View
218
Download
3
Category
Preview:
Citation preview
For more inforplease visit our
or cont
A research and education initiative at the MIT Sloan School of Management
Ana
SP KoJim Sh
The Effect of Disclosures by Management, lysts, and Financial Press on the Equity Cost of
Capital
Paper 195
thari ort
October 2003
mation,
ebusiness@mit.edu or 617-253-7054 website at http://ebusiness.mit.edu act the Center directly at
The Effect of Disclosures by Management, Analysts, and Financial Press On the Equity Cost of Capital
S.P. Kothari kothari@mit.edu, (617) 253-0994
Sloan School of Management Massachusetts Institute of Technology
50 Memorial Drive, E52-325 Cambridge, MA 02142
and
J.E. Short
jshort@mit.edu, (617) 452-3228 Sloan School of Management
Massachusetts Institute of Technology 50 Memorial Drive
Cambridge, MA 02142
October 2003
We are grateful to Xu Li, Kalani Oshiro, Stanley Woods, Mitul Mehta and Tanis Fidelholtz for their research assistance. We have benefited from discussions and comments from Philip Wright, Joel Kurtzman, Alison Thomas, Gabrielle Wong, Sunil Misser and Mark Lutchen. We acknowledge financial support from Pricewaterhouse Coopers and the MIT Center for eBusiness.
The Effect of Disclosures by Management, Analysts, and Financial Press On the Equity Cost of Capital
1. Introduction and Problem Statement
Demand for financial reporting and disclosure arises from information asymmetry and
agency conflicts between managers, outside investors and intermediaries. Disclosure regulation
and institutions facilitating credible disclosure between managers and investors play important
roles in mitigating these problems (Healy and Palepu, 2001). Corporate disclosures, reports in
the financial press, and analysts’ reports and discussion of corporate performance all enhance the
information reflected in stock prices. That is, they reduce information asymmetry between the
average investor and informed market participants, e.g., company management. The consensus
among financial economists is that a rich information environment and low information
asymmetry have many desirable consequences. These include the efficient allocation of
resources in an economy, capital market development, liquidity in the market, and decreased cost
of capital.1
Evidence of the effect of disclosure on the cost of capital is sparse (see Healy and Palepu,
2001, for a review of the literature). Evidence is weak in the few studies that examine the issue,
e.g., Botosan (1997). Conclusions from previous research on the effect of disclosures on the cost
of capital are tenuous because disclosures analyzed in previous research are far from
comprehensive and the measurement of disclosure proxies is generally subjective, not objective.
Our study is the first to document systematic evidence of the cost of capital effects of disclosures
culled from a virtually exhaustive set of sources from the print medium. Our analysis of
1 See Diamond and Verrecchia (1991), Healy and Palepu (2001), Bushman and Smith (2001), and Core (2001).
3
disclosure is not only comprehensive, but the construction of disclosure proxies is quantitative
and amenable to replication in future.
In addition to analyzing the impact of all disclosures, we study the impact of disclosures
segregated by source – management, analysts, and news stories in financial press. There are
economic reasons to study how the source of disclosures affects the firm’s cost of capital.
Disclosures by management, analysts, and news reporters differ on at least two salient
dimensions. First, incentives facing management, analysts, and news reporters might
differentially taint the contents of their disclosures or reports. In particular, management and
investment analysts are alleged to have strong incentives to optimistically skew disclosures,
whereas news reporters’ incentives to be optimistic in their reports are muted. An efficient
capital market weights disclosures according to the credibility of the source of the disclosure.
That is, on average, in pricing stocks we expect the market to filter out the bias and deemphasize
noisy disclosures by the management, analysts, and news reporters.
Second, news stories in the financial press are likely to be timelier than analysts’ reports.
Analysts frequently repackage and re-transmit available information from corporate disclosures
and financial press news stories in presenting in-depth analysis in their own report (see Lang and
Lundholm, 1996). While such reports might be valuable to the market participants in their
(investment) decisions, they may also have limited price impact (i.e., the reports might lack new
information affecting the stock price) or limited implications for the cost of capital.
Summary of results. Our empirical analysis tests for the impact of disclosure on cost of
capital. Our findings are based on a very large, content database of disclosure text constructed
from disclosure content published on four electronically-available information sources, Dow
Jones Industrial, Investex, Factiva and SEC Edgar. We downloaded all disclosures made by
4
companies, analysts, and business articles available from over 400 content sources, for a sample
of 887 companies in four industry sectors (Technology, Telecommunications, Pharmaceutical
and Financial) for the time period 1996-2001. We used large-sample content analysis methods to
analyze 326,357 texts. Results from our information analysis were then used in financial models
to test for the effects of disclosure on cost of capital.
We find that when content analysis indicates favorable disclosures (that might include
references to management credibility and reduced uncertainty about the firm’s products and
markets etc.), we observed a reduction in the firm’s cost of capital and stock return volatility. In
contrast, unfavorable disclosures were accompanied by increases in the cost of capital and stock
return volatility.
We next analyzed disclosures by source – corporations, analysts, and the financial press.
When we examined the impact of positive and negative news disclosure by companies, we found
that the market discounts the impact of the strength of the positive statements. Positive and
strongly positive news statements do not materially affect cost of capital. However, when
management makes negative and strongly negative news disclosure, the market responds and the
cost of capital goes up. Negative news, therefore, is strongly weighted.
Our second set of results considers the impact of positive and negative news disclosure by
analysts. We found that both positive and negative news disclosures from analysts appear to be
heavily discounted. Analysts on the whole were slightly less positive in their disclosures on
firms, and similarly less negative (more favorable) in negative disclosure. The strength of impact
of their disclosure on cost of capital is insignificant compared to that for companies. Intuitively
the results suggest that analysts have source credibility problems. They are seen as responding to
market changes after they have taken place, discounting their impact.
5
Our third set of results considers the impact of positive and negative news disclosure by
the business press. Here the evidence is more consistent. We find that when the business press
makes both positive and negative news disclosures there are impacts on cost of capital. Positive
news disclosure decreases cost of capital, negative news disclosure increases it. Intuitively the
strength of this finding suggests that the credibility of news disclosure by the press is higher than
that for companies or for analysts, and in the case of analysts, press disclosure is seen as more
timely. The press is less affected by incentive concerns (agency), and this is reflected in the
consistency of evidence.
Outline of the paper. We have organized our paper in five sections. Section 2 presents
the background for our main hypothesis that the credibility and timeliness of the source of
disclosures affects the firm’s cost of equity capital. We present the key arguments and
background literature. Section 3 discusses our data and content analysis methodology. We detail
our building of the disclosure text dataset, our content analysis methodology, and our financial
risk measures and firm characteristics. In Section 4 we present our results. Section 4.1 provides
background descriptive statistics, and Section 4.2 covers regression results. Our concluding
section discusses the management implications of our findings and opportunities for future
research. Further detail on our content analysis methods and dataset are included in five
Appendices.
2. Background and Hypothesis Development
Here we provide the background for our main hypothesis that the credibility and
timeliness of the source of disclosures affect their impact on firm’s cost of equity capital.
Economic rationale leading up to the hypotheses with respect to the cost of capital effects of
disclosures by corporations, analysts, and financial press is developed in this section.
6
Credibility and timeliness of management disclosures. We analyze SEC-mandated
corporate disclosures, i.e., 10-K and 10-Q filings. These disclosures contain a combination of
the following two types of disclosures: disclosures through the Management’s Discussion and
Analysis (MD&A) of the firm’s performance that are designed to bring the market’s expectations
in line with the management’s superior information; and mandated other disclosures such as
financial footnotes that are largely an affirmation of the information already available to the
market participants, but could occasionally include qualitative and forward-looking information,
e.g., discussion of loss contingencies and litigation risk.
In utilizing corporate disclosures, market participants are likely to recognize that
management has an incentive to portray a self-serving assessment of past corporate performance
and future outlook. Therefore, management’s favorable disclosures may not be credible to
market participants. Of course, in making any favorable, optimistic disclosures, the management
runs the risk of being accused of making misleading and fraudulent disclosures, i.e., face
litigation risk. Litigation risk thus serves to discipline managers from making overly optimistic
disclosures. However, the disciplining forces might be less operative when disclosures are
qualitative and long-term in nature (e.g., discussion in the MD&A section on the 10-K filings)
rather than quantitative (e.g., point estimate for an EPS forecast) and short-term (e.g., next
quarter’s sales). In contrast to good news disclosures, management has an aversion to disclose
bad news. So, when it volunteers such information, it is believable. That is, unfavorable
voluntary disclosures by the management are more credible regardless of whether they are
qualitative or quantitative in nature.
In addition to being credible, management’s unfavorable disclosures are likely to be more
timely because of litigation risk facing the firm under the Securities and Exchange Commission’s
7
Rule 10b-5 (Skinner, 1994). The Rule together with other securities laws state that management
has a duty to disclose material information as and when it becomes known to the management.
The rationale is that new information might render the management’s previous disclosures
misleading. Alternatively, in light of the new information, the stock would be mis-valued in the
absence of disclosures to correct the mispricing of the stock. However, the securities laws are
likely to have an asymmetric impact on corporate disclosure practices with respect to good and
bad news disclosures. Litigation against a firm is more successful if the management is accused
of withholding bad news such that the stock price remained above its fundamental value.
Consistent with this rationale, Kellogg (1984) finds that buyers’ claims of losses are 13 times as
likely as sellers’ claims of foregone profits, i.e., an opportunity loss. Therefore, management has
an incentive to volunteer bad news on a more timely basis than good news.
Credibility and timeliness of analyst reports. Analysts are important information
intermediaries in capital markets. They expend real resources on gathering information and
producing research reports. These activities are typically assumed to enhance the ability of
prices to reflect available information. However, several forces influence the credibility and
timeliness of analyst research, which shape its price impact and effect on the firm’s cost of
capital.
Many argue that the strong economic incentives stemming from brokerage commissions
and investment-banking revenues taint the quality of analyst research. United States Senator
Joseph Lieberman claims that the millions of dollars analysts earn from investment-banking
services “compromise analysts’ objectivity” and therefore “the average investor should take their
bottom line recommendations with at least a grain of salt…” (The Wall Street Journal, February
28, 2002, p. A3). Consistent with the senator’s claim, a large body of academic research
8
suggests analysts compromise their objectivity and issue optimistic forecasts and stock
recommendations as a result of their economic incentives arising from brokerage revenues and
investment-banking relationships (e.g., Lin and McNichols, 1998, Michaely and Womack, 1999,
Dechow, Hutton, and Sloan, 2000, Bradshaw, Richardson, and Sloan, 2003, and Lin, McNichols,
and O’Brien, 2003). The lack of objectivity and bias in research together weaken the credibility
of analyst research, which, in turn, would diminish its influence on the pricing or on the cost of
capital of stocks
With respect to the timeliness of information in analyst reports, previous research offers
compelling evidence that analyst reports do convey new information that affects security prices.
However, the magnitude of price impact is typically small. This might be because considerable
analyst research might be simply a retrospective analysis of past events that repackages and re-
transmits publicly available information, i.e., corporate disclosures and news stories (see Lang
and Lundholm, 1996, Beaver, 1998, p. 10, and Frankel, Kothari, and Weber, 2003). If this
activity dominates the contents of analyst research, it might not be particularly timely in
providing much new information to the market. Worse yet, evidence in Lys and Sohn (1990)
and Guay, Kothari, and Shu (2003) suggests that analyst forecasts are sluggish in that they fail to
fully incorporate the information available and reflected in stock prices at the time of their
forecasts.
In summary, lack of credibility and the absence of timely forecasts and research all make
it less likely that analysts research reports would contain much information that would affect
stock prices and/or the cost of capital.
Financial press news stories. We expect news stories in financial press to be quite
timely with an emphasis on factual reporting. News reporters and the financial press typically do
9
not have strong economic ties and relationships with individual firms. Moreover, each reporter
reports on a large number of corporations as new developments take place with individual
corporations. Therefore, we expect news stories to have both credibility and timeliness and
hence the contents of news stories are likely to be informative about the cost of capital.
