The Economic Implications of Corporate Financial Reporting

1
The Economic Implications of Corporate Financial
Reporting
September 28, 2004 University of Chicago

• John R. Graham
• Duke University, Durham, NC USA
• Campbell R. Harvey
• Duke University, Durham, NC USA
• National Bureau of Economic Research, Cambridge,
  MA USA
• Shiva Rajgopal
• University of Washington, Seattle, WA USA

2
Graham/Harvey/Rajgopal Corporate Reporting
Background

• In 1995, Duke and Financial Executives
  International make a deal to conduct a quarterly
  CFO survey
• The deal allows for some special academic
  surveys outside of the quarterly survey that
  would use the FEI e-mail list

3
Graham/Harvey/Rajgopal Corporate Reporting
Background

• 1. Graham and Harvey conduct a survey on capital
  structure and project evaluation
• Theory and Practice of Corporate Finance
  Evidence from the Field appears in JFE 2001
• 2. Brav, Graham, Harvey Michaely survey on
  dividend and repurchase policy
• Payout Policy in the 21st Century forthcoming
  in JFE 2004
• 3. Graham, Harvey and Rajgopal survey on
  corporate financial reporting

4
Graham/Harvey/Rajgopal Corporate Reporting
Methodology

• General goals our research program
• To examine assumptions
• To learn what people say they believe
• To provide a complement to the usual research
  methods archival empirical work and theory

5
Graham/Harvey/Rajgopal Corporate Reporting
Methodology

• Approach contrasts with Friedmans (1953) The
  Methodogy of Positive Economics
• Goals of positive science are predictive
• Dont reject theory based on unrealistic
  assumptions
• Also, rejects notion that all the predictions of
  a theory matter to its validity goal is narrow
  predictive success

6
Graham/Harvey/Rajgopal Corporate Reporting
Methodology

• Alternative view, Daniel Hausman (1992)
• No good way to know what to try when a
  prediction fails or whether to employ a theory in
  a new application without judging its
  assumptions.

7
Graham/Harvey/Rajgopal Corporate Reporting
Narrow goals

• Insight on following issues
• Importance of reported earnings and earnings
  benchmarks
• Are earnings managed? How? Why?
• Real versus accounting earnings management
• Does missing consensus indicate deeper problems?
• Consequences of missing earnings targets
• Importance of earnings paths
• Why make voluntary disclosures?

8
Graham/Harvey/Rajgopal Corporate Reporting
Strengths and limitations

• Strengths
• Surveys enable us to ask decision-makers specific
  qualitative questions about motivations
• Less of a variable specification problem
• Complements large sample analyses
• A unique angle to confront theories with data
• Limitations
• Questions may be misunderstood
• Truthful responses?
• Non-response bias
• Friedman (1953)

9
Graham/Harvey/Rajgopal Corporate Reporting
Comparison to archival empirical work

• Limitations to existing research include
• Earnings management and voluntary disclosure hard
  to measure
• Rank ordering among various motivations difficult
• Variable with least measurement error may
  dominate
• Same r.h.s. variables can proxy for different
  economic motivations (e.g., size)
• Often a narrow focus on one motivation

10
Graham/Harvey/Rajgopal Corporate Reporting
Method

• Survey and Interview Design
• Draft survey instrument refereed by both
  finance and accounting researchers as well as
  experts in survey design
• Interviewed structured to adhere to best
  scientific practices of interviews, e.g. Sudman
  and Bradburn (1983)
• IRB certification for human subject research

11
Graham/Harvey/Rajgopal Corporate Reporting
Sample

• 401 usable survey responses
• response rate of 10.4
• 25 response rate at a practitioner conference
• 8 response rate to Internet survey
• Interview 20 CFOs
• 40-90 minutes in length
• More give and take than in the survey
• Interviewed firms are much larger, more levered
  and more profitable than the average Compustat
  firm.
• Relative to Compustat firms
• Surveyed firms are larger, more levered, greater
  dividend-yield, fewer firms report negative
  earnings
• Similar B/M and positive P/E

