A smcmsnc momma: mama T0 MEASUREMENT or MERGER success Thesis for the Degree of Ph. 9., MICHIGAN STATE UNIVERSETY DONALD HENRY WORT 1 97 3 MiChign h - ' University This is to certify that the thesis entitled A STOCHASTIC DOMINANCE APPROACH TO MEASUREMENT OF MERGER SUCCESS presented by Donald Henry Wort has been accepted towards fulfillment of the requirements for Ph.D Business - Finance degree in ill/MW; Major professor Date OZ//\‘§/9 i? 0-7639 - V magma BY ‘5" :: HUM? 5 SUNS' 3' 800K BINDERY INC. LIBRARY BINDERS nun-emu Ian-mu- ABSTRACT A STOCHASTIC DOMINANCE APPROACH TO HEASUREHENT OF HERGER SUCCESS BY Donald Henry Hort The purpose of this study is to determine whether or not large corporate mergers have been generally successful in increasing the wealth of the merging flrms' stockholders. There have been several studies in recent years dealing with merger profitability. but for various reasons, an unambiguous resolution of the problem has not yet been achieved. The general methodology consists of a comparison of the fre- quency distribution of aggregate market value for a population of weighted combinations of acquiring and acquired firms for a period of time prior to merger to the frequency distribution of aggregate market value for the same papulation of firms subsequent to merger for a time period of the same length. The test papulation of thirty- two large mining and manufacturing firm mergers represents the re- sult of a rigid elimination process designed to yield only firms for which the ”merger effect" can be relatively isolated from other in- dividual firm effects. Although it would be preferable to measure merger success directly in terms of increased stockholder utility, a link to a Donald Henry Hort tangible market-based measure of comparison is clearly necessary. The total market value of common stockholders' equity was selected as the measure because it takes into consideration the premerger to post- merger change in the lgxgl_of stockholders' wealth as well as the pre- merger to postmerger change in the growth rate of this wealth position. The stochastic dominance comparison criterion was selected rather than the more familiar mean-variance criterion because: (I) Stochastic dominance compares complete distributions rather then estimated parameters of the distributions. (2) Recent studies of common stock market price distri- butions have indicated the unreliability of variance computations for these distributions. (3) Stochastic dominance can be used to measure stock- holders' evaluation of merger performance without specifying either their utility functions beyond non- satiety (and general risk aversion for second degree dominance) or the statistical distribution of the performance measure with which it is assumed to be directly related. The market value data used for each merging firm in the test pepulation are sixty weekly observations for each of the premerger and postmerger distributions--a total of one hundred and twenty obser- vations per merger. A transition period of approximately one year is allowed between the end of the premerger period and the beginning of the postmerger period to avoid including the relatively erratic price Donald Henry Hort behavior that is often found to exist between a merger's announcement and its eventual completion. The results of this study indicate both first and second degree dominance for the aggregate postmerger market value distribution. In other words, the aggregate wealth of the shareholders of merging firms was greater after the merger than it was before the merger. This is an 25_ggg£ evaluation and is only directly applicable to the merger pepulation and related time period that are specified in this study. However. with qualification, it can be stated that based on the re- sults found, mergers contribute to the aggregate wealth position of the participating firms' stockholders. This is of particular interest because of the fact that most prior studies have indicated that mer- gers are not profitable, except to stockholders of acquired firms for which excessive premiums have been paid. A comparison was also made of premerger and postmerger market value distributions for each individual test merger. If it is assumed that investors recognize the diluting effects of shares that are like- ly to soon become outstanding as a result of various contractual con- version arrangements, the individual merger results confirm those of the aggregate distribution comparison. On the other hand, if it is assumed that investors only consider officially outstanding shares when setting the share price at which they are willing to trade in the mar- ket, a majority of the individual mergers show premerger dominance. The results of a comparison of stockholder return (as opposed to market value) distributions were indeterminate in terms of first Donald Henry Hort degree stochastic dominance, but showed second degree dominance for the premerger aggregate distribution. While return distributions are not sufficient as a total measure for merger success, these re- sults do indicate that the aggregate postmerger dominance found in this study is a result of a market value level increase rather than an increase in the rate of return to stockholders. A STOCHASTIC DOMINANCE APPROACH TO MEASUREMENT OF MERGER SUCCESS BY Donald Henry Hort 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 1973 C’ . Efcj.:L.w?uf? d§>C0pyright by DONALD HENRY HORT l973 ACKNOWLEDGHENTS There are several people who have contributed either direct- ly or indirectly to the successful completion of this thesis. During the course of my program at Michigan State University, those who help- ed most in my reaching the point of writing a thesis include Dr. James Don Edwards and Dr. Gardner H. Jones, both of whom were very instru- mental in arranging the necessary financial support. The Earhart Foundation was also of considerable help in this respect. Hy faculty advisers, Dr. Hyles S. Delano and Dr. Alden C. Olson, were always available and helpful in shaping a pregram best suited for my personal goals. Special mention must also go to Dr. C. Robert Carlson, who as both fellow student and friend has helped to provide the moral support always necessary for undertakings such as these. Those who have contributed most directly to this thesis are my committee members Dr. Hyles S. Delano, Dr. Gardner H. Jones, and Dr. Ronald H. Marshall. Their guidance, c00peration and consideration has been greatly appreciated. The considerable contribution of Dr. Hyles S. Delano must be emphasized. Without the Opportunity to work with him, this particular thesis would not have been written. My life has been enhanced in many ways by my parents and I thank them again for the special encouragement and support which they have never failed to offer. The debt l owe to my wife, Karen, cannot be adequately expressed. She has unhesitatingiy integrated her goals with mine in a way that has truly made them ”ours.“ Without her love and understanding and that of our children, Brian, Robert and Kelly, there would be no need for these acknowledgments. TABLE OF CONTENTS Page ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . ............... vl Chapter I INTRODUCTION . . . . . . ........ . . . . . . . l Goal. . ....................... l Purpose . . . . . . ............... . . 2 Scape of Study . . . . . . ......... . . . . 3 Rationale for Methodology . ....... . ..... l7 ll PRIOR RESEARCH OF MERGER PROFITABILITY . ..... . . 20 Ill METHODOLOGY . . . . . .......... . . . . . . . 2h The Model ...................... 2h Axioms for the Data Preparation Model ..... . . . 26 Data Sources. . . . . . . . . . . . . . . . . . . . . 27 Merger Selection-Criteria . . . . . ......... 3i Market Value Distributions. ........ . . . . . 32 IV RESULTS AND CONCLUSIONS . . ............ . . 35 v LIMITATIONS, IMPLICATIONS AND RECOMMENDATIONS . . . . . 56 Limitations . . ............... . . . . 56 Implications. . ................ . . . 58 Recommendations ......... . . . . . . . . . . 58 Appendices A STOCHASTIC OOMINANCE ................. 60 3 PERFORMANCE MEASUREMENT MODEL . . . . ......... 70 B 'BL'OGMPHY O O O O O O O O O O O O ....... O I O O O O O 75 LIST OF TABLES Table Page 1 EXAMPLE lLLUSTRATlON OF MARKET ADJUSTMENT TECHNIQUE. . . 6 2 SUMMARY OF STOCHASTIC DOMINANCE COMPARISON FOR INDIVIDUAL MERGERS ................. . . 36 3 RESULTS OF INDIVIDUAL STOCHASTIC DOMINANCE COMPARISONS NAIVE METHOD OF COMPUTING OUTSTANDING SHARES . . . . . . A0 A RESULTS OF INDIVIDUAL STOCHASTIC DOMINANCE COMPARISONS SOPHISTICATED METHOD OF COMPUTING OUTSTANDING SHARES . . H2 5 STOCHASTIC DOMINANCE COMPARISON OF AGGREGATE PREMERGER AND POSTMERGER MARKET VALUE DISTRIBUTIONS - NAIVE METHOD #4 6 STOCHASTIC DOMINANCE COMPARISON OF AGGREGATE PREMERGER AND POSTMERGER MARKET VALUE DISTRIBUTIONS - SOPHISTICATED METHOD . . . . . . . . . . . . . . . . . . A7 7 STOCHASTIC DOMINANCE COMPARISON OF AGGREGATE PREMERGER AND POSTMERGER RETURN DISTRIBUTIONS - NAIVE METHOD . . . 50 8 STOCHASTIC DOMINANCE COMPARISON OF AGGREGATE PREMERGER AID POSTMERGER RETURN DISTRIBUTIONS - SOPHISTICATED HE THOD C O O C O O O O O O O O C O O O O O O O O O O O O S 3 A-I OBSERVED PORTFOLIO RETURNS - HYPOTHETICAL EXAMPLE . . . 66 A-Z COMPUTED DATA MATRIX . . . . . . . . . . ........ 68 LIST OF FIGURES Figure Page i Verbal Description - Data Preparation Model . . . . . . 28 2 Symbolic Description - Data Preparation Model . . . . . 29 A-i E-V Ordering . . . . . . . . . . . . . . . . . . . . . 6h A-2 FSD Ordering . . . . . . . . . . . . . . . . . . . . . 65 A-3 Graphical Representation of Stochastic Dominance Comparison-Hypothetical Example . . . . . . . . . . . . 69 B-l Performance Measurement Model . . . . . . . . . . . . . 