Earnings Quality

Patricia M. Dechow Catherine Schrand

We thank the Association of Investment Management and Research (AIMR) for funding

Accounting Quality Scandals

Enron special purpose entities to hide debt WorldCom capitalization of expenses Tyco disclosure of loans Qwest International network capacity swaps Xerox long-term leases and cookie jar reserves Waste Management capitalization rules

HIH underestimation of insurance liabilities Harvey Norman: franchisees revenue (Boston Chicken) Parmalat cash balance overstated

What has caused the observed decline in earnings quality?

Poor corporate governance Management compensation packages Decline in auditing quality Movement away from manufacturing towards the creation of intangible assets that are difficult to value Philosophical change in accounting standard setting Lack of scrutiny by financial analysts Growth of institutional investors with no interest in firms business State of the economy

Outline of monograph

Define earnings quality (Chpt 2) Provide evidence from research on financial statement analysis to assess earnings quality (Chpts 3-6) Provide evidence on red flags that suggest potential earnings quality issues:
Incentives for earnings management (Chpt 7) Corporate governance (Chpt 8)

Provide evidence of the role of voluntary disclosures in predicting earnings(Chpt 9)

What is earnings quality?

Earnings quality is contextual depends on user

Standard setters Auditors Compensation committees Debt holders Investors Financial analysts

All would agree fraudulent reporting is low quality We adopt a financial analyst perspective

What is earnings quality?

Objective of the financial analyst

1. 2.

Forecast earnings Make stock recommendations

1.

2.

Earnings are higher quality: The more predictable and easier to forecast; AND The better they map into intrinsic value

1. Earnings predictability

What factors make earnings easier to predict compared to cash flows?

The higher persistence ( close to one)

Et+1 = Et + t

The fewer transitory components The less volatile

On average, earnings is superior to current cash flows on these dimensions

Persistence of earnings vs cash flows

0.9 0.8 Estimated Persistence Parameter 0.7 0.6 0.5 0.41 0.4 0.30 0.3 0.2 0.1 0.0
Sales Operating Operating Pretax Income Earnings Income Income Income After Before Before Before Depreciation Special ItemsExtraordinary Depreciation Items CFO FCF CFF CFI

0.85 0.76 0.76 0.72 0.71 0.71 0.65

0.25

2. Earnings and intrinsic value

Do earnings map into value better than do cash flows?

Measuring intrinsic value

Current stock price Actual future cash flows Residual income model value (using future prices)

Stock prices yes but functional fixation? Future cash flows mixed but its a noisy measure Intrinsic value generally yes

Suggests that on average, earnings map more easily into value than do current cash flows.

What affects earnings predictability across firms?

Stage of firms life cycle (steady state vs growing or declining)

Sources of competition particularly for growing firms Transitory components and losses for declining firms Financial statement analysis of I/S and B/S components

Quality of forecasts and estimates made by management imbedded in accruals (intentional and unintentional estimation error) Accounting standards followed by the firm Managements voluntary disclosures

Stage of life cycle and persistence: Sales growth

0.7 0.6

0.5 Sales Growth Rate 0.4

0.3

0.2

0.1

0 1 -0.1 Year Relative T o Portfolio Formation Year -0.2 2 3 4 5 6

Source: Nissim and Penman (2001)

Stage of life cycle and persistence: RNOA growth

Figure 4.2: Return on Net Operating Asset (RNOA) over time. Deciles formed on the magnitude of RNOA.
Source: Nissim and Pehman (2001). 0.4

0.35

0.3 RNOA 0.25

0.2

0.15

0.1

0.05

0 1 -0.05 Year Relative To Portfolio Formation Year -0.1 2 3 4 5 6

0.45

0.4 Health Services, Hospitals etc. Petroleum Pipelines 0.35 Water Supply 0.3 Profit Margin Pulp Mills Transportation 0.25 Electric Services Gold and Silver Ores 14% Iso-Return on Net Operating Asset Line

0.2 Communications 0.15 Accounting Services Business Services, Photocopying, Art Mobile Homes Hotels, Lodging Mobile Homes Dealers Durable Goods, Wholesale Grocery Stores Wholesale Grocers 0 1 2 3 4 Asset Turnover Industry Median ISO 14% Line 5 6 7 8 9

0.1 Warehouse Storage 0.05 Airport, Terminal Services

Soliman (2003) : This figure plots 248 three- digit SIC industry groups from 1970- 2001. The sample
consists of 77,199 firm-year observations. The points represent median profit margin and median asset turnover by industry. The Iso- RNOA line represents the profit margin and asset turnover combination that would result in an RNOA (return on net operating assets) of 14%.

