Quarter 1962 1963 1964 1965 1966 '1967 First .0237 .0169 .0069 .0085 .0199 .0321 Second .0322 .0114 .0185 .0080 .0271 .0518 Third .0276 .0227 .0265 (.0059) .0352 .0563 Fourth .0332 .0167 .0254 .0512 .0900 .0724 ( ) Denotes Net Loss/Net Sales also fluctuates markedly over the various quarters of the year (see Table 4-5). Again, as in the case of the N.I./N.S. ratio, the fourth quarter tax/net income before tax (tax/NIBT) ratio departs significantly from that of the other three quarters, as well as from the yearly percentages. In fact, for each of the years studied the fourth quarter tax/NIBT ratio is less Table 4—5 HUYCK CORPORATION FEDERAL INCOME TAX/NET INCOME BEFORE TAX Period 1962 1963 1964 1965 1966 1967 lst Quarter .539 .549 .507 .524 .651 .510 2nd Quarter .553 .488 .538 .518 .531 .499 3rd Quarter .548 .513 .513 .469 .564 .479 4th Quarter .466 .269 .058 .376 .513 .437 Year .522 .477 .414 .411 .545 .473 71 than the ratio for any of the other three quarters of the year as well as less than the annual ratio. The scatter diagram and line of best fit, in Figure 4-5, show the extent of this deviation. The sum of the Figure 4-5 QUARTERLY TAX/NIBT RATIOS HUYCK CORPORATION Li66 .68 ~64 1962 1963 .60 .56 .52' .48 .44 .40 Tax NIBT.32 .28 .24 .20 .16 .12 .08 .04 1 2 3‘4212 3 41.2 3 4 1 2234-1 2 3 4 l 2 3 4 62 63 64 65 66 67 Time 72 deviations from the trend line for the various quarters is presented in Table 4—5.l. The extent of the fourth quarter deviation is indicated by the fact that the sum of the de- viations of the fourth quarter is greater than the sum of the first three quarters combined. Table 4—5.1 HUYCK CORPORATION SUM OF DEVIATIONS OF QUARTERLY TAX/NIBT RATIOS FROM LINE OF BEST FIT Sum of Deviations from Quarter Line of Best Fit First .3750 Second .2230 Third .2150 Fourth .8440 Other Income The tendency for fourth quarter data to deviate from that of the other three quarters is also true of "other income." In every year except 1964 I'Other income" reported in the fourth quarter was substantially higher than reported in any of the other three quarters of the year. In fact in 1962, 1963, 1965 and 1967 the "other income" reported in the fourth quarter was greater than the ”other income” for the 73 other three quarters combined. This fluctuation is illus- trated in Figure 4-6. The sum of the deviations from the trend line for the various quarters indicates that "other income" deviates much more in the fourth quarter than in any of the other three quarters (see Table 4-6). Figure 4-6 QUARTERLY "OTHER INCOME" HUYCK CORPORATION 4 . $35,000 1967 m 25,000 1 964 20,000 Other Income 15,000 . liiiillx d 2. 10,000 5.000 3. ' 3. 1 2 3 4 1 2 3 4212 3 4 1 213111 2 3 4 l 2 3 4 62 63 64 65 66 67 Time 74 Table 4-6 HUYCK CORPORATION SUM OF DEVIATIONS OF QUARTERLY "OTHER INCOME" FROM LINE OF BEST FIT Sum of Deviation from Quarter Line of Best Fit First $34,514 Second 33,556 Third 42,920 Fourth 67,660 Cubic Corporation As in the other illustrations, the data of Cubic Cor- poration reveal that fourth quarter results are quite differ- ent from those of the other three quarters of the year. In each of the six years (1961-1967) for which data are avail— able the fourth quarter net income is either higher than that of any of the other three quarters or the net income (net loss) is lower than that of any of the other three quarters. The fourth quarter's net income to net sales percentage also deviates substantially from the annual net income to net sales percentage and from the other quarters' percentages (see Table 4—7). The extent Of the deviation is indicated by the dispersion around a line of best fit as illustrated in Figure 4-7. 75 Figure 4—7 QUARTERLY N.I./N.S. RATIOS CUBIC CORPOR TION .09 .08 .07 .06 .05 (.01) (.02) (.03) (.04) (.05) 4. 12341234412341234142341234 62 63 64 65 66 67 Time 76 Table 4-7 CUBIC CORPORATION NET INCOME AS A PERCENT OF NET SALES Period 1962 1963 1964 1965 1966 1967 lst Quarter .0502 .0449 .0400 .0429 .0700 .0734 2nd Quarter .0446 .0619 .0524 .0341 .0436 .0649 3rd Quarter .0353 (.0069) .0550 .0167 .0532 .0644 4th Quarter .0198 (.0570) .0711 .0897 .0853 .0710 Year .0381 .0077 .0572 .0507 .0656 .0689 ( ) Denotes Net Loss/Net Sales Depreciation The depreciation charged against revenue on a quar- terly basis also varies Significantly (see Table 4—8 and Figure 4—8). While there is no explanation given for these Table 4-8 CUBIC CORPORATION DEPRECIATION Period 1962 1963 1964 1965 1966 1967 lst Qtr. $20,000 $28,000 $37,000 $31,000 $30,000 $ 6,000 2nd Qtr. 30,000 32,000 32.000 29,000 15.000 40.000 3rd Qtr. 56,000 84,000 34,000 34,000 32,000 51,000 4th Qtr. 21,000 23,000 14,000 54,000 43,000 96,800 Total 127,000 167,000 117,000 148,000 120,000 193,800 $100.000 90.000 80.000 70.000 60,000 De re— cia ion 50.000 40.000 30,000 20.000 10.000 wide variations, to determine any specific reasons, it does not appear that it results from an attempt to match depreciation expense against sales. cause(s) for such fluctuations the effect of such variations 77 Figure 4—8 QUARTERLY DEPRECIATION EXPENSE CUBIC CORPORATION 1 2 3 4 122334 1 2 3 4 1.2 3 4 1 2 3 4 1 2 3 4 62 63 64 65 \ 66 Time Although it may be impossible to determine the on net income is apparent. and the lack of detail makes it impossible 78 Great Basin Petroleum An even more extreme example, which may lead users to question the reliability of quarterly data, is indicated by the reported earnings of Great Basin Petroleum. The data for the fourth quarter of 1961 (as calculated by subtracting the accumulated first nine months data from the annual data) re- veals that net income was greater than sales. In fact, fourth quarter reported sales were $58,200 while reported net income was $170,000. This "unusual" phenomenon was due to a decrease in total accumulated expenses of $111,900 between the nine month year-to—date figures and the annual figures. There was no reason given for this reduction. This extreme example sug- gests that the reported net income in the first three quar- ters was probably understated