72 vi CHAPTER I INTRODUCTION In recent years, the profitability of corporate mergers has been a subject of an increasing volume of academic study. While most researchers have concluded that mergers have, in general, not been significantly profitable, their results have not satisfactorily laid the question to rest. §__L The goal of this study is to measure the degree of success (or failure) of corporate mergers in terms of their contribution to the maximization of shareholder wealth. This measurement is accomplished by comparing the frequency distribution of aggregate market value for a population of weighted combinations of acquiring and acquired firms for a period of time prior to merger to the frequency distribution of aggregate market value for the same population of finms subsequent to merger for a time period of the same length. Generalizations concern- ing merger success are made primarily for the aggregated results of the individual mergers studied, rather than for the results of each individual merger. However, results of the latter type are used for purposes of comparison with the results of previous merger profit- ability studies. PURPOSE The purpose of this study is to fill a void in the field of merger research, viz., the direct measurement of merger success using the premerger and postmerger data of an aggregate of individual mer- ger firms in such a way that both return and risk are taken into account. Until quite recently, prior research of merger results has been limited mostly to such approaches as measurement of the degree of diversification provided by mergers in general or to comparisons of postmerger returns of "merging firms" with the returns over the same period of "non-merging firms." Use of the aggregated market value distributions in this study provides an indication of whether or not the combinations of securi- ties available to common stock investors after the mergers are superior to combinations available before the mergers. This type of comparison is more important in terms of overall merger performance than measurement of the degree of diversification provided by the mergers because a similar degree of diversification could have been obtained by the investors themselves by rearranging the combination of securities in their portfolios. Thus, an attempt is made here to determine whether the merger movement has contributed to the overall economic welfare of investors. The measurement technique used in this study (stochastic dominance) is also an important factor in achieving the stated pur- pose because, as used, it is a method of measuring stockholder evalu- ation of merger performance without specifying either the utility function of investors beyond general risk aversion or the statistical distribution of the performance measure with which it is assumed to be directly related. SCOngOF STUDY A merger is defined herein as any combination of acquisition involving the purchase or transfer of ownership of a company that was previously under separate control. Total market value of common equity is defined as market price per share multiplied by the total number of shares outstanding.l Only mergers effecting a substantial increase in the total asset size of the acquiring firm are considered. As an arbitrary standard, a fifty percent increase in the size of the total assets of 2 is used as a minimum ”substantial" increase. Small- the acquiring firm scale merger transactions would not be as likely to have a measurable effect on the market value distribution of the acquiring firm even if such an effect occurred. Partial mergers (acquisition of a part of another firm) and multifirm mergers (more than two firms combining) are not used. Because of limitations of data availability, the test mergers involve only acquired manufacturing or mining companies which had assets of at least SlO million at the time of acquisition. The merger lFor reasons that are discussed later, results are also ob- tained using shares outstanding plus shares being held in both speci- field and unSpecified treasury stock reserves. 2Mergers in which the total assets of the acquiring firms are less than 50 percent of the total assets of the acquired firm are also‘ eliminated from consideration. I. data, other than the market value data, was obtained from the Federal Trade Commission's Statistical Report No._], "Large Mergers in Manu- facturing and Mining, i9h8—l970.” This data was obtained by the FTC from public sources, such as The Wall Street Journal, Moody's Indus- trial Manual, Standard anngoors Corporation Records, and prospectuses filed with the Securities and Exchange Commission. All types of mergers-~vertical, horizontal and conglomerate-- are included together in the test pOpulation. No specific considera- tion is given (nor was it necessary, given the methodology used) to the method of accounting (purchase or pooling), the method finan- cing, the exchange ratio, or the particular reasons for mergers. As to this last point, merger "success" in this study is defined as re- lating exclusively to the maximization of shareholder wealth, not- withstanding any other goals of the combining firms' managements.3 Use of market value distributions rather than shareholder return (market value change) distributions is suggested by the nature of the measurement attempted--i.e., a measurement of the valuation of firm performance before and after merger. There are two submeasures of "success" involved: (l) the overall premerger to postmerger change in market value level and 3For an interesting discussion of management goals vs. stock- holder goals in a merger context, see Samuel R. Reid, Mergers, Managers, and the Economy, McGraw-Hili, lnc., l968. (2) the overall premerger to postmerger change In market value growth.“ Use of shareholder return distributions would adequately represent change in performance in terms of market value growth, but would lg- nore the change in market value level. While comparison of mean market value levels would ignore changes In growth, use of market value distributions in conjunction with the stochastic dominance technique allows consideration of both level and growth by comparing the market value distributions themselves rather than selected para- meters. The extent to which the market values of the aggregate pre- merger and postmerger pepulations have been affected over time by changes in the general economy are corrected by dividing each market value observation for each firm by the Standard and Poors SOO Compo- site Index value corresponding to the same date. An example of this adjustment technique is illustrated in Table I. Removal of the general time-related trend in stock market prices is necessary to make comparable the firm market values within and between premerger and postmerger time periods. This is accomp- lished by the above procedure because market index percentage changes l‘I-‘or a more detailed discussion, see U. E. Reinhardt, "Con- glomerate Earnings for Share: Immediate and Post-Merger Effects," Acgggnting Review, XLVII (April, l972), 360-370. TABLE I EXAMPLE ILLUSTRATION OF MARKET ADJUSTMENT TECHNIQUE s s P Adjusted Firm Market Index Firm Observation Valge. SMillions Value Market Value I IOOO IOO I0.00 2 IIOO IOS IO.H8 3 l300 llS il.30 h IZOO IIO IO.90 include the general time-related trend. Removal of specific time- reiated trends would not be appropriate because market-adjusted growth over the test period is one of the performance components being measured. Removal of between-distribution time trends Is accomplish- ed by the conversion of all the market value observations to a common base. This adjustment technique might be unsatisfactory if the ad- justed values obtained for the individual firms were then used to measure the success of the individual mergers involved. It is well known that the individual firms may have a typical percentage change relationship with the percentage change of "the market” ranging any- where between (-I)-to-(+l) and (+l)-to-(+I). This market relation- ship measure is usually referred to as the beta value. To measure this beta value for each firm would require regressing each firm's market values on the selected market index for a period of time prior to both the premerger and postmerger time periods. This would not be practicable because market value distributions do not meet the require- ments for use of ordinary least squares techniques. (The preperties of market value distributions are discussed in more detail in Chapter III). Other problems with computing beta values are: (l) The stability of beta values over time Is questionable. (2) The proper time interval for the computation of the beta values has not yet been resolved. Fortunately, there is empirical evidence that the diversity of in- dividual beta values within a security portfolio of the size repre- sented by the aggregations used in this study (32 firms) is not Im- portant. The beta value for a randomly selected common stock port- folio of 32 different firms is very likely to approximate l.0.5 The choice of the Standard and Poors 500 Stock Index was made because: I (i) This index is considered to be broad enough to serve as general standard for stock price movements. (2) This index is based on market value aggregations similar to the ones used for the test firms. Although removal of industry effects might also be beneficial, this is not a practicable procedure for merger analysis because, along with the ordinary difficulty of categorizing firms by industry in a meaningful way, there is also the problem represented by the fact that firms often move from one uncertain category to another by merging. 5John L. Evans and Stephen H. Archer, "Diversification and the Reduction of Dispersion: An Empirical Analysis,” Journal of Finance, XXIII (December, l968), 76l-769. The market value data used for each firm are sixty weekly observations for each of the premerger and the postmerger distribu- tions--a total of one hundred and twenty observations per merger. However, the premerger period begins twenty-four months prior to the effective merger date, and ends approximately ten months prior to the effective merger date. The postmerger period begins two months after the effective merger date and ends approximately fourteen months after the effective merger date. This results in a gap of approxi- mately one year between the end of the premerger period and the be- ginning of the postmerger period. The reason for allowing this time gap is that common stock prices are known to frequently behave errati- cally during a transition period beginning with the time that Inves- tors first recOgnize the merger attempt by revising their expectations of future performance, and ending with the time that investors are aware that the merger has been completed with some known terms of consideration (used in the legal sense) and have some initial im- pression of the newly combined organization. Prior researchers have estimated the premerger part of the transition period described above as typically beginning about six to eight months before the effective merger date,6 making the ten months allowed likely to be adequate for most cases. The selection of two months after the effective date to 6Thomas F. Hogarty, "The Profitability of Corporate Mergers," Journal of Business, XLIII (July, l970), 3l7-327. represent the postmerger part of the transition period is relatively arbitrary. it is selected because-it "seems” to be a reasonable length of time after the merger data for the criteria described above to be essentially fulfilled. The effect of individual firm events, other than the test merger, on the market value distributions within the total test period will be ignored, except for the occurrence of other individual mergers or combinations of mergers which