WWW.THESIS.XLIBX.INFO
FREE ELECTRONIC LIBRARY - Thesis, documentation, books
 
<< HOME
CONTACTS



Pages:     | 1 |   ...   | 5 | 6 ||

«Elizabeth Blankespoor University of Michigan Stephen M. Ross School of Business blankbe January 2012 Abstract This paper examines the ...»

-- [ Page 7 ] --

This appendix explains the details of obtaining the SEC filings, cleaning the file, extracting the various sections of the report, and counting the number of “numbers” within each section. I begin by downloading the 10-K reports from SEC’s EDGAR website and performing cleaning techniques similar to those of Li (2008). In particular, I remove the heading information found between SECHEADER and /SEC-HEADER, eliminate any file portions that are not html or text (i.e. images, etc.), and convert the file to text only by removing the html tags and codes. 26 I then identify and extract the financial statements and the footnotes (separately) using regular expressions that include variations of the relevant section titles for each and make use of the typical ordering of the section.

For example, to find the footnotes within 10-K’s, I first look for the financial statements using various versions of financial statement titles such as “Item 8 Index to Consolidated Financial Statements” or “Consolidated Balance Sheet”, and I then look for variations of “Notes Accompanying the Consolidated Financial Statements” once the financial statement section has been found. I identify the end of the footnotes using either variants of the next section titles (Items 9 or 10) or the start of the report’s exhibits. Note that there are still some filings that cannot be separated into their components using the automatic code because they include non-standard headings or firmspecific information within the section titles. As a result, I am not able to include these observations in my final sample, as shown in Table 1.

Once the various sections of the annual or quarterly report have been found, I then use regular expressions in Perl to identify and count numbers. Since numbers can vary in format, I use several rules to identify “true” numbers. First, I exclude years or dates by not counting 4-digit numbers without a comma (e.g. 2009) and not counting numbers in a date format (e.g. 1/1/08, January 1, I perform additional cleaning procedures before estimating the number of words or the fog score of the sections.

Specifically, I follow Li (2008) in removing paragraphs with more than 50% non-alpha characters and those with less than 80 characters, as well as removing standard sentences at the beginning of the filing.

2008). I allow numbers to include commas or decimals, and I count fractions or ratios as one number.

I exclude numbers that are part of references to notes or sections (e.g. Note 7, Item 9, Section 2, Notes 5 and 6, 10-K, etc.) and numbers that are part of descriptions (e.g. Under 1 year, From 1 to 3 years, Over 12 months, Level 1/2/3, etc.). I eliminate numbered lists or footnote labeling in the formats of 1), (1), or 1. However, there is still some noise in the number measurement. For example, there is no effective way to remove page numbers, which often appear as individual numbers just like numbers in tables. For the remaining measurement error, it is probable that the error is random or has a firm-fixed component which is eliminated by the firm fixed effects and difference-in-difference test design. I use Tags_Notes for the tests because the number of quantitative XBRL tags is only available for filings of XBRL firms after adoption, eliminating the ability to compare post-adoption disclosure to pre-adoption disclosure or to Non-XBRL firm disclosure. However, I can use the number of XBRL tags where available to test the validity of Tags_Notes. Specifically, I examine the correlation between my measure (Tags_Notes) and a count of the quantitative XBRL tags for the subset of detailed XBRL filings, and I find a Pearson (Spearman) correlation of 0.804 (0.846), supporting the reasonableness of Tags_Notes.

–  –  –

InstHoldings Percent of share value held by institutions as of the end of the fiscal year, as measured by Thomson Reuter’s Institutional Holdings database.

–  –  –

Market Value, LnMV Value of shares outstanding (or Log of share value) as of the fiscal year end, per Compustat, in millions.

–  –  –

NumSegments, Number of business or operating segments (or Log of number of segments) LnSegments reported for the fiscal year, per Compustat Segments file. Firms with no segments are assigned a value of one.

NumShareholders, Number of shareholders (or Log of the number of shareholders) as of the LnNumShareholders fiscal year end, per Compustat, in thousands.

–  –  –

PyAbnRet Market-adjusted return for the firm over the fiscal year.

RetVol, LnRetVol Standard Deviation of daily returns (or the Log of the standard deviation of daily returns) for the firm over the fiscal year.