and, therefore, misleading to the users. Reasons for Fluctuations in Quarterly Data Letters were written to the top financial Officers of selected firms attempting to determine the reasons for the substantial variations in their firm's reported quarterly data. Inquiry was made concerning (1) reasons for the fluc— tuations in the N.I./N.S. ratio; (2) the effect of year-end adjustments; (3) the reliability of quarterly data for use by security analysts; (4) the quarterly provision for federal IIII---------___________________________________________________:j:_____ 79 income tax; and (5) information regarding quarterly deprecia- tion and miscellaneous income. This inquiry was made although it was realized that the response to these questions might be limited and that any responses might be biased because of the personal involvement of the respondents. In spite of these limitations, the response was satisfactory, although lacking in specific details. In general, the financial officers re- sponding, noted that there were troublesome areas in quar- terly reporting but they were also of the opinion that quar— terly data were quite useful. In answer to the question "why does the fourth quar- ter net income/net sales ratio vary as much as it does from annual?” the Treasurer Of Stepan Chemical stated: In many cases the fourth quarter is not separately reported and, in effect respresents the difference between the annual figures and the previously reported nine months. As such, it in- cludes the year end adjustments which, depending upon the specific company, may be insignificant or quite material. In the case of Stepan Chemical, the year end adjustments in several of the years which you have studied were quite material. As a result, the N.I./N.S. ratio did vary substan- tially from the annual ratio. In a reply to this same question Howard Ridley, Con- troller of Crowley Milner and Company stated that the following 1Letter dated JUne 12, 1968 from W. W. Meier, Vice President and Treasurer, Stepan Chemical Company. 80 were reasons why the net income to net sales ratio varies as much as it does in the fourth quarter: Year-end Adjustments (1) Inventory book to physical adjustment. Our aetailers) shrinkage percentage has been on the rise in the past several years and is a most significant adjustment compared to sales and net income. National average for retailers probably is around l.6%—l.8% of sales. We provide for shrinkage, but usually the actual shrinkage is more than we provided for. (2) LIFO inventory adjustments made at year-end. This adjustment also cannot be determined until year-end. Again, we provide for this but usually not enough. The Treasurer of Atlas Corporation stated: . . . I wish to advise you that Atlas' fourth quar- ter results over the past few years have differed from the other three quarters during such years for various reasons, none of which were known at the time of mailing of the first three quarterly re- ports for each year. . . . In 1966 and 1967 we had some year end adjustments which threw the fourth quarter results out of line with the other quarters of such years. . . The possible distortion Of fourth quarter results due to year-end adjustments was noted by several respondents. These responses include the following: . . . the year end adjustments frequently do distort fourth quarter results. The obvious solution is to 1Letter dated JUne 19, 1968 from Robert T. Marquart, Vice President-Finance and Operations, Crowley, Milner and Company. 2Letter dated June 12, 1968 from Walter G. Clinchy, Vice President and Treasurer, Atlas Corporation. 81 more accurately anticipate events and give them realistic accounting treatment in prior quarters. In many cases this may be possible, but obviously there are also other situations in which this just is not feasible. Some companies on an in— ternal basis separate the operating results of the fourth quarter from the year end adjustments but this would not be a feasible solution in terms of public distribution or reporting of information. Another stated: As the calendar year and fiscal year near a close, there is a determined effort to exceed prior years sales and profits, thus, an acceleration takes place in the last quarter resulting in higher sales and earnings. A conservative accounting policy is maintained during the year to prevent year-end surprise adjust- ments, such as a major inventory adjustment. The Controller of Crowley, Milner and Company stated: I believe that . . . adjustments . . . do and can significantly affect our fourth quarter and year-end results. Inventory and reserve account (inventory, contingency, warranty, etc.) adjustments probably are the two greatest year—end adjustment problems for most companies. The only practical alternatives I can see is spending more time and money to continually analyze the prOblem areas which require significant year-end adjustments. Such as — monthly or quarterly physical inventories, quarterly or semiannual reserve account computations, etc. Letter from W. W. Meier. 2Letter dated June 20, 1968 from P. R. Loomis, Vice President/Finance, Cubic Corporation. 3Letter from Robert T. Marquart. 82 Frank Palmer, Controller of Huyck Corporation, stated that in addition to possible fourth quarter variations due to final determination of the tax expense: . . . other adjustments, which are reasonably com— mon in industry, are those required to correct for the annual physical inventory and to adjust for routine accruals such as vacation pay and repair costs and several other such spotty costs which we attempt to spread out as well as we can over the year.1 In regards to the effect of adjustments on fourth quar- ter results the Treasurer of Huyck Corporation stated that, “Experience teaches that usually adjustments will cancel each other out so that the net effect is close to zero. In some years, however, everything seems to go to pluses and in other years to minuses, and particularly in the latter an account- . . . "2 ing life 18 not a happy one. The replies indicated that the apprOpriate tax rate to be applied is the same for each of the quarters (as well as the year) but gave no reasons for noted fluctuations. They also failed to give any specific reasons for the wide variations (in certain cases) in depreciation expense and miscellaneous income between quarters. 