would cause a 50 percent total asset expansion. To the extent that other individual factors (such as. announcements of new internally generated product lines, management changes, and product obsolescence) are not related to the merger, but have long-run effects on market value distributions during the test period, measurement of success for an individual case could be con- founded. However, it is assumed herein that such extraneous indivi- dual effects are not systematically related such that the aggregate data would also be confounded. The random variable in the distributions being studied is the market value of common stockholders' equity. The market value at time t, V is defined as the product of the common stock market price per t' share at time t, Pt, and the total number of perceptibly outstanding shares of common stock at time t, "t3 i.e., Vt - Ptnt. In one of the test runs, the market value at the ex-cash-dividend period for each firm is adjusted by adding an amount equal to the tOtal dollars of cash dividends paid. Previous research has verified that market price is usually lowered at the ex-dividend date by at least a substantial l0 percentage of the cash divident paid.7 By "perceptibly outstanding" shares is meant the sum of the shares currently outstanding and those that are perceived by investors to be likely to become outstanding at some imminent, albeit uncertain, point in time. Accountants use such a conceptual measure of shares outstanding to compute ”primary" and ”fully-diluted" earnings per share figures for reporting purposes. However, the computational methods suggested in the Accounting Principles Board APB No. l58 are not used in this study for the following reasons: (i) The APB No. l5 methods do not include unspecified treasury stock holdings, which also represent issued shares which could be (and often are) publicly resold at any time. (2) The methods used to decide which convertible securi- ties to use in the computation of "primary” common stock equivalents are quite controversial.9 Even the number of shares used for "fully-diluted” earningsper 7See Durand and May, "The Ex-Dividend Behavior of American Telephone and Telegraph Stock," Journal of Finance, XV, (l960), I9-3l. QAccounting Principles Board of the American Institute of Certified Public Accountants,“Earnings Per Sharef'gginion No.gl§, (AICPA, I969). 9See W. Frank and J. Weygandt, ”Convertible Debt and Earnings Per Share: Pragmatism vs. Good Theory," The Accounting Review, XLV (April, l970), 280-289. share figures are "based on neither the probability of conversion or exercise nor on their lmminence. They are rather, computations based on arbitrary rules and assumptions, without evidence that either computation is necessarily relevant for investment decisions.”'0 (3) While the suggested methods might be acceptable for a study covering a time period beginning I969 because es- timates of the number of "primary" and "fully-diluted" shares have since that time been an item of information readily accessible to investors, they would not be of such value in this study because they were not in general use during seven of the ten years that are covered. Two other methods of computing perceptibly outstanding shares are usedin this Study. The first method is called the ”naive" method and uses only the currently outstanding shares, thus representing the minimum value for this measure (if one ignores the possibility of pend- ing stock repurchases). The sources used to obtain the number of out- standing shares are the quarterly volumes of ISL Price Lists for the New Yorkll and American'2 Stock Exchanges. Although daily observations loEldon S. Hendriksen, Accountin Theor , Revised Edition (Nomewood, lll.: Richard D. Irwin, Inc., I970). 553. HInvestment Statistics Laboratory, ISL Daily Stock Price Index, New York Stock Exchan (New York: Standard and Poors Corporation, I9 2-70 . 'zlnvestment Statistics Laboratory, ISL Daily Stock Price Index, American Stock Exchangg (New York: Standard and Poors Corporation, '9 2-70 e I2 of market prices are available in these volumes, the number of out- standing shares is updated on a quarterly basis. Thus, the first method employs quarterly updated figures for shares outstanding, ex- cept for Interim updatings occurring because of stock Splits and stock dividends. While the first method of computing the number of perceptibly outstanding shares assumes that investors are generally naive and un- perceptlve, the second method makes the directly Oppostie Implicit assumption, i.e., that investors are very sOphisticated and per- ceptive. Therefore, the second method is called the "sephisticated" method and assumes that stockholders are net only aware of common stock share equivalents through Imminent conversions, exercises of options, etc., but actively take these share equivalents Into account in their market price determinations. The number of share equivalents at each market price observation is not computed because of data un- availability. The specific procedure used is as follows: (i) From annual volumes of Mggdy's Industrial Manual, data is obtained for the number of shares outstanding, the total number of unspecified treasury shares, and the total number of treasury and/or unissued shares speci- fied as reserves for such contingencies as conversion of canvertible securities, exercise of warrants, exer- cise of Options, and accumulation for acquisitions. This data was obtained for each test firm for a period extend- ing from the last reporting date prior to the test period to the first reporting data after the test period. I3 (2) These share figures are summed for each firm at each reporting date and used to represent the perceptibly outstanding shares for the sOphisticated method. Since the figures are only available for all the test firms on an annual basis, the updating Includes only the test period beginning data (using the perceptibly outstanding shares from the last previous re- porting date) and any reporting dates that occurred within the test period. However, interim updatings are used at stock split dates, stock dividend dates, and convertible security issuance dates. There are only two of the thirty-two firms studied that have a convertible issue outstanding without a treasury stock reserve for its conversion contingency. In both of these cases the convertible issue involved is a postmerger carryover originated by the acquired firm. The rationale for using managements' estimates of share equivalents, represented by their specified treasury stock reserves, seems clear. It is in managements' best Interest to maintain treasury stock reserves that would adequately cover any imminent or potentially imminent contractual demands. The classification of unspecified treasury stock as perceptibly outstanding shares is not as clearly relevant, but is used in this study in order to estimate a maximum number of shares that might be considered imminently outstanding. Unissued authorized shares which are not Specified as reserves do not meet the imminence requirement and are not used. Announcements of new common stock issues and announcements of common stock repurchase plans would certainly affect the number of perceptibly outstanding shares during the period between such an announcement and its completion. IA However, because of the difficulty of obtaining accurate first- announcement dates for all of the firms studied and the likelihood that the ignoring of this Information would not have a systematic effect on the comparison of premerger and postmerger market value distributions, these announcements are not considered in the compu- tations. The market price data is obtained from the ISL Price Lists for the New York and American Stock Exchanges. Friday closing prices are used as the weekly market price observations. When Friday closing prices are not available, the just previously available daily closing price is used. For consistency, the Friday closing values for the Standard and Poors 500 Composite Index are used in the market adUust- ment process described earlier. These index values are also obtained from the ISL Price Lists. The total test period over which market values are collected for this study is from I962 to l97l--a ten year period which shows considerable diversity of common stock price movements. Since each point in the premerger and postmerger aggregate market value distributions represents a summation of market values occuning at different points in time, the adjustment for general market index movements Is clearly necessary. However, this necessity is alleviated somewhat by the fact that both the premerger and the postmerger data aggregate points cover a common time period begin- ning in I96h and ending in l969, which represents sixty percent of the total period. IS The stochastic dominance criterion (which is further described in Appendix A) is the technique used to compare the aggregate pre- merger and postmerger market value distributions. The aggregate pre- merger distribution consists of a series of sixty weekly market value observations during the previously defined premerger test period. Each observation is adjusted for general stock market-related and general time-related common stock price movements. Each point in this distribution is a sum of sixty-four individual firm premerger market values (thirty-two acquiring firms plus thirty-two acquired firms). Although the actual sixty-week test periods for these firms differ in time from merger-to-merger, the time period is Identical for acquiring and acquired firms within each merger. The common bases for each of the sixty-four summands making up a single point in the aggregate premerger distribution are: (I) Each is an observation taken at a specified number of weeks prior to one of the test mergers. (2) Each is an observation that has been adjusted for the change that has occurred in a general stock market index since the last previous observation. This adUustment includes allowance for general time-related trends. The aggregate postmerger distribution is similarly defined except that observations for only thirty-two firms make up each distribution point. Each merger firm is the postmerger counterpart of the premerger com- bination of acquiring and acquired firms from which it derived. As will be discussed more fully in Chapter III, both first and second degree stochastic dominance criteria are applied to the I6 aggregate distributions being compared. The first degree stochastic dominance (hereafter referred to as FSD) criterion iteratively com- pares the cumulative frequency dlstributions from lowest to highest market values and signifies dominance for one of the distributions If its cumulative frequency is always less than or equal to (with at