12
Graham/Harvey/Rajgopal Corporate Reporting
Sample

• Firm characteristics (self reported)
• Agency
• CEO age, tenure, education
• Inside ownership
• Size
• Revenues
• Number of employees
• Growth opportunities
• P/E
• Growth in earnings

13
Graham/Harvey/Rajgopal Corporate Reporting
Sample

• Firm characteristics (self reported)
• Free cash flow effects
• Profitability
• Leverage
• Informational effects
• Public/private
• Which stock exchange
• Industry
• Credit rating

14
Graham/Harvey/Rajgopal Corporate Reporting
Sample

• Firm characteristics (self reported)
• Financial reporting practices
• Number of analysts
• Do they give guidance?
• Ticker symbol!
• Demographic correlations in Table 1
• Note positive relation between whether you give
  guidance and number of analysts (Lang and
  Lundholm TAR 1996)

15
Corporate Financial Reporting
Performance measurements (earnings, cash flows)
Sec 3.1,Table 2
Voluntary disclosure
Timing Sec 6.3 Table 13
Why disclose? Sec 6.1,Table 11
Earnings benchmarks Sec 3.2, Table 3
Earnings trends
Why not disclose? Sec 6.2, Table 12
Why meet benchmarks? Sec 3.3, Table 4
What if miss benchmarks? Sec 3.4, Table 5
Why smooth earnings? Sec 5.1, Table 8
How to meet benchmarks Sec 4.1, Table 6 Value
sacrifice to meet benchmarks Sec 4.2, Table 7
Value sacrifice for smooth earnings Sec 5.2,
Table 9
Fig. 1 Flowchart depicting the outline of the
paper
16
Corporate Financial Reporting
Performance measurements (earnings, cash flows)
Sec 3.1,Table 2
Voluntary disclosure
Timing Sec 6.3 Table 13
Why disclose? Sec 6.1,Table 11
Earnings benchmarks Sec 3.2, Table 3
Earnings trends
Why not disclose? Sec 6.2, Table 12
Why meet benchmarks? Sec 3.3, Table 4
What if miss benchmarks? Sec 3.4, Table 5
Why smooth earnings? Sec 5.1, Table 8
How to meet benchmarks Sec 4.1, Table 6 Value
sacrifice to meet benchmarks Sec 4.2, Table 7
Value sacrifice for smooth earnings Sec 5.2,
Table 9
Fig. 1 Flowchart depicting the outline of the
paper
17
Graham/Harvey/Rajgopal Corporate Reporting
Motivation
DeGeorge, Patel, Zeckhauser, JB 1999
18
Graham/Harvey/Rajgopal Corporate Reporting
Earnings benchmarks
Responses to the question How important are
following earnings benchmarks? based on a
survey of 401 financial executives.
19
Graham/Harvey/Rajgopal Corporate Reporting
Earnings benchmarks

• Conditional Consensus is relatively more
  important for
• Firms with more analysts
• Firms that give guidance
• Large firms
• More levered firms
• Table 3

20
Corporate Financial Reporting
Performance measurements (earnings, cash flows)
Sec 3.1,Table 2
Voluntary disclosure
Timing Sec 6.3 Table 13
Why disclose? Sec 6.1,Table 11
Earnings benchmarks Sec 3.2, Table 3
Earnings trends
Why not disclose? Sec 6.2, Table 12
Why meet benchmarks? Sec 3.3, Table 4
What if miss benchmarks? Sec 3.4, Table 5
Why smooth earnings? Sec 5.1, Table 8
How to meet benchmarks Sec 4.1, Table 6 Value
sacrifice to meet benchmarks Sec 4.2, Table 7
Value sacrifice for smooth earnings Sec 5.2,
Table 9
Fig. 1 Flowchart depicting the outline of the
paper
21
Graham/Harvey/Rajgopal Corporate Reporting Why
meet earnings benchmarks?
Responses to the statement Meeting earnings
benchmarks helps based on a survey of 401
financial executives.
22
Graham/Harvey/Rajgopal Corporate Reporting Why
meet earnings benchmarks?

