Event studies Irene Karamanou, Ph'D

1
Event studiesIrene Karamanou, Ph.D

• Event studies
• Measuring the market reaction to an event.
• Explaining the cross-sectional variation in
  market reaction
• Why event studies?
• Easy to identify
• Standard methodology
• Strong motivations

2
Event studiesIrene Karamanou, Ph.D

• Two types of event studies
• Event dates are contemporaneous e.g.,
  Regulation effects (policy implications)
• Event dates differ cross-sectionally e.g.,
  Earnings announcements, restatement
  announcements, changes in analyst
  recommendations, voluntary adoptions

3
Event studiesContemporaneous events H1

• Why would the market move in response to a new
  regulation? Benefits vs. costs
• H1a Positive (negative) market reaction on the
  events that increase (decrease) the probability
  of adoption
• H1b Positive overall market reaction to the
  events that led to the adoption of the law
• Easy to test
• Issue of measuring the abnormal return
• Event windows
• Note Measuring expectations of the effect

4
Contemporaneous events- H1Armstrong, Barth,
Jagolinzer, Riedl 2007
5
Event studiesContemporaneous events H1

• Expect a positive Overall market reaction to all
  events that increased the probability of adoption
• Event window used 3 days
• Method Value weighted market-adjusted returns
• CARi, t Rp Rm where i event , t -1, 0,
  1
• Portfolio returns based on market value
• Market DJ Global Index
• Results 16 event CAR is positive and significant
• Investors expect net benefits to IFRS adoption
  in Europe
• Question Foreign vs. local investor reaction?

6
Event studiesContemporaneous events H1

• Li, Pincus and Rego, 2006
• Market reaction to SOX and earnings management
• Benefits Investor Protection by improving the
  quality, accuracy and reliability of corporate
  disclosures
• Costs of implementation
• Marosi and Massoud foreign firms exiting US,
  US firmsderegistering from SEC
• Piotroski and Srinivasan foreign listings
  moving to LSE

7
Event studiesContemporaneous events H1

• Li, Pincus and Rego, 2006
• Eight events relating to SOX passage
• H1 Note that because the Act affects all
  publicly traded firms, we cannot control for
  market-wide stock price effects in computing
  stock returns, doing so could control for the
  SOX-related effects we seek to estimate.
  However, this means that contemporaneous
  market-level events unrelated to SOX could
  confound our analysis (p.7)
• Market Return a SßiD e
• N252 (2002 returns)
• All event average and important specific events
  are significant in the right direction

8
Event studiesContemporaneous events- H1

• Regulations do not have to affect all firms
• Overall effect easier to document
• Schipper and Thompson (1983, JAR) The impact of
  merger-related regulations on the shareholders of
  acquiring firms
• 1969 reform act imposes tax penalties on
  financing acquisitions with debt and increases
  the capital gains tax from 25-30 ? decreasing
  after tax returns to sellers in acquisitions
• Hypothesis Negative reaction around the events
  that led to adoption of reform act for firms with
  on-going acquisition programs.
• Results consistent with hypothesis.

9
Contemporaneous events- H1Measuring abnormal
returns on event dates

• Standard methodology
• Rit a b Rmt
• K9, N106 firms Tdaily returns 1984-1989
• Econometrically weak
• 106 firm specific regressions of above model
• 106 market model regressions pre-event
• Ri a ßRm ?
• R

10
Event studiesContemporaneous events-H1

• Preferred methodologySchipper and Thompson
  (1983)
• Rp is the return of portfolio p on day t (p1,2)
• N number of days in the period
• Accounts for cross-sectional heteroskedasticity
  and contemporaneous correlation
• Disadv Restricts coefficients

11
Event studiesContemporaneous events H2

• Hypothesis 2 The market reaction differs
  cross-sectionally based on firm characteristics
• Armstrong, Barth, Jagonlinzer and Riedl, 2007
• CARj,e ß0 ß1 ADRj,e ß2 Turnoverj,e ß3
  Closely_Heldj,e ß4 Herfj,e ß5 Codej,e ß6
  Big4j,e ß7 Bankj,e ?j,e
• CARj,e market adjusted (market is global index)
  3-day return for firm j and event e.

12
Event studiesContemporaneous events H2

• Armstrong, Barth, Jagonlinzer and Riedl, 2007
• Expectations
• ADR (kind of ADR?)
• Info asymmetry proxies
• Turnover (shares traded/outstanding shares) 1 if
  greater than median and 0 otherwise
• Closely held shares
• Herfindahl index (industry concentration the
  closer to 1)
• Code Law They expect a negative ?
• Big 4 They expect a positive coefficient ?