least one point less than) the other distribution. Investors' utility is assumed In this study to be directly related to that por- tion of their wealth which can be measured by the market value of their common stock holdings. Thus, If investors are assumed to pre- fer more wealth to less, the only utility function specification necessary for the F50 criterion is that of monotonically increasing utility with increasing market value. Because the F50 criterion com- pares the entire distributions, no parametric specifications of the market value distributions are required. In other words, the shape or type of the market value distributions is Irrelevant. The second degree stochastic dominance (hereafter referred to as $80) criterion lteratively compares the areas under the cumulative frequency distributions from lowest to highest market values and signifies dominance for one of the distributions if its area is al- ways less than or equal to (with the area being less at least one point) the other distribution. The FSD criterion does not allow a dominance determination if the cumulative distributions being com- pared cross at any point. Such crossings occur due to differences in market value variability between the distributions. For example, for distributions having equal mean market values, the one with the lower variability would have a smaller area under its cumulative '7 distribution. Thus, by allowing the comulative distributions to cross and by determining dominance in terms of the areas under these distributions, general variability is taken into consideration. Assuming that general market value variability is an acceptable mea- sure of the general riskiness of the shareholders' wealth positions, it can be seen that the $50 criterion adds the utility function speci- fication that investors are risk averse in a general sense, i.e., they prefer less risk to more risk. Since the $50 criterion also requires no parametric specifications, there is still no need to specify the shape or type of the market value distributions. RATIONALE FOR HETHOQQLQSY To conclude this introductory chapter, a summarized rationale will be offered for the methodology used in this study, i.e., the comparison of market value distributions by means of stochastic dominance criteria. The selection of market value as the measure for comparison has previously been shown to be necessary given the defini- tion of merger success being used. It can be further shown that this measure is also sufficient for the purpose of shareholder wealth maximization through an explanation of the "market value rule” used in a recent work by Eugene Fama and Merton Miller.’3 The market value rule for making investment decisions is defined as the maximization of the market value of those securities of the firm that are outstanding at the time the investment decision is to be made. This rule is the basis of virtually all financial decision theories and, as used by Fama l3Eugene F. Fame and Merton H. Miller, The Theorx,of Finance (New York: Holt, Rinehart and Winston, l972). l8 and Miller, implies a separation of investment and financing decisions through assumption of perfect capital markets. More importantly, through the combination of market value maximization and the perfect market assumption, the investment decision is effectively separated from the requirement of specifying stockholders' utility functions. By employing decision models designed to maximize the market value of common equity, the management can effectively leave the utility satis- faction decisions up to the individual shareholders. That is, given the stockholders' wealth, the firm's investment decisions do not affect the consumption--investment opportunities that are available to the stockholders In the market. Therefore, the only thing that the firm can affect by their investment decisions is the stockholders' wealth, as represented by the market value of their equity. The problem with the application of the market value rule to an empirical situation is the fact that capital markets are not per- fect. However, there is considerable evidence that capital markets are reasonably efficient,lu which is another way of saying that common stock prices fully reflect all available information. This is assumed to be sufficient for the separation of the investment and financing decisions but is not assumed to be necessarily sufficient for the separation of investment decisions from the necessity to specify stock- holder utility.functions. This latter requirement of specifying "is“ Eugene r. Fame, "Efficient Capital Markets: A Review of Theory and Empirical Work," Journal of Finance, XXV (Hay, I970), 383- hi7. l9 utility functions is rendered unnecessary by use of the first and second degree dominance criteria described above. Thus, the compari- son of premerger and postmerger common equity market value distribu- tions by means of stochastic dominance criteria can be used to provide a valid measure of merger success in terms of the maximization of both the wealth and the utility functions of common shareholders. CHAPTER II PRIOR RESEARCH 0F MERGER PROFIIABILITY Many merger research studies have considered such questions as the effect of mergers on competition, the determination of ex- change ratios, and proposed reasons for merging. However, the only studies that will be considered here are those relating to measure- ment of merger performance. Until fairly recently, there had not been many studies in this area, and those that had been done had usually contrasted postmerger performance measures for merging and nonmerging firms in general, rather than comparing performance data before and after merger for the specific firms involved. in l966, Alberts and SegallI pointed out that there had been a noticeable lack of empirical research into the profitability of growth by merger. in fact, they were unable to find a single published study of any kind on the results of post-world ll mergers. Since that time, however, there have been a number of published studies dealing to some extent with enpirical profitability results from more recent time periods. The most consistent conclusion of these studies has been that mergers have not generally been successful for the acquiring firms. Both accounting-and market-based profitability measures have indicated this to be true. However, there has been evidence presented that stock- holders of firms that were acquired during this period have benefited 'Hilliam‘w. Alberts and Joel E. Segall, eds., The Corporate Merggr (Chicago: The University of Chicago Press, l966 . 20 2| significantly due to merger effects.2 The combination of these re- sults found for acquiring and acquired firms has led to the apparent- ly logical conclusion that premiums paid by acquiring firms in order to consumate mergers have been excessive and therefore detrimental to Investors holding shares of their stock prior to the merger effect on the value of these shares. This summary statement of results of prior merger research does not represent an irrefutable body of evidence. Merger success has not been measured within an integrated risk/return framework in which total synergistic effects can be properly evaluated. Total synergy is defined herein as the measure of performance superiority for the merged firm over the combined measure of performance of the acquiring and acquired firms before the merger. Except for a study by Gort and Hogarty,3 synergy measurements have been inappropriately attempted by either comparing postmerger results with the premerger results of the acquiring firm only or by comparing postmerger results of merged firms with some postmerger control sample. Although Sort and Hogarty prOperly measured synergistic effects by comparing combinations of acquiring and acquired firms premerger with the merged firms postmerger, they did not incorporate a risk mea- sure into their analysis. 2For example, see Stanley Block, "The Merger impact on Stock Price Movements," MSU Business Topics, Vol. l7, No. 2 (Spring, l969), 7-l2. 3Michael Sort and Thomas F. Hogarty, "New Evidence on Mergers," gournal of Law and Economics, Xlil (April, l970), l67-l8h. 22 A few researchers did specifically analyze risk as well as return in their merger evaluations. A study by Lev and Mandelker“ is a notable example. However, they limited their premerger data to that of the acquiring firms. In addition, they like others who have attempted to include a risk measure, represented riskiness by the standard deviation of the periodic return measurements. While this may be adequate for accounting return measures, recent examinations of common stock price change distributions have indicated that variance and standard deviation are unreliable measures of the variability of these distributions? Since only market-based return measures can be used in decision models for which the objective Is the maximization of shareholder wealth, this does represent a serious limitation to the measurement of merger success. With the exception of the previously mentioned Cort and Hogarty6 study, an earlier study by Hogarty,7 and a more recent work by Anson, Blandenburg, Portner and Radosevich,8 little attention has been accorded to changes in the market valuation level of common “Baruch Lev and Cershon Mandeiker, "The Microeconomic Conse- quences of Corporate Mergers," Journal of Business, XLV (January, l972), 85-10“. 5For a more detailed discussion, see Chapter III. 6Cort and Hogarty, 22. cit. 7Thomas F. Hogarty, "The Profitability of Corporate Mergers," Journal of Business, XLlli (July, I970), 3l7-327. 