Market Reaction to E-Commerce Impairments Evidenced by Website Outages Joseph H. Anthony* Wooseok Choi** Severin Grabski* *Department of Accounting and Information Systems Michigan State University **Department of Accounting, California State University at Los Angeles
41
Embed
Market Reaction to E-Commerce Impairments Evidenced by Website Outages Joseph H. Anthony* Wooseok Choi** Severin Grabski* *Department of Accounting and.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Market Reaction to E-Commerce ImpairmentsEvidenced by Website Outages Joseph H. Anthony*Wooseok Choi**Severin Grabski*
*Department of Accounting and Information SystemsMichigan State University **Department of Accounting, California State University at Los Angeles
Presentation
Introduction & Research Question Research Approach Prior Research Literature Hypotheses Regression Models Results
Last summer, on-line auction site eBay Inc. unwittingly became the latest poster child for Web-site crashes, as it endured a host of outages, the worst of which took the site offline for nearly 22 hours on June 10. Bidders and sellers were angry, and investors sent the company’s stock down more than 25% in the two business days after the problems began, slashing nearly $6 billion off its market value.
Wall Street Journal: November 22, 1999
Systematically investigate the impact of website and other e-commerce related outages on economic returns as measured by the stock market
“Self-Inflicted”, not
“Hacked”
Research Objective
Direct Measures of Loss Due to Website/ e-commerce Outages Repeated outages resulted in loss of 10% of
customer base (McKnight 1997) Hour of web downtime results in $50,000 in
lost sales (Woods 2000)
Unfortunately, data is generally not available
Alternative Costs of Website/e-commerce Outages TD Waterhouse fined by SEC (Simon 2001) TicketMaster - Prioritized business units
Ticketing Online Personals Cityguide
Lost revenues from ticketing is real Might result in permanent loss of customer
(Fonseca 2001)
Again, data is generally not available
Alternative Costs of Website/e-commerce Outages
CIOs “overspent” on security (Yager 2002) Spend average of $3.6M on Security Average cost of security breach - $193,000
Might be missing other costs, the potential decline in the market value of the firm
Other Costs of Website/e-commerce Outages
Hacker Attacks (Ettredge and Richardson 2002) Resulted in negative abnormal stock returns
BUT---
Firms examined were only in the same industry as “hacked” firms, they were not hacked!!
Other Costs of Website/e-commerce Outages
Security Breaches (Campbell et al. 2003) Resulted in negative abnormal stock returns Market discriminated between types of attacks
Significant negative reaction to unauthorized access to confidential data
No significant reaction when not involving confidential data
Other Costs of Website/e-commerce Outages
Software Vulnerabilities – Cost to software developers (Telang and Wattal 2005) 18 firms, 146 announcements (1999-2004) Resulted in negative returns of .6% stock price per
disclosure Average loss $.86B per vulnerability announcement
More negative impact w/o patch (.8%) More severe flaws have more negative impact Confidentiality breach resulted in greater
decline than other breach types (.75%)
Event Studies Investors process information about expected and
unexpected events and consider these events in the valuation of shares
Event studies examine the residual price change of a sample of firms for a window of time on either side of an identifiable “event,” such as announcements of earnings, stock splits and dividends, cash dividends, earnings forecasts, or changes in accounting methods. The influence of economy-wide and, if necessary, additional factors such as industry-wide information on stock prices is extracted to obtain a residual return. The expected value of the residual price changes, not conditioning on the event, is zero. (Beaver 1998, p.133)
Prior Accounting & Finance Research
Beaver (1968) & Ball and Brown (1968) – stock returns and accounting earnings
Aharony and Swary (1980) – quarterly dividend announcements
Eccher et. al. (1996) & Barth et. al. (1996) – fair value disclosures by banks
Anthony (1987) – expected and unexpected news releases
Prior IT Research Dos Santos et al. (1983) - innovative uses of IT
rewarded Subramani and Walden (1999) - E-commerce
initiatives Krishnan and Sriram (2000) - estimates of Y2K
compliance costs Chatterjee et al. (2001) - announcement of CIO
position creation Im et al. (2001) – IT investment announcements
E-Commerce Firm Valuation Measures of website usage are value
relevant, they provide incremental explanatory power for stock prices (Trueman et al. 2000) Number of unique visitors Number of page views
Bartove et al. (2002) and Rajgopal et al. (2003), also provide research results related to the valuation of internet/e-commerce firms.
Hypotheses
H1: There will be a significant negative association between an outage announcement and a firm’s stock returns.
H2: Firms with a high percentage of on-line revenue will have a significantly more negative association in the stock returns than firms with a low percentage of on-line revenue.
Internet Firm Types (Barua et al. 1999, 2000) Infrastructure Providers
Provide the backbone and basic Internet services
Commerce Providers Provide goods and services to businesses (B2B)
and individuals (B2C) over the Internet (either as an intermediary or directly)
Outage Type
Two basic types of outages reported by the news services E-mail
The e-mail function of a website fails, but the website itself works well without shutting down
Non E-mail outage (Website) The website is completely shut down or some important
functions other than an e-mail failure (e.g., stock trading functions)
Website Outage “The suit, filed in the Santa Clara County Superior Court by
the Alexander law firm of San Jose, Calif., is seeking unspecified damages for investors who claim they missed out on making money in the stock market because of the outage.
"E*Trade customers were unable to trade or obtain access to their online accounts on Feb. 3, 1999 for in excess of one hour, on Feb. 4, 1999, for in excess of two and one half hours, and for approximately one-half hour on Feb. 5, 1999," the Alexander suit said. "As a result of this 'virtual' lockout, class members lost potentially millions of dollars in damages." One E*Trade customer, Dar Hay of Memphis, Tenn., said he lost close to $12,000 last week when he was unable to cancel an order to buy 350 shares of brokerage firm Siebert Financial Corp.” (Reerink 1999b).
Hypotheses
H3a: The negative stock market impact of an e-mail type outage will be greater than that of a non e-mail type outage (website) for infrastructure providers
H3b: The negative stock market impact of a non e-mail type outage (website) will be greater than that of an e-mail type outage for commerce providers
Hypotheses
H4: Long outages (12+ hours) are more negatively associated with stock returns than short outages (1 hour or less)
H5: The frequency of outages is negatively associated with stock price changes.
“Traditional” Event Study
Event date = first day reported by press Used a 4-day window
-1 to +2 -1 because the outage could have occurred past trading
but was picked up prior to the opening of trade the next day
+2 since some outages were longer than 24 hours Similar results obtained for 2 and 3-day windows
Sample Selection Started with firms in “Internet 500” (ZDNet
Interactive Week Special Report 1999) Had at least one outage Eliminated firms not “primarily internet”
Internet revenues > 50% total revenues for 1997, 1998 or 1999 (retained 4 infrastructure firms – ATT, IBM, MCI, Sprint)
Stock return data available for 240 day estimation period
19 firms, 86 outages
Sample Firms (Internet Classification)
Amazon.com (C) Excite, Inc.* (I) America Online (I) IBM (I)