13
Event studiesContemporaneous events H2

• Li, Pincus and Rego, 2006 SOX
• H2 Firms with previous earnings management will
  earn more positive (negative) abnormal returns on
  the events that increased (decreased) probability
  of adoption
• CAR EM, AUD Dependence, Non audit fees, size,
  BM, beta, industry
• Each event separately. Compute t-test for the
  mean (exclude negative events)
• Results consistent with expectations

14
Event studiesContemporaneous events H2

• The economic consequences of Regulation Fair
  Disclosure Irani and Karamanou
• Reg FD -- Passed to combat Selective disclosure
• Abercrombie Fitch warned a WS analyst about
  disappointing earnings before making info. public
• With the opening of the trading session on NYSE
  the stock of Electronic Data Systems Corp.
  plummeted Some analysts admitted having the bad
  news the night before the public release.
• GM up 3.2 and Lehman Brothers up 6.8 after
  company officials informed certain analysts of
  positive news

15
Event studiesContemporaneous events H2

• Whenever an issuer or a person acting on its
  behalf discloses material nonpublic information
  to certain enumerated persons the issuer must
  make public disclosure of the information either
  simultaneously if the act was intentional or
  promptly if it was not intentional.
• Enumerated persons sell-side analysts, buy-side
  analysts, large institutional investment
  managers.

16
Event studiesContemporaneous events H2

• Expected benefits
• Effects similar to insider trading ? less liquid
  markets
• Fairer disclosure increased investor confidence
  and liquidity reduced cost of capital.
• Reduce bias in analyst forecasts
• Expected costs
• Chilling of information
• The playing field will be more level, but it
  will be empty

17
Event studiesContemporaneous events H2

• To examine the anticipated effects of Reg. FD
  captured by the market reaction to the events
  that led to its adoption.
• Motivation debate and policy implications
• Assumption Market reaction a good indication of
  markets expected long-term effects.
• But markets initial assessment may not reflect
  actual effects
• Timely first indication -- actual long-term
  effects need for longer time-series data

18
Event studiesContemporaneous events H2

• H1 The greater the net benefits from elimination
  of s.d. the more positive is the market reaction
  to events leading to the adoption of FD
• Benefits are lower for firms with rich
  information environments
• SIZE (-), ANALYST FOLLOWING (-), DISPERSION ()
• Benefits are greater if greater propensity of
  selectively disclose information
• Following (), Dispersion (), Institutional
  ownership ()

19
Event studiesContemporaneous events H2
20
Event studiesContemporaneous events H2
21
Event studiesContemporaneous events H2

• Method
• Standard methodology
• Obtain CAR on events (e.g. market model as
  before)
• Regress CAR on set of variables separately for
  each event
• Issue with contemporaneous events
  cross-sectional heteroskedasticity and
  cross-correlation

22
Event studiesContemporaneous events H2

• Sefcik and Thompson (1986 JAR)
• Construct portfolio returns for each day in the
  period using as weights each one of the
  explanatory variables
• F 1, SIZE, FOLL, DISP, IO3
• Run p5 equations. NNumber of days
  (1/1/99-31/12/00)
• 3 event dates
• Accounts fully for heteroskedasticity and
  cross-correlation

23
Event studiesContemporaneous events H2
24
Event StudiesFirm specific event dates

• Not all events happen on the same day
• Earnings announcements
• Cross-listing announcements
• IFRS voluntary adoption announcements
• Restatements
• Repurchases (share buy backs)
• Mergers and Acquisitions
• Analyst recommendation changes

25
Event StudiesFirm specific event dates

• H1 The event results in positive (or negative)
  abnormal returns
• (EA, Restatements, MAs, Recommendation
  downgrades)
• (short vs. long window studies)
• Note that cross correlation not much of concern
  in this setting. Why?
• Miller, 1999 (JFE)
• Examines the announcement of the intention to
  cross list in the US
• 181 announcements, 35 countries
• Period 1985-1995

26
Event StudiesFirm specific event dates H1

• Miller 1999
• Market model adjusted returns -125 to -26
• mean abnormal returns on each of the three days
  around the event (across sample firms)
• Positive and significant return on -1,0
• Tests frequency of the observations with positive
  returns (significant for day -1)
• Segmentation argument
• Other?