8H. igor Ansoff, Richard C. Blandenburg, Fred E. Portner, and Raymond Radosevlch, Ac uisition Behavior of U. S Manufacturin Firms, ififib-lfiéfi, (Nashville: Vanderbilt University Press, l97i 23 stockholders' investment. Most other studies have concentrated solely on a comparison of stockholder return measures. it is pointed out in Chapter l of this study that both level and growth rate of mar- ket value must be compared for an adequate measure of the total bene- fit accruing to common stockholders of merging firms. The methodology and general approach to the measurement of merger success outlined in the succeeding chapter of this study is designed to measure the total synergy effect of large corporate mer- gers within a framework which simultaneously considers both the level and the variability of total benefit to the common stockholders of the merging firms. The primary emphasis of this merger success mea- surement is concerned with the comparison of the performance of an aggregate premerger papulation with the same population's performance after merger; thus hopefully providing an answer to the question as to whether mergers in general have been successful in terms of contribu- ting positively to the overall market value of common stockholders' investment. CHAPTER III METHODOLOGY THE MODEL The model used in this study is separated into two sub- models, the Data Preparation Model (Figures i and 2) and the Perfor- mance Measurement Model (Figure B-l in Appendix B). The intent of the data preparation process is to remove the systematic nonmerger- related trends from both the premerger and the postmerger distribu- tions so that the aggregation and subsequent comparison of these dis- tributions can be validly accomplished. The actual market values used in the model are computed by multiplying the sum of the market price per share at time t (Pt) and the dividend per share received during the period from t-l to t (at) by the number of shares perceptibly outstanding at time t ("t)' vt - (Pt + Dt)(Ht) in the terminology used In the Data Preparation Model, an important axiom ls: V 'V . +V bij ler bije where v is the total market value of the common stockholders' equity: b signifies premerger period; i signifies the merger identification number, i - l, 2, ...32; j signifies the observation number, j - l, 2, ...60; r signifies the acquiring firm; 2h 25 e signifies the acquired firm. This axiom is not necessary for the primary comparison of the aggre- gate adjusted distributions, but is is necessary for the secondary comparison of individual merger distributions. its use assumes that the expected postmerger market value is equal to the sum of the ac- quiring and acquired firms' market values. E (vA) - vBr + v80 This also represents the aggregate market value of the two firms be- fore the merger and, as used in this study, is a measure of the aggregate wealth and thus the aggregate utility of the firms' stock- holders before the merger. The conclusion drawn from the individual comparisons must then be limited to statements concerning the aggre- gate utility of the stockholders Involved in each merger analyzed. Although it is true that the dominant distribution in a stochastic dominance comparison is also preferred by gash stockholder regardless of his utility function, it is not possible to generalize to other pairs of combinations. Limitation of the potential investment combinations to the market-weighted aggregate in a world of imperfect capital markets and heterogeneous investor expectations, also precludes the feasibility of setting up a realistic gx’ggtg decision model for the individual investor.l Thus 3522;; performance measurement on an aggregate basis is the proper focus for this study. 'Even if the assumptions of the capital asset pricing model could be accepted for this empirical study, it is not clear that this model can be transformed from the mean-variance framework, in which Its relevance has been established, to the stochastic dominance framework. 26 Similarly, conclusions drawn from the comparison of the aggregate premerger and postmerger adjusted distributions must be limited to statements concerning the aggregate utility of all in- vestors in the test populations. Returning to the computation of the actual premerger market values, vbijr " (Pbijr)("bijr)' '"d '( )( ). vhue Pbije+dbije "we The comparison of the aggregate premerger and postmerger adUusted distributions is accomplished through the use of first degree (F50) and second degree (550) stochastic dominance. An intuitive explana- tlon of how and why these criteria can be used to determine prefer- ence ordering with unspecified (F50) and minimally specified (550) investor utility functions and unspecified market value distribu- tions can be found in Chapter i. A detailed stochastic dominance performance measurement submodel is illustrated in Appendix B. The input data used are the aggregate premerger and postmerger adjusted market value distributions which are the output of the data prepara- tion submodel (See Figure 2). The comparison process for the indi- vidual mergers is identical and is not separately illustrated. Allgflfi FOR THE DATA PREPARATION MODEL l. Removal of market-related and time-related trends from both premerger and postmerger market value distributions will make 27 these distributions directly comparable. Any systematic differences which then show up in the distributions can be attributed to the mer- ger occurrences. 2. The tatal market value of the premerger equivalent of the merged firm is the sum of the market values of the acquiring and acquired firms. DATA SOURCES Lack of adequate data concerning mergers has been noted by at least one researcher 2 to be a factor which helps explain what has been until recently a dearth of empirical research into merger pro- fitability. (A survey of merger research can be found in Chapter ii). However, one source which has proved to be quite useful is the one used in this study--the annual Federal Trade Commission (FTC) statisti- cal report on large manufacturing and mining mergers, which includes data extending back to l9h8. For a merger to be included in this re- port, the acquired firm must be involved in mining or manufacturing and possess total assets of at least ten million dollars at the time of acquisition. All of the data included must also be available in public sources. For the l9h8-l970 period, the FTC estimates that over 70 percent of the total number of all large mergers and over 86 percent of total acquired assets are included. 2Samuel R. Reid, Mergers, Managers, and the Economy, New York: McGraw-Hill, inc., l968, p. 20. 28 .mec_w cameos NM ecu co mummeemma ecu com ae:_m> nexeme ceuuaaveiesmu vce uexcme __mce>o >_xeex >ux_m co_u:a_cun_a noun: ecenno >_xeez om use me some on no:_m> uuxeme nEe_w cameos Nm 0;» 55m _ mee_m cameos NM 0;» mo mummoemmm on» omm one one >n oemmsou nac_m cameos «m mo some com ne:_m> nexcme ueumzqveies_u centuuxcms __mcu>o >.xoos.>ux_m omm A| m cm... >3 uemacoucaou v cameos mu. zu_3 Ee_m 3232.133 edema—efcooeoeunoxm _e:e_>_uc_ 3mm e33 +05 mo :03 im>cewno xov:_ new use >3 mco_ue>comno e:_m> boxces owm. use we some oe_>_a _ —i‘ ee._c cameos an #0 some LOw ue:_m> nexcms uo>comao >.xoez >ux_m neo_»=e_.uw_o .mauu uexcee counsavmioe_a ccmiuuxcms __mee>o >_xeuz xux_m co_u:n_cum_a eeune.ccemao >_xoos ow use we come um mu:_m> uoxcme Ec_c euc_nEOu NM use esm _ .mee_m coc_n50u «m mo some cow mo:_m> uoxams moumamcmioa_u can tuoxcme __meo>o x_xuus xux_m .esr _m:u_>_vc. A, .oumu damn one mo meo_u im>comno xocc. mam ecu >3 mco_um>cemno o:_e> uexcma cum. one »o some ev_>_a _ mec_u Acoc_:uum can mc_c_:wumv coc_oEou Nm mo some ecu mo:_m> aumeE cm>eemno >_xee: >ux_m meo_u:n_cum_o .eauu(rc .esezsmee. .ovo: :o_umeeeoum mama i co_ue_cumoa .mnce> _ 0L:m_u 00....Ne—IH ——0 ‘3' mo~0_ Hun-K SeeseNe—IH ——0 3:0 :3: DD 29 I I N ~m....~._i_ ..e sec __e> ~m....~..i_ ... tee ..es s _ 11_m> . . . ~.e> . . . ..em. . fim.e> . . . H.n> . . . ..e> on I a u u u ow I m u u “ fi_m> . . . q_m> . . . ._m> I (5 due) . . . Hma> . . . __a> I n> rw_m> . . . q_m> . . . __mD. iw_a> . . . n.n> . . . __aoL pee wee ~m ... .N.. i _ ..e sec . an . . ..~._ a _ ..e Lee cm I a As_m> . . . .a_m> . . . .__m>v I_<> cm I a Am_a> . . . .qmn> . . . ..mn>v I_o> aces: acme: an I . A .(S . . . . _(5 . . . . .(5v I <5 NM I . A .u) . . . . .0) . . . . _n>v I o> eco_u:a_cue.o .mauu< emmweeawoe .m:c_>_ce_ weo_u:n_cum_a .eauu< cmueeeece .e:v_>_ce_ N ee=m_m mummecmm< enamecmm< A, . an e 8... 3 a . coe_ceoaou xeel_ee «m... use 7. um... use .... 8:13... :. .5.— ...-x .. Fe. 8....~._L :- ..8 sex i E... P _ , . o a . . . :ex . . . :ngll Jflex . . . fax . . . 2.5 3 n H u u u n 31.2}.-. ...-x... ......(x Siesta.-. ...ex... :extex «m I _ a_ex . . . q_ex . . . ..ex um I _ m.ex . . . H.4x . . . ..ax I I can A 11 l 1 tea Nazi... :e .8“. 2.3%.... :e .8 ow I a Am.mx . . . .q_mx . . . ...exv I_.z .mcm_m .mceceu :— cemceeecm canteens; moimuio_ Lox—m: eom_nce: .uc...cc_ Lemmas: m— eeesabeeee. 323.3%... 3-2-3 .8 52:3 .88 :8 N. 38.5.... 3.553%... 3-3-8 32.2. 5.9.28 e83»... .m... : 32.533... Ems-Cemete- ccimeo nae—.23 a sauce: «9:95... ccmccmum o. 388...... .3389... 3-8-2 .88 .8: 3:23. 3.258: one; a Lemceeeem cemceeecm ocim0i_o ..o c_e_mcu_¢ mc_c_mu¢ u_uce_u< m eumc_aceuecc_- eume_eceummc_ ooi.oi_o w.>ceseca mc_3ecm .uomm< m Lensesueoe Lemcoeuwoe moi_ci__ .ecoo ceuu.¢ u_u:ecee Le_memce o eumc_5ceuecc_ uuee_eeeuecc_ mwicMimc co_um_>< u._e:ee¢ ce__e: m__:uc_mu m eumc_5ceuemc_ euec_aceuocc. moi—ciao ..o mesa m_ccom._mu co ..o co_e= : 32.53%... 3.553%... 3-3-8 .8 3...... 3:3: 8.28.... .m... m Lemceaece Lumeoaeee :oim_imo me_euw:mc_ Lea-seek mu_cocuue_m cmEmuo: N camcoaumom cemcuauaoe eons—iuo .oeum eao< com. mxm_ceue. . 8» 8:3 8e 8:3 3451 6.. i.eaw_o acme_eoo i.eua_a ucec.eoe co_ue_oeou ae.u cec_:cu< aL_u mc.c_:vu< Lance: mugs; ozezhgo 92.5580 ...0 aim: was. mzcm.¢_c:_ no mbaamuc m maooz-ouoo4 .u:_ cooLOuU_> .m Leagues»; LomLoEsLm mm-ounoo .Eo:u new ..o mac: sac—Oguom ovngoe< on ou~:_suouovc_ ou~c_egouovc. awn—m-No .aLOu ou~>o_u >Louuoo ..uoz v_:oo mu Lomhoeumom cane—ELOHovc. mo.m~-mo .voLm amen gaguc< m:_:_: oc_:mc:m mu LumLoEumom LomgosumOm mouse-co .Lo L_< omaocmc_umo: unannoum cou_gus< mu Lomgosogm gumgosoLm mo-_-oo .nLoQ ungoacgot .QLOu .uu ow LomuoEun0¢ ouoc_egouov:_ mo.~o-_ .ALOu ugo___504 mucuaozh «.3004 mm .3323; .3388...— 8-Nonmo .3: :32...— muuavo: .322 an somuosogm Lomgoeogm mason-mo ~12 acum u_umo_u .QLOQ oungos< mu :3 5:3 8... Saw-.... . .fldd 6: -_guu_o acac_eoa -..uu_o acocmsoa :o_uo_aaou 35.; vo..:cu< 55.; m:_g_:uu< meuoz vo:c_u:ouanm m4csou .38 .8292 .ou 0.x uo...< .0500 0.00m 000.0h mom a ..c xuogsozm Bonn-um . 330.00.. exogn ..< .>.z Lox.o) cou.n.0= .ou c000..u nuoaoc 00.nsugu 530....) a caucogh .a.oo .m.x .o__o=m ..o v.0.uzu.¢ n.>.030.a .050u Louu.¢ no.u..>< u__a=ao¢ ..o 0.3. .0u «noun .0..0:x uo_.um=uc. ...-.... —8um 80< .vc. u.0u .vc. guoa .00.!0su .xuoqu .0500 030¢ugoxcan .ou >¢an=u u..uuo.m oucn..0¢ .00 ..qx.< v:050.o :0. 3.3.3.02 .0:m.m .0gocou .uc. ..vc. gouache .agou tum coax... .m.= um:.m.0m vgovcnum oxu.:.03 ono.u m=.c..o¢ o.u=._u< m:.30.n .00nn< u.u0550m 50.000». .020.. u..:ug.0u o.:.0m..0u .0 ..o '-0...: me.u.oem .m.= «0.005uuo.m coemmoz can. oxa..0uc. —N NN ”NM-3’ mommmo—Nmamxotscomg Lomuoaanom bong-lunom mmINN-c. Longueuuom somgoauuom monom-mo 503060.... 503030.... 3.. . «no. Lomgoaanom howsoEunom moumouoc .3305 «0.. .3325 «0.. 8-. gnu. Lem-.03 «0.. 509.953.. 3.3-“. gomgoaumom somgoeuaO. mo-m.-. .3985 «0.. 503053.. 3-3-8 .3305... 509360.... 3-3an 0u0:.agouovc. ouuc.egouovc_ swummuo. 0u~:.lgouooc. 0u0=.eLouovc. monuuumo Lomgoeuaom homgoauuom mou.c.mo 32.8300... 83.533... 3-3-3 LomuoauoO. gumboluoom women-N. 500505090. somuoaunom woumou.o ouac.agououc. au-c.e.ououc. oo-.o..o Longueuaom Lomgoauaom mwo.o-.. 0.0:.agouovc. ouoc.sgouovc. mw-cm-mo . mm:.oaoo Lombosunom homLQSuaom ouoc.a.ouovc. ouoc.a.ououc. mw-suamo Lomgologm gomuosou. :mlw.-mo Lomgoaunom Lomgoauaom :o-.-~o ...