Persistence is affected by nonoperating components

Model ROEt-1 only Coefficient (Std. Dev) 0.042 0.664 (0.004) (0.023) Full Model Coefficient (Std. Dev) 0.029 (0.005) Intercept ROE Gross margin SG&A Depreciation Interest Minority interest Non-operating income Income taxes Discontinued operations Extraordinary items Adjusted R2 42.3% (0.029)

Operating income components

0.636 0.639 0.629 0.621 0.615 0.548 0.445 0.003 0.153 49.0%

(0.022) (0.022) (0.073) (0.019) (0.078) (0.037) (0.028) (0.028) (0.058) (0.023)

Source: Fairfield, Sweeney, and Yohn (1996)

Persistence is predictable using I/S and B/S

Signal Inventory Accounts Receivable Capital Expenditures Gross Margin Selling and Admin Effective Tax Rate Inventory Method Audit Qualification Labor Force Coefficient Measurement (predicted to be negative) (* for significant) -0.017* Inventory - Sales 0.009* (wrong sign) Accounts Receivable - Sales 0.005 (wrong sign) Industry CAPX - Firm CAPX -0.031* Sales - Gross Margin -0.010 (insignificant) SG&A - Sales (Average ETR for past 3 years -0.594* ETRt) x EPSt 0 for LIFO, 1 for FIFO or other -0.006* 0 for unqualified, 1 for qualified 0.014 (insignificant) -0.026* -1*(% in sales per employee)

Source: Abarbanell and Bushee (1997): regress the future change in earnings on current red flags

Persistence is a function of accruals

Sloan (1996)
Earningst+1 = + Earningst + t+1 , and Earningst+1 = + 1Accrualst + 2Cash from Operationst + t+1. (3.1) (3.2)

Sloan obtains the following result using 40,679 firm- years from 1962 1991: Earningst only Coefficient (t-statistic) 0.015 0.841 (32.57) (303.98) Accruals and cash flows Coefficient (t-statistic) 0.011 (24.05)

Model Intercept Earnings Accruals Cash from operations

0.765 0.855

(186.53) (304.56)

Various decomposition of accruals can provide more insights into earnings persistence

Decomposing accruals into inventory, A/R, A/P, tax payable, etc Decomposing accruals into discretionary vs nondiscretionary using Jones model Decomposing accruals based on extent of estimation error Taxable income versus book income Volatility of cash flows versus magnitude of accruals Nature of the business and operating cycle will also affect the level of accruals

Persistence is a function of accruals

1.1 Estimated Persistence Parameter 1 0.9 0.8 0.7 0.6 0.5 1 2 3 4 5 6 7 8 9 10 Rank of current accruals (difference between earnings and CFO) Persistence of Earnings Persistence of CFO

Source: Dechow and Ge (2003) The sample consists of 98,624 firm-years from 1987 to 2002. Decile 1 consists of firms with the most negative accruals. Decile 10 consists of firms with the most positive accruals. Current accruals are calculated as the difference between earnings and CFO. The persistence coefficients of earnings and operating cash flows (CFO) are estimated from the following regressions: Earningst+1 = at + bEarningst + et CFOt+1 = at + bCFOt + et

Persistence is a function of both current and long-term accruals

Fig1b. Persistence of Earnings and FCF Based on Ranks of Total Accruals 1.1 1 0.9 0.8 0.7
Persisten

Persistence of Earnings Persistence of FCF

0.6 0.5 0.4 0.3 0.2 1 2 3 4 5 6 7

10

Ranks of Total Accruals

Persistence of Earnings

Persistence of FCF

Source: Dechow and Ge (2003)

What causes investors to misunderstand the nature of the quality of accruals?

Manipulation of long-term and short term accruals Estimation error in extreme accruals that reverses Misunderstanding of future growth

Declining returns to scale Over-optimistic growth forecasts by analysts for high accrual firms (particularly for loss firms)

Misunderstanding of temporary components

High accrual are glamour stock , low accrual are value stock so driven partly by market sentiment

Persistence is affected by accounting standards used

What is income?

Income is an indicator of performance about an enterprise and its management; OR Income is an enhancement of wealth or command over economic resources.

Standards moving away from performance evaluation towards balance sheet focus (measuring assets and liabilities)

Income is an enhancement of wealth

Accounting rules focus on valuing assets and liabilities (the balance sheet)

Fair value accounting

Securitization Equity and debt securities available for sale Asset impairments Goodwill rules Deferred taxes

More transitory components Frequency of losses increases Balance sheet approach implies a decline in earnings quality under our definition

More difficult to forecast earnings Harder to map earnings alone into value

Frequency of special items has increased over time

0.025
Special Items/Total Operating Expenses

0.020 0.015 0.010 0.005 0.000

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

Q1 Q2 Q3 Q4

Source: Bradshaw and Sloan (2002)

Adjusted R-square 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% 5%

Source Francis and Schipper 1999

Year

Relation between earnings and returns:

19 52 19 54 19 56 19 58 19 60 19 62 19 64 19 66 19 68 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94

Relation between book value and market value

80% 70% Adjusted R-square 60% 50% 40% 30% 20% 10% 0%

19 53 19 55 19 57 19 59 19 61 19 63 19 65 19 67 19 69 19 71 19 73 19 75 19 77 19 79 19 81 19 83 19 85 19 87 19 89 19 91 19 93
Year

Source Francis and Schipper 1999

How has the market coped with decline in earnings quality? Rise of pro forma earnings
Source: Bradshaw and Sloan (2002)
PANEL A: QUARTER 1
0.020

PANEL B: QUARTER 2
0.020

0.015

0.015

0.010 Street GAAP 0.005

0.010 Street GAAP 0.005

0.000

0.000

-0.005 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

-0.005 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

Panel C: Quarter 3
0.020

Panel D: Quarter 4
0.020

0.015

0.015

0.010 Street GAAP 0.005

0.010 Street GAAP 0.005

0.000

0.000

-0.005 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

-0.005 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

Are proforma higher quality than GAAP?

Proform a adjustm ents reported in 1,149 press releases (source Bhattacharya, Black, Christensen, and Larson (2002)) 25%

20% Percentage

15%

10%

5%

0% Depreciation and amortization Stock compensation costs Merger and acquisition costs R&D costs and w rite-offs on purchased in-process R&D Gains and losses on sales Extraordinary items and discontinued operations Alteration of number of shares outstanding for EPS Other adjustments

Are proforma higher quality than GAAP?

Source: Bradshaw and Sloan (2002)
Panel A: Earnings Response Coefficient
4.0

3.5

3.0
Earnings Response Coefficient

2.5
Street GAAP

2.0

1.5

1.0

0.5

0.0

04

02

04

02

04

02

04

02

04

02

04

02

04

02

04

02

04

02

04

02

04

02 97

86

87

87

89

88

88

89

90

90

91

91

92

92

93

93

94

94

95

95

96

96

Year-Quarter

97

04

Pro forma more likely to beat analyst forecasts

EPS Above versus Below Mean Analyst Forecast 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0%

77.5% 61.3% 38.7% 22.5%

80.1%

19.9%

GAAP

IBES

Pro Forma

EPS at or above Mean Forecast

EPS below Mean Forecast

Are items excluded from proforma really transitory?

80.0% 70.0% 60.0% Hedge Return 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% -10.0% 1988 1990 1992 1994 1996 1998 Total Exclusions Other Exclusions

Source: Doyle, Lundholm and Soliman (2003) Results suggest that items are not as transitory as investors think

Outline of monograph

Define earnings quality (Chpt 2) Provide evidence from research on financial statement analysis to assess earnings quality (Chpts 3-6) Provide evidence on red flags that suggest potential earnings quality issues:
Incentives for earnings management (Chpt 7) Corporate governance (Chpt 8)

Provide evidence of the role of voluntary disclosures (Chpt 9)

A ggregate Earnings Managem en

10

15

20

25

30

Austria G reece Korea (South) Portugal Italy Taiwan Switzerland Singapore G erm any Japan Belgium Hong Kong India Spain Indonesia Thailand Pakistan Netherlands Denm ark M alaysia France Finland Philippines United Kingdom Sweden Norway South Africa Canada Ireland Australia United States

Earnings Managem ent A round the W orld (Leuz, Nanda and W ysocki, 2003)

What accounts are managed most?

80% 70% 60% 50% Percentag 40% 30% 20% 10% 0% Overstate revenue and accounts receivable/Understate bad debt allowance Understate A/R discounts/allowances other than bad debts Overstate inventory/Understate COGS Overstate intangibles and other long-term assets Understate an expense (other than COGS) Misstatement of accrued liabilities Overstate security valuation Overstate PPE

The sample is collected from the Accounting and Auditing Enforcement Releases from AAER 1 to AAER 1745. The releases that involve manipulations of annual financial statements are identified. The final sample consists of 294 separate firms that had manipulated 426 different accounts. Firms can manipulate more than one account, so the percentages add up to more than 100%.