• Stock price motivation
• 86 of CFOs say builds credibility
• 80 maintain or increase stock price

23
Graham/Harvey/Rajgopal Corporate Reporting Why
meet earnings benchmarks?

• Stakeholder motivations
• Firms enhance reputation with stakeholders, such
  as customers, suppliers, creditors
• Conditional analysis shows this is important for
  small, tech, inside dominated, young and not
  profitable

24
Graham/Harvey/Rajgopal Corporate Reporting Why
meet earnings benchmarks?

• Employee bonus
• Survey evidence not significant
• Interviews suggest that internal targets more
  important for managers (stretch and budget
  greater than consensus)

25
Graham/Harvey/Rajgopal Corporate Reporting Why
meet earnings benchmarks?

• Career concerns
• External reputation very important
• This motivation was prominent in interviews.
  Executive labor market important. Failure to
  deliver on targets inhibits intra-industry
  mobility.

26
Graham/Harvey/Rajgopal Corporate Reporting
Consequences of missing benchmarks
Responses to the statement Failing to meet
benchmarks based on a survey of 401 financial
executives.
27
Graham/Harvey/Rajgopal Corporate Reporting
Consequences of missing benchmarks

• Uncertainty
• Uncertainty about future prospects is thought to
  be priced

28
Graham/Harvey/Rajgopal Corporate Reporting
Consequences of missing benchmarks

• Cockroach problem
• You have to start with the premise that everyone
  manages earnings
• If you cant come up with a few cents, there must
  be some previously unknown serious problems at
  the firm
• If you see one cockroach, you immediately assume
  there are hundreds behind the walls, even though
  you have no proof that this is the case

29
Graham/Harvey/Rajgopal Corporate Reporting
Consequences of missing benchmarks

• Mitigation of negative reaction
• Explain miss is due to specific accounting
  accrual
• Miss quarterly but confirm annual guidance
• Nonfinancial indicators suggest good future
  performance
• Other factors
• Conference call becomes negative investors
  become defensive

30
Corporate Financial Reporting
Performance measurements (earnings, cash flows)
Sec 3.1,Table 2
Voluntary disclosure
Timing Sec 6.3 Table 13
Why disclose? Sec 6.1,Table 11
Earnings benchmarks Sec 3.2, Table 3
Earnings trends
Why not disclose? Sec 6.2, Table 12
Why meet benchmarks? Sec 3.3, Table 4
What if miss benchmarks? Sec 3.4, Table 5
Why smooth earnings? Sec 5.1, Table 8
How to meet benchmarks Sec 4.1, Table 6 Value
sacrifice to meet benchmarks Sec 4.2, Table 7
Value sacrifice for smooth earnings Sec 5.2,
Table 9
Fig. 1 Flowchart depicting the outline of the
paper
31
Graham/Harvey/Rajgopal Corporate Reporting
Actions taken to meet benchmarks
Near the end of the quarter, it looks like your
company might come in below the desired earnings
target. Within what is permitted by GAAP, which
of the following choices might your company make?
32
Graham/Harvey/Rajgopal Corporate Reporting
Actions taken to meet benchmarks

• Real versus accounting actions
• 80 would reduce discretionary spending, RD,
  maintenance, advertising
• 55.3 would delay starting a new project even if
  it entailed a small sacrifice in value
• Not as much support for accounting actions

33
Graham/Harvey/Rajgopal Corporate Reporting
Actions taken to meet benchmarks

• Real versus accounting actions
• Little research on real actions
• Dechow and Sloan (JAE 1991) Bartov (TAR 1993)
  Bushee (TAR 1998), RD or asset sales
• Roychowdhury (WP 2003) over produce and sales
  discounts to meet targets

34
Graham/Harvey/Rajgopal Corporate Reporting
Actions taken to meet benchmarks

• Real versus accounting actions
• Significantly more likely to say they are taking
  real rather than accounting actions
• In contrast, most of the work on earnings
  management has focused on accruals