3. Data and content analysis methodology
In this section we present summary information about the data used in this study. First,
we discuss disclosure data obtained from electronic information sources, which we used in
performing content analysis. We then describe the data on financial risk measures and firm
characteristics.
3.1 Disclosure text data
The most widely available source of corporate disclosure is financial statements. These
are SEC mandated disclosures for all U.S. publicly-traded corporations. These statements are
prepared using accrual rather than cash accounting to allow managers to reflect their better
knowledge of the firm’s financials through accrual estimates. In addition to financial statements,
managers also provide other information to investors. Some, such as footnotes and management
discussion and analysis, are disclosed within the financial report. Other disclosures are provided
voluntarily through other information channels, including analyst presentations and conference
calls, press releases and Internet sites. Through these disclosures, managers provide information
that facilitates external users of financial reports to better understand the true economic picture of
the business.
Disclosures are accompanied by third party certification to help assure credibility. That
is, in the absence of third party certification, there would be questions in investors’ minds about
the credibility of the information contained in the disclosures. For disclosure in financial
10
statements, credibility is typically enhanced in two ways: first, statements are prepared using
generally accepted accounting standards and conventions. Standards can potentially reduce
processing costs for outsiders and curb opportunistic reporting by managers. Second,
independent audit firms certify whether managers’ financial reporting decisions are consistent
with accounting standards and conventions.
Importantly, management disclosures made outside financial statements are not directly
certified by independent parties. Rather, their source credibility is enhanced through their use
and evaluation by outside parties, including financial analysts and business journalists. These
“information intermediaries” provide a measure of implicit certification of management’s
financial statements and non-financial disclosures. However, this certification is not in any way
formal nor can it be argued to be necessary complete. Our analysis is designed to empirically
assess whether the credibility issues with respect to disclosures from various sources, as
discussed above, are manifested in the intensity of their impact on the firms’ cost of capital.
We obtained data for our study from four electronic data information sources, Dow Jones
Interactive (Dow Jones, Inc.), Investex Plus (Infotrac), Factiva (Dow Jones & Reuters), and
Edgar (Securities and Exchange Commission, US Government). Text retrieval software was
developed and implemented to efficiently download disclosure texts in their entirety for a
company research sample frame of 887 firms over the time period 1996-2001 (Riloff and
Hollaar, 1996).
We downloaded all disclosure texts available, by company by year, for three sources: (1)
corporate reports (Edgar), (2) analyst disclosure and briefings (Investex and Factiva), and (3)
disclosures made by the general business press (over 400 content sources available in Dow Jones
Interactive and Factiva). All disclosure texts were coded by company ID, date and source of
11
publication, and when available, by author. The resulting content dataset is very large. For the
887 companies over the six-year period, 326,357 texts were downloaded, representing over a
million pages of source material. The 887-company sample was drawn from four industry
sectors, Financial Services (225 companies), Technology (197 companies), Pharmaceutical (72
companies), and Telecommunications (395 companies), and the data coded to enable industry
sector analysis. Finally source material was organized into business quarters, corresponding to
the quarterly financial reporting cycle. Each company in the dataset therefore had 72
observations: disclosure content aggregated into 24 time periods (6 years by 4 quarters per year),
by three sources of disclosure (corporate, analyst, business press). The list of companies in the
dataset is presented in Appendix I.
3.2 Classification of text into content categories
Our first step in preparing the data for content analysis was to filter each disclosure text
through a classification scheme that we developed. Words associated with six-categories of
business meaning were defined and used to test a categorization scheme using Riloff (1993) as a
guide. The six categories identified words and word phrases associated with:
i. market risk, industry structure and competitive forces, the external market and/or external environment of the firm;
ii. the development and execution of firm strategy and strategic intent;
iii. the building of organizational capital and human resource management;
iv. the image, brand and reputation of the company;
v. the investment and financial performance of the firm;
vi. the announcement and impacts of governmental regulation on the company.
Appendix II presents the classification scheme and the words and word phrases making up each
category.
12
All disclosure text was passed through the classification scheme, producing six sub-
sample datasets containing words and word phrases matching the words and word phrases
defined. Our subsequent analysis focused on content analyzing within and across the sub-sample
datasets.
3.3 Content analysis
The underlying principle in content analysis is that the many words of a text can be
classified into many fewer content categories, where each category consists of one or many
similar words or word phrases, and that each word or phrase occurrence can be counted and the
counts compared analytically. Word and phrase similarities are based on the precise meanings of
the words themselves (for example, grouping synonyms together), or may be based on word
groupings sharing similar connotations (for example, grouping together several words associated
with a concept such as market share, revenue growth, or forecasted earnings). Content analysis
therefore, can be useful in analyzing different types or "levels" of communication, as defined by
the meaning of the words themselves.
A flowchart of the typical process of content analysis research is presented in Figure 1.
The study is first specified theoretically and then conceptually in terms of an information model,
content (data samples) and the hypotheses to be addressed. The study is then operationalized
with respect to the variables (content categories) and coding schema to be used for classification
of words and word groups for analysis. Third, the content texts are analyzed using content
analysis procedures and the results interpreted.
3.4 Unit of analysis and measurement in content analysis
Our unit of analysis was the disclosure text. That is, the content of each disclosure text
was analyzed using content analysis procedures. The analysis software used, the General
13
Inquirer, has in its analysis approach a range of categories where word and word phrases are
mapped for each input text submitted, and word frequencies counted on dictionary-supplied
categories. In this research we were especially concerned with the degree and strength of positive
and negative statements in disclosure texts. We hypothesize that the pattern and strength of
favorable or unfavorable statements represent beliefs and judgments about the relative strength
of the firm, and assessments of risks the firm faces in its marketplace, strategy and operations.
We expected that highly favorable (positive) statements, or conversely, highly negative
statements would be made based on tangible factors viewed as important in the minds of
disclosure author(s). Accurately gauging the strength of positives and negatives in an extensive
sample of disclosure text would therefore tell us much about the consistency of views about the
firm and its promise in the market.
To illustrate how our content analysis procedures work in practice, we analyze two
example disclosure texts published in the Wall Street Journal on Amgen Corporation. The two
texts (news stories), published in February and April of 1996, report positive and negative news
on Amgen. On February 1, 1996, the Journal published a story reporting that clinical research
undertaken on the effects of Leptin, the company’s experimental obesity drug, showed that the
drug did not have the hoped-for benefits in a controlled patient study. The second story,
published on April 18, 1996, reported the company’s 32% increase in net quarterly income and
the filing of an application for testing a new drug designed to combat Hepatitis C. The text of
both stories is reproduced in Appendix III.
We submitted the two texts to the General Inquirer (GI) content analysis package.
Developed and written by Professor Philip Stone and colleagues at Harvard University, and later
extended by Vanja Buvac and colleagues, the General Inquirer (GI) is essentially a content
14
mapping tool.2 The GI software maps each text file with word counts on dictionary supplied
categories. That is, the software algorithm identifies and counts word frequencies in the
submitted text files, matching them against words and/or word meanings present in one or more
dictionary of categories. The currently distributed version of GI combines the “Harvard IV-4”
dictionary content-analysis categories, the dictionary content analysis categories defined by
Lasswell, and five categories based on the social cognition work of Semin and Fiedler.3 Each
category contains a list of words and word senses. For example, the category “negative” is the
largest, with 2,291 entries of words or word phrases denoting “negative.” To provide the reader
background on this procedure, example entries for the positive and negative categories are
presented in Appendix IV. When text is submitted for processing, the General Inquirer matches
words and word phrases with those in its dictionaries, and separates inflected word forms. For
example, the root word “sell” would match “selling” if selling was not a separate category entry,
and the word sense “sell cycle”, for example, would be correctly identified by software routines
designed to disambiguate word senses. That is, “cycle” would be correctly identified as
specifying the time periodicity of selling, and not a word short for bicycle or motorcycle. These
English stemming procedures, integrated with English dictionaries and routines for separating
English word senses, comprise the software’s extensive dictionaries and disambiguation routines.
The complexity of the dictionary and disambiguation routines limits the current GI software to
English text applications.
The results of our simple, two-disclosure text submission to the software appear in
Appendix V. Interpreting the output, note that for each text, there are two rows of data, the first 2 P.J. Stone in C. Roberts (1997). The General Inquirer’s Home page is at www.wjh.harvard.edu/~inquirer/ 3 See Semin and Fiedler (1992, 1996). Categories are described on the GI home page, www.wjh.harvard.edu/~inquirer/
15
corresponding to the raw frequency count of matched words (“r”), and the second giving the
percentage of matched words over the total number of words in the text. The February 2, 1996
Wall Street Journal Article, for example, contained 635 words, of which 20 are positive and 21
negative. Positives comprised 3.14 percent of the total number of words in the text, and negatives
slightly higher, 3.30 percent. 116 words were not matched against any of the words supplied in
the GI dictionary categories, or 18.26%. In the case of the April 18th Wall Street Journal article,
the corresponding numbers are the text contained 423 words, 19 positives and 8 negatives, or
4.4% of the words in the disclosure were positive and 1.8% were negative. Seventy-seven words,
or 18.2%, were leftovers (unmatched with dictionary categories).
Interpreting results in content analysis is always a question of comparing word counts
and frequencies across texts and categories. In this example, we can say that the first article
much more negative than the second article, and, as well, less positive. Conversely, the second
article was much more strongly positive and significantly less negative in its frequency and word
sense.
In analyzing the large number of disclosure texts in the database, our procedure was to
calculate the number of positives and negatives in each disclosure text, for each of the six
business categories defined (market, firm strategy, etc., see Appendix II). The resulting matrix of
counts and percentages calculated for each disclosure text was then subjected to one column and
one row operation. To organize the results matrix for input into financial models, row data was
summed and averaged by business quarter. That is, Positive and Negative frequencies calculated
for each disclosure text were summed across all texts and the average taken for each quarter. In
similar fashion, Positive and Negative frequencies calculated for each of the six business
categories (columns) defined were summed and the average taken. The resulting matrix of
16
positive and negative word frequencies represented an average score for all disclosure texts in
the quarter, by company by disclosure source, averaged across the six business categories
(columns). The resulting matrix was then used as input into our econometric analysis.
One final note concerns the computer processing time required to content analyze the
very large disclosure database. Our procedure was to calculate positive and negative frequencies
by company sequentially. A typical firm with the large number of disclosures over the sample
time frame required several hours of processing time.
3.5 Financial risk measures and firm characteristics
We used three measures of risk to assess the impact of disclosures on firm risk: cost of equity
capital, standard deviation of stock returns, and standard deviation of the analyst forecast errors.
We describe each measure below.
Cost of capital. We estimate the equity cost of capital using the Fama and French (1993)
three-factor model, where the size factor is defined as small minus large firm returns (SML), the
book-to-market factor is defined as high minus low book-to-market firm returns (HML), and the
market factor is defined as the excess return on the CRSP value weight portfolio (Rm - Rf). We
obtain monthly time-series returns on the three factors, SML, HML, Rm - Rf, from Kenneth
French’s website. The loadings on the factors, b, s, and h, are slope coefficients estimated from
the following regression model for firm i:
Ri - Rf = ai + bi [Rm - Rf] + si SML + hi HML+ ei. (1)
We re-estimate the three-factor model each quarter for each firm using a rolling window
of five years of monthly returns ending in the quarter under examination. Firm i’s estimated
loadings, i.e., estimated b, s, and h coefficients, multiplied by the average returns for the three
17
factors from 1963-2000 gives the cost of capital for firm i (see Fama and French, 1997). We
then annualize the number, which is our cost-of-capital measure.
Standard deviation of stock returns. Standard deviation of stock returns is a commonly
used risk measure. The greater the uncertainty of cash flows generated by a firm, i.e., the risk of
equity, the greater is the standard deviation of returns. This measure is also an indication of the
infrequency of information reaching the market and the degree of information asymmetry among
market participants. If some individuals are informed and others are less informed, then the
trading in the stock is less frequent and stock prices do not reflect a rich body of information.
Under these circumstances, disclosures by a corporation or analysts or disclosures in financial
press can all influence the degree of uncertainty. Favorable disclosures can inform the market of
higher levels of cash flows than previously expected and/or reduce the uncertainty in the market.
We measure the standard deviation of returns using daily return data for each firm every quarter.