-...mmw.co.uaa on. no.8mm. ..mqum, -.Luu.a u:0:.50a -..ua.o acoc.loo c0.u0.nlou I... 005.:cu< Jan! a..u mc.g.:vu< Lease: uu¢¢um uz_a:¢th:o oz.h:mtou no oozhu: amh(u_hm.zm0m mzom.¢¢&zou uuz¢z.xoa u.hm¢:uphm .¢:a.>.nz. no whaamux : u.n§p #3 0000000000 0000000000 00-00-00 .0000 ..o 0_._u=.0 0.0000000 u00co_0< 00 000000000 000000000 00-.0.~0 .00 0...>0z-00000 .00. 000000o.> .0 000002000 000008000 00-00-00 .0000 000 ..0 000: 000.00000 00000e< on 0000000000 0000050000 00-.0-00 .0000 00.>0.0 >000000 ..002 0.000 00 0000.0000000. 0000.0000000. 00-00-00 .0000 0000 00000< 00.0.: 00.00000 0N 0000050000 0000000000 00.00-00.00 0.< 00:0000.0003 00000000 000.00§< «a 0000.0000000. 0000.5000000. 00.0.uoo .0000 00000000: .0000 .00 00 0000000000 0000080000 00-00-0. .0000 000...000 0000000» 0.3000 mu 0000.3000000. 0000.0000000. 00-00-00 .00. 000.00 00000000 00...< :0 000000000 000000000 00-00-00 002 0000 0.000.u .0000 0000000 n0 .000 500% {30W Jfll -.000.0 0000.300 -.000.0 0000.500 mmww0.0§00 00.0 000.000< 00.0 00.0.:00< 00000: 03528.-.. 0020 .roo .000 ozw «m 0.000 0.000“ mm. 00. «0 0.00m 0.000 mm. mm. «o m.m~m 0.500 0N. mm. «o 0.00m m.0mo «N. Na. an 0.:0m 0.m00 0m. ma. «m 1.00: 0.uwm 0m. . . «m 3.0m: u..:m mm. 0”. «m 3.0cn 0.000 mm. 0.. 00 «.03m n.40m mm. mm. «o ~.04~ m.osm mm. mm. «o o.omm 0.00m mm. mm. «m 0.0aw m.:mm ma. mo. an n.m0N 0.00m 00. 0m. 00 0.00N «.mam 00. 00. «0 0.000 0.00m 00. mm. «m :.00« m.no: m0. mm. «m n.~o« 0.0:: Ma. 00. an ~.m:« 3.00: Na. mm. «m 0.000 0.~:n 00. mm. «0 ”.mad 3.0mm 00. mm. o a.««« N.nwn 00. mm. «m :.:0« o.man 00. mm. «m ~.~0« :.oon no. 00. an 0.00 0.0MN so. am. am 0.00 c.00m 00. on. m 0.00 0.m0« 00. 00. «o 0.00 5.00" nu. ma. «0 @.0m 0.000 00. mu. «m 0.0: o.mm« mo. mu. «m n.0n :.h~« mo. N0. «0 m.o~ 0.««« ma. 00. «m m.@~ 0.000 me. 00. «0 «.MN m.mm 00. 00. «m 0.30 «.00 no. 0:. «u 0.0« «.ms no. me. an 0.0 0.0: ma. mu. «m 0.0 0.00 00.0 n. m 0.0 c.0w no.0 no. «m 0.0 0.0 00.: N0. .tno .wmn hm“ mo ”0 an .0 fmmmunnrnu wUZQ/m.n 0" '4 .J t —‘ C K *0 O f \ \ \ \ \ C" H 0" c.) (.1 -( max. uxxd cm\« 0@\« omxa omxu ' ( .w01nvuc»w ~00 omxa :00000 nwxfi «000m0 cox“ vmomma mmmnma cmxa m30~m0 ofdmfm -mxu «ntwma c.0mw0 0nm~m0 :04mm0 ~00mmd 00m-0 amxa mswwma nw\a mmmmmu moNNma mmdwwa «owawn «:00ma mmmamd om\« «amuma ooxa 00:0Na 00\« mosawu ocmdmd 0mxa 0000~0 omxa wmmomd omxa r000~u 00x" mmoowa mooowu 0mxu 0«:0mu omxu 0:00mn owxa amwawu mcuowu omxn 000000 owxa «30000 000000 000000 cox“ 000000 cmxa «0:000 omxu cmmsuu 0 7x unfi00> ootht m>.(z u mzo.#=0.¢h0_n m:0¢s hm¥¢ hug-«(t «99995.99. 9.? 9999.969.— whtauawa< no 299325999 99:32:99 92.929me 9 w... 9_¢z a m:9.h:9.¢hm.9 29:59: cuumuzhmom 92¢ 999992989 wh(uu¢99< 59 :99.¢<5xou muz(z_t99 u_hm(:uohm 5 395. 5| .9 4¢Dc5 .. rt. .. 5c. .. wt. .. 9c. .. mc. .. «9. ~. 1‘. .. no. .. «c. .. we. .. mc. .. 5;. .. 5r. .. 5:. .. 5c. .. me. .. cc. .. «m. .. wm. .. mu. .. mm. 9. rm. 9. 19. 9. «m. 9. 5w. 9. 59. 9. 5w. 9. 5w. 9. um. 9. mm. 9. 59. 9. 5w. :0 cm. 9. :3. o. I... 9. 5o. 9. 5c. 9. we. 9. .¢. :0 F990 9. we. 9099—9909335 39¢... 54. c¢\. c¢\. rcx. c(\. ccx. 99\~ cc\. cc\. ccx. crx. c9\. cO\— crx. :95. co\. ccx. ccx. ocx. 9:5. cox. cox. cox. cox. ccx. ocx. crx. 99\. cc\. 99\. 99\. c9\. cO\. 90\. cO\. CO\~ c9\. CC\. cc\. €9\. 99\. 90\. 0.. 1.. r.. 9.. m.. 9.. 9.. :.. 3.. 3.. 9.. 9.. ~.. ... ... ... 9.. 9.. 9.. 09. 99. 09. 99. 99. 99. :9. n9. no. no. N9. .9. .9. .9. 99. 99. 59.: ~9.| 59.: 99.: n9.. i‘.l11|‘lll 52 .dacu .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 4(300 .0 .9300 .0 Adacu .(DOw .0 .4300 .0 .0 .0 .0 .0 ..(2Ch .u 9.. 99.. e. co. m. 59. v. mo. m. mo. w. 90. 0‘. NO. 9. No. w. 90. m. «r. 9. >9. r. 9«. 9. pm. 9. ur. N. ax. N. we. N. we. N. «a. N. N9. N. 99. N. :9. N. 99. N. 99. N. 99. N. as. N. c». N. «p. N. rs. N. pp. N. vs. N. as. N. m». .. as. .. Ns. .. Np. .. 9s. .. so. .. cc. .. so. .. pa. .. so. 99.. 99.. 99.. 99.. «a co. co. vase—accents n.9(h 90\. cr\~ ccx. 90\. 9£\. c¢\. 9.x. ccx. cox. cox. 9(\. 99\. cc\. 94\. 9c\. 90x. 90\. cc\. c¢\. 9(\. 90\. 90\. 90\. cox. cc\. 90\. co\. 99\. 90\. 90\. 90\. 90\. 90\. 90\. 90\. 90\. 90\. ccx. 90\. oO\. 90\. 99.. N». no. 99. am. am. 9m. .e. .4. .:. 9:. 9:. on. 09. um. cm. on. mm. mm. en. fin. .n. .n. on. ON. ON. (N. hN. bN. ON. mN. 0N. QNo NN. .N. .N. .N. .N. .N. .N. 9N. 0.. 9.. 9.. ¢.. m.. c.. 9.. N.. ... 9.. 09. 99. #9. 99. m9. :9. 99. N9. .9. 99. 00 No 00 mo :0 no No .0 oo 00 «0 ha 00 mm «0 nu ma .0 90 53 F. ll...1Hl-.l:! I N9 .9 .. 9. «a. Na. :rx. :9.- N9 .9 .. 9. N9. an. «ex. wc.- an N9 .9 .. 9. 99. 9«. cc\. c9.- 99 m9 .9 .. 9. N3. 9.. 95\. 09.- on N9 .9 .. 9. Na. 9.. cC\. (9.0 mm N9 .9 .. 9. 99. 9.. 99\. 99.- 99 N9 .9 .. 9. an. 9.. 99\. 99.- 99 N9 .9 .. 9. pm. 9.. cox. 99.- 99 N9 .9 .. 9. pm. 9.. c:\. 79.- .9 N9 .9 .. 9. pm. 9.. .«x. 39.- 99 N9 .9 .. 9. mm. 9.. cox. 9..- on N9 .9 .. 9. 99. N.. 9c\. 9..- «N N9 .9 .. 9. mm. 9.. 99\. ...- 9N N9 .9 .. 9. pm. 9.. 9:\. N..- am N9 .9 .. 9. pm. «9. ccx. N..- 9N N9 .9 .. 9. N9. 99. 90\. n..- 9N N9 .9 .. 9. N9. 99. co\. :..- 9N N9 .9 .. 9. N9. mc. 99\. 2..- mm 99 .9 9. 9. N9. 99. co\. m..- .N N9 .9 9. 9. 99. «9. 9c\. 9..- 9N N9 .9 9. 9. «N. 99. cc\. ~..- 0. N9 .9 9. 9. 9N. 99. 99\. 9..- 9. N9 .9 9. 9. 9N. .9. cox. 9N.- 9. N9 .9 9. 9. mN. N9. :c\. .N.- 9. N9 .9 9. 9. 9N. «c. 9r\. NN.- m. N9 .9 9. 9. NN. N9. 9(\. NN.- 9. N9 .9 9. 9. 9N. «9. 9L\. «N.- 9. N9 .9 9. 9. 9.. «c. 99x. mN.- N. N9 .9 9. 9. 9.. «9. 9c\. m~.- .. N9 .9 9. 9.9 m.. N9. 90\. 9N.- 9. N9 .9 9. 9.9 m.. 99.9 9(\. 9N.- a N9 .9 9. 9.9 m.. 99.9 ccx. .n.- 9 N9 .9 9. 9.9 9.. 99.9 99\. .n.- 9 N9 .9 9. 9.9 9.. 99.9 cc\. .n.- 9 N9 .9 9. 9.9 :9. 99.9 99\. en.- m N9 .9 9. 9.9 99. 99.9 cox. 9:.- 9 N9 .9 9. 9.9 99. 99.9 99\. oc.- n N9 .9 9. 9.9 99. 99.9 99x. No.- N .9 9.9 9.9 N9. 99.9 cox. mo.- . .199 .999 92m .19: .099 .9. N9 N9 .9 .9 9 9 (- zcm.z¢1:C9 99242.19: 9.9m91999m :99 :9:.<> cm»<.:9.¢u 903w: cubs-9.3.325 -mzo_...:n_¢...m_9 2.59.3. «3.52.599 92< cmammtmzm mhtam¢00< 99 293.3539 uuz- o and/(,- J‘s->- o This is identical in form to an efficiency criterion based on some knowledge about the third moment of a quadratic utility function by Levy and Hanoch.S Their criterion was derived from the existence of a maximum value Xm for which the quadratic concave function is non- decreasing, and assuming that all observed values of X are smaller than Xm for all individuals' utility. Thus, the use of TSO would require either the assumption that all Xn:f XI“ of a quadratic utility function or that the X distributions of both f and 9 have finite values for the means and variances and would require some knowledge concerning the Sum. 6h degree of skewness. For a pairwise comparison, it would only be necessary to know which distribution had the lower skewness coeffi- cient.6 Because of these added restrictions, TSO will not be used as an efficiency criterion for this study. The following example7 will be presented to distinguish be- tween the E-V and the SO efficiency approaches by showing graphically an ordering of three portfolios, A, B and C by both E-V and F50. E(x) LO ____“._..mE—--——~E I I i . i i I I 0.5 5 l ' ""‘a l I I ' l o L 5 ; V(X) 0.08 0.l9 i.34 Figure A-I E-V ordering Under the E-V approach, A and 8 would be in the efficient set, while C would be excluded because it has the same mean as B but has a higher variance. (See Figure A-l). But the F50 approach shows C to be clear- t ly superior to A and determines the efficient set to be 8 and C. It 5mm. 7From R. Burr Porter and Jack Caumnitz. "Stochastic Dominance vs. Mean-Variance Portfolio Analysis," Workin Pa r No. , University of Kansas, School of Business, Lawrence: (December l97OI, pp. 5-6. ‘- A‘_‘_—-‘-— 6S CIHULAT IVE PROBABILITY l.0 —— ~— - Figure A-Z FSD Ofdering can be seen in Figure A-2 that the worst outcome that can be experi- enced with C is better than the best possible outcome that could be obtained with A. Therefore, no rational investor would choose A over C--an example of how the SO approach produces choices which are more consistent with rational behavior than those produced by the EV approach. Using the same example above, it can also be seen that using the SSD criterion, 8 would eliminate C because the area under F(XO) is consistently less than the area under F(Xc). This could also be determined by comparing the non-common portions of the areas under the distributions, at and € , As long asezfl, then f(XB) O f(Xc). _ —‘—_—-—“.-*WA 66 A more detailed example of an empirical application8 will also be presented in the context of a pairwise comparison of two portfolios, f and 9. Monthly Investor returns D. + Pi will be used Po to deveIOp frequency distributions on which the F50, SSD, and T50 tests can be applied. Class intervals of return could have been P used for a more extensive example instead of all available observa- i tions. The observed rates of return are indicated in Table A-l. TABLE A-i OBSERVED PORTFOLIO RETURNS - HYPOTHETICAL EXAMPLE fig w: . % Return Portfolio Period I Period 2‘. Period 3 Period’h f 6 2 l.8 7 g 6 3 I 3 it is necessary to approximate the true underlying functions by means of finite discrete sets of sample observations. First, the sample I observations are arranged in ascending numerical order. The distribu- tions must be monotonically increasing regardless of chronological or- der. Even if two or more observations have the same numerical value, for consistency in labelling, each observation is considered to be distinct. These ordered data should be combined (but identified) data 8From Porter, Wart and Ferguson, "Efficient Algorithms For Conducting Stochastic Dominance Tests on Large Numbers of Portfolios," Working Paggr No. #2, University of Kansas, School of Business, Lawrence: (September, l97ll pp. 6-7. 67 from both distributions. Given K distinct observations, for each portfolio each observation will occur with a relative frequency f(Xl) I l/K. The corresponding cumulative distribution function F'(Xn) is generated directly by summing the sample frequencies for all X;. For comparison of f(X) and g(X) there will be a total of N I 2K distinct observations. If the i th observation belongs to portfolio f, then f(X‘) I l/K and g(X') I O. The SD criteria compu- tations for discrete functions are: 52 F, (xn) .631 f(X‘) n - l, 2,. . . u c, Hal-ii; g(X‘) n-I,2,. . .N gig F2 (xn) - é Fl(xi-I)(xi - xH) n - I, 2, . . . N where F2(X') I O n 620:") - I); I;'(xH)(xi - xM) n - I, 2, . . . N where 62(X') I 0 11°. 0 F3(Xn) - I/2 i2; Ezui) + 50“] (xi - XH) n - 2. 3. - - - Hmr. F3(x') - O n 63(xn) - I/Z 'Z-z 62(Xl) ‘I' 62(Xi_li] (Xi 'xi_') n 2 2s 3s a . . N where 63(X') I 0 Using the test data: K I A, N I (2)(h) I 8 TA COMPUTED DATA MATRIX 68 OLE A-Z .