1Letter dated May 13, 1968 from Frank L. Palmer, Controller, Hnyck Corporation. 2Letter dated January 23, 1968 from Frank C. Lowe, Treasurer, Huyck Corporation. 83 In spite of the effect of year-end adjustments, which lead one to question the reliability of published quarterly data, the respondents felt that quarterly data are adequate for use by security analysts. The Vice President of Finance of the Cubic Corporation was very emphatic concerning their adequacy. He stated, "Quarterly data are certainly reliable for financial analysts."1 The Treasurer of Stephan Chemical Company was not quite as emphatic and stated: If a company makes a conscientious effort to report accurately on a quarterly basis, I think such information is useful and in most cases reli- able for financial analysts. I would hOpe that such individuals are experienced enough to recog- nize some of the potential limitations of interim information, but the potential importance of such information probably encourages analysts to want to accept quarterly data as being accurate. In some cases there is indication that some companies use this situation unfairly. . . . I might comment as a matter of philosophy that some of the interim reporting (quarterly) of publicly owned companies does serve the obvious desirable purpose of providing a recurring com- munication as to the status of the company. Un- fortunately, the financial community generally reacts so strongly to such information that manage- ment is under pressure to report "desirable" results. If it is assumed that annual figures which are cer- tified will be reasonably stated, any Optimistic or perhaps unduly favorable figures released will have only short term benefits. Nevertheless, whether intentional or otherwise, interim reporting is not always accurate and does pose problems for both the conscientious management and the investor. 1Letter from P. R. Loomis. 2Letter from W. W. Meier. 84 The Controller of Crowley, Milner and Company simi- larly noted the limitations of quarterly data and stated: . . . I believe quarterly figures are useful to the extent that they are understood. A person must realize that significant seasonal fluctua— tions will greatly affect profit. You just can't take one of our quarters and multiply it timestur tn) get. the yearly profit figure. Also, most companies (especially large companies) are con- sistent in their accounting policies and therefore valid comparisons can be made. . . Walter G. Clinchy, the Treasurer of Atlas Corporation states that, "despite the deviations in several of our fourth quarter results, I still believe that for business as a whole quarterly data are reliable enough for use by financial ana— lysts."2 The Controller of the Huyck Corporation stated: As you have suggested, the problem of preparing meaningful quarterly reports is a trying one. One must realize that the shorter the period for which a report is prepared the less accurate and meaningful it can be. Also, one must realize that the time con- sumed and the cost of the preparation of audited data at the end of the year are rather prohibitive for interim reports. Finally, there is the ever—present problem of the risk of misleading shareholders with interim reports. I recognize that it is just as misleading to understate as it is to overstate. I believe you will find that all corporate officers responsible for submission of interim statements feel a strong sense of responsibility to produce the best practical data, and I also suspect that most would prefer to err on the conservative side where preci— sion is not practical. I do believe that interim 1Letter from Robert Marquart. 2Letter from Walter G. Clinchy. 85 statements, accepted for what they are, can be a useful tool for knowledgeable readers in spite of the acknowledged drawbacks.l Summary The above comments seem to summarize adequately the Opinions of the financial officers concerning quarterly re- ports. That is, that there are limitations and drawbacks inherent in quarterly data and that adjustments sometimes do affect significantly their reliability. In spite of this, it is their Opinion that quarterly data can be useful to analysts if they use such data with consideration for their limitations. 1Letter from Frank L. Palmer. CHAPTER V EMPIRICAL ANALYSIS Introduction The selected firms” data presented in Chapter IV pro- vide specific examples of wide fluctuations in quarterly data. The responses from the financial officers confirmed that ma— terial year—end adjustments are not unusual for their firms and have been responsible for wide fluctuations in reported quarterly net income. These material year=end adjustments Often throw the fourth quarter's data out of line from that Of the other three quarters and lead the user to question, justifiably, the realiability of quarterly data. As such adjustments affect income statement items other than net sales, any significant effect will be evident in net income and therefore in the net income to net sales ratio. The first section of this chapter will investigate the quarterly N.I./N.S. ratio for a sample of firms to deter- mine if such fluctuations are a common occurrence. The sec— ond section will examine the fluctuations in the quarterly provision for federal income taxes and test if such fluctuations 86 87 lead to the reporting of desirable results. Selection of Firms Studied and Source of Data There are 400 different U. S. industrial firms which were listed on the American Stock Exchange (A.S.E.) between 1962 and 1968 inclusively. One hundred and ten of these firms were randomly selected and their published financial data analyzed. Letters were written to these 110 firms requesting their published