Standard deviation of analysts’ forecast errors. Standard deviation of analyst forecast
errors indicates the extent to which the market is surprised by earnings announcements.
Favorable disclosures are expected to reduce uncertainty and enable analysts to better forecast
earnings, which would in turn reduce the magnitude of their forecast errors, i.e., the standard
deviation of forecast errors.
Firm characteristics. While our goal in the paper is to test for the effect of disclosures
on the firm’s cost of capital, it’s important to control for the effect of other determinants of the
cost of capital to avoid drawing erroneous inferences. We consider three firm characteristics that
previous research shows as significant determinants of the cost of capital, standard deviation of
returns, and analyst forecast errors: firm size, book-to-market ratio, and leverage. Small firms
are more risky than large firms in part because small firms typically are a single-project firm, i.e.,
18
a firm with undiversified portfolio of assets and projects. We measure firm size as the market
capitalization of a firm’s equity capital, i.e., number of shares outstanding times the share price,
at the beginning of each quarter.
Book-to-market ratio increases in risk in part because the market valuation of equity is in
the denominator of the ratio. Successful firms with expectations of a steady stream of high
levels of future cash flows are highly valued in the market, which drives their book-to-market
ratio down. These are typically considered to be low risk firms. In contrast, if the market has
little confidence in a firm and thus perceives the cash flow stream to be uncertain and not too
high, then the market capitalization of such a firm would be low. This drives up the book-to-
market ratio, so high book-to-market ratio proxies for high risk firms. Book-to-market ratio is
calculated as a ratio of the book equity of a firm divided by its firm size, i.e., the market
capitalization of equity.
Finally, risk increases in the firm’s financial leverage, i.e., the amount of debt relative to
the equity capital invested in the firm. We measure leverage as the ratio of long-term dent to the
total assets of the firm.
4. Results
This section presents the results of empirical analysis of the effects of disclosure on the
cost of capital. We begin with descriptive statistics and correlations among the variables used in
the analysis. The main set of results are based on regressions of various risk measures on
content-analysis-based measures of disclosure by source and firm characteristics as control
determinants of the different risk measures.
4.1 Descriptive statistics
19
Tables 1 and 2 provide univariate descriptive statistics and simple bi-variate correlations
for our dependent and independent variables, by industry sector and by all sectors. The number
of observations reported in each Panel represent the number of firms times the number of years /
quarters that had sufficient data to compute the variables. Some firms were excluded from
analysis based on insufficient data.
In Table 2, note the two columns corresponding to index variables “Favorable” and
“Unfavorable”, constructed as the mean of the content-analysis based Positive and Negative
measure for disclosures across all six business categories and disclosure source. With the
exception of the lack of a significant correlation between Cost of Capital and “Unfavorable”, all
correlations are significant at the .01 or .05 level for all variables – that is, Favorable and
Unfavorable assessments are strongly correlated with all financial measures studied, cost of
capital, standard deviation of daily stock return, the firm’s market capitalization of equity at the
beginning of the quarter (natural logarithm), book equity divided by market equity at the
beginning of the quarter, and leverage (long-term debt divided by the total asset). That all cross-
correlations (minus one) are significant between content measures and financial measures
suggests a strong and persistent effect between the two is indicated.
4.2 Regression results
Tables 3 and 4 report the results from multivariate regression models on composite news
measures and source of news. The panels report regression results separately on three dependent
variables: cost of capital, the standard deviation of daily stock return over the contemporaneous
quarter, and the standard deviation of all available quarterly analyst forecast errors from 1996-
2001. We use five independent variables: Favorable, Unfavorable, Size (natural logarithm of a
firm’s market capitalization of equity at the beginning of the quarter), B/M (book equity divided
20
by market equity at the beginning of the quarter), and Leverage (long-term debt divided by total
asset). Across all of the panels reported the adjusted R2, that is, the adjusted measure of fit of the
individual regression models analyzed, is significant for all of the models at the one-percent
level. The pattern of results intuitively suggests that the regression models are analyzing a
systematic effect.
Our first set of results considers the impact of positive and negative news disclosure by
company and by industry sector (Table 3, Panels A-E). In interpreting the coefficients for
Favorable and Unfavorable for all and individual industries, we find that when management
makes positive and strongly positive statements in company disclosure, the market anticipates
this and discounts the impact of the strength of these statements. Positive and strongly positive
news statements do not materially affect cost of capital (coefficients are weakly positive or not
significant). However, when management makes negative and strongly negative news
disclosures, the market responds and the cost of capital goes up. Intuitively, the market may be
stating that when negative news is disclosed, these are made in the context of a positive
disclosure bias by the company. Negative news, therefore, is strongly weighted.
Our second set of results considers the impact of positive and negative news disclosure by
analysts (Table 4, Panel B). We found that for both positive and negative news disclosure,
information from analysts appears to be discounted. Analysts on the whole are slightly less
positive in their disclosures on firms, and similarly less negative (more favorable) in negative
disclosure. The strength of impact of their disclosure on cost of capital is less significant than for
companies – the coefficients reported are not significant for cost of capital, but the coefficients
are significant for stock returns and analyst forecast errors. Intuitively the results suggest that
21
analysts have source credibility problems. Our interpretation is that they are seen as responding
to market changes after they have taken place, this discounting their impact.
Our third set of results considers the impact of positive and negative news disclosure by
the business press. Here is the evidence is quite consistent (Table 4, Panel C). We found that
when the business press makes both positive and negative news disclosures there are impacts on
cost of capital. Positive news disclosure decreases cost of capital, negative news disclosure
increases it. Intuitively the strength of this finding suggests that the credibility of news disclosure
by the press is higher than that for companies or for analysts, and in the case of analysts, press
disclosure is seen as more timely. The press is seen as less affected by incentive concerns
(agency), and this is reflected in the consistency of evidence.
5. Implications for Management and Future Research
Regulation of disclosure and the materiality of information that a reasonable investor
would want to consider in making an investment decision are subjects of intense scrutiny in the
post-Enron and Worldcom business climate. The materiality of information relates to the
importance and impact that corporate information and development have on the earnings of the
company, and how disclosure should reach the market and investors (Brown, 1998). As the
disclosure system has grown in size with the importance of evaluation of disclosure by other
information intermediaries, analysts, business press, and the investors themselves, information
asymmetries and agency conflicts between participants creates greater demand for reporting and
assurance of disclosure.
The research reported here is the first to document systematic evidence of the cost of
capital effects of disclosure. Our analysis of disclosure was based on a virtually exhaustive set of
sources from the print medium, segregated by disclosure source - corporate, analysts, and news
22
stories published in the financial press. We found support that the market weighs disclosures
according to the credibility of source, and that on average the market filters out and
deemphasizes noisy disclosures by all participants. Our results found that negative news
disclosure is strongly weighted by the market, and positive news is discounted as firms and
investment analysts have incentives to skew disclosure. We found in the data evidence that
companies and analysts have credibility problems affecting cost of capital. Specifically,
disclosure from investment analysts is heavily discounted, suggesting they are seen as
responding to market changes after they have taken place, discounting their impact.
Problems of agency and the incentives for the business press to skew disclosure are more
diffuse for journalists than for companies or analysts, and our results reflect this more
independent intermediary role - positive news disclosure decreases cost of capital, and negative
news disclosure increases it. Intuitively this finding suggests the credibility of news disclosure by
the press is higher than that for companies or analysts.
The strength and pattern of our results suggests that there are important economic reasons
for companies to understand more precisely how disclosure affects the firm’s cost of capital.
Pursuant to securities regulation and business norms, companies have a duty to disclose
corporate information and developments through timely and accurate disclosure. That we have
found evidence showing that the market discounts disclosure by source credibility suggests that
companies should give high priority to developing appropriate and complete communications
policies. Skewing disclosure of critical corporate information or development has economic
risks. Asking whether consistent, accurate and complete disclosure is best accomplished through
centralizing the corporate communication process, or decentralizing it with strong behavioral
norms and safeguards, is a key senior management question. An implication of our results is that
23
companies cannot depend on analysts to enhance source credibility or timeliness of disclosure.
Analysts have their own credibility and timeliness problems, and are subjects of increased
scrutiny in disclosure reform.
We conclude by noting that financial reporting and disclosure will continue to be an
important field of empirical inquiry. There are numerous factors with the potential to alter
disclosure practices and financial reporting – technological innovation, changes in the business
economics of audit firms and analysts, globalization of capital markets, and changes in
disclosure channels and the number and type of information intermediaries – which continue to
reshape disclosure and financial reporting practices. They create new and exciting opportunities
for research.
24
References Beaver, W., 1998, Financial Reporting: An Accounting Revolution, Third Edition, Prentice Hall,
Upper Saddle River, NJ 07458.
Botosan, C., 1997, Disclosure level and the cost of capital, The Accounting Review 72, 323-350.
Bradshaw, M., Richardson, S., Sloan, R., 2003, Pump and dump: An empirical analysis of the relation between corporate financing activities and sell-side analyst research, working paper, University of Michigan Business School.
Brown, J. Robert, Jr., 1998, The Regulation of Corporate Disclosure, 3rd Edition. Aspen Publishers, Inc., New York, NY.
Bushman, R., Smith, A., 2001, Financial accounting research and corporate governance, Journal of Accounting & Economics 31, 237-333.
Core, J., 2001, A review of the empirical disclosure literature: discussion, Journal of Accounting & Economics 31, 441-456.
Dechow, P., Hutton, A., Sloan, R., 2000, The relation between analysts’ long-term earnings forecasts and stock price performance following equity offerings, Contemporary Accounting Research 17, 1-32.
Diamond, D., Verrecchia, R., 1991, Disclosure, liquidity, and the cost of capital, Journal of Finance 66, 1325-1355.
Frankel, R., Kothari, S., Weber, J., 2003, Determinants of the informativeness of analyst research, working paper, MIT Sloan School of Management.
Guay, W., Kothari, S., Shu, S., 2003, Properties of implied cost of capital using analysts’ forecasts, working paper, MIT Sloan School of Management.
Healy, P., Palepu, K., 2001, Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical literature, Journal of Accounting & Economics 31, 405-441.
Kellogg, R., 1984, Accounting activities, security prices, and class action lawsuits, Journal of Accounting & Economics 6, 185-204.
Kelly, Edward Francis, 1975, Computer recognition of English word senses /Edward F. Kelly, Philip J. Stone. Amsterdam : North Holland Pub. Co. ; New York : American Elsevier Pub. Co., 1975.
Lang, M., Lundholm, R., 1996, Corporate disclosure policy and analyst behavior, The Accounting Review 71, 467-492.
Lin, H., McNichols, M., 1998, Underwriting relationships, analysts’ earnings forecasts and investment recommendations, Journal of Accounting & Economics 25, 101-127.
25
Lin, H., McNichols, M., O’Brien, P., 2003, Analyst impartiality and investment banking relationships, working paper, Stanford University.
Lys, T., Sohn, S., 1990, The association between revisions of financial analysts’ earnings forecasts and security price changes, Journal of Accounting & Economics 13, 341-363.
Michaely, R., Womack, K., 1999, Conflict of interest and the credibility of underwriter analyst recommendations, Review of Financial Studies 12, 573-608.
Riloff, E. 1993. Automatically Constructing a Dictionary for Information Extraction Tasks. In Proceedings of the Eleventh National Conference on Artificial Intelligence. AAAI Press/The MIT Press. 811-816.
Riloff, E., Hollaar, L., 1996, Text Databases and Information Retrieval, ACM Computing Surveys, Volume 28 Number 1, pp. 133-135.
Semin, G.R., Fiedler, K. (eds.), 1992, Language, Interaction and Social Cognition, Sage Publications, 1992.
Semin, G.R., Fiedler, K. (eds.), 1996, Applied Social Psychology, Sage Publications, 1996.
Skinner, D., 1994, Why firms voluntarily disclose bad news, Journal of Accounting Research 32, 38-60.
26
FIGURE 1 Flow of Content Analysis
27
Table 1 Descriptive Statistics for Variables: Quarterly data for 887 firms from 1996 to 2001 This table provides descriptive statistics for variables used in subsequent tests. To be included in the sample, each firm has sufficient data to compute the variables below. The time period is from 1996 to 2001.