35
Graham/Harvey/Rajgopal Corporate Reporting
Actions taken to meet benchmarks

• Why real versus accounting actions?
• Aftermath of Enron-Worldcom along with S-Ox
• Any hint of accounting questions could have
  devastating effect on stock prices
• More willing to admit to real actions
• Auditors cant second guess real actions

36
Graham/Harvey/Rajgopal Corporate Reporting
Sacrificing long-term value
Hypothetical scenario Your companys cost of
capital is 12. Near the end of the quarter, a
new opportunity arises that offers a 16 internal
rate of return and the same risk as the firm. The
analyst consensus EPS estimate is 1.90. What is
the probability that your company will pursue
this project in each of the following scenarios?
Actual EPS if you do not pursue the project Actual EPS if you pursue the project The probability that the project will be pursued in this scenario is (check one box per row) The probability that the project will be pursued in this scenario is (check one box per row) The probability that the project will be pursued in this scenario is (check one box per row) The probability that the project will be pursued in this scenario is (check one box per row) The probability that the project will be pursued in this scenario is (check one box per row) The probability that the project will be pursued in this scenario is (check one box per row)
0 20 40 60 80 100
2.00 1.90
1.90 1.80
1.80 1.70
1.40 1.30
37
Graham/Harvey/Rajgopal Corporate Reporting
Sacrificing long-term value
Probability of accepting project
38
Graham/Harvey/Rajgopal Corporate Reporting
Sacrificing long-term value
Only 45 would take the project for sure even
if they are projected to meet consensus
Table 7
39
Graham/Harvey/Rajgopal Corporate Reporting
Sacrificing long-term value

• Reminiscent of Brav, Graham, Harvey and Michaely
• Sacrifice positive NPV projects before cutting
  dividends

40
Graham/Harvey/Rajgopal Corporate Reporting
Sacrificing long-term value
Repurchases Dividends
41
Graham/Harvey/Rajgopal Corporate Reporting
Other insights on meeting benchmarks

• Interviews
• 18/20 interview mentioned trade off of short-run
  earnings and long-term optimal decisions
• Investment banks offer products that create
  accounting income with negative cash flow
  consequences

42
Graham/Harvey/Rajgopal Corporate Reporting
Other insights on meeting benchmarks

• Guidance
• Goal of guidance is to meet or exceed consensus
  every quarter
• Analysts complicit in game of always meeting or
  exceeding
• Large positive surprises lead to ratchet-up
  effect
• Asymmetric

43
Graham/Harvey/Rajgopal Corporate Reporting
Other insights on meeting benchmarks

• Break out of the game
• Why not declare that you will not play the
  earnings management game?

44
Corporate Financial Reporting
Performance measurements (earnings, cash flows)
Sec 3.1,Table 2
Voluntary disclosure
Timing Sec 6.3 Table 13
Why disclose? Sec 6.1,Table 11
Earnings benchmarks Sec 3.2, Table 3
Earnings trends
Why not disclose? Sec 6.2, Table 12
Why meet benchmarks? Sec 3.3, Table 4
What if miss benchmarks? Sec 3.4, Table 5
Why smooth earnings? Sec 5.1, Table 8
How to meet benchmarks Sec 4.1, Table 6 Value
sacrifice to meet benchmarks Sec 4.2, Table 7
Value sacrifice for smooth earnings Sec 5.2,
Table 9
Fig. 1 Flowchart depicting the outline of the
paper
45
Graham/Harvey/Rajgopal Corporate Reporting
Smoothing
96.9 and 20/20 interviews prefer smooth
earnings over more volatile holding cash flows
constant
46
Graham/Harvey/Rajgopal Corporate Reporting
Smoothing
Responses to the question Do the following
factors contribute to your company preferring a
smooth earnings path?
47
Graham/Harvey/Rajgopal Corporate Reporting
Smoothing

