Corporate Disclosures and Financial Intermediaries. Nino Papiashvili Institute of Finance Ulm University

Transcription

1 Corporate Disclosures and Financial Intermediaries Nino Papiashvili Institute of Finance Ulm University 1

2 Introduction Managers have superior information on their firms expected future performance Financial reporting and disclosure are potentially important means for management to communicate firm performance to outside investors A flow of information that is accurate, efficient and fair contributes to a well-functioning capital market that satisfies the needs of all users and enhances economic growth and stability Firms provide disclosure through regulated financial reports, including the financial statements, footnotes, management discussion and analysis and other regulatory filings In addition, some firms engage in voluntary communication, such as management forecasts, analysts presentations and conference calls, press releases, internet sites and other corporate reports 2

3 Introduction Investors face an incomplete information problems when they make investments in business ventures Managers typically have better information than savers about the value of business investment opportunities => information asymmetry Once investors have invested, managers have incentives to expropriate their savings, creating an agency problem Possible solutions to information asymmetry and agency conflicts: Optimal contracts between entrepreneurs and investors that provide incentives for full disclosure of private information to mitigate the misvaluation problem Information intermediaries who engage in private information production to uncover managers superior information or misuse of firm resources The credibility of management disclosures is enhanced by regulators, standard setters, auditors, financial analysts and other capital market intermediaries 3

4 Benefits of Disclosure (a) Improved stock liquidity disclosure reduces information asymmetries among managers, informed and uninformed investors => investors can be relatively confident that any stock transactions occur at a fair price, increasing liquidity (b) Reduced cost of capital when disclosure is imperfect, investors bear risks in forecasting the future payoffs from their investment. If this risk is non-diversifiable, investors will demand an incremental return for bearing the information risk ( c) Increased information intermediation Expanded disclosure enables financial analysts to create valuable new information, such as superior forecasts and buy/sell recommendations 4

5 Motivation for Disclosure Forces that affect managers disclosure decisions: a) The cost of financing Managers who anticipate making capital market transactions (e.g. issue public debt or equity or to acquire another company) have incentives to provide voluntary disclosure to reduce the information asymmetry problem, thereby reducing the firm s cost of external financing b) Market for corporate control Given the risk of hostile takeover and job loss accompanying poor stock and earnings performance, managers use corporate disclosures to reduce the likelihood of undervaluation c) Equity based compensation Managers interested in trading their stock holdings have to meet restrictions imposed by insider trading rules and disclose insider information Prior to the expiration of their stock option awards managers want to correct any perceived undervaluation problem If stock prices are not a precise estimate of firm values equity based compensation is less efficient form of remuneration as managers will demand additional compensation to reward them for bearing any risk associated with misvaluation => Firms that use stock compensation extensively are therefore likely to provide additional disclosure to reduce the risk of misvaluation 5

6 Motivation for Disclosure d) The threat of shareholder litigation Litigants and courts rationally focus on whether there were delays in bad news announcements => less incentives to time the disclosure of good and bad news e) Signalling management talent Firm s market value is a function of investors perceptions of its managers ability to anticipate and respond to future changes in the firm s economic environment f) Threats of transparency Firms might have an incentive not to disclose information that will reduce their competitive position 6

7 The Role of Financial Intermediaries in the Disclosure Process Auditors provide investors with independent assurance that the firm s financial statements conform to GAAP Research shows that capital providers require firms to hire an independent auditor as a condition of financing => regard auditors as enhancing credibility that adds value for investors Financial analysts collect information from public and private sources, evaluate the current performance of firms that they follow, make forecasts about their future prospects, and recommend that investors buy, hold or sell the stock They add value in the capital market => their earnings forecasts are more accurate than time-series models of earnings (incorporate more timely firm and economy news); Evidence shows that analysts earnings forecasts and recommendations affect stock prices Bond-rating agencies provide upgrades and downgrades that generate new information to investors and affect stock market prices 7

8 The Role of Financial Intermediaries in the Disclosure Process Financial analysts generate valuable new information through their earnings forecasts and stock recommendations However, there are systematic biases in financial analysts outputs, potentially arising from the conflicting incentives that they face: analysts motivated by investment banking considerations and trading profits, generate overoptimistic recommendations that cause prices to be higher than a firm s fundamental value Two main reasons why analysts bias their reports: To cultivate access to management => important source of private information about the firm Generate investment banking business for their brokerages: (i) is a good way to increase annual paychecks (ii) by pleasing management with optimistic reports an analyst increases the probability that her brokerage will be chosen for the subsequent equity offering (SEO), merger or debt issue Both arguments suggest that affiliated analysts research reports are more favourable than those issued by unaffiliated analysts Calls for regulation from SEC => favourable forecasts and recommendations from the investment bank s analysts prior to an offering is precluded 8