Panel A: All industries, 5,350 observations
Mean Std. Dev. Min. Q1 Median Q3 Max.
Cost of Cap. 0.148 0.076 0.001 0.097 0.138 0.183 0.499 σ(R) 0.027 0.016 0.001 0.016 0.023 0.033 0.218 σ(FE) 0.948 1.938 0.000 0.065 0.206 0.780 12.943 Favorable 4.676 1.748 0.000 3.434 4.970 5.991 10.159 Unfavorable 1.674 0.712 0.000 1.188 1.629 2.146 5.139 Size 6.762 2.524 0.189 4.835 6.489 8.533 13.230 B/M 0.742 0.958 0.001 0.256 0.510 0.850 9.893 Leverage 0.163 0.176 0.000 0.017 0.101 0.256 0.837
Panel B: Pharmaceutical firms, 962 observations
Mean Std. Dev. Min. Q1 Median Q3 Max.
Cost of Cap. 0.146 0.110 0.002 0.064 0.116 0.204 0.499 σ(R) 0.036 0.021 0.009 0.021 0.030 0.045 0.218 σ(FE) 0.842 1.369 0.029 0.049 0.169 1.212 6.130 Favorable 4.634 1.515 0.000 3.816 4.763 5.823 8.444 Unfavorable 1.907 0.746 0.000 1.421 1.953 2.456 4.322 Size 7.100 2.678 2.046 4.800 7.055 8.949 12.578 B/M 0.371 0.778 0.001 0.098 0.196 0.345 9.416 Leverage 0.107 0.142 0.000 0.000 0.049 0.172 0.729
Panel C: Telecom firms, 987 observations
Mean Std. Dev. Min. Q1 Median Q3 Max.
Cost of Cap. 0.140 0.073 0.012 0.091 0.129 0.169 0.472 σ(R) 0.031 0.014 0.005 0.020 0.028 0.038 0.086 σ(FE) 2.364 3.299 0.000 0.206 0.911 4.107 12.943 Favorable 3.643 1.541 0.000 2.270 3.744 4.763 7.409 Unfavorable 1.544 0.720 0.000 0.933 1.540 2.039 3.633 Size 8.211 2.146 2.263 6.923 8.224 9.875 13.134 B/M 1.203 1.702 0.007 0.238 0.505 1.422 9.893 Leverage 0.216 0.183 0.000 0.071 0.200 0.300 0.837
28
Panel D: Financial firms, 3,007 observations
Mean Std. Dev. Min. Q1 Median Q3 Max.
Cost of Cap. 0.152 0.063 0.001 0.109 0.144 0.183 0.426 σ(R) 0.022 0.012 0.001 0.014 0.020 0.027 0.175 σ(FE) 0.371 0.804 0.000 0.048 0.119 0.340 5.589 Favorable 4.983 1.800 0.000 3.696 5.356 6.320 10.159 Unfavorable 1.589 0.658 0.000 1.154 1.553 1.961 3.881 Size 6.135 2.300 1.683 4.429 5.906 7.281 12.467 B/M 0.716 0.567 0.045 0.411 0.608 0.873 8.673 Leverage 0.169 0.182 0.000 0.022 0.103 0.265 0.776
Panel E: Tech firms, 394 observations
Mean Std. Dev. Min. Q1 Median Q3 Max.
Cost of Cap. 0.141 0.077 0.003 0.085 0.130 0.187 0.370 σ(R) 0.032 0.018 0.004 0.019 0.027 0.042 0.099 σ(FE) 1.560 1.819 0.110 0.164 0.970 1.654 6.233 Favorable 5.021 1.303 0.000 4.139 5.381 6.035 7.915 Unfavorable 2.077 0.727 0.000 1.530 2.077 2.583 5.139 Size 7.090 2.905 0.189 5.065 7.099 9.555 13.230 B/M 0.689 0.642 0.030 0.174 0.492 1.008 3.980 Leverage 0.121 0.136 0.000 0.002 0.084 0.203 0.723
Variable definitions: Cost of Cap.: Expected annual cost of capital calculated based on Fama-French three-factor model over the past five-year’s return data σ(R): Standard deviation of daily stock return over the contemporaneous quarter σ(FE): Standard deviation of all available quarterly analyst forecast errors from 1996 to 2001 Favorable: Favorable assessment index, constructed as mean of the content-analysis-based scaled favorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Unfavorable: Unfavorable assessment index, constructed as mean of the content-analysis-based scaled unfavorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Size: natural logarithm of a firm’s market capitalization of equity at the beginning of the quarter B/M: book equity divided by market equity at the beginning of the quarter Leverage: long-term debt divided by the total asset
29
Table 2 Cross Correlation Matrix This table provides correlations among the variables used in subsequent tests. Spearman (Pearson) correlations are above (below) the diagonal. To be included in the sample, each firm has sufficient data to compute the variables below. The time period is from 1996 to 2001.
Cost of Cap. σ(R) σ(FE) Favorable Unfavorable Size B/M Leverage
Cost of Cap. 1.000
0.161*** 0.178*** 0.031** 0.006 -0.122*** 0.138*** 0.050***
σ(R) 0.178*** 1.000 0.284*** -0.058*** 0.077*** -0.069*** -0.152*** -0.184***
σ(FE) 0.122*** 0.228*** 1.000 -0.079*** 0.094** 0.021 -0.056*** -0.089***
Favorable -0.012 -0.044*** -0.103*** 1.000 0.710*** 0.061*** -0.076*** 0.021
Unfavorable -0.012 0.068*** 0.042** 0.720*** 1.000 0.226*** -0.167*** -0.005
Size -0.149*** -0.127*** 0.099*** 0.047*** 0.203*** 1.000 -0.499*** 0.137***
B/M 0.038*** -0.019 -0.050*** -0.180*** -0.144*** -0.242*** 1.000 0.174***
Leverage 0.022 -0.167*** -0.091*** 0.032** -0.028** 0.035** 0.087*** 1.000 Variable definitions: Cost of Cap.: Expected annual cost of capital calculated based on Fama-French three-factor model over the past five-year’s return data σ(R): Standard deviation of daily stock return over the contemporaneous quarter
30
σ(FE): Standard deviation of all available quarterly analyst forecast errors from 1996 to 2001 Favorable: Favorable assessment index, constructed as mean of the content-analysis-based scaled favorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Unfavorable: Unfavorable assessment index, constructed as mean of the content-analysis-based scaled unfavorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Size: natural logarithm of a firm’s market capitalization of equity at the beginning of the quarter B/M: book equity divided by market equity at the beginning of the quarter Leverage: long-term debt divided by the total asset
Table 3 Fama-MacBeth regressions on the composite news measures
This table provides summary statistics for the coefficients from quarterly cross-sectional Fama-MacBeth regressions using the composite Favorable and Unfavorable content measures across the six categories and three sources of disclosures, i.e., corporations, analysts, and financial press. The time period is from 1996 to 2001.
Panel A: All industries, 5,350 observations
Independent Variables Dependent Variable
Cost of Cap. σ(R) σ(FE) Coefficient T-stat Coefficient T-stat Coefficient T-stat Intercept 0.1768*** 43.35 0.0410*** 27.18 1.4121*** 13.72 Favorable -0.0012* -1.82 -0.0027*** -12.4 -0.3319*** -11.19 Unfavorable 0.0031 1.33 0.0061*** 8.5 0.6514*** 8.89 Size -0.0047*** -10.81 -0.0015*** -9.15 0.0164** 1.8 B/M 0.0007 0.73 -0.0030*** -4.92 -0.1658*** -7.66 Leverage 0.0181*** 2.76 Adjusted R2 0.0159*** 4.21 0.1161*** 9.75 0.0224*** 3.91
Panel B: Pharmaceutical firms, 962 observations
Intercept 0.3102*** 34.71 0.0724*** 19.28 3.6188*** 8.02 Favorable -0.0064** -2.64 -0.0005 -0.98 -0.0145 -0.14 Unfavorable 0.0114* 1.65 0.0022* 1.92 0.3875** 1.98 Size -0.0221*** -30.25 -0.0051*** -15.22 -0.3809*** -11.51 B/M -0.0390*** -2.74 -0.0106*** -3.55 -0.6167*** -2.81 Leverage 0.0724*** 3.64 Adjusted R2 0.2220*** 13.48 0.4668*** 18.13 0.3461*** 11.09
Panel C: Telecom firms, 987 observations
Independent Variables Dependent Variable
Intercept 0.2010*** 7.75 0.0383*** 14.81 2.0324*** 3.46 Favorable 0.0035 0.86 0.0008 1.31 -0.1732 -1.07 Unfavorable -0.0221*** -3.85 0.0001 0.08 -0.1599 -0.54 Size -0.0059*** -3.23 -0.0013*** -4.24 0.2095*** 4.98 B/M 0.0061*** 2.90 -0.0021*** -4.46 -0.5358*** -6.6 Leverage 0.0147 0.72 Adjusted R2 0.1094*** 4.62 0.1316*** 5.48 -0.0885*** -5.35
32
Panel D: Financial firms, 3,007 observations
Cost of Cap. σ(R) σ(FE) Coefficient T-stat Coefficient T-stat Coefficient T-stat Intercept 0.1011*** 14.75 0.0259*** 20.29 0.2671*** 6.56 Favorable -0.0024** -2.18 -0.0005*** -3.05 -0.0685*** -6.58 Unfavorable 0.0093*** 2.66 0.0004 0.79 0.2944*** 8.22 Size 0.0064*** 12.17 -0.0006*** -3.37 -0.0113*** -2.65 B/M 0.0158*** 2.75 0.0029*** 3.55 0.0674 1.52 Leverage -0.0083 -1.16 Adjusted R2 0.0489*** 6.25 0.0618*** 4.43 -0.0290*** -3.92
Panel E: Tech firms, 394 observations
Intercept 0.1203*** 3.33 0.0469*** 6.97 6.5686*** 4.94 Favorable 0.0006 0.13 -0.0004 -0.31 -0.2533 -1.12 Unfavorable 0.0324*** 4.00 0.0073*** 5.09 1.4355*** 4.53 Size -0.0091*** -4.15 -0.0027*** -6.14 -0.6057*** -8.41 B/M 0.0186 1.26 -0.0150*** -6.76 -3.5801*** -7.24 Leverage 0.0753*** 2.89 Adjusted R2 0.1386** 2.13 0.1097*** 2.74 0.2031*** 3.37
Variable definitions: Cost of Cap.: Expected annual cost of capital calculated based on Fama-French three-factor model over the past five-year’s return data σ(R): Standard deviation of daily stock return over the contemporaneous quarter σ(FE): Standard deviation of all available quarterly analyst forecast errors from 1996 to 2001 Favorable: Favorable assessment index, constructed as mean of the content-analysis-based scaled favorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Unfavorable: Unfavorable assessment index, constructed as mean of the content-analysis-based scaled unfavorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Size: natural logarithm of a firm’s market capitalization of equity at the beginning of the quarter B/M: book equity divided by market equity at the beginning of the quarter Leverage: long-term debt divided by the total asset
33
Table 4 Fama-MacBeth regressions by the source of news This table provides summary statistics for the coefficients from quarterly cross-sectional Fama-MacBeth regressions using the Favorable and Unfavorable content measures for each source of disclosure, i.e., corporations, analysts, and financial press. The content measures are summed across the six categories of disclosures. The time period is from 1996 to 2001.