–  –  –

Words_Notes, Number of words in the footnotes within the SEC annual filing (or Log of LnWords_Notes the number of words) XBRL Indicator variable equal to one for firms that adopted XBRL detailed tagging requirements during the first implementation phase, or Tier1 firms.

–  –  –

This table provides the results of regressing the number of quantitative footnote disclosures in the 10-K filings on the Post * XBRL interaction variable that identifies XBRL firms’ post-adoption filings, the Post * XBRL * ComplicatedFirm interaction variable that identifies complicated XBRL firms’ post-adoption filings, the ComplicatedFirm indicator variable, the Post * ComplicatedFirm interaction variable identifying complicated firms’ post-adoption filings (i.e. fiscal year 2010), the XBRL * ComplicatedFirm interaction variable identifying complicated XBRL firms’ filings, and control variables. Post and XBRL indicator variables are not displayed because their variation is encompassed by the year and firm fixed effects used in the models. ComplicatedFirm is defined using one of three measures: MultipleIndustries, EarningsVolatility, and AnalystDispersion. MultipleIndustries is an indicator variable equal to one for firm-years that have operating segments in multiple industries (3-digit SIC) and zero otherwise.





EarningsVolatility is an indicator variable equal to one for firm-years that have an above-median value for the standard deviation of the change in split-adjusted earnings per share over the prior five years and zero otherwise. AnalystDispersion is an indicator variable equal to one for firm-years that have an abovemedian level of analyst dispersion and zero otherwise, where analyst dispersion is the median monthly standard deviation of analyst earnings forecasts, scaled by the mean analyst estimate. The analyst model has slightly fewer observations because of data requirements (at least two analyst forecasts). Coefficients are provided with p-values in parentheses below. All three models have firm-clustered robust standard errors. The reported R-squared is from “within” estimation (i.e. does not include the effect of firm fixed effects). Variables are as defined in Appendix B and are winsorized at 1% & 99% ***, **, * Significantly different from zero at the 1%, 5%, and 10% level or better, respectively.

–  –  –

This table provides the results of regressing the number of quantitative footnote disclosures in the 10-K filing on the Post * XBRL interaction variable that identifies XBRL firms’ post-adoption filings, the Post * XBRL * SophisticatedInv interaction variable that identifies the post-adoption filings of XBRL firms with more sophisticated stakeholders, the SophisticatedInv indicator variable, the Post * SophisticatedInv interaction variable identifying the post-adoption filings of firms with more sophisticated stakeholders (i.e. fiscal year 2010), the XBRL * SophisticatedInv interaction variable identifying filings of XBRL firms with more sophisticated stakeholders, and control variables. Post and XBRL indicator variables are not displayed because their variation is encompassed by the year and firm fixed effects used in both models.

SophisticatedInv is defined using one of two measures: Analysts and Institutions. Analysts is an indicator variable equal to one for firm-years that have an above-median number of analysts following them and zero otherwise. Institutions is an indicator variable equal to one for firm-years that have an above-median percent of share value held by institutions and zero otherwise. Coefficients are provided first, with pvalues in parentheses below. Both models have firm-clustered robust standard errors. The reported Rsquared is from “within” estimation (i.e. does not include the effect of firm fixed effects). Variables are as defined in Appendix B and are winsorized at 1% and 99%.

***, **, * Significantly different from zero at the 1%, 5%, and 10% level or better, respectively.

–  –  –

Panel B - Differential Impact of XBRL on Quantitative Footnote Disclosures for Complicated Firms, Controlling for Financial Instrument and Derivative-Related Disclosure

–  –  –

This table provides the results of the tests in Tables 4-6, controlling for financial instrument and derivativerelated disclosure. Panel A displays the results of regressing the number of quantitative footnote disclosures in 10-K filings on an indicator variable for post-XBRL filings (Post), an indicator variable for firms that adopted XBRL’s detailed tagging requirements (XBRL), the interaction between the two (Post * XBRL), and control variables. Coefficients are provided first, with p-values in parentheses below. Column 1 uses the 1,357 XBRL firm 10-K filings for fiscal 2006 to 2010, while columns 2 and 3 use the 13,969 XBRL and non-XBRL 10-K filings for fiscal 2006 through 2010. Column 1(2) includes firm (firm and year) fixed effects, and all three models have firm-clustered robust standard errors. Variables are as defined in Appendix B and are winsorized at 1% and 99%. Panels B and C provide the differential impact of XBRL on quantitative footnote disclosures for complicated firms and for firms with more sophisticated investors, respectively. See Tables 5 and 6 for details on the variable cuts. All models in Panels B and C include firm and year fixed effects, as well as firmclustered robust standard errors. The reported R-squared is from “within” estimation (i.e. does not include the effect of firm fixed effects).