— Value Observations of i 2 3 A 5 6 J 3 xn I.0 l.8 2.0 3.0 3.0 6.0 6.0 7.0 f(xn) 0 1/5 I/u 0 0 I/u 0 III I- g(Xn) 1/5 0 0 I/u I/h 0 I/u 0 r,(xn) o .25 .50 .50 .50 .75 .75 l.OO Gl(Xn) .25 .25 .25 .50 .75 .75 l.OO l.OO F2(Xn) o 0 .05 .55 .55 2.05 2.05 2.80 E5 62(Xn) 0 .20 .25 .50 .50 2.75 2.85 3.75 F3(Xn) 0 0 .005 .305 .305 5.205 5.205 6.630 63(Xn) 0 .080 .125 .500 .500 5.375 5.375 8.625 Sample Calculations:9 F2(X3) - O(l.8 - I) + .25(2 - l.8) + 5(3 - 2) + .5(3 - 3) + .5(6 - 3) + .75(6 - 6) + .75(7 - 6) G3(X8) - 0 + .05 + .5 + 0 1.5 + 0 + .75 - 2.80 I .S (.2 + O) (l.8 - l) + (.25 + .2) (2 - 1.8) + (.5 + .25) (3 - 2) + (.5 + .5) (3 - 3) + (2.75 + .5) (6 - 3) + (2.75 + 2.75) (6 - 6) + (2.75 + 2.75) (7 - 6) I .5 .16 + .09 + .75 + o + 9.75 + o + 6.5 - .5(I7.25) - 8.625 Results: FSD or SSD. f(X) D g(X) by TSD but not by 9F (X8) is the area under the cumulative frequency distribution of portfolTo f return observations when n I 8 observations. G3 (X8) IS the integral of the area under the cumulative frequency distribution of portfolio 9 return observations when n I 8 observations. .OO -75 .SO .25 69 CUMULATIVE PROBABILITY r--- I 5] (Xn) I r----'--— - _.____. I l 4: I r--- -_.I Fl (Xn) I O l 2 3 4 5 6 7 % Return Figure A-3 Graphical Representation of Stochastic Dominance Comparison-Hypothetical Example *‘—._- ‘- “...—.— APPENDIX 8 PERFORMANCE MEASUREMENT MODEL AXIOMS FOR Tfl§;PERFORMANCE MEASUREMENT MODEL 1. Common stockholders' utility (U) is directly related to that portion of their wealth (W) that can be measured by the total market value of their common equity holdings (V). That is, I U I f (W) and W I f (V) So, V 4 W ————-+U max max max 2. Non—satiety is assumed. That is, the shareholders pre- fer more wealth (in terms of market value) to less wealth. 3. Aggregate utility is the sum of the utility of all the individual stockholders involved. N “Aggregate ' 2:% Ui' '- N is the total combined number of stockholders of both the acquiring and acquired firms before the merger and is the total number of stock- holders in the merged firm after the merger. 4. Success of a merger or a group of mergers is defined in terms of common stockholders' aggregate utility. Where UA is the 70 L‘A‘. ‘Mw_ 7i stockholders' aggregate utility relating to the premerger distribu- tion and U8 is the stockholders' aggregate utility relating to the postmerger distribution, (UA D U3) --- Merger Success (UB D UA) --- Merger Failure if neither distribution dominates the other, no determination of mer- ger success or failure can be made. 5. Since aggregate stockholders' utility is measured in terms of the aggregate market value of their common equity holdings. merger success is measured directly using stochastic dominance comparisons of premerger and postmerger aggregate market value distributions (VB and VA respectively). (VA D VB) --- Merger Success (V. D VA) --- Merger Failure 6. General riskiness of the shareholders' wealth valuations can be measured by the general variability inherent in the market value distributions. I.“ [1‘1 .Amvm co.u:n.cu In.e >uc03v099 0>.um.09 9099asumoa 0:. one Amy. co.u:e n.9um.e >ucezvu99 0>.um.09 90m905094 0;. 2909 o. Ewen some Lem u\. 0:.m> ecu cm.mmm .. :m zu.3 mc.cc.mom ..e I s. no.909 Lemcesumoa 0;» 9o An I :. 90.909 909903099 0:. 9059.0 no.9.cm.m : one co.u:n.9un.m cemeaeunoe pee 909905099 oec.eEOu 0:. c.3u.) 06:..cmme 9o 90o9o 0:9 «0.9.cm_m e 090;: a I cu. I ..an At...» . .. 5...; . . .N....m. :5... ., 0v:..cmme mc.nmocuc. o. mc.ecouue 90e9009 one mco.u:n.9um.e.3 ....9 a ....a .a m a one .900. be 909 .m..9.vmav.u. 9. tee a m a :- .6. .... 9.0.... u 9. .eumc.590u .000. n. «000000 000 90:90 on. «0.0:.500 00.900.9um.0 90:0.0: .5 m m ..e 909 .m..9 I .mv.9 9. 90 ... u n 0300 909 .3... v.3... 0:0 0 u a 0509 909 .m..9 v.0..9 9. .ooumu.oc. m. 09:..09 909902 ac. 099 x. ....u a ....9 .s m a 0:0 .000. an 909 A... JV.» .... 9. tom .. m m ..m 90.. A». uflAnv 9 9. H. .# ON. . . . .N >0 00.90u.mu am9 0:» >.NN< .. I E some 909 Amv. 9 wee Anv.9 000.000.9um.e >00000099 0>.ue.:e:0 0:0 mc.ceoeoo >.0>.u090.. Ile ow. I z . . . .N.. I e ....m Amv.m 00.930.9um.v >00030099 0>.um.:eeu 0;. 3909 o. «0.00030099 0>.ue.09 000:» see >.0>.9090.. —Ifi E I Asmv— G on I t Anv.9 00.900.9un.0 >00000099 0>.ue.:a:u 0:. 3909 o. m0.u:0:0099 0>.ue.09 000:» 2:0 >.e>..eLeu. ._ _IH -31... 0.9 E I Ash..9 New I a .a..m . . .AWMVQ . . . .A—mvm I Anya 00 I 5 Afimvw . . . 9A.mv& . . . .A—flV‘ a I .... A? 0000.0009II.I9 0909.9 L ..u:. n. 0000000 90m903 -:.E90000:. n. 0000000 0:0 90000 000 .00000 .000 _ 0:0 0mm >0 Amv~9 o .mku 0000:.300 :0.u:0.9um.v 9000.0: .0000u.v:. m. 09: "a u a 0:0 0000. 00 909 .0 u m ..0 909 Amvuu IAuv~u 9. 90 .u..09 90m905 0:0 0mm >0 Amku 0 Anv~9 Auvumvamvuu 9. 0:0 .. u m _ .... w ... 0.9.00 909 Anya... vAnvu... 0:0 ... u .9 0:0 0.900. 00 909 A33 Immvwh. r :0 909 39.0.3? 9. 0 a m m 8.3 909 39.9.33 9. _ _ 9. .23 .. m. ... :0 909 .38 «.3 ... 9. 0 . »H7 \La aw om. ... .N.. I E 0000 909 Amvuu 0:0 Auvmu m:0.u:0.9um.0 0090 0>.u0.0e:u 00a 0:.900500 >.0>.u w” :0900. >0 :0.90u.90 0mm 00» >.00< 1: m In! _ om I a Asmvma ... .AHwVNQ ... .A.mv~u I Amv~u_ cm I a Ahmvmu ... .Aanku ... .A.mv~9 I Amvmu .Iufi. . a, - .HHHWIJ N... o I 2...ku 090...... .ow. I t .N.. I ... ATP .. ..nZ.....3.u WW I 1.3: a .. a 0.... o I ....N... 80.... .8. .. x .N.. - ... .....m - 5......3... W .. GEN... .:0.u0>9om00 0:0.>090 000 00:.m 0:.0> 90x90e :. 0m:000 000 0:0 0:.00 :0.a0>90000 m:0.>090 000 00 >u:0:0099 0>.00.0200 000 90 0000090 00» m:.EE:m >.0>.u0900. >0 0:.00 :0.» c0>90m00 0000 um Am..u 0:0 Any—9 0:0.000.9um.0 >0:0:0099 0>.u0.0&:0 00» 900:: Amvmu 0:0 Auvw9 90 000.000.9um.0 000 0:.590000 000:.u:003u_uo 090m.9 BIBLIOGRAPHY Books Alberts, Hillian w. and Segall, Joel E. (eds.) The Corggrate Herggr, Chicago: University of Chicago Press, l966. Ansoff, H. igor; Blandenburg, Richard 6.; Partner, Fred E. and Radosevich, Raymond, Acquisition Behavior of U. S. Hanu- facturing FirmsI lgflb-lgbg, Nashville: Vanderbilt Univer- sity Press, i97i. Beveridge, N.i.B., The Art of Scientific investigation, New York: Vintage Books. l950. Cootner, Paul N.. (ed.) The Random Character of Stock Market Prices. The MIT Press, 1935. Deiienbarger, Lynn E. Jr., Common Stock Valuation in industrial MergersI Gainesvilie: University of Florida Press, i933. Fame, Eugene and Miller, Merton, The Theory of Finance, New York: Holt, Rinehert and Hinston. i972. Feller, Hilliee,.An introducgion to Probability Theory and its Aggiications, Vol. i. 2nd Edition, New York: John Wiley S Sons, inc., l968. Francis, Jack C.. investment Analysis and Nanagggent. New York: Francis. Jack C. and Archer, Stephen H., Portfolio Analysis, Engiewood Cliffs. N.J.: Prentice-Hail, inc., i97i. Freund, John E. and Williams. Frank J., Modern Business Statistics, Englewood Cliffs, N.J.: Prentice-Hall, inc., i95 . Friend, irwln; Biume, Marshall; and Crocket, Jean, Mutual Funds and Other institutional investors: A New Perspective, New York: FEErau-Hm, inc., 1970. Granger, Clive N. J. and Norgenstern, Oskar. Predictability of Stock Market Prices, Lexington: D. C. Heath and Company, l970. 75 76 Hadar, Josef, Mathematical Theory of Economic Behavior, Reading: Addison-Wesley Publishing Cowany, l97l. Hendriksen, Eldon 5., Accounting Theory, Revised Edition, Momewood, lli.: Richard D. irwin, Inc., l970. Kelly, Eamon M., The Profitability ofygrowth Through Mergers, The Pennsylvania State University, i937. Latane, Henry A. and Tuttie, Donald L., Security Analysis and Port- folio Management, New York: The Ronald Press Company, i970. Levy, Naim and Sarnat, Marshall, investment and Portfolio Analysis, New York: John Wiley 0 Sons, inc., l972. Lynch, Harry N.. Financial Performance of Conglomerates, Boston: Division of Research, Graduate School of Business Administra- tion, Harvard University, i97i. Nelson, Ralph 5., Mer r Movements in American industr i8 -l 6, Princeton: Princeton University Press, i959. Reid, Samuel R., Merggrs, Managers and the Egggggy, New York: McCraw- Mill, inc., l9 . Rudner, Richard 5., Philos h of Social Science, Englewood Cliffs, N.J.: Prentice-Nail, inc., i933. Sharpe, William F., Portfolio Theory and Capital Markets, New York: McGraw-Mili Book Company, i970. Spurr, William A. and Bonini, Charles P., Statistical Analysis for Business Decisions, Homewood, iii.: Richard D. irwin, inc., 19 7. Weston, J. Fred, The Sc and Methodoio of Finance, Englewood Cliffs, N.J.: Prentice-Hail, inc., l9 Aggyggi Articles Berry, Charles N.. ”Corporate Growth and Diversification," Journal of Law and Economics, XIV, No. 2 (October, l97i). 371-333. Block, Stanley, "The Merger impact on Stock Price Movements," gsg figsiges; ngics, Vol. 17, No. 2 (Spring, 1969), 7-l2. 77 Bosiand, Chelcie, "Stock Valuation in Recent Mergers: A Study of Appraisal Factors," Trusts and Estates (June-JuiynAugust. '955) I 5 '6.5690 Boyle, Stanley E., "Premerger Growth and Profit Characteristics of Large Conglomerate Mergers in the United States: l9h8-i968," in St. John's Law Review, Conglomerate Mergers and Acquisitions, Uh (Spring, I970), Special issue, i52-l70. Conrad, Gordon R. and Plotkin, lrvlng M., "Risk/Return: U. 5. industry . Pattern,” Harvard Business Review (March-April, l960), 90-iii. “B Coombs, C. H.; Raiffa, M., and Thrall, R. M., "Some Views on Mathe- matical Models and Measurement Theory," Psychological Review, 6i, No. 2 (i954). Curiey, Anthony J., "Conglomerate Earnings per share: Real and Transi- togy Growth," Accounting Review, XLVI, No. 3 (July, l97l), 519- £5 52 . Curiey, Anthony J., "Conglomerate Growth: The Ostrich Effect," Account- ing Review, XLVII, No. 2 (April, i972). 37l-37b. Dewing, Arthur 5., "Statistical Test of the Success of Consolidation," uar erly Journal of Economics (December, i92i), Bh-lOI. Durand, David and May, ”The Ex-Dividend Behavior of American and Tele- phone and Telegraph Stock," Journal of Finance, XV (i960), i9- 3|. Evens, John L. and Archer, Stephen M., "Diversification and the Reduc- tion of Dispersion: An Empirical Analysis," Journal of Finance. XXiii (December, i968), 76i-769. Fame, Eugene F., ”Portfolio Analysis in A Stable Paretian Market,” Management Science, ll (January, l965), “Oh-hi9. Fame, Eugene F., "The Behavior of Stock Market Prices,” Journal of Business (April, i965), 3h-l05. Fame, Eugene F., “Efficient Capital Markets: A Review of Theory and Empirical Work," Journal of Finance, xxv (May, 1970). 383-hi7. Fieiitz, Bruce D., "Stationarity of Random Data: Some implications for the Distribution of Stock Price Changes,” Journal of Financial and Quantitative Analysis (June, i97i), i025-lO3EZ Frank, Werner G., and Weygandt, Jerry J., "Convertible Debt and Earnings per Share: Pragmatism vs. Good Theory," Accounting Review (April, 1970), Ial'lage 78 Gort, Michael, "An Economic Disturbance Theory of Mergers," Quarterly Journal of Economics, 83 (November, i969), 62h-6h2. Gort, Michael and Hogarty, Thomas F., "New Evidence on Mergers," Journal of Law and Economics, Xiii (April, i970), i67-l8h. Hadar, Josef and Russell, William R., "Stochastic Dominance and Diversification," Journal of Economic Theory, 3 (l97l), 288- 305. Hadar, Josef and Russell, William R., "Rules for Ordering Uncertain *5 Prospects," Economic Review, 49 (March, i969), 25-39. ' - Hanoch, Giora and Levy, Helm, "The Efficiency Analysis of Choices involving Risk," Review of Economic Studies. 