quarterly and annual statements for the period 1958 to 1968. While a majority of the firms responded to the letter, very few sent all the information requested. To ob- tain quarterly information not sent by the firms, a trip was made to the American Stock Exchange Library in New York where COpies of the published data of firms listed on that exchange are filed. In most cases there were no quarterly data avail— able for periods prior to 1962 (when the A.S.E. first began requiring quarterly reports). Twentyuthree of the selected firms filed no quarterly reports prior to 1968, or their quarterly reports contained inadequate data for meaningful analysis. In other cases certain quarterly reports were missing from the files. In these instances another letter was sent to the firm in question requesting the data for the missing quarter(s). If this request still failed to secure the required data it was Obtained from Moody's Industrial News, 88 a semi-weekly publication, which reports the published quar- terly financial information of most firms. In most cases firms did not report data for the quar- ter only. They reported year-to—date figures instead. That is, the first quarter report would cover the first three Inonths of the fiscal year, the second quarter report would <:over the first six months of the fiscal year, and the third (quarter report would cover the first nine months of the fiscal :year. In these cases data for the second quarter were obtained 13y subtracting the first quarter (first three months) data from the second quarter (first six months) data. Similarly, ‘the third and fourth quarter's data were obtained by finding 'the difference between the year—to—date figures. From these available data an investigation was made 1:0 determine if fourth quarter data presented a different Jgoicture of the firm's performance than did the other three <23uarters' data. A difference could result from a firm's 'Ifinanagement being over-conservative or over-optimistic during ‘tzhe first three quarters; it could result from a firm's man- {Eagement being unable to adequately measure revenue or expenses (Ruth.the given effort) during the first three quarters; or fit could result from apprOpriate adjustments being made only Eat year-end Closing or when required by the annual audit. in. 89 In any case the fourth quarter's reported results may be quite different from those of the first three quarters and this difference would be reflected in the N.I./N.S. ratio. .Method Used to Investigate Quarterly Results In determining whether there is a variation in re- Inorted quarterly results for a particular firm the difference IDetween the annual N.I./N.S. ratio and the quarterly N.I./N.S. 1:atio for each quarter of that year was calculated. After (Setermining these differences for every year in the study the :Eollowing approach was used: 1. The points were assigned to each quarter de— pending upon the difference between the annual N.I./N.S. ratio and the quarterly N.I./N.S. ratios ignoring whether the difference was positive or negative. For example: Points Assigned To Quarter 4 The farthest from annual percentage 3 The third closest to annual percentage 2 The second closest to annual percent— age 1 The closest to annual percentage The following illustration indicates the points that would be assigned to the various quarters given the N.I./N.S. ratios: 90 Difference Between Annual & Quarterly Points Period N.I./N.S. Ratio Assigned lst Quarter 19x1 .055 .003 1 2nd Quarter 19x1 .064 .012 4 3rd Quarter 19x1 .047 .005 2 4th Quarter 19x1 .042 .010 3 Annual .052 2. The total points assigned to each quarter for each firm, using the above criteria, were accumulated. 3. The number of firms in which the first quarter deviates the most from annual, the second devi— ates the most from annual, etc., was determined using the above criteria. 4. The quarterly data were statistically tested to determine if the fourth quarter data deviate more from annual than do any of the other three quarter's data. In this study the annual N.I./N.S. ratio is considered t:o be the ”normal” ratio. Using the point assignment proce- Ciure presented above, the quarter which has the most points EEtssigned to it was considered to be the quarter which expew Ilz‘ienced the greatest deviation from normal. In determining the points assigned to a particular .25 The probability of finding x observations in one clas— sification and N - x observations in the other classification is given by the formula: P(X)=()1:)PXQn_X Where: P = Proportion of cases expected in one of the classifications Q = 1-P = PrOportion of cases eXpected in the other classification N! <3) Xi (n-x)i However, in the present case we are not concerned with finding the probability of having a certain number of observa- tions falling within a classification, given the expected pro~ portion, but rather with the probability of having a certain number, or more, observations falling within a classification, 94 given the expected proportion. In other words we sum the prObability of the observed value with the probabilities of 1 values even more extreme. Fortunately, there are tables which present these probabilities.2 Results In determining if the results given in Table 5—1 are likely to have come from a population with the before men- tioned characteristics, the following values are relevant: P = .25 Q = .75 N = 87 The classification of the data is presented in Table 5-l.2 Table 5-l.2 GREATEST DEVIATION FROM NORMAL Z)of First, Second and Third Quarters Fourth Quarter Total Frequency 40 47 87 l . . . Sidney Siegel, Nonparametric Statistics, 37-38. 