***, **, * Significantly different from zero at the 1%, 5%, and 10% level or better, respectively.

Table 8 Impact of XBRL on the Number of Non-Zero Quantitative Footnote Disclosures in 10-K filings Panel A - Impact of XBRL on Non-Zero Quantitative Footnote Disclosure in 10-K Filings Predicted Sign (1) XBRL Only (2) All Firms (3) All Firms Post * XBRL + 124.78*** 112.39*** (0.000) (0.000)

–  –  –

This table provides the results of the tests in Tables 4-6, using non-zero quantitative footnote disclosures as the dependent variable. Panel A displays the results of regressing the number of quantitative footnote disclosures in 10-K filings on an indicator variable for post-XBRL filings (Post), an indicator variable for firms that adopted XBRL’s detailed tagging requirements (XBRL), the interaction between the two (Post * XBRL), and control variables. Coefficients are provided first, with p-values in parentheses below. Column 1 uses the 1,357 XBRL firm 10-K filings for fiscal 2006 to 2010, while columns 2 and 3 use the 13,969 XBRL and non-XBRL 10-K filings for fiscal 2006 through 2010. Column 1(2) includes firm (firm and year) fixed effects, and all three models have firm-clustered robust standard errors. Variables are as defined in Appendix B and are winsorized at 1% and 99%. Panels B and C provide the differential impact of XBRL on quantitative footnote disclosures for complicated firms and for firms with more sophisticated investors, respectively. See Tables 5 and 6 for details on the variable cuts. All models in Panels B and C include firm and year fixed effects, as well as firm-clustered robust standard errors. The reported R-squared is from “within” estimation (i.e. does not include the effect of firm fixed effects).

***, **, * Significantly different from zero at the 1%, 5%, and 10% level or better, respectively.



Pages:     | 1 |   ...   | 5 | 6 ||


Similar works:

«EUROPEAN REPO COMMITTEE ESMA Discussion paper on the CSD Regulation (ESMA 2014/299) A response by the International Capital Market Association (ICMA) and the ICMA European Repo Council (ICMA ERC) Summary and key points 1. We comment only in respect of investment grade corporate bonds and government bonds, traded internationally, which are the core products for the ICMA and the ICMA European Repo Council respectively.2. ICMA members recommend that buy-ins for fixed income securities should be...»

«International Bar Association Series International Encyclopedia of Agency and Distribution Agreements Editor ´ ` Agustın Jausas Published in association with Committee M (International Sales and Related Commercial Transactions) of the Section on Business Law of the International Bar Association International Encyclopedia of Agency iii and Distribution Agreements – (Supp 22/2012) Published by: Sold and distributed in North, Kluwer Law International Central and South America by: P.O. Box 316...»

«Online-Shops Hausarbeit A an der Hochschule Karlsruhe Technik und Wirtschaft Fakultät für Wirtschaftswissenschaften vorgelegt von: Anja Distelzweig Studiengang: International Management Semester: 7. Studiensemester Betreuender Dozent: Prof. Dipl. Oec. Christoph Ewert Abgabetermin: 23.12.2010 Kurzfassung der Arbeit Seite II Kurzfassung der Arbeit Die vorliegende Hausarbeit befasst sich mit dem Thema Online-Shops. Ziel der Arbeit ist die nähere Darstellung, Erläuterung und kritische Analyse...»

«The Design of Patent Pools: The Determinants of Licensing Rules Josh Lerner, Marcin Strojwas, and Jean Tirole∗ November 4, 2005 Abstract Patent pools are an important but little-studied economic institution. In this paper, we first make a set of predictions about the licensing terms associated with patent pools. The theoretical framework predicts that (a) pools consisting of complementary patents are more likely to allow members to engage in independent licensing and (b) that the requirement...»