Incentives for earnings management

Extensively studied by academics

Boost stock price

Cheaper to raise new financing, acquire other firms Boost value of stock option grant/insider trade Keeps interest in the stock high (affects liquidity) Reputation of management Earnings based bonuses Debt covenants, unions, etc

Meet analysts forecasts

Contractual incentives

Regulatory incentives loan loss reserve, link to taxes

Corporate governance and the opportunity to engage in earnings management

100%

Percentage of each sample

90% 80% 70% 60%

SEC enforcement sample

50% 40% 30% 20% 10% 0% Company has Company has an independent outside blockholder that owns more than 5 percent of the stock The company has a big six auditor CEO is also Chairman of the Board CEO is also the founder of the company More than 50 percent of the Board are insiders

Control sample

an audit committee

Figure 8.1: Comparison of governance structures for firms that violate GAAP to a control group, Source: Dechow, Sloan, and Sweeney (1996)

Earnings quality and audit firm

Research generally indicates firms report lower discretionary accruals when

The auditor has industry expertise The auditor has performed the audit for a longer time period When a big 4-6-8 auditor is used (client selection may drive this)

But, are not more likely provide qualified opinions on high accrual firms even though more of these firms end up restating earnings and having SEC enforcement actions. Audit changes are a red flag

Have low accruals prior to switch more normal level of accruals afterwards Auditor too conservative or firms want to boost earnings Firms that switch are more likely to have debt covenant violations, higher leverage, declining earnings (poor performance)

Earnings quality and audit firm

Does non-audit services reduce earnings quality?

Public disclosures on fees only recently disclosed in US Generally a little bit is Okay, too much is bad Tax advisory services may not hurt as much as consulting?

- firms are less likely to restate earnings when they provide tax advisory services (Kinney, Palmrose, and Scholz 2003 private source of data) Frankel, Johnson, Nelson (2003): yes higher discretionary accruals, more likely to meet analyst forecasts More on this tomorrow!

Do analysts play a monitoring role?

Not much direct evidence on analysts playing a positive monitoring role in markets Unfortunately evidence suggests
Do not fully adjust forecasts for the fact that high accrual firms are likely to have a decline in EPS Parrot what management tell them Forecasts start optimistic and end pessimistic (so management can beat the forecast)

Outline of monograph

Define earnings quality (Chpt 2) Provide evidence from research on financial statement analysis to assess earnings quality (Chpts 3-6) Provide evidence on red flags that suggest potential earnings quality issues:
Incentives for earnings management (Chpt 7) Corporate governance (Chpt 8)

Provide evidence of the role of voluntary disclosures (Chpt 9)

Can voluntary disclosure improve the predictability of earnings?

35.00% 30.00% Percentage of firm-quarters 25.00% 20.00% 15.00% 10.00% 5.00% 0.00%
-0 .5 1 -0 .4 5 -0 .3 9 -0 .3 3 -0 .2 7 -0 .2 1 -0 .1 5 -0 .0 9 -0 .0 3 0. 03 0. 09 0. 15 0. 21 0. 27 0. 33 0. 39 0. 45 M or e

Deviation from consensus forecast No Guidance Guidance

Source: Hutton (2003). Sample consists of NIRI members (pre REG. FD)

Management guidance tends to be pessimistic (reduce analyst forecast of EPS)

Proportion of Pessimistic Management Forecast
0.6 0.5 Proportion 0.4 0.3 0.2 0.1 0 1993-1997 1998-2001 Year Pre RegFD Post RegFD

Proportion of On Target Management Forecast

0.27 0.26 Proportion 0.25 0.24 0.23 0.22 0.21 0.2 1993-1997 1998-2001 Year Pre RegFD Post RegFD

Proportion of Optimistic Management Forecast

0.5 0.4 Proportion 0.3 0.2 0.1 0 1993-1997 1998-2001 Year Pre RegFD Post RegFD

Source: Cotter, Tuna, and Wysocki (2003) A pessimistic forecast of EPS is below actual reported EPS, an optimistic forecast is above actual reported EPS.

Management forecasts improve analyst forecast accuracy

0.1 0.05

0 1 2 3

Earnings Surprise

-0.05

-0.1

-0.15

-0.2

-0.25

Analysts prior forecasts

Earnings Surprise based on: Managers forecasts

Analysts posterior forecasts

Source: Feng (2003) The graph plots the median of earnings surprises based on analysts prior forecasts, managers forecasts and analysts posterior updated forecasts for each group.

Voluntary disclosure and predictability of earnings

Generally more disclosure is associated with lower costs of capital (advantage of voluntary disclosure) Better MD&A disclosure in annual report is associated with lower forecast errors, less dispersion of errors Across country results indicates that countries with better disclosure have more accurate analysts Firms disclose more when they have financing needs and less when there are higher potential proprietary costs or litigation costs. Firms are more likely to disclose bad news early (before earnings announcement).

Do management act in their self interest when providing guidance?

Evidence that management release bad news before option grants Evidence that managers like to release all the bad news early but leak out the good news so they can have positive earnings surprise

Summary

Provide an analysts perspective on earnings quality Careful financial statement analysis can improve the ability to forecast earnings Quality of earnings depends on the nature of the business as well as the integrity and quality of management Corporate governance and auditors play a role in improving earnings quality Important to understand managements incentives who are they trying to please? What do they want? Are the overly concerned with stock price and meeting forecasts? Recent accounting rules generally have not improved earnings quality from the perspective that we adopt.