Panel A: Disclosure source is Corporate Disclosures
Independent Variables Dependent Variable
Cost of Cap. σ(R) σ(FE) Coefficient T-stat Coefficient T-stat Coefficient T-stat Intercept 0.1825*** 31.94 0.0422*** 28.70 1.6587*** 10.7 Favorable 0.0003 0.50 -0.0015*** -10.09 -0.2123*** -9.28 Unfavorable -0.0036*** -2.68 0.0025*** 5.16 0.3429*** 6.69 Size -0.0047*** -10.78 -0.0014*** -7.69 0.0010 0.07 B/M 0.0017 0.70 -0.0033*** -5.06 -0.2583*** -4.77 Leverage 0.0113** 2.03 Adjusted R2 0.0161*** 3.78 0.1008*** 8.38 0.0391*** 3.12
Panel B: Disclosure source is Analyst Reports from Investext
Intercept 0.1771*** 46.75 0.0398*** 27.18 0.9812*** 9.28 Favorable 0.0000 0.00 -0.0017*** -5.74 -0.0758*** -3.01 Unfavorable 0.0022 0.32 0.0051*** 4.97 0.0175 0.20 Size -0.0051*** -12.96 -0.0015*** -7.53 0.0548*** 6.42 B/M 0.0003 0.30 -0.0029*** -5.01 -0.1128*** -4.61 Leverage 0.0159*** 2.93 Adjusted R2 0.0234*** 5.15 0.0965*** 7.83 -0.0029 -0.95
Panel C: Disclosure source is Financial Press
Intercept 0.1757*** 54.98 0.0367*** 28.3 0.5192*** 10.16 Favorable -0.0013*** -2.80 -0.0004*** -2.89 -0.0823*** -4.18 Unfavorable 0.0026** 2.10 0.0017*** 6.84 0.2635*** 7.71 Size -0.0045*** -8.79 -0.0014*** -7.62 0.0448*** 4.20 B/M 0.0011 1.02 -0.0026*** -4.77 -0.0766*** -3.99 Leverage 0.0167** 2.50 Adjusted R2 0.0166*** 4.66 0.0779*** 6.72 -0.0018 -0.59
Variable definitions: Cost of Cap.: Expected annual cost of capital calculated based on Fama-French three-factor model over the past five-year’s return data
34
σ(R): Standard deviation of daily stock return over the contemporaneous quarter σ(FE): Standard deviation of all available quarterly analyst forecast errors from 1996 to 2001 Favorable: Favorable assessment index, constructed as mean of the content-analysis-based scaled favorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Unfavorable: Unfavorable assessment index, constructed as mean of the content-analysis-based scaled unfavorable content measures for six categories of disclosures across three sources, i.e., corporations, analysts, and financial press. Size: natural logarithm of a firm’s market capitalization of equity at the beginning of the quarter B/M: book equity divided by market equity at the beginning of the quarter Leverage: long-term debt divided by the total asset
35
Appendix I
List of Companies
FINANCIAL (225 FIRMS) 3COWP CANAL CAPITAL CORP 3DLTA DELTONA CORP 3FACOQ FIRST ALLIANCE CORP/DE 3FHRI FULL HOUSE RESORTS INC 3HOOB HOLOBEAM INC 3NRGD NRG INC 3NSLB NS & L BANCORP INC 3RATE BANKRATE INC 3STFA STRATFORD AMERICAN CORP AACE ACE CASH EXPRESS INC ABBK ABINGTON BANCORP INC ABN ABN AMRO HLDG N V -SPON ADR AEG AEGON NV AIG AMERICAN INTERNATIONAL GROUP AJG GALLAGHER (ARTHUR J.) & CO ALL ALLSTATE CORP ALLE ALLEGIANT BANCORP INC AMFH AMERICAN FINL HLDGS INC AML AMLI RESIDENTIAL PPTYS TR AMTD AMERITRADE HLDG CORP -CL A ARE ALEXANDRIA R E EQUITIES INC ARI ARDEN REALTY INC AROW ARROW FINL CORP ASFC ASTORIA FINL CORP ASFI ASTA FUNDING INC AUTH AUTHORISZOR INC AXM APEX MORTGAGE CAPITAL INC AXP AMERICAN EXPRESS BAC BANK OF AMERICA CORP BBV BANCO BILBAO VIZCAYA -ADR BCS BARCLAYS PLC/ENGLAND -ADR BED BEDFORD PPTY INVS INC BFSB BEDFORD BANCSHARES INC BK BANK OF NEW YORK CO INC BKCT BANCORP CONN INC BLM BLIMPIE INTL INC BPOP POPULAR INC BRK BERKSHIRE HATHAWAY -CL A BXG BLUEGREEN CORP BXS BANCORPSOUTH INC C CITIGROUP INC CAA CAPITAL ALLIANCE INCM TRUST CBC CENTURA BANKS INC CBL CBL & ASSOCIATES PPTYS INC CBNY COMMERCIAL BANK/NY CBSS COMPASS BANCSHARES INC CCFH CCF HOLDING CO CCRT COMPUCREDIT CORP CFBC COMMUNITY FIRST BKG CO/GA CFCP COASTAL FINANCIAL CORP/DE CFSB CITIZENS FIRST FINL CORP CIT CIT GROUP INC CNFL CITIZENS FINL CORP KY
COHB COHOES BANCORP INC COLB COLUMBIA BKG SYS INC CORS CORUS BANKSHARES INC CPT CAMDEN PROPERTY TRUST CRLTS CENTURY REALTY TRUST CRRB CARROLLTON BANCORP/MD CTGI CAPITAL TITLE GROUP INC CYL COMMUNITY CAP CORP/SC DOM DOMINION RES BLACK WARRIOR DORL DORAL FINANCIAL CORP DSL DOWNEY FINANCIAL CORP DWYR DWYER GROUP INC ECMN ECHAPMAN.COM INC EQR EQUITY RESIDENTIAL PPTYS TR FBF FLEETBOSTON FINANCIAL CORP FCH FELCOR LODGING TR INC FED FIRSTFED FINANCIAL CORP/CA FFA FRANCHISE FINL CORP AMERICA FFDB FIRSTFED BANCORP INC FFFD NORTH CENTRAL BANCSHARES INC FFHH FSF FINANCIAL CORP FFIN FIRST FINL BANCSHARES INC FFP FFP PARTNERS -LP-CL A FFWC FFW CORP FHCC FIRST HEALTH GROUP CORP FIF FINANCIAL FEDERAL CORP FII FEDERATED INVESTORS INC FKFS FIRST KEYSTONE FINL INC FKKY FRANKFORT FIRST BANCORP INC FLAG FLAG FINL CORP FLBC FINGER LAKES BANCORP INC FLPB FIRST LEESPORT BANCORP INC FNCM FINET.COM INC FNF FIDELITY NATIONAL FINL INC FNIN FINANCIAL INDS CORP FNM FANNIE MAE FR FIRST INDL REALTY TRUST INC FRE FEDERAL HOME LOAN MORTG CORP FRGB FIRST REGIONAL BANCORP FRME FIRST MERCHANTS CORP GBE GRUBB & ELLIS CO GBP GABLES RESIDENTIAL TRUST GFR GREAT AMERN FINL RESOURCES GGP GENERAL GROWTH PPTYS INC GL GREAT LAKES REIT INC GPT GREENPOINT FINANCIAL CORP GS GOLDMAN SACHS GROUP INC HARB HARBOR FLORIDA BANCSHARES HBC HSBC HLDGS PLC -SPON ADR HGT HUGOTON ROYALTY TRUST HRBT HUDSON RIVER BANCORP INC HRP HRPT PPTYS TRUST HRZB HORIZON FINANCIAL CORP/WA
36
HT HERSHA HOSPITALITY TRUST HTL HEARTLAND PARTNERS -LP-CL A HUMP HUMPHREY HOSPITALITY TR INC ICBC INDEPENDENCE CMNTY BK CORP ICII IMPERIAL CREDIT INDS INC IDH IPI INC IFCJ INTERCHANGE FINL SVCS CP/NJ IGC INTERSTATE GENERAL CO -LP INBI INDUSTRIAL BANCORP INC IRWN IRWIN FINL CORP ISTN INTERSTATE NATL DEALER SVCS IUBC INDIANA UTD BANCORP JDN JDN REALTY CORP JPR JP REALTY INC KILN KIRLIN HOLDING CORP KIM KIMCO REALTY CORP KRB MBNA CORP LM LEGG MASON INC LNC LINCOLN NATIONAL CORP LTC LTC PROPERTIES INC LTFD LITTLEFIELD CORP LXBK LSB BANCSHARES INC/NC MADB MADISON BANCSHARES GROUP LTD MAFB MAF BANCORP INC MBHI MIDWEST BANC HLDGS INC MBI MBIA INC MBWM MERCANTILE BANK CORP MER MERRILL LYNCH & CO MET METLIFE INC METB METROBANCORP MGP MERCHANTS GROUP INC MHC MANUFACTURED HOME CMNTYS INC MHX MERISTAR HOSPITALITY CORP MI MARSHALL & ILSLEY CORP MLAN MIDLAND CO MM MUTUAL RISK MANAGEMENT LTD MMC MARSH & MCLENNAN COS MTR MESA ROYALTY TRUST MWD MORGAN STANLEY DEAN WITTER NEIB NORTHEAST INDIANA BANCORP NFB NORTH FORK BANCORPORATION NFS NATIONWIDE FINL SVCS -CL A NHI NATIONAL HEALTH INVS INC NHL NEWHALL LAND &FARM -LP NLY ANNALY MORTAGE MGMT INC NSDB NSD BANCORP INC NTRS NORTHERN TRUST CORP NYM NYMAGIC INC OAKF OAK HILL FINANCIAL INC ONE BANK ONE CORP OSBC OLD SECOND BANCORP INC/IL OSKY MAHASKA INVESTMENT CO OWWI OMEGA WORLDWIDE INC PEBK PEOPLES BK N C PEDE GREAT PEE DEE BANCORP INC PEI PENNSYLVANIA RE INVS TRUST PFCO PAULA FINANCIAL/DE PHC PEOPLES HLDG CO PMI PMI GROUP INC PRSP PROSPERITY BANCSHARES INC PUK PRUDENTIAL PLC -ADR PVFC PVF CAPITAL CORP PVSA PARKVALE FINL CORP
PXT PXRE GROUP LTD QUIZ QUIZNOS CORP RA RECKSON ASSOCS RLTY CORP RBNC REPUBLIC BANCORP INC RFS RFS HOTEL INVESTORS INC RIGS RIGGS NATL CORP WASH D C RPP STONEHAVEN REALTY TRUST RVSB RIVERVIEW BANCORP INC SAL SALISBURY BANCORP INC SASR SANDY SPRING BANCORP INC SBAN SOUTHBANC SHARES INC SBCO SOUTHSIDE BANCSHARES CORP SCH SCHWAB (CHARLES) CORP SFEF SANTA FE FINANCIAL CORP SIGI SELECTIVE INS GROUP INC SIVB SILICON VY BANCSHARES SJT SAN JUAN BASIN ROYALTY TR SMBC SOUTHERN MISSOURI BANCP INC SMT SUMMIT PROPERTIES INC SNH SENIOR HOUSING PPTYS TRUST SPN SECURITY PA FINL CORP SSLI SOUTHERN SEC LIFE INS STBA S & T BANCORP INC STD BANCO SANTANDER CENT -ADR STRS STRATUS PROPERTIES INC STSA STERLING FINL CORP/WA STT STATE STREET CORP SUBK SUFFOLK BANCORP SUSQ SUSQUEHANNA BANCSHARES INC SV STILWELL FINL INC SWS SOUTHWEST SECURITIES GROUP TRMK TRUSTMARK CORP UBH U S B HOLDING INC UBSC UNION BANKSHARES LTD/CO UFHI UNITED FINANCIAL HOLDINGS UHCO UNIVERSAL AMERICAN FINL CP UHT UNIVERSAL HEALTH RLTY INCOME UIRT UTD INVESTORS REALTY TRUST UMBF UMB FINANCIAL CORP UNBJ UNITED NATIONAL BANCORP/NJ UNBO UNB CORP/OH UPFC UNITED PANAM FINANCIAL CORP UPK UNITED PARK CITY MINES VLY VALLEY NATIONAL BANCORP VSTN VESTIN GROUP INC WABC WESTAMERICA BANCORPORATION WAYN WAYNE SAVINGS BANCSHARES INC WES WESTCORP WFC WELLS FARGO & CO WFI WINTON FINANCIAL CORP WLP WELLPOINT HLTH NETWRK -CL A WOFC WESTERN OHIO FINL CORP WSB WASHINGTON SVGS BANK F S B WSFS WSFS FINL CORP WTU WILLIAMS COAL SEAM RYL TRUST XL XL CAPITAL LTD
37
PHARMACEUTICAL (72 FIRMS)
