Ecker Information Risk and Long-Run Performance of Initial Public Offerings

2009
ISBN: 978-3-8349-8117-2
Verlag: Betriebswirtschaftlicher Verlag Gabler
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 132 Seiten, eBook

Frank Ecker examines the performance of U.S. initial public offerings (IPOs) from 1980 to 2002. He links positive and negative abnormal returns to the deviation of the realized information risk from the expected information risk. The author proposes effective measures for a long-term profitable investment strategy in IPOs.

Dr. Frank Ecker promovierte bei Prof. Dr. Hellmuth Milde am Lehrstuhl für Geld, Kredit und Finanzierung der Universität Trier. Er ist Assistant Professor of Accounting an der Duke University, Fuqua School of Business, Durham, USA.

Zielgruppe

Research

Autoren/Hrsg.

Ecker, Frank

Weitere Mitwirkende

Milde, Prof. Dr. Hellmuth

Olsson, Prof. Dr. Per

Inhaltsverzeichnis

1;Foreword;6
2;Preface;9
3;Contents;10
4;List of Tables;12
5;List of Figures;14
6;Symbols and Abbreviations;15
7;1. Introduction and Motivation;16
8;2. Valuation under Information Risk;22
8.1;2.1 Classification of information risk;22
8.2;2.2 Empirical measurement of information risk;23
8.3;2.3 Conceptual consequences of the introduction of information risk;27
9;3. Derivation of a Returns-Based Measure of Information Quality;29
10;4. Abnormal Returns Measurement and Hypotheses Development;39
10.1;4.1 Methodological issues in long-term abnormal returns measurement;39
10.2;4.2 Explaining abnormal IPO performance;57
11;5. Tests with Abnormal Portfolio Returns;65
11.1;5.1 Construction of the IPO sample;65
11.2;5.2 Calendar-time portfolios from the full IPO sample;69
11.3;5.3 Persistence tests;75
11.4;5.4 Deviation tests;82
12;6. Robustness Tests;99
12.1;6.1 Varying the calendar-time approach;99
12.2;6.2 Firm-speci.c tests;100
12.3;6.3 Further Robustness Tests;107
13;7. Concluding Remarks;111
14;Appendix;113
15;Bibliography;141

and Motivation.- Valuation under Information Risk.- Derivation of a Returns-Based Measure of Information Quality.- Abnormal Returns Measurement and Hypotheses Development.- Tests with Abnormal Portfolio Returns.- Robustness Tests.- Concluding Remarks.

Leseproben

3. Derivation of a Returns-Based Measure of Information Quality (S. 15-16)

In this chapter, I present the development and validation of a returns-based earnings quality metric, termed e-loading. This chapter is an abbreviated version of Ecker, Francis, Kim, Olsson, and Schipper (2006).11 In short, e-loading is the .rm-speci.c slope coe.cient on an information quality mimicking factor from an asset pricing regression and thus a returns-based representation of the perceived information quality.

There are two main reasons why e-loadings are superior to the existing metrics in the literature and make the study of information risk in an IPO setting feasible. First, e-loadings can be calculated for a comprehensive sample of firms and do not require a time-series of accounting data. The latter is simply not attainable for IPO firms. Second, e-loadings can be adjusted to measure information quality in event time. Changes in information quality can therefore be measured in a timely manner. In contrast, traditional earnings quality metrics are calculated over a certain and usually lengthy time period. For example, the calculation of accruals quality is over seven years at a minimum.

When a change from one period to the other occurs, only the last of the seven annual observations actually re.ects this change in earnings quality. Thus, traditional earnings quality metrics behave like slow moving averages and are not able to immediately indicate changes. For a metric to be less static, the frequency of independent observations must be higher. Independent accruals quality observations can only be assessed every seven years, requiring fourteen consecutive years of accounting data. Accruals quality – the underlying construct of e-loadings – goes back to Dechow and Dichev (2002).

It is the standard deviation from a regression of current accruals on past, present and future cash .ows from operations. The intuition for this measure is as follows. The evaluation of current accruals only requires cash .ow data with one lead and one lag term, as the operating cycle of most firms is below one year. Thus, there is no empirical justi.cation why present current accruals should persist for longer than one year. Conversely, cash deferrals in the prior year should map into the present year’s earnings. On the whole, current accruals in the present year should map into cash .ows from operations in the last, the present or the next year. If this mapping is not accurate, an error will occur capturing the portion of current accruals which does not correspond to a cash .ow in the adjacent .scal years.

But to the extent that this error in accruals is constant and therefore predictable, investors can learn from past observations and will not perceive this error as a ’risk’, independent from its sign and magnitude. It would simply be an indication of a bias in current accruals. What should be perceived as risk is its variability. Consequently, the accruals quality metric is the standard deviation over error terms from consecutive years.

Besides the three cash flows, McNichols (2002) adds the two explanatory variables, taken from the modi.ed Jones (1991) model. The basis model is transformed into a model of current accruals being explained by past, present and future cash flows from operations, as well as the change in sales and the gross property, plant and equipment. From a purely econometric standpoint, this adjustment increases the model’s explanatory power.

Über Autor(innen)

Dr. Frank Ecker promovierte bei Prof. Dr. Hellmuth Milde am Lehrstuhl für Geld, Kredit und Finanzierung der Universität Trier. Er ist Assistant Professor of Accounting an der Duke University, Fuqua School of Business, Durham, USA.