• Reasons
• Lowers risk increased predictability lower
  risk premium
• Clear from survey and interviews that CFOs
  believe that this risk is priced
• Possible link to literature on estimation error,
  disagreement in asset pricing, information risk
  premium, and behavioral literature on risk versus
  uncertainty

48
Graham/Harvey/Rajgopal Corporate Reporting
Sacrificing value for smoothing
Responses to the question How large a sacrifice
in value would your firm make to avoid a bumpy
earnings path?
49
Graham/Harvey/Rajgopal Corporate Reporting
Other insights on smoothing

• Interviews
• Volatile earnings will create trading incentives
  for speculators, hedge funds and legal vultures
• Volatile earnings mean that you will have a
  number of misses which CFOs want to avoid
• Smoothing example

50
Graham/Harvey/Rajgopal Corporate Reporting
Marginal investor
Responses to the statement Rank the two most
important groups in terms of setting the stock
price for your company
51
Graham/Harvey/Rajgopal Corporate Reporting
Marginal investor

• Price setters
• Institutional investors
• Analysts have important short-term impact
• Retail investors important because they are
  potential customers and are less likely to flip
  stock

52
Graham/Harvey/Rajgopal Corporate Reporting
Marginal investor

• Critique of analysts, institutions
• Young, do not have sense of history
• Contagion bandwagon effect important given
  relative performance measurement
• Quantitative hedge funds issue sell signal if you
  miss irrespective of fundamental information
• CFOs believe idiosyncratic risk is priced

53
Corporate Financial Reporting
Performance measurements (earnings, cash flows)
Sec 3.1,Table 2
Voluntary disclosure
Timing Sec 6.3 Table 13
Why disclose? Sec 6.1,Table 11
Earnings benchmarks Sec 3.2, Table 3
Earnings trends
Why not disclose? Sec 6.2, Table 12
Why meet benchmarks? Sec 3.3, Table 4
What if miss benchmarks? Sec 3.4, Table 5
Why smooth earnings? Sec 5.1, Table 8
How to meet benchmarks Sec 4.1, Table 6 Value
sacrifice to meet benchmarks Sec 4.2, Table 7
Value sacrifice for smooth earnings Sec 5.2,
Table 9
Fig. 1 Flowchart depicting the outline of the
paper
54
Graham/Harvey/Rajgopal Corporate Reporting
Voluntary disclosure

• Types
• Conference calls, meetings, press releases, and
  disclosure of more than mandated information in
  regulatory filings
• Healy and Palepu (2001) say that motivations for
  voluntary disclosure important unresolved
  question for future research

55
Graham/Harvey/Rajgopal Corporate Reporting
Voluntary disclosure

• Drivers
• Information asymmetry
• Increased analyst coverage
• Corporate control contest
• Stock compensation
• Management talent
• Limitations of mandatory disclosure

56
Graham/Harvey/Rajgopal Corporate Reporting
Voluntary disclosure

• Contraints
• Litigation risk
• Proprietary costs
• Political costs
• Agency costs
• Setting a precedent that may be hard to maintain

57
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure
Survey responses to the question Do these
statements describe your company's motives for
voluntarily communicating financial information?
58
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Information asymmetry Information risk
• Diamond Verrecchia (1991) voluntary disclosure
  reduces asymmetry between informed and
  uninformed, increases liquidity.
• 81.9 agree only 4.3 disagree
• Related 56.2 agree that predictability of
  companys future prospects is enhanced

59
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Information asymmetry Information risk
• Interviews distinguish between information risk
  and inherent risk
• Believe that both command a risk premium
• Releasing bad news quickly can be beneficial in
  reducing information risk

60
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Information asymmetry Reputation
• 92.1 agree with reputational benefit for
  transparent reporting (scores the highest)
• Interviews
• Correct investors misperceptions
• Create an environment of trust so strategic
  actions more easily taken in the future
• Trust may be important in gaining access to
  future capital

61
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Information asymmetry Cost of capital
• While only 39.3 point to cost of capital, the
  information risk is linked to cost of capital
• P/E lift 42 might be similar to the cost of
  capital
• Interviews
• A number mentioned reducing analysts
  disagreement and linked that to cost of capital