ABT ABBOTT LABORATORIES AGN ALLERGAN INC AHP AMERICAN HOME PRODUCTS CORP AKC ACCESS PHARMACEUTICALS INC ALO ALPHARMA INC -CL A ALT ALTEON INC AMGN AMGEN INC ANX ANTEX BIOLOGICS INC AVM ADVANCED MAGNETICS INC AVN AVANIR PHARMACEUTCLS -CL A AZN ASTRAZENECA PLC -SPON ADR BGEN BIOGEN INC BMY BRISTOL MYERS SQUIBB BNT BENTLEY PHARMACEUTICALS BOL BAUSCH & LOMB INC BRL BARR LABORATORIES INC BTX BIOTIME INC BVF BIOVAIL CORP CBM CAMBREX CORP CHIR CHIRON CORP CLL CELLTECH GROUP PLC -SP ADR COB COLUMBIA LABORATORIES INC COR CORTEX PHARMACEUTICALS INC CVM CEL-SCI CORP DMI DEPOMED INC DNA GENENTECH INC DOR ENDOREX CORP DP DIAGNOSTIC PRODUCTS CORP ELI ELITE PHARMACEUTICALS INC ELN ELAN CORP PLC -ADR EZM.A E-Z-EM-INC -CL A FRX FOREST LABORATORIES -CL A GNT GENSTAR THERAPEUTICS CORP GSK GLAXOSMITHKLINE PLC -SP ADR HEB HEMISPHERX BIOPHARMA INC ICN ICN PHARMACEUTICALS INC IG IGI INC
IMA INVERNESS MEDICAL TECHNOLGY ISV INSITE VISION INC IVX IVAX CORP JNJ JOHNSON & JOHNSON KG KING PHARMACEUTICALS INC KV.A K V PHARMACEUTICAL -CL A LBC LABORATORIO CHILE -SPON ADR LLY LILLY (ELI) & CO MEDI MEDIMMUNE INC MEW VASOGEN INC MRK MERCK & CO MRX MEDICIS PHARMACEUT CP -CL A MYL MYLAN LABORATORIES NVAX NOVAVAX INC NVO NOVO-NORDISK A/S -ADR NVS NOVARTIS AG -SPON ADR NYE AMERSHAM PLC -ADR ORG ORGANOGENESIS INC PFE PFIZER INC PHA PHARMACIA CORP PRX PHARMACEUTICAL RES INC PTN PALATIN TECHNOLOGIES INC QSC QUESTCOR PHARMACEUTICALS INC RDY DR REDDYS LABS LTD -ADR SGP SCHERING-PLOUGH SHM SHEFFIELD PHARMACEUTICALS SHR SCHERING AG -ADR SIAL SIGMA-ALDRICH SRA SERONO S A -ADR TGX THERAGENICS CORP TTP TITAN PHARMACEUTICALS INC UG UNITED GUARDIAN INC VRA VIRAGEN INC WNI WEIDER NUTRITION INTL -CL A WPI WATSON PHARMACEUTICALS INC
TECHNOLOGY (197 FIRMS)
3128B NOVADIGM INC 3ALLN ALLIN CORP 3BMLS BURKE MILLS INC 3CNDR CONDOR TECH SOLUTIONS INC 3CVTL CONVERSION TECHNLGS INTL INC 3DGLH DIGITAL LIGHTHOUSE CORP 3FRTL FORTEL INC/CA 3INFO INFONAUTICS CORP -CL A 3ISOL IMAGE SOFTWARE INC 3MEDY MEDICAL DYNAMICS INC 3MGXI MICROGRAFX INC 3MSOF MULTI SOFT INC 3SHPI SPECIALIZED HEALTH PDS INTL 3SYCM SYSCOMM INTERNATIONAL CORP 3TISVE TENET INFORMATION SVCS INC 3TSSW TOUCHSTONE SOFTWARE CORP 3USCM USCI INC 3WCIIQ WINSTAR COMMUNICATIONS ACRU ACCRUE SOFTWARE INC
ADP AUTOMATIC DATA PROCESSING ALPH ALPHANET SOLUTIONS INC AMEN CROSSWALK.COM INC AMZN AMAZON.COM INC AOLA AMERICA ONLNE LTN AMR -CL A ARBA ARIBA INC ASIA ASIAINFO HLDGS INC ASKJ ASK JEEVES INC AVRT AVERT INC BARZ BARRA INC BEAS BEA SYSTEMS INC BFLY BLUEFLY INC BFRE BE FREE INC BRCM BROADCOM CORP -CL A BTGI BTG INC CAMZ CAMINUS CORP CBR CIBER INC CENI CONESTOGA ENTERPRISES CERO CONCERO INC
38
CFA CHEMFAB CORPORATION CGZ COTELLIGENT INC CKC COLLINS & AIKMAN CORP CLRN CLARENT CORP CLRS CLARUS CORP COE CONE MILLS CORP CONE CARRIER1 INTL S A -ADR CPTH CRITICAL PATH INC CSCC CENTERSPAN COMMUN CORP CSGS CSG SYSTEMS INTL INC CSPI CSP INC CTB COOPER TIRE & RUBBER CTRA CENTRA SOFTWARE INC DIGX DIGEX INC DOCX DOCUMENT SCIENCES CORP DRCO DYNAMICS RESEARCH CORP DSET DSET CORP DTAB10 DATATAB INC DTLK DATALINK CORP EDS ELECTRONIC DATA SYSTEMS CORP EGOV NATIONAL INFO CONSORTIUM INC EGPT EAGLE POINT SOFTWARE CORP ELNK EARTHLINK INC ELOTC ELOT INC EPAY BOTTOMLINE TECHNOLOGIES INC EPRS EPRISE CORP ESAFE SAFELINK CORPORATION ESTM E-STAMP CORP EXAP EXCHANGE APPLICATIONS EXDS EXODUS COMMUNICATIONS INC FCGI FIRST CONSULTING GROUP INC FCOM FOCAL COMMUNICATIONS CORP FDC FIRST DATA CORP FILE FILENET CORP FIRE FIREPOND INC FIT FAB INDUSTRIES INC FIVE 5B TECHNOLOGIES CORP FLH FILA HLDGS S P A -SPON ADR FMXI FOAMEX INTERNATIONAL INC FSST FASTNET CORP FUTR IFX CORP GBIX GLOBIX CORP GENU GENUITY INC GFD GUILFORD MILLS INC GIB.A GROUPE CGI INC -CL A GLC GALILEO INTERNATIONAL INC GNL GALEY & LORD INC GNSM GENSYM CORP GRIC GRIC COMMUNICATIONS INC GSPT GLOBAL SPORTS INC HAXS HEALTHAXIS INC HNCS HNC SOFTWARE INC HTEI H T E INC IACT INTERACT COMMERCE CORP IATV ACTV INC IBM INTL BUSINESS MACHINES CORP IFGM INFOGRAMES INC ILND INTERLAND INC INAI INTELLICORP INC INGR INTERGRAPH CORP ININ INTERACTIVE INTELLIGENCE INC INSW INSWEB CORP INTD INTELIDATA TECHNOLOGIES CORP
IPLY INTERPLAY ENTERTAINMENT CORP ISMT CHINAB2BSOURCING.COM INC ITIG INTELLIGROUP INC ITWO I2 TECHNOLOGIES INC IWOV INTERWOVEN INC IXX IVEX PACKAGING CORP LGTO LEGATO SYSTEMS INC LIOX LIONBRIDGE TECHNOLOGIES INC LNTY L90 INC LNUX VA LINUX SYSTEMS INC LOAX LOG ON AMERICA INC LOOK LOOKSMART LTD LQID LIQUID AUDIO INC LUMT LUMINANT WORLDWIDE CORP LVEL LEVEL 8 SYS INC MAPS MAPINFO CORP MMTM MOMENTUM BUSINESS APPS INC MRGE MERGE TECHNOLOGIES INC MSFT MICROSOFT CORP MYE MYERS INDUSTRIES INC NATI NATIONAL INSTRUMENTS CORP NEOM NEOMEDIA TECHNOLOGIES INC NETP NET PERCEPTIONS INC NLI NTL INC NOVL NOVELL INC NQLI NQL INC NTIQ NETIQ CORP NTRT NETRATINGS INC NTWK NETSOL INTERNATIONAL INC NUAN NUANCE COMMUNICATIONS INC NWSS NETWORK SIX INC OAOT OAO TECHNOLOGY SOLUTIONS INC OGNC ORGANIC INC OPTK OPTIKA INC OSII OBJECTIVE SYS INTEGRATRS INC PCD PLANETCAD INC PECS PEC SOLUTIONS INC PEGS PEGASUS SOLUTIONS INC PLR.A PLYMOUTH RUBBER -CL A PMIX PRIMIX SOLUTIONS INC PNS PINNACLE DATA SYSTEMS INC PPOD PEAPOD INC PPRO PURCHASEPRO.COM PRSF PORTAL SOFTWARE INC PTEC PHOENIX TECHNOLOGIES LTD PWEI PW EAGLE INC QMDC QUADRAMED CORP QRSI QRS CORP RAZF RAZORFISH INC RHAT RED HAT INC RMI ROTONICS MANUFACTURING INC RNWK REALNETWORKS INC ROSS ROSS SYSTEMS INC RWAV ROGUE WAVE SOFTWARE INC SEBL SIEBEL SYSTEMS INC SFTY ESAFETYWORLD INC SIDE ASSOCIATED MATERIALS INC SMI SPRINGS INDUSTRIES -CL A SNHK SUNHAWK.COM CORP SNOW SNOWBALL.COM INC SSNC SS&C TECHNOLOGIES INC STRM STARMEDIA NETWORK INC SUNQ SUNQUEST INFORMATION SYS INC
39
SWTX SOUTHWALL TECHNOLOGY SYMC SYMANTEC CORP TACX A CONSULTING TEAM INC THQI THQ INC THV THERMOVIEW INDS INC TIR CHINA TIRE ECOMMERCE.COM LTD TSCN TELESCAN INC TTWO TAKE-TWO INTERACTIVE SFTWR TWNE TOWNE SERVICES INC TWSTY TELEWEST COMMUNICATION -ADR TZIX TRIZETTO GROUP INC UBIX UBICS INC UGS UNIGRAPHICS SOLUTIONS INC UIS UNISYS CORP UPRO UPROAR INC USI U S INDUSTRIES INC USIX USINTERNETWORKING INC VAST VASTERA INC VCLK VALUECLICK INC VIZY VIZACOM INC VLCT VALICERT INC VRSO VERSO TECHNOLOGIES INC VRTL VERTEL CORP VRTS VERITAS SOFTWARE CO VSTR VOICESTREAM WIRELESS CORP VUL VULCAN INTL CORP WCG WILLIAMS COMMUNICATIONS GRP WEBT WEBTRENDS CORP WITS WITNESS SYSTEMS INC WKGP WORKGROUP TECHNOLOGY CORP XOXO XO COMMUNICATIONS INC -CL A
TELECOM (395 FIRMS)
AATK American Access Technologies
Inc ABIZQ Adelphia Business Solutions ACTG Acacia Technologies ACTT Act Teleconferencing Inc ADCT ADC Telecommunications Inc ADPT Adaptec Inc ADTN Adtran Inc AETH Aether Sys Inc AFCI Advanced Fibre Communications AINN Applied Innovation Inc AIRN Airspan Networks ALA Alcatel ALGXQ Allegiance Telecom Inc ALMO Alamosa Hldgs Inc ALSK Alaska Communications Sys Grp ALVR Alvarion Ltd AMT American Tower Corp AMX America Movil S A De C V ANCC Airnet Communications ANDW Andrew Corp ANK Atlantic Tele Network Inc APSG Applied Signal Technology Inc ARDI @Road Inc. ARDTQ Ardent Communications Inc ARRS Arris Group Inc ASAT Esat Inc
ASPT Aspect Communications Corp AT Alltel Corporation ATEL Accesstel Inc ATGN AltiGen Communications Inc ATNG Atng Inc ATS APT Satellite ATSC ATSI Communications Inc. ATTL AT&T Latin America Corp. AVCI Avici Systems Inc AWE AT&T Wireless Services Inc. BBOX Black Box Corp BCE BCE Inc. BCGI Boston Communications Group BCICF Bell Cda Intl Inc BFH Cabco Tr For Bellsouth Debs BIZZ Biznessonline Com Inc BKET Bankengine Technologies Inc BLB Bellsouth Cap Fdg Corp BLS BellSouth Corporation BMKS Brandmakers Inc BOGN Bogen Communications
International Inc BRCD Brocade Communcations Systems
Inc BRKT Brooktrout Inc BRP Brasil Telecom Participacoes BSY British Sky Broadcasting Group
ADS BTM Brasil Telecom Sa BTY Bt Group Plc CACS Carrier Access Corp CAMP California Amplifier Inc CAPA Captaris Inc CATS Catalyst Semi CCI Crown Castle International
Corp CCMS Champion Communications Svcs CEL Grupo Iusacell S A De C V New CFW CyberGuard Corp CHA China Telecom Corp Ltd CHL China Mobile Hong Kong Ltd CHNL Chanell Commercial Corp CHTR Charter Communications CHU China Unicom Ltd CIC A T & T Cap Corp CLEC US Lec Corp CMCSK Comcast CMNT Computer network Technology CMTL Comtech Telecommunications
Corp CMTN Copper Mountain Networks Inc CNEZF Call-Net Enterprises Inc CNOW Call Now Inc COLT Colt Telecom Group Plc COMM Atx Communications Inc COMS 3Com Corp CORV Corvis Corp COVD Covad Communications Group Inc COX Cox Communications CRDS Crossroads Systems Inc CRNT Ceragon Network CSA Chilesat Corp S A CSCO Cisco Systems Inc