2The tables used in this study are included in: Tables Of The Cumulative Binomial PrObability Distribution (Harvard University Press, Cambridge, Massachusetts, 1955) which were compiled by the staff of the Computation Labora- tory. 95 The binomial test was used to determine if it is reasonable to believe that fourth quarter N.I./N.S. experi- ences the greatest deviation in 25% of the firms in the pOpu- lation given that 47 firms, of the 87 firms studied, reported data in which the fourth quarter deviated the most. The tables indicate that the chance of 47 out of 88 observations falling into a classification, given the prob— ability Of .25 is less than .00001.1 Therefore, the conclu- sion is that reported fourth quarter data indicate different performance than do the data of the first three quarters and that the greatest deviation from normal (as represented by the reported N.I./N.S. ratio) occurs in the fourth quarter in more than 25 percent of the firms in the population. Using the tables of the binomial distribution it can be stated with 95 percent confidence that between 42 and 65 per- cent of the U.S. firms listed on the American Stock Exchange between 1962 and 1968 inclusively reported the greatest devi- ation from normal (as represented by the reported N.I./N.S. 2 ratio) as occurring in the fourth quarter. This holds 1Tables of the Cumulative Binomial Probability Dis— pribution, p. 156. (As the tables include figures for 86 and 88 observations, but not 87, the figures for 88 were used. This gives a more conservative result than would have been the case if 87 were available) See Carl A. Bennett and Norman L. Franklin, §£atistical Analysis in Chemistry and the Chemical Industry (John Wiley & Sons, Inc., New York: 1954) 603-605. 96 although there appears to be no reason to expect the devia- tion from normal of the fourth quarter to be greater than the deviation of any of the other three quarters. It should also be noted that there is no evidence that the reported earnings Of the first three quarters of a year (when there is more opportunity for management) are overstated. Assuming that a "higher" N.I./N.S. ratio is more likely to be the result of an overstatement rather than an understatement of net income, and that a "lower'I N.I./N.S. ratio is more likely to be the result of an understatement rather than an overstatement of net income, the findings indicate that the fourth quarters do not tend to be understated. There were 305 instances in which the fourth quarter ratio was either higher or lower than any of the other three quarters of the year. In 171 (56%) of these cases the fourth quarter ratio was the highest and in 134 (44%) was the ratio the lowest. As the fourth quarter data are the difference between audited annual figures and accumulated nine months data, these re— sults imply that the first three quarters' data are not overu stated and, if they are biased, tend to be understated. Results of Inquiry Splitting the Firms' Fiscal Year In the prior analysis there could be a question raised as to a develOping trend biasing the results. That is, if the 97 firms' expenses were either increasing or decreasing in rela- tionship to net sales the results could affect the N.I./N.S. ratio with the fourth quarter being affected more than the other quarters. This could result in the fourth quarter deviating more than the other three quarters because of the facts and not unreliable reporting. In order to eliminate this potential bias, the fiscal year was not used to calculate "normal” but instead, the third and fourth quarters of one fiscal year and the first and second quarters of the follow— ing fiscal year were accumulated and considered to be "normal". This is illustrated in Figure 5-1. Figure 5-1 SPLITTING THE FISCAL YEAR 19x1 19x2 19x3 1st 2ndv 3rd 4th lst 2nd 3rd 4th 1st 2nd 3rd 4th Q Q Q Q Q Q Q Q Q Q Q Q u u u u u u.% u. 7 u u u u u a a a a a a [a 1 a a a a a r r r r r r Tr r r r r rf vL—~\Quarterly data accumulated in xs/ determining "normal" ratios Using the same criteria as in the earlier test (but Changing the period used for calculating the ”normal" ratio) 98 the analysis again indicates that there is a substantial difference between the quarters' reported data. The results using the above "normal” period are presented in Table 5-2 for the 87 firms for which sufficient data were available. Table 5-2 NUMBER OF FIRMS IN WHICH THE VARIOUS QUARTERS EXPERIENCE THE GREATEST DEVIATION FROM "NORMAL" SPLITTING THE FIRMS' FISCAL YEARS WHEN CALCULATING "NORMAL" NUmber of Firms in Which the Quarter Deviation is Greatest Percent Third 16 19 Fourth 45 51 First 10 11 Second lg _12 Total 7 .199 Results of Binomial Test The data are classified in Table 5—2.1. Again the bi— nomial test is used to determine if it is reasonable to believe that fourth quarter N.I./N.S. experiences the maximum devia- tion in 25% of the firms in the pOpulation given that 45 firms, of the 87 firms studied, reported fourth quarter N.I./N.S. which deviated more from normal than any of the other quarters. 99 Table 5-2.1 GREATEST DEVIATION FROM NORMAL ZDOf First, Second and Third Quarters Fourth Quarter Total Frequency 42 45 87 The probability of 45, out of 88 observations, falling into a classification given the probability of .25 is less than .00001.1 Again the conclusion is that reported fourth quarter data indicate different performance than do the data of the first three quarters and that the greatest deviation from normal (as represented by the N.I./N.S. ratio) occurs in the fourth quarter in more than 25 percent Of the firms in the pOpulation. It can be stated with 95 percent confidence that between 40 and 63 percent Of the firms in the pOpulation will report the greatest deviation from normal (as represented by the N.I./N.S. ratio) as occurring in the fourth quarter.2 ‘ lAgain N=88 was used because N=87 is not available in tables. 2See Carl A. Bennett and Norman L. Franklin, Statistical Analysis In Chemistry, 603—605. 