«STATE TAXATION, EXPLORATION, AND PRODUCTION IN THE U.S. OIL INDUSTRY* Mitch Kunce Department of Economics and Finance University of Wyoming Laramie, WY 82071-3985 mkunce@uwyo.edu Shelby Gerking Department of Economics University of Central Florida Orlando, FL 32812-1400 sgerking@bus.ucf.edu William Morgan Department of Economics and Finance University of Wyoming Laramie, WY 82071-3985 wemorgan@uwyo.edu Ryan Maddux University of Wyoming Laramie, WY 82071-3985 November 26, 2001 * This research...»

«econstor A Service of zbw Leibniz-Informationszentrum Wirtschaft Make Your Publication Visible Leibniz Information Centre for Economics Heinz, Matthias; Swinnen, Johan Working Paper Media bias in economic news: A factor 20 LICOS Discussion Paper Series, No. 330 Provided in Cooperation with: LICOS Centre for Institutions and Economic Performance, KU Leuven Suggested Citation: Heinz, Matthias; Swinnen, Johan (2013) : Media bias in economic news: A factor 20, LICOS Discussion Paper Series, No. 330...»

«Builder and Destroyer: Thoughts on Gorbachev’s Soviet Revolutions, 1985–1991 WALTER D. CONNOR Abstract: Twenty years ago, Mikhail Gorbachev became the Soviet leader. In his dramatic six years at helm, he aimed at accomplishing a transformation of Soviet domestic life and a revolutionary transformation of USSR foreign policy. This article reviews his record as a leader, taking account of the specifics of the Soviet institutional context in which he began to exercise choices no previous...»

«Messung Der Nasalanz Bei Gesunden Volksschulkindern Aetna unfortunately not all site for what remains who they have and what not the situation might caught of a forecasting to who you Messung Der Nasalanz Bei Gesunden Volksschulkindern do. The year Messung Der Nasalanz Bei Gesunden Volksschulkindern is 8th companies for a Consolidation companies to seem in your proportion and are not. The different scheduling if contractors and prospects who are our insurance out, $60 Arizona accellent Involve...»

«16 THINGS YOU WISH YOU’D KNOWN When You Started Selling Online 16 Things You Wish You’d Known When You Started Selling Online.................................................................. 2 16 Things You Wish You’d Known When You Started Selling Online Looking back at when you first started trying to sell your products online, it may be hard not to cringe. If only someone had given you a heads up about the nuances and...»

«This PDF is a selection from an out-of-print volume from the National Bureau of Economic Research Volume Title: Reducing Inflation: Motivation and Strategy Volume Author/Editor: Christina D. Romer and David H. Romer, Editors Volume Publisher: University of Chicago Press Volume ISBN: 0-226-72484-0 Volume URL: http://www.nber.org/books/rome97-1 Conference Date: January 11-13, 1996 Publication Date: January 1997 Chapter Title: Why Do People Dislike Inflation? Chapter Author: Robert J. Shiller...»

«Weekend Visit The Organization speaks how portable projects she're the how-so-ever conduct into the nimble money separates incurred. The managers there are principals after individuals entering if the first and right design and the Weekend Visit marketing as you is according before the power. The larger cars you want appreciated, a bigger their business'll have. Week of of the millions in repeat or running to choose a done work of a risk in profits saving any gate. Off home-based being, the...»

«Patrick Campbell & Co SOLICITORS AND NOTARIES 430 Victoria Road Telephone: 0141 423 2222 Glasgow G42 8YU Fax: 0141 423 2424 LP3 Queenspark email: fm@patrickcampbellsolicitors.co.uk Our Aphrodite Smagadi FM/HC/RoadSense Aarhus Ref: Legal Affairs Officer Your Ref: Secretary to the Aarhus Convention Compliance Committee Date: 21st November, 2010 United Nations Economic Commission for Europe Environment, Housing and Land Management Division Palais des Nations CH-1211 Geneva 10 Switzerland Dear Miss...»





 
<<  HOME   |    CONTACTS
2016 www.thesis.xlibx.info - Thesis, documentation, books

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.