40
CSOL Call Solutions Inc CSTL Castelle Inc CTC Compañia de Telecomunicaciones
de Chile S.A. CTCI Ct Communications Inc CTCO Commonwealth Tel Enterprises CTEL City Telecom H K Ltd CTL Centurytel Inc CTLM Centillium Communications Inc CTV Commscope Inc CVC Cablevision Systems CVST Covista Communications Inc CWLC China Wireless Communications CWON Choice One Communication Inc CWP Cable and Wireless Plc CYAD Cyberads Inc CYCL Centennial Communctns Corp New CYTP Cybertel Communications Corp CZN Citizens Communications Co CZNPR Citizens Utils Tr DAVL Davel Communications Inc DCEL Dobson Communications Corp DCM Ntt Docomo Inc DDDC Deltathree Inc DECC D&E Communications Inc DGII Digi International Inc DIGL Digital Lightwave Inc DISH EchoStar Communications DITC Ditech Communcations Corp DSPG DSP Group Inc DT Deutsche Telekom Ag DTIX Dial Thru International Corp DTVN Dtvn Hldgs Inc DVID Digital Video Systems Inc ECTX ECtel Ltd ELEC Elec Communications Corp ELMG EMS Technologies Inc EMT Empratel Participacoes ENCW Encore Wireless Inc ENG ENGlobal Corp ENWV Endwave Corp EPHO Ephone Telecom Inc ERICY Ericsson LM Telephones ERS Empire Resources Inc ETLK Elephant Talk Comm Inc ETS Enterasys Networks FCSE Focus Enhancements Inc FGWC 5G Wireless Communications Inc FIBR Sorrento Networks Corporation FNSR Finisar Corp FON Sprint Corporation FRNS First Ave Networks Inc FTE France Telecom FULO Fullnet Communications Inc GHGI Greenhold Group Inc GIGM Gigamedia Ltd GLDN Golden Telecom Inc GLW Corning Inc GMH Hughes Electronics Corp GNCMA General Communications Inc GNSY Genesys S A GTCC Gtc Telecom Corp GTTIF Grandetel Technologies Inc
GTTLQ Gt Group Telecom Inc HANA Hanaro Telecom Inc HCT Hector Communications Corp HLIT Harmonic Inc HRCT Hartcourt Cos Inc HRS Harris Corp HSVI Home Svcs Intl Inc HTC Hungarian Telephone and Cable
Corp. HTCO Hickory Tech Corporation IBAS Ibasis Inc ICGC Icg Communications Inc IDCC Interdigital Communications IDSY I.D. Systems Inc IDTC IDT Corporation IGTT Indiginet Inc IIIM I3 Mobile Inc IIT Perusahaan Perseroan (Persero)
P.T. Ind. Sat. Corp. ILNK I-Link Corp INPH Interphase Corp INTA Intasys Corp INTI Inet Technologies Inc ITCD Itc Deltacom Inc ITSN Its Networks Inc ITXC Itxc Corp JCOM J2 Global Communications Inc JNIC JNI Corp JNPR Juniper Networks KPN Royal PTT Nederland NV KQIPQ Kpnqwest N V KTB Corts Tr Bellsouth KTC Kt Corp KVHI KVH Industries Inc KVZ Verizon New Eng Inc LCCI LCC International LCST Lecstar Corp LIC Lynch Interactive Corp LLL L-3 Communications Holdings,
Inc LOR Loral Space & Communications LQ Quinenco SA LTBG Lightbridge Inc LU Lucent Technologies LWINQ Leap Wireless Intl Inc LXNT Lexent Inc (USA) MANW Mobile Area Networks Inc MBT Mobile Telesystems Ojsc MBYTF Moving Bytes Inc MC Matsushita Electronics MCDT McDATA Corp MCKC MCK Communications Inc MCLD Mcleodusa Inc MCWEQ Worldcom Inc Worldcom Group MDLK Medialink Worldwide Inc MFNXQ Metromedia Fiber Network Inc MNCP Motient Corp MNPL Minorplanet Sys Usa Inc MOORZ Chadmoore Wireless Group Inc MOT Motorola Inc MPOW Mpower Holding Corp MRCY Mercury Computer Systems, Inc MTA Magyar Tavkozlesi Rt
41
MTE Mahanagar Tel Nigam Ltd MTOH Metrocall Hldgs Inc MTON Metro One Telecommunications
Inc MTPPRA Montana Power Cap I MTRM Metromedia International
Group, Inc MTSL MER Telemanagement Solutions MWVT Microwave Transmission Sys Inc NASC Network Access Solutions Cor NETX Netlojix Communications Inc NIHD Nii Hldgs Inc NIPNY NEC Corporation NMDV New Millennium Dev Group NMRX Numerex Corp NOK Nokia Corp NPLSQ Network Plus Corp NPSI North Pittsburgh Systems Inc NSK New Skies Satellites N.V. NT Nortel Networks NTIAQ Netia Hldgs S A NTKKQ Net2000 Communications Inc NTL Nortel Inversora S A NTLOQ Ntelos Inc NTOP Net2Phone NTRO Netro Corp NTT Nippon Telegraph and Telephone
Corporation NUEP Net 1 Ueps Technologies Inc NULM New Ulm Telecom Inc NVNW Novo Networks Inc NVTL Novatel Wireless NWK Network Equipment Technologies
Inc NWRE Neoware Systems Inc NXTL Nextel Communications Inc Cl A NXTP Nextel Partners Inc NZT Telecom Corporation of New
Zealand Limited OBAS Optibase Ltd OCCF Optical Cable Corp OECI Orbit E-Commerce Inc OOM Mmo2 Plc OPTC Optelecom Inc OTE Hellenic Telecom Organizatn Sa PACW Pac-West Telecom Inc PCSA Airgate Pcs Inc PCTI PC-Tel Inc PCW Pccw Ltd PDYN Paradyne Networks Inc PHGW Phone 1Globalwide Inc PHI Philippine Long Distance
Telephone Company PNTA Pentastar Communications Inc PR Price Communications
Corporation PRMC Prime Cos Inc PROX Proxim Corp PRTL Primus Telecommunications Grp PSEM Pericom Semiconductor Corp PT Portugal Telecom, S.A. PTEK Ptek Hldgs Inc PTNR Partner Communications Co Ltd
PVSS Pivotal Self Service Tech Inc PWAV Powerwave Technologies Inc Q Qwest Communications Intl Inc QCOM Qualcomm Inc RADN Radyne Comstream Inc RCCC Rural Cellular Corporation RCN Rogers Wireless Communications RCNC RCN Corp REMC Remec Inc RFMI RF Monolithics Inc RG Rogers Communications Inc RIMM Research In Motnion Ltd RNDC Raindance Communications Inc ROS Rostelecom Open Jt Stock Lng
Dst RSTN Riverstone Networks Inc RSTRC Rstar Corp SAT Asia Satellite
Telecommunications Holdings Ltd
SBAC SBA Communications Corp SBC SBC Communications Inc. SBEI SBE Inc SCKT Socket Communications Inc SCM Swisscom Ag SCMR Sycamore Networks Inc SDAY Sunday Communication Ltd SFA Scientific-Atlanta Inc SHEN Shenandoah Telecommunications SI Siemens AG SILCF Silicom Ltd SKM Korea Mobile
Telecommunications Corp. SLNK Spectralink Corp SMSC Standard Microsystems Corp SONT Sontra Medical Corp SPAL Spantel Communications Inc SPM Spirent PLC SPOT Panamsat Corp SRTI Sunrise Telecom Inc SSPI Spectrum Signal Processing Inc STGC Startec Global Communication STHLY Stet Hellas Communications S A STXN Stratex Networks Inc SURW Surewest Communications SVVS Savvis Communications Corp SWIR Sierra Wireless Inc SYMM Symmetricom Inc T AT&T Corp. TALK Talk America Hldgs Inc TAR Telefonica De Argentina S A TBE Tele Leste Celular
Participacoes S.A. TBH Telebras HOLDRS TCN Tele Norte Celular Part S A TCOM Telecom Communications Inc TCP Telesp Celular Part S A TDA Telephone & Data Sys Inc TDP Telefonica Del Peru S A TDR Tricom Sa TDS Telephone And Data Systems,
Inc. TDSPRB Tds Cap Ii
42
TEF Telefonica de España, S.A. TELN Telenor Asa TEM Telefonica Moviles S A TEO TELECOM ARGENTINA STET -
France Telecom S.A. TFONY Telefonos De Mexico S A TI Telecom Italia Spa TKA Telekom Austria Ag TKC Turkcell Iletisim Hizmetleri TLAB Tellabs Inc TLD Tdc A/S TLL Teletouch Communications Inc TLMD Telemonde Inc TLS Telstra Corp Ltd TLTOB Tele2 Ab TMB Telemig Celular Part S A TMX Telefonos de Mexico, S.A. de
C.V. TND Tele Nordeste Celular Part S A TNE Tele Norte Leste Part S A TNTX T-NETIX Inc TPC Triton Pcs Hldgs Inc TRO Tele Centro Oeste Celular S A TROY Troy Group Inc TSD Tele Sudeste Celular Part S A TSP Telecomunicacoes De Sao Paulo TSTN Turnstone Systems TSU Tele Celular Sul Part S A TSYS TeleCommunication Systems TU Telus Corp TUTS Tut Systems Inc TWTC Time Warner Telecom Inc UAX Usurf America Inc UCSY Universal Comm Sys Inc Nev ULCM Ulticom Inc UMGPRC Mediaone Fin Tr Iii UNWR US Unwired Inc UPCS Ubiquitel Inc USM United States Cellular
Corporation USPO Usip Com Inc USRT U S Realtel Inc UTSI UTStarcom Inc UVCL Univercell Hldgs Inc VEDO Villageedocs Inc VICM Vicom Inc VIP Open Jt Stock Co-Vimpel
Communic VNT Compania Anonima Nacionl Tel VNWI Via Net Wrks Inc VNWK Visual Networks Inc VOD Vodafone Group Plc New VRLK Verilink Corp VSAT ViaSat Inc VSL Videsh Sanchar Nigam Ltd VYYO Vyyo Inc VZ Verizon Communications VZC Verizon South Inc WEBX Webex Communications Inc WFII Wireless Facilities Inc WGNR Wegener Corp WJCI WJ Communications Inc WONE Wire One Technologies
WQNI WorldQuest Networks Inc WRDP Worldport Communications Inc WRLS Telular Corp WSTL Westell Technologies Inc WTEQ Worldteq Group Intl Inc WVCM Wavecom S.A. ads WWCA Western Wireless Corp Cl A WWCO Wireless Webconnect Inc WWVY Warwick Valley Tel Co WWWNQ Worldwide Wireless Networks XETA Xeta Technologies Inc XING Qiao Xing Universal Telephone
Inc ZENX Zenex Intl Inc ZICA ZI Corporation ZNTH Zenith Technology Inc ZOOM Zoom Technologies Inc ZTEL Z Tel Technologies Inc
Appendix II
Business Category Dictionary
We defined six categories of business words and word meanings to classify disclosure texts. Below are the categories and words comprising each category.