100 This holds although there appears to be no reason to expect the deviation from normal of the fourth quarter to be greater than the deviation of any of the other three quarters. Splitting the Fiscal Year and Eliminating Seasonal Data: Because total fixed expenses do not vary with produc— tion or sales volume it is possible that these charges, as a percentage of quarterly sales, may vary significantly. This would cause the N.I./N.S. ratios to differ between quarters depending upon the amount of sales of the particular quarter. This would be true if annual non—manufacturing fixed expenses are material and are divided equally between quarters. To eliminate this possible bias the identical analysis was conducted eliminating those years in which fourth quarter sales indicated seasonality. The criterion for determining seasonality is borrowed from Green's paper, ”Interim Reporting: Directing Costing and Seasonalizing." Any quarter which ree ported sales of greater than 30 percent of annual sales (or less than 20 percent of annual sales) Green defined as being subject to seasonality factors.1 This criterion was applied to every year in the study to determine those fourth quarters l . . . . . DaVId Green, Jr., Interim Reporting: Direct Costing and Seasonalizing," p.7. 101 which were of a seasonal nature. If they were classified as seasonal the year under review was eliminated from consideraw tion. Results Adjusting for Seasonality Using the same time span for the normal period as in the preceding test (i.e., the third and fourth quarter of one year and the first and second quarter of the following year) and eliminating the years in which the fourth quarter is of a seasonal nature gives results similar to the preceding. That 'is, there is a significant difference between the quarters' reported data. After eliminating seasonal data, as defined above, eighty firms remained in the sample. The results are presented in Table 5-3. Table 5~3 NUMBER OF FIRMS IN WHICH THE VARIOUS QUARTERS EXPERIENCE THE GREATEST DEVIATION FROM "NORMAL" SPLITTING THE FIRMS' FISCAL YEARS WHEN CALCULATING "NORMAL" AND ELIMINATING SEASONAL DATA Number of Firms in which the Quarter Deviation is Greatest Percent Third 14 17 Fourth 39 49 First 9 11 Second 18 '_g§ Total 80 100 102 ReSults of Binomial Test The data are classified in Table 5-3.1. Although the percentage of firms in which the fourth quarter reflects the greatest deviation is not quite as large as in the previous two cases, the results of the test are very similar. As before, Table 5—3.1 GREATEST DEVIATION FROM NORMAL ZDOf First, Second and Third Quarters Fourth Quarter Total Frequency 41 39 80 the test indicates that it is not reasonable to believe that fourth quarter N.I./N.S. experiences the greatest deviation in 25% of the firms in the pOpulation, given that 39 firms out of the 80 firms studied, reported fourth quarter N.I./N.S. which deviated more from normal than did any of the other three quarters. In fact, the tables indicate that the chance of this happening given the probability of .25, is again less than .00001. The conclusion is that, after eliminating seasonal data, the reported fourth quarter data indicate different performance than do the data of the first three quarters and 103 that the greatest deviation from normal (as represented by the N.I./N.S. ratio) occurs in the fourth quarter in more than 25 percent of the firms in the population. It can be stated with 95 percent confidence that between 37 and 61 percent of the firms in the population will report the greatest deviation from normal (as represented by the N.I./N.S. ratio) as occur— ring in the fourth quarter.1 This holds although the firms' Operations do not appear to provide any reason for expecting the deviation from normal of the fourth quarter to be greater than the deviation of any of the other three quarters from normal. Summary These findings confirm that fourth quarter reported results (as measured by the N.I./N.S. ratio) do deviate from those of the other three quarters of the year although there appears to be no reason for believing that the firms' actual Operating activities cause this deviation. This is true even if those years in which the fourth quarter is subject to sea- sonal sales activity are eliminated. These findings verify that the problems that are inherent in quarterly reporting lSee Carl A. Bennett and Norman L. Franklin, Statistical Analysis in Chemistry, 603-605. 104 do affect the reported results and that quarterly statements are limited in their reliability. .Managing Quarterly Data The Treasurer of the Stepan Chemical Company noted that fourth quarter results ”represent the difference between the annual figures and the previously reported nine months."1 This implies that they represent the difference between the reported data for the first nine months (which were unaudited and vulnerable to "management") and audited annual figures (which are generally assumed to be reliable). If the first nine months data were "managed" we might expect to find the fourth quarter data deviating from the data of the first three quarters. As this deviation was indicated by the facts pre— sented earlier, the question of managing quarterly data will now be investigated. Motives for Managing Net Income Every company is aware of the market price of its stock and likewise every Company is aware of the influence that reported net income can have on the market price. It is generally conceded that fluctuating earnings are not valued as highly as stable earnings, given the same expected earnings, 1Letter from W. W. Meier. 105 ceteris paribus. For this reason some hold that management uses the flexibility allowed by accounting methods to manage reported income. Hammel and Hodes, in discussing the factors that in— fluence price-earnings ratios, stated: . . .Companies with a history of highly volatile earnings tend to trade at lower price—earnings multiples than other comparable companies whose growth in earnings have been more stable around a basic trend. The empirical evidence shows that this . . . holds even when the volatile and more stable companies have achieved a similar trend growth rate over a long period of years. . .1 Myron Gordon stated as a theorem that ”. . . manage- ment should within the limits of its power, i.e., the latitude allowed by accounting rules, (1) smooth reported income and, (2) smooth the rate of growth in income. . ."2 In elabor- ating on why management would ever reduce reported income Gordon stated: It may be wondered, how does the downward smoothing of the rate of growth benefit the man— agement of a corporation? Accounting practices that reduce the reported rate of growth when a corporation is highly successful create "hidden reserves" and allow reporting a higher rate of growth than otherwise during "the seven lean years". . . 