36 (l969), 335- 3H6. Hanoch, Giora and Levy, Haim, "Efficient Portfolio Selection With é? Quadratic and Cubic Utility," Journal of Business (April, 1970), gal l8l-189. . Haugen, Robert A. and Udell, Jon G., "Rates of Return to Stockholders of AcQuired Companies," Journal of Financial and Quantitative Anal- ysis, Vii, No. l (January, i972 , l387-i39 . Hogarty, Thomas F., "The Profitability of Corporate Mergers," Journal of gusiness, “3 (July, l970), 317-327. Howard, Ronald A., "The Foundations of Decision Analysis," iEEE Transactions on S stems Science and C bernetics, Voi. SSC-h, No. 3 (September, 8655, 2ii-2i9. Jennings, Edward M., ”An Empirical Analysis of Some Aspects of Common Stock Diversification," Journal of Financial and Quantitative Analysis (March, i97i), 797-3lh. Jones, Charles P. and Simkowitz, Michael, ”Abstract: An Exploratory Analysis of the Simultaneity of Stock Price Movements," Journal of Financial and Quantitative Analysis, VII, No. 2 (March, .972 s I 9" 72- Joyce, Jon M., and Vogei, Robert C., "The Uncertainty in Risk: ls Variance Unambiguous?" Journal of Finance, XXV (March, l970), '27‘l3he V King, Benjamin F., "Market and industry Factors in Stock Price Be- havior," Journal of Business, 39 (January, i966), l39-i90. 79 Lev, Baruch and Mendeiker, Gershon, "The Microeconomic Consequences of Corporate Mergers," Journal of Business, #5 (January, 1972), 85-109. Levy, Helm and Hanoch, Giora, "Relative Effectiveness of Efficiency Criteria for Portfolio Selection," Journal of Financial and Quantitative Analysis (March, i960), 63-76. Levy, Helm and Sarnat, Marshall, "Diversification, Portfolio Analysis and the Uneasy Case for Conglomerate Mergers," Journal of Finance, xxv (September, I970). 795-807. Levy, Helm and Sarnat, Marshall, "Alternative Efficiency Criteria: An Empirical Analysis," Journal of Finance, XXV (December, 1970), 1153-1158. Levy, Helm and Sarnat, Marshall, WA Note on Portfolio Selection and investors' Wealth," iournai of Financial and uantitative Analysis (January, 1971), 639:652. Levy, Helm and Sarnat, Marshall, "TWo-Period Portfolio Selection and investors' Discount Rates," Journal of Finance, XXVI (June. 1971). 757-762. Leweiien, Wilbur G., “A Pure Financial Rationale for the Conglomerate Merger? Journal of Finance, XXVI (May, 1971), 521-537. Lintner, John, "Expectations, Mergers and Equilibrium in Purely Competitive Securities Markets," American Economic Review, LXI (May, i97i), lOI-lli. Livermore, Shaw, "The Success of industrial Mergers," Quarterly Journal of Economics (November, 1935). 68-96. Lorie, James H. and Halpern, Paul, "Conglomerates: The Rhetoric and the Evidence," Journal of Law and Economics, XIII (April, I970) . lug-I660 McDonald, J. G. and Fisher, A. K., "New issue Stock Price Behavior," Jgurnai of Finance, XXVII, No. 1 (March, 1972), 97-102. Meiicher, Ronald W. and Harter, Thomas R., ”Abstract: Stock Price Movements of Firms Engaging in Large Acquisitions," Journal of Financial and Quantitative Analysis Vii, No. 2 (March, 1972) , ili69-ili75 . Mead, Walter J., instantaneous Merger Profit As A Conglomerate Merger Motive," Western_§conomic Journal, 7 (December, 1969), 295-306. 80 Michaeison, Jacob B., and Gorshay, Robert C., "Portfolio Selection in Financial Intermediaries: A NewiApproech:'Jour 01 of Financial andegantitative Analysis”(June, 1967), 166-199. Mueller, Dennis C., "A Theory of Conglomerate Mergers," Quarterly Qgggggi gf Eggnomics, 83 (November, 1969). 693-659. Mokkelbost, Per 8., "Unsystemetic Risk Over Time,” Journal of Financial and Quantitative Analysis (March, 1971), 735-796. Mumey, G. A. and Korkie, R. M., ”Balance Sheet Additivity of Risk Measures,” qugnai of Financial and Quantitative Analysis North, D. Werner, VA Tutorial introduction to Decision Theory,” E IEEE Transactions on S stems Science and C bernetics, Vol. : SSCI’, No. 3 (Septenber, 196-5). 1' . Porter, R. Burr and Geumnitz, Jack E., “Stochastic Dominance vs. Mean- 2 Variance Portfolio Analysis: An Empirical Evaluation," Ameri- can Economic Review (June, 1972), h38-446. Praetz, Peter 0., "Australian Share Prices and the Random Walk Hypothesis,” Australian Journal of Statistics, ii, N0. 3 (1969)s '23-'39- Quirk, James P. and Seposnlk, Rubin, ”Admissibility and Measurable Utility Functions," Review of Economic Studies, 29 (1962), ”IO-"+6. Ramenathan, K. V. and Rappeport, Alfred, "Size, Growth Rates and Merger Valuation," Accounting Review, XLVI, No. A (October, 197'). 733-745. Reid, Samuel R., ”A Reply to the Weston/Mansinghka Criticisms Dealing With Conglomerate Mergers," Qggrnal of Finance, XXVI (Septem- ber, 197'). 937-9h5- Reinhardt, U. E., "Conglomerate Earnings per Share: immediate and Post-Merger Effects," Accounting:Review, XLVii, N0. 2 (April, 1970). 350-370. Rothschild, Michael and Stiglitz, Joseph E., ”increasing Risk 11: Its Economic Consequences," Journal of Economic Theory, 3 (1971), 66-8h. Russell, William R. and Smith, Paul E., "Taxation, Risk-Taking and Stochastic Dominance," The Southern Economic Journal (April, 1970) l025 433. 81 May, Robert G., "The Influence of Quarterly Earnings Announcements On investors Decisions As Reflected in Common Stock Price Changes," Unpublished dissertation, Michigan State University, 1970. Meiicher, Ronald W. and Hempel, George M., "Difference in Financial Characteristics Between Conglomerate Mergers and Horizontal or Vertical Mergers," Unpublished paper, l97i. Meiicher, Ronald W. and Rush, David F., “Conglomerate Firm Performance and the Contribution of Major Acquisitions," Unpublished paper, 1972. Oveson, Richard M., "Regression Parameter Estimation by Minimizing Absolute Deviations: Confidence intervals and Monte Carlo Re- sults," Unpublished paper, 1971. Poindexter, Eugene 0., "The Profitability of industrial Merger," Un- published dissertation, Syracuse University, 1970. Porter, R. Burr, "Stochastic Dominance vs. Mean-Variance as a Basis for Capital Budgeting Decisions Under Risk," Working Paggr No. 36. School of Business, University of Kansas, Lawrence, October, 1970. Porter, R. Burr, "The Application of Stochastic Dominance Rules to the Financial Decision Problems of the Firm," Unpublished paper, l972. Porter, R. Burr, HA Comparison of Stochastic Dominance and Stochastic Programming As Corporate Financial Decision Models 'Under Un- certainty," Unpublished paper, 1972. Porter, R. Burr and Geumnitz, Jack E., "Stochastic Dominance vs. Mean- Variance Portfolio Analysis," Working Paggr No. 32, School of Business, University of Kansas, Lawrence, December, 1970. Porter, R. Burr and Wart, James R. and Ferguson, Donald L., "Efficient Algorithms for Conducting Stochastic Dominance Tests on Large Numbers of Portfolios," Working Paper No. 59.Schooi of Business, University of Kansas, Lawrence, September, 1971. Pinches, George E., ”Financing Corporate Mergers and Acquisitions With Convertible Preferred Stock,” Unpublished dissertation, Michigan State University, 1968. Stevens, Donald L., ”A Multivariate Analysis of Financial Characteris- tics of Acquired Firms in industrial Mergers," Unpublished dissertation, Michigan State University, 1972. 82 Samuelson, Paul A., "Efficient Portfolio Selection for Pareto-Levy investments," Journal of Financial and anntitative Analysis (June, 1967), 107-122. Shick, Richard A., "The Analysis of Mergers and Acquisitions," Journal of Finggce, XXVII, No. 2 (May, 1972), 995-502. Smith, Keith V. and Schreiner, John C., VA Portfolio Analysis of Conglomerate Diversification," Journal of Finance, XXIV (June, 1969), 912-927. Smith, Keith V. and Schreiner, John C., ”A Reply," Journal of Finance, XXV, No. 9 (September, 1970), 915-916. Van Horne, James C., "New Listings and Their Price Behavior," igurnal of Finance, XXV, No. A (September, 1970), 783-79“. Westerfield, Randolph, HA Note on the Measurement of Conglomerate Diversification," figurnal of Figgnce, XXV, No. A (September, 1970). 909-914. Weston, J. Fred and Mansinghke, Surendra K., ”Tests of the Efficiency Performance of Conglomerate Firms,” Journal of Finance, XXVI (September, 1971), 919-936. Whitmore, G. A., "Third-Degree Stochastic Dominance," American Economic Review (June, 1970), 957-959. Wippern, Ronald F., "Utility implications of Portfolio Selection and Performance Appraisal Models," Journal of Financial and Quantitative Analysis (June, 1971), 913-92 . Working, Hoibrook, HA Random Difference Series For Use in the Analysis of Time-Series," Journal of the American Statistical Associ- ation, 29 (193k), ll-Zh. Un ubii hed Material Geumnitz, Jack E., "The Influence of industry Factors in Stock Price Movements,” Working Paper No. #2, School of Business, Univer- sity of Kansas, Lawrence, June,.l971. Mason, R. Mai and Goudzwaard, Maurice 0., ”Growth Patterns of Highly Diversified Firms: A Sample Survey, 1962-1967," Unpublished paper, 1970. 83 Stitch, Robert 5., “Aggregate Analysis of Merger Performance, 1998- 1969," Unpublished paper, 1972. Treadway, Peter T., “An Explanation of the Conglomerate Merger Move- ment--The Conglomerate As a Result of Forces Arising in the Capital Markets," Unpublished dissertation, University of North Carolina at Chapel Hill, 1971. Weston, J. Fred and Dunn, Michael F., "CAP and the Measurement of Business Risk," Unpublished paper, 1971. Zdanowicz, John 5.. "Investment Risk considering lntraperiod Liquida- tion: A Re-evaiuation of Price Variance," Unpublished disser- tation, Michigan State University, 1971. Other Sources Accounting Principles Board of the American institute of Certified Public Accountants, "Earnings Per Share ," Oginion No, 15, New York, 1969. 4 Federal Trade Commission, Egonomic Report on anglomerate Merggrs, Hearings on Antitrust and Mon0poiy, Committee of the Judi- ciary, U.S. Senate, 91st Congress, ist Union, Part 0A, USGPP, 1969. Federal Trade Commission, "Large Mergers in Manufacturing and Mining, 1998-1970," Statistical Rgpgrt No. 7, Bureau of Economics, March, 1971. Federal Trade Commission, "Current Trends in Merger Activity, i970," Statigtical Reggrs N9. 8, Bureau of Economics, March, 1971. investment Statistics Laboratory, iSL Dailnytock Price index, American Stock Exchange, New York: Standard and Poors Corpor- investment Statistics Laboratory, iSL Daily Stock Price index, New York Stock Exchange, New York: Standard and Poors Corporation, 19 2-1970. Mood s industrial Manual Moodys investors Service, Inc., New York, 1965-1971. National Industrial Conference Board, Mergers in Industry New York: National industrial Conference Board, 1929.