62
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Information asymmetry Liquidity
• Motivation especially for small firms

63
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Increased analyst coverage
• Bhushan (1989a,b) and Lang and Lundholm (1996)
• 50.8 agree
• More agreement with small and insider dominated
  firms

64
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Stock price motivation
• 48.4 use disclosure to try to correct
  undervalued stock

65
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Stock compensation
• Managers want to reduce contracting costs with
  employees where there is information asymmetry,
  otherwise employees will demand a risk premium
• No support, half disagree

66
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Management talent signaling
• Trueman (1986)
• More support for small firms plus other questions
  suggest that this is important

67
Graham/Harvey/Rajgopal Corporate Reporting
Motivations for voluntary disclosure

• Limitations of mandatory disclosures (new)
• 72.1 say that voluntary corrects gaps in
  mandatory
• Interviews
• Some mandatory confuse rather than enlighten
• Some of our own footnotes related to off-balance
  sheet items and securitizations are so complex,
  even I dont understand them.
• Quarterly mandatory disclosures lack timeliness
• Mandatory ignores intangibles

68
Graham/Harvey/Rajgopal Corporate Reporting
Constraints on voluntary disclosure
Survey responses to the question Limiting voluntary communication of financial information helps
69
Graham/Harvey/Rajgopal Corporate Reporting
Constraints on voluntary disclosure

• Precedent (new)
• The most popular response with 69.6 agreeing
• Most important for insider dominated firms
• Start a practice that you might want to abandon
  later

70
Graham/Harvey/Rajgopal Corporate Reporting
Constraints on voluntary disclosure

• Litigation costs
• Threat of litigation makes managers disclose bad
  news quickly
• 46.4 agree especially important for young and
  tech

71
Graham/Harvey/Rajgopal Corporate Reporting
Constraints on voluntary disclosure

• Proprietary costs
• Might jeopardize firms competitive position
• 58.8 agree
• More agreement with small firms and those with
  few analysts

72
Graham/Harvey/Rajgopal Corporate Reporting
Constraints on voluntary disclosure

• Agency costs
• We know that career concerns and external
  reputation important for meeting benchmarks
• Information may be limited to reduce the chance
  of undue focus by stakeholders
• Not much support for this agency cost angle

73
Graham/Harvey/Rajgopal Corporate Reporting
Constraints on voluntary disclosure

• Political costs
• Disclosure may be limited to avoid unwanted
  attention of regulators
• No support on average but this question, in
  particular, is difficult to interpret

74
Graham/Harvey/Rajgopal Corporate Reporting
Good news versus bad news
Survey responses to the question Based on your
company's experience, is good news or bad news
released to the public faster?
75
Graham/Harvey/Rajgopal Corporate Reporting
Good news versus bad news
Survey responses to the question Do the
following statements describe your company's
motives related to the timing of voluntary
disclosures?
76
Graham/Harvey/Rajgopal Corporate Reporting
Conclusions

• Consensus earnings factors into decisions
• Strong desire to meet benchmarks cockroach
  problem
• It is routine to sacrifice long-term value to
  meet these benchmarks
• Meeting benchmarks is important both for the
  firms stock price and managers reputation and
  mobility
• Agents optimizing over short-term horizon

77
Graham/Harvey/Rajgopal Corporate Reporting
Conclusions

• Having predictable smooth earnings is thought to
  both reduce the cost of capital and enhance
  manager reputation
• Voluntary disclosure is an important tool in
  managers arsenal
• Disclosure can potentially reduce information
  risk and enhance a managers reputation

78
Graham/Harvey/Rajgopal Corporate Reporting
Future research

• Another survey??
• We are thinking of administering the identical
  survey before it is published to non-management
  members of Boards of Directors.
• Also
• Detection of Financial Earnings Management
• Detection of Real Earnings Management
• We have the tickers for 107 firms many of which
  admit to both financial and real earnings
  management