1John E. Hammel and Daniel A. Hodes, "Factors Influ- encing Price—earnings Multiples," Financial Analysts Journal, Vol. 23, No. 3 (May—June, 1967) 91. 2Myron J. Gordon, "Postulates, Principles and Research in Accounting," p. 262. 106 . . . To elaborate, insofar as two or more alternative bases of valuation are allowed for a transaction, management will make the choice as follows: If the choice is just relevant to the current year's income, the choice will be the one that raises (lowers) income if it is below (above) the trend line for the year. If the valuation choice will influence income for a number of years in the future, the prediction is that a corporation with a high (low) rate of growth in income will make the choice that lowers (raises) the rate of growth in income. Hepworth also elaborated on the desirability of re— porting a stable income over time. He said: A less tangible, but perhaps more fundamentally important type of advantage of a relatively stable level of periodic income lies in the area of manage— ment relations with investors and workers. Certainly the owners and creditors of an enterprise will feel more confident toward a corporate management which is able to report stable earnings than if consider— able fluctuations of reported earnings exist. . . A sharp increase in reported profits is very likely to produce the feeling in the minds of the members of the working force that they should participate to a greater extent in such profits, with existing demands for wage increases, strikes and general industrial unrest . . . It would seem that the main- tenance of a relatively stable level of periodic in— come might do much to reduce the effects of "waves of optimism and pessimism" on the level of business activity. Chambers also commented on the effect that alternative accounting methods could have on the price of a firm's stock. 1Ibid. 2 u 0 I Samuel R. Hepworth, "Smoothing Periodic Income," p. 34. 107 The stated, ”. . . suppose that there are alternative sets of rules vflfirfllnay be adOpted and abandoned at the Option of management. It will then be possible at some times to influ- encmazfavorably the valuation of stockholders, actual and po- tential, and at other times to influence favorably the valu— 1 O O ations of bondholders, actual and potential. “ The aforementioned views concerning the motives for managing income are applicable to both quarterly and annual earnings. Seidler and Benjes stated: "It has always been widely understood that some companies may have a definite interest in presenting reported incomes which are either higher or lower than income which might he arrived at from completely 2 unbiased vieWpoints." They also noted that management of income is much easier in quarterly reports than it is in annual reports. As these quarterly reports are unaudited the firm has some Opportunity to "manage" most items used to calculate net income. While no one accuses management of completely ignoring the facts in quarterly statements, Thomas Holton makes the accusation that quarterly earnings are stabilized 1Raymond J. Chambers, Accounting, Evaluation and Economic Behavior, Prentice—Hall, Inc., Englewood Cliffs, New Jersey, 1966), p. 281. 2Seidler and Benjes, "Credibility Gap," p. 111. 108 by'amccounting techniques although noting that he is unable to prove this.1 One problem that a researcher faces in attempting to investigate "management" of net income on a quarterly basis is the lack of detail in quarterly reports. This is true in the statements themselves as well as in footnotes and textual 'material. However, the provision for federal income tax is one very potent means of "managing" income and it is often present in the quarterly income statement. The following section will investigate the fluctuations in the federal in- come tax to net income before taxes ratio for evidence of "management." Federal Income Taxes and ”Managing" Net Income As no independent audit is performed on quarterly statements there is more Opportunity for management to ”manage" quarterly income than annual income. The quarterly provision for federal income tax provides management with one Opportunity for managing reported earnings. However, if a firm does not use interperiod tax allocation (base the tax expense on the current year's published financial data) but charges as 1Thomas L. Holton, "Discussion of the Predictive Power of First-Quarter Earnings Reports: A Replication," Empirical Research in Accounting: Selected Studies 1966, 38. 109 expense the amount currently payable to the government, it may be difficult to determine if the tax provision is used to ”manage" reported income. It is possible, although un- likely, that significant fluctuations in the quarterly tax ratios which appear to "manage" net income may be caused by the difference between financially reported net income before taxes and taxable income per the tax return.1 The fact that this is a possibility required that the analysis be broken down according to (1) firms that allocate taxes: and (2) firms that do not allocate taxes. A firm's performance is often evaluated by comparing currently reported quarterly net income with reported net in— come for the same quarter of the prior year. ”Most interim reports are presented by comparing the interim results for a 1The following is an extreme example of such possible fluctuations within a given year: Per Financial Statements Per Tax Return Net Income Tax Net Income Tax Quarter Tax Erpense Before Taxes NIBT Before Taxes NIBT First $105,000 $250,000 .42 $210,000 .50 Second 120,000 250,000 .48 240,000 .50 Third 140,000 300,000 .467 280,000 .50 Fourth 130,000 400,000 .325 260,000 .50 Year $495,000 $1,200,000 $990,000 :1. .413 = .50 Although the Tax/NIBT per the financial statements varied widely the Tax/NIBT per the tax return was stable over the four quarters. 