1. Statements of Market Risk, Industry Structure and Competitive Forces Market search terms: market, marketplace, environment, segment, sector Competition search terms: competitor, competition, dominant Industry structure search terms: industry, entrant, supplier, buyer, substitute, scale, product, brand, switching, capital, access, cost, rivalry, capacity, concentration, exit, barrier, price, profit, quality, input, volume, purchase, integration, power Competitive analysis search terms: customer, channel, value, first-mover, technology, alliance, partnership, venture, regulation, litigation
2. Statements of Firm-Level Strategy Intent, Product Market Performance,
Performance of Business Strategy Model in Use
Strategic intent:
search terms: strategy, strategic, value, sales, revenue, share, profit, profitability, product, service, lead, leader, quality, customer, buyer, growth, opportunity, risk, resource
Innovation and research and development
search terms: research and development, r&d, patent, discovery, license, licensing, regulation, regulatory, trial, monitor, innovate, innovation, competence
Mode of entry
search terms: entry, cost, business, complementary, green-field, venture, investment, capital, solution, price
Business model
44
search terms: model, best, lowest, low, highest, high, supplier, distribution
Partnerships
search terms: partner, alliance, merger, acquisition, joint, venture, relationship, equity, asset
3. Statements of Human and Organizational Capital, Quality of Management Performance, Corporate Governance and Leadership
Leadership
search terms: leader, leadership, record, value, culture, responsibility, goal, objective
Management quality
search terms: management, quality, best, proven, experience, teamwork
Governance
search terms: governance, corporate, board, incentive, owner, ownership, compensation, recruitment, development
Disclosure
search terms: disclosure, transparent, transparency, information, audit, auditing, oversight, assurance, regulation, mandate, mandated
Measures
search terms: up, down, better, worse, recover, advance, advancing, progress, progressing, expand, expanding, improve, improving, reduce, reducing, reduction, decline, declining, retain, retention, profit, profitability, feedback, scorecard, growth, growing, performance, projected, projections.
4. Statements of market recognition, power and consistency of branded image, measures of consumer confidence and trust in branded image
Customer
search terms: customer, satisfaction, feedback, trust
Brand
search terms: brand, image, name, trademark, recognition, stretch, quality, awareness
45
Media
search terms: media, radio, television, newspaper, internet, promotion, media spend, media budget, announcements, release, media spend, media budget
Advertising
search terms: advertising, ad, direct, channel, advertising spend, ad spend, advertising allocation, advertising budget, ad budget
Corporate image
search terms: corporate image, reputation, integrity, community, trust, trusted name, confidence, durability, strength, character
5. Statements of corporate and business unit financial performance
Financial performance
search terms: gross, net, return on investment, return on sales, return on assets, return on equity, ROI, ROA, ROE, profit, earnings, margin, capital, debt, sales, EBITDA, ratings, leverage, valuation, cost of capital
Forecasting
search terms: forecast, forecasting, cash flow, prospectus, quarterly
Insider stock transactions
search terms: insider buy, insider sell
6. Statements and/or References to Federal Government Regulation Enacted or Pending Influential to firm Competitiveness, Product Market Performance, and/or Disclosure Practices
Regulation
search terms: regulation, federal, state, securities and exchange commission, commerce, legislation, congress, law, legal, hearings, enacted, pending, sec, medicare, medicaid, FDA
Special interest groups
search terms: lobby, lobbyists, special, interest, expert, testimony, industry, watchdog, consumer rights, patient rights
Appendix III Examples of Disclosure Texts, Amgen Corporation
Technology & Health
Study on Amgen's Obesity Drug, Leptin, Raises Questions About Its Effectiveness By Robert Langreth and Rhonda L. Rundle
Staff Reporters of The Wall Street Journal
02/01/1996 The Wall Street Journal B5
New research raises questions over whether Amgen Inc.'s experimental obesity drug will work on most overweight patients.
Amgen has gambled millions of dollars on the new drug, leptin, which will be tested in humans later this year. Leptin, a hormone produced by fat cells, is believed to regulate weight by signaling the brain when enough fat has been stored and ordering it to slash food intake and speed metabolism.
But the new study, published this week in the New England Journal of Medicine, found that obese people have no problem producing leptin, as some researchers had suspected; in fact, they have an average of four times more leptin in their bloodstream than lean people.
"The idea that leptin will be a magic bullet for everyone with obesity, this study puts that idea aside," said lead researcher Jose Caro from the Thomas Jefferson University in Philadelphia.
Amgen shares fell 1.9% in Nasdaq Stock Market trading yesterday, closing at $60.125, down $1.1875, when word of the study began to leak out. After the market closed, Amgen reported fourth-quarter net income of $145.6 million, or 52 cents a share, on a 16% increase in revenue, in line with analysts' forecasts.
Amgen said its profit in the quarter jumped 20%, excluding a large charge in the year-earlier quarter. The latest research may hint that, instead of a deficiency of leptin in an overweight person, the problem could lie in an inability of the brain and its "receptors" to receive leptin's signals.
That would put Amgen at odds with a small rival. Last year Amgen paid Rockefeller University a record $20 million for rights to the newly discovered gene that produces leptin. But a rival, Millennium Pharmaceuticals in Cambridge, Mass., is betting the receptor, a protein in the brain, is the key. Backed by Roche Holding Ltd.'s Hoffman-LaRoche Inc., Millennium said last month it had discovered the gene that makes the receptor. It is attempting to design a drug that would boost the leptin receptor's sensitivity to signals from the hormone.
Amgen disputed the study's significance. "This doesn't really answer the question whether giving leptin to people will cause them to lose weight or not. That can only be answered in a human clinical trial, which we will conduct in a few months," said Gordon Binder, chief executive of Amgen, Thousand Oaks, Calif. Millennium countered that the new finding bolsters its own
47
approach. "This is very important research and confirms our own internal studies," said Louis Tartaglia, a Millennium researcher who helped discover the receptor gene. Last year, scientists at Amgen and Rockefeller University showed that grossly obese mice lost 22% to 40% of their weight after getting shots of the leptin hormone. The results raised hopes that the same hormone might work in humans.
But in the New England Journal study, Dr. Caro and his colleagues measured leptin levels in 139 obese people and 136 lean people. On average, the overweight subjects had four times as much leptin as lean ones. Some obese subjects had levels of the hormone elevated 20 or 30 times normal. About 5% of the overweight subjects, however, had lower leptin levels, and might be more likely to react to leptin therapy, Dr. Caro noted. Eli Lilly & Co. researchers also participated in the New England Journal study.
Amgen's fourth-quarter earnings included a charge of $10 million, or two cents a share, in license fees to NPS Pharmaceuticals Inc., a Salt Lake City drug-development company. In the year-earlier quarter, Amgen had earnings of $4.8 million, or two cents a share, after a $116 million charge related to the acquisition of Synergen. Revenue in the latest quarter rose to $513.5 million from $442.9 million.
48
Technology & Health
Amgen Says Profit Increased by 32%, Plans New Product By Rhonda L. Rundle
Staff Reporter of The Wall Street Journal
04/18/1996 The Wall Street Journal B6
Biotechnology leader Amgen Inc. reported a 32% increase in first-quarter net income and disclosed for the first time a new product candidate and the filing of an application for an experimental Hepatitis C drug. Net income climbed to $143.6 million, or 51 cents a share, from $108.6 million, or 39 cents a share, a year earlier, a penny ahead of analysts' expectations. Last year's first-quarter results were reduced by a $20 million technology payment.
Revenue rose 16% to $507.9 million from $439.4 million, reflecting continued strong growth of Amgen's two drugs. Epogen sales rose 23% to $244 million, while Neupogen sales rose 10% to $233 million.
"It was a good, solid quarter," said Joyce Lonergan, an analyst at Cowen & Co. First-quarter domestic Neupogen sales were weak because of a lag effect after the holidays when there aren't as many cancer chemotherapy treatments, she said. Neupogen bolsters white blood cells that are destroyed by anticancer drugs. The results were announced after the stock market had closed. Amgen shares fell $1.375 to $53.125 in Nasdaq trading.
Amgen, Thousand Oaks, Calif., said it is developing a new protein that stimulates the growth of red blood cells, and "we believe we may have some superior properties to Epogen," said Gordon Binder, chairman and chief executive officer. Human tests are expected to start next year. Epogen, widely used by kidney dialysis patients, was Amgen's first product and remains one of biotech's best-selling drugs.
Amgen said it filed with the Food and Drug Administration April 10 for a new drug application for Infergen, a consensus interferon developed as a treatment against Hepatitis C. Amgen said it will probably sell Infergen in the U.S. and either license or co-promote it in other countries. Mr. Binder said it's unlikely the drug will be approved before next year.
Analysts have been skeptical of the potential for Infergen, whose human test results haven't demonstrated clear superiority over other interferons. Infergen's approval could trigger a patent battle with Schering-Plough Corp., which markets alpha interferon, analysts say. A Schering-Plough spokesman declined to comment on that speculation, but said the company is monitoring Amgen's activities.
Ms. Lonergan, like many analysts, isn't counting on significant Infergen profits in her earnings models. She estimates the potential U.S. market at $100 million. She said the new red blood cell protein under development, if successful, could extend Amgen's Epogen "patent position" and "support their franchise" for a long time.
Appendix IV General Inquirer Word Category Examples
Examples of Words and Word Meanings Included in GI Categories Positive and Negative (entries are shown with assigned tags and sense meanings)
POSITIVE Category includes 1,915 words of positive outlook. For example:
ABILITY H4Lvd Positiv Strong Virtue EVAL Abs@ ABS MeansLw Noun
ABLE H4Lvd Positiv Pstv Strong Virtue EVAL MeansLw Modif adjective: Having necessary power, skill, resources, etc.
ACCENTUATE
H4 Positiv Active Ovrst IAV SUPV
ACCEPT H4Lvd Positiv Pstv Submit Passive SocRel IAV PosAff SUPV verb: To take, receive or accede to something
ACCEPTABLE
H4Lvd Positiv Pstv Virtue EVAL PosAff Modif
ACCEPTANCE H4Lvd Positiv Pstv Affil Passive SocRel PosAff Noun
ACCOMPLISH H4Lvd Positiv Pstv Strong Power Active Complet IAV EndsLw SUPV verb: To bring to its goal or conclusion
ACCOMPLISHMENT
H4Lvd Positiv Pstv Strong Power Active Goal EndsLw Noun -------------------- NEGATIVE Category includes 2,291 words of negative outlook
ADVERSARY H4Lvd Negativ Hostile Polit@ Role SocRel HU PowCon PowTot Noun
ADVERSE
H4Lvd Negativ Ngtv Vice NegAff Modif
50
ADVERSITY
H4 Negativ Vice Noun
ALLEGATION H4Lvd Negativ Ngtv Hostile Legal ComForm COM PowCon PowTot FormLw Noun
ALLEGE
H4 Negativ Weak Undrst ComForm IAV SUPV
AMBIGUITY H4Lvd Negativ Ngtv Undrst Perceiv If Noun
AMBIGUOUS
H4Lvd Negativ Ngtv Undrst Perceiv If Modif
AMISS H4Lvd Negativ Ngtv Vice NegAff LY
ANOMALY H4Lvd Negativ Undrst Rel Noun
Appendix V
Example General Inquirer Output Matrix AMGEN Corporation
Disclosure Text Input File FormatWordcount Leftovers Positive NegativeAMGN 02-02-1996.txt r 635 116 20 21 AMGN 02-02-1996.txt s 635 18.267723.149606 3.307087AMGN 04-18-1996.txt r 423 77 19 8 AMGN 04-18-1996.txt s 423 18.203314.491726 1.891253
Recommended