27
Event StudiesFirm specific event dates H2

• H2 The market reaction to the event is related
  to firm characteristics
• Ball and Brown (1968, JAR)
• Market returns and accounting earnings on
  announcement
• Returns are associated with unexpected earnings
  with a Drift
• UE Random walk
• Analyst forecasts (mean vs. median, period,
• deflators)

28
Event StudiesFirm specific event dates H2

• Miller 1999
• 3-day firm specific CAR
• Regressed on OTC, NYSE, Emerging economy dummies
• T-stats adjusted for heteroskedasticity
• Results
• More positive reaction if exchange listed (vs.
  OTC. vs. 144A), and if emerging economy
• Alternative explanations?

29
Event StudiesFirm specific event dates H2

• The Valuation Effects of Firm Voluntary Adoption
  of International Accounting Standards
• Irene Karamanou and George Nishiotis
• RQ What are the valuation effects of the
  corporate decision of a firm to voluntarily adopt
  IAS?
• We investigate the price reaction at the
  announcement of the IAS adoption and conduct a
  cross sectional analysis of the abnormal returns.
  
• Constant regulatory environment

30
Event StudiesFirm specific event dates

• Increased disclosure, reduces cost of capital,
  increases firm value
• IAS is superior to local GAAP
• (Ashbaugh and Pincus, 2001 Barth et al., 2005
  Ding, Hope, Jeanjean and Stolowy, 2005)
• H1 There are positive abnormal returns around
  the announcement of voluntary IAS adoption.

31
Event StudiesFirm specific event dates

• Sample construction
• 564 firms from Global Vantage using IAS for at
  least 1 year in the period 1989-1999
• Kept the ones that switched from local GAAP to
  IAS voluntarily Additional
• 53 firms from IASB list of fiscal years1999/2000.
• 171 potential firms
• Searched for announcement dates (Lexis-Nexis,
  Factiva, email).
• Sufficient time series of daily returns for the
  underlying stocks and their national indices.
• 54 firms

32
Event StudiesFirm specific event dates
33
Event StudiesFirm specific event dates

• Method
• Market model per firm (-150 -26)
• Daily returns denoted in local currency,
  Datastream
• Market index fromDatastream
• Abnormal returns are averaged across firms ?
  Abnormal return on each day
• If multiple announcements the earliest clear
  announcement is used

34
Event StudiesFirm specific event dates
35
Event StudiesFirm specific event dates

• Explaining firm abnormal returns
• H2 the greater a firms undervaluation prior to
  the IAS adoption announcement, the higher the
  abnormal return. signaling and bonding
• Tobins q ( or )
• H3 the higher the growth opportunities prior to
  the IAS adoption announcement, the higher the
  abnormal return. bonding effects
• Sales growth ( )
• H4 the poorer the information environment prior
  to the IAS adoption announcement, the higher the
  abnormal return. investor protection pre
  adoption,
• Foll, (-) DIAS () AntiDir ( -) Effic (
  )

36
Event StudiesFirm specific event dates
37
Event StudiesFirm specific event dates
38
Event StudiesFirm specific event dates

• Espahbodi, H., E. Strock, H. Tehranian. 1991.
  Impact on equity prices of pronouncements related
  to nonpension postretirement benefits. Journal of
  Accounting and Economics 14 323-346
• Issues
• Confounding events
• Low R2s
• Returns should be computed manually Adjust for
  splits and stock dividends
• Market adjusted returns

39
Event studies

• Armstrong, C., M. E. Barth, A. D. Jagolinzer, E.
  J. Riedl. 2007. Market reaction to the adoption
  of IFRS in Europe. Working paper, Harvard
  Business School
• Ball R. and P. Brown. 1968. An empirical
  evaluation of accounting income numbers. Journal
  of Accounting Research 158-178
• Miller, D. P. 1999. The market reaction to
  international cross-listings evidence from
  Depositary Receipts. Journal of Financial
  Economics 51 103-123
• Li, H., M. Pincus, S. O. Rego. 2006. Market
  reaction to events surrounding the Sarbanes-Oxley
  Act of 2002 and earnings management. Working
  paper, University of California, Irvine.
• Schipper, K. and R. Thompson. 1983. The impact of
  merger-related regulations on the shareholders of
  acquiring firms Journal of Accounting Research
  21184-221
• Sefcik, S. E. and R. Thompson. 1986. An approach
  to statistical inference in cross-sectional
  models with security abnormal returns as
  dependent variable. Journal of Accounting
  Research 24 316-334