110 particular period of one year with the same period of the pre- ceding year."1 The comparison between these periods is empha- sized in both the financial and textual portions of quarterly reports and provides management with an incentive to report net income as high as (and preferably higher than) income for the same quarter of the prior year. There appears to be no reason for the federal income tax/net income before tax (tax/NIBT) on a quarterly basis to 2 deviate erratically from an annual basis. Accordingly the lNeubig, ”Effect of Periodicity," p. 34. 2The letters from the financial officers (see Chapter IV) stated that the estimated effective annual tax rate is applied in each quarter of that year. This procedure was verified in a discussion_with.a partner of one of the large certified public accounting firms. Another acceptable method for computing taxes is given by formulae 1 and 2 below. If formula 1 was used for computing quarterly taxes, and if the income sub ect to taxes varies Significant y between the four quarters O the year, the tax ratio could also vary. 1 Quarterly__ Quarterly Income (Quarterly Income r Tax _ ‘22 (Subject to Taxes + °26 Subject to Taxes .. 6’2"O Provision 2. Aggial = 22 Annual Income + .26 Annual Income Provision ° Subject to Taxes Subject to Taxes ‘ 25.000 The following is an example of such possible variations: Quarter Income Subject To Taxes Tax TangIBT First $200,250 $ 94,495 .4719 Second 200,250 94,495 .4719 Third 300,250 142,495 .4746 Fourth 300,250 142,495 .4746 Annual $1,001,000 $473,980 .4735 While the quarterly tax/NIBT ratio may vary in cases such as the above the effect is generally limited. In the following analysis it is assumed that the annual tax rate is applicable in each quarter of that year. 111 sign test can be used to determine if this percentage varies depending upon the difference between the net income for the quarter under review and for the same quarter of the preceding year. In this analysis an assumption is made that one of management's objectives is to minimize any unfavorable differ— ence between the current quarter's income and income for the same quarter of the prior year. The sign test will be used to determine if the quar- terly tax/NIBT varies from annual tax/NIBT: 1. To "improve" net income in situations where the application of the annual tax rate would have resulted in net income less than that of the same quarter of the prior year. 2. To "reduce” net income in situations where the application of the annual tax rate would have resulted in net income being greater than that of the same quarter of the prior year. In determining whether it "improves” or "reduces" re- sults, a comparison has been made between quarterly tax/NIBT and annual tax/NIBT. In this analysis, the federal income tax will be considered to ”improve" if the tax charged (credited) on a quarterly basis increases net income after taxes over, or reduces net loss after taxes under, what it would have been if the annual tax percentage had been used. Similarly, the federal income tax will be considered to ”reduce" if the tax charged (credited) on a quarterly basis 112 reduces net income after taxes under, or increases net loss after taxes over, what it would have been if the annual tax percentage had been used. The criteria for determining whether the quarterly tax/NIBT "improves" or "reduces" net income after taxes is presented in Figure 5-2. Statistical Test to be Used The statistical test used in determining if the fluc— tuations in the quarterly tax/NIBT ratio affect quarterly net income in the manner stated above is the sign test. This is a nonparametric test "which is based on the signs of the ob- served differences (that is, whether they are positive or negative) instead of their actual magnitudes."l This test is especially applicable to the present case because it does not require that the observations be independent.2 In the pro— posed test, the Observations are not independent as the amount of tax charged in one period (and therefore the tax/NIBT) may affect the amount of tax charged in a subsequent period (and therefore the tax/NIBT). For example, if the first quarter income statement shows only a $50,000 tax provision, but actually $75,000 should have been charged, one, two, or 1John E. Freund, Modern Elementary Statistics, (Pren— tice—Hall, Inc., Englewood Cliffs, New Jersey, 1967) 317. 21bid. 113 OEOUOH Dos\owcomxm xmu .ouo “cows soon pug ommucoouom xmu Hmocom map ma sown o>m£ UHSOB DH umnz moons moxmu Houmm OEOUQH no: ooscon Op 0Q HHH3 uaomon 03D OHDMH xmu THOMOQ OEOUCH MOO\mmcwmxo xmu amazon map CMSD prmmum mH oflumh xmn whomon AN sssaoo .H some "mgmdB Qfimm OH 303 :0>OHQEN: OD cows ma xme I mmmq :moocou: on own: ma xme I umpmono zooooonz on com: me xme Ono mmoa no Monmouw Honpflm Huouumov ozu NH :ooDUOH: OD poms ma xme 03D mmma Ho umpmouo moguam sooscou: OD coho we x09 Ono mmoH HO HODMOHO Honpflm QOQCmIOH mOXMB whommm wmoq D02 omommxm NOB OLD MH :0>OHQEN: on Cows ma xme map mwoa no Houmouo Hoauflm couscous OD cows we xme I mqu :m>ou@Efl: on com: me xme I Houmono :o>OHmEH: on com: me xme 0:“ mmoa no Houmono nonuflm :o>OHmEH: Op cows we Noe Ono mmoa Ho Houmono HOSDHm Q...mlnfi moxme mnowmm OEOUCH uoz DHOOHO NOB map wH :OOSUOH: :ooooouz :o>OHmEa: smoocon: on com: on own: ....wH on poms ma xme OD cows ma xmB ma xma I wmoq ma x09 I mmoq moxma onommm may mmoa HO Ono mmOH Ho =o>onmfiflz :o>OHmEN: mmoq uoz HOHMUHO HOSHHM HOHMOHU HOQUflm OD Ummfi mfi Op UOmD wH #HUOHU XMB xme I noumono NOB I umpmouo OLD MH :o>ouQEH: so>onmfiflr ....wH :o>OHmEH: :moopon= ma WMBCWmWwoq ma mmemeMmoq moxmB ouomom on own: ma xme on own: mfl xme . . oEOocH uoz on» mmoa Ho opp mmoH HO :oonoon: :ooscmnz Monmouo Honuflm Houmono Hmfluflm OD cows we Op cows ma mmcmmwm NOB xme I nonmono xme I Monmouo may MH mflXfiB mmommm mmXomm2H N48 MEOUZH Q