9 Role of Information Intermediaries in Corporate Governance Do analysts serve as external monitors to managers or do they put excessive pressure on managers? With training in finance and substantial industry background knowledge, analysts track corporate financial statements on a regular basis. They usually interact directly with management and raise questions on different aspects of earnings numbers through earnings release conferences One of the primary earnings targets that managers try to achieve is to meet analysts forecast consensus. Firms that miss analyst forecasts usually suffer significant declines in their stock price Empirical Question: Do firms followed by more analysts manage their earnings more or less? 9

10 Analyst Coverage and Earnings Management Analysts possess several distinctive characteristics that could make them effective monitors against earnings management: Analysts are expected to provide information in the interest of not only current shareholders but also prospective investors as well as other participants in the market, unlike internal governance devices that are designed to protect current shareholders interests Analysts usually have financial training in finance and accounting as well as background in the industries they cover; they also have the resources to go through tedious financial statements and complicated footnotes Analysts track firms on a regular basis, continuously scrutinizing management behavior and financial reporting irregularities this level of oversight is not usually afforded to other gatekeepers If analysts play the role of external monitor, a firm s earnings management behavior should decrease as the number of analysts following the firm increases 10

11 Data and Variables The sample consists of 33,127 firm-year observations from 1988 to 2002 Estimation of earnings management - discretionary accruals (DA) Earnings have two major components: cash flow and accruals (accounting adjustments) Accrual Basis of Accounting: (i) Revenues are recognized when earned without any regard to the timing of cash collection (=>change in Unearned Revenue); (ii) Expenses are recognized when incurred (=> change in accrued Expense) Total Accruals = Net Income minus Cash Flow from Operations Accruals requires managers judgment and estimation => more vulnerable to manipulation Some accrual adjustments are necessary and appropriate and they need to be applied on a regular basis non-discretionary accruals (NDA) 11

12 Data and Variables 1. Estimate total accruals from cross-sectional regressions of total accruals on changes in revenues and on property, plant, and equipment (PPE) 2. Derive Non-Discretional Accruals (that can be explained by the previously estimated coefficients ) 3. Discretionary Accruals (DA) = Total Accruals (TA) Non-Discretionary Accruals (NDA) Positive DAs suggest income-increasing manipulations, while negative DAs indicate income decreasing manipulations => Use Absolute Discretionary Accruals 12

13 Data and Variables Main control variable is analyst coverage - the number of analysts who made forecasts about firm s earnings in any given year The number of analysts exhibits a negative correlation with the level of discretionary accruals and total accruals while showing a positive correlation with the level of nondiscretionary accruals 13

14 Tests of Comparison between High and Low Analyst Coverage Firms firms with coverage have: higher market valuation better performance lower level of discretionary accruals Firms with a higher level of coverage have: higher market valuation and performance lower level of discretionary accruals than firms with a lower level of coverage Consistent with the hypothesis that analysts act as external monitors of managers and reduce earnings management behavior 14

15 The Effect of Analyst Coverage on Earnings Management Methodology - Ordinary least squares regressions: 1. Control for the factors that affect accruals and could also affect firms earnings management => get residual coverage : 2. Estimate the effect of analyst coverage on earnings management with the estimated residual coverage as the main proxy for analyst coverage = a component of analyst coverage that is uncorrelated with firm size, past performance, growth, external financing activities, or volatility of business: 15

16 The Effect of Analyst Coverage on Earnings Management The coefficient on residual coverage is negative, indicating that a higher level of coverage is associated with a lower level of earnings management But the magnitude of coefficients are small and only of marginal significance Endogenous analyst coverage: If analysts intentionally choose to cover firms with less earnings management: It could drive the negative relation between analyst coverage and earnings management 16

17 The Effect of Analyst Coverage on Earnings Management Instrumental Variable Analysis with: Expected coverage caused by the change in the size of a brokerage house as instrument When a brokerage house reduces its size, it employs fewer analysts and tends to drop some of its existing coverage to reduce total workload => unrelated to earnings management but affecting the analyst coverage=exogenous variable Inclusion the S&P 500 index dummy as instrument More analysts following than to their non S&P 500 peers Not directly related to earnings management 17

18 The Effect of Analyst Coverage on Earnings Management 2SLS tests Expected coverage caused by change in size of a brokerage house as instrument The coefficients on instrumented residual coverage are negative and significant and of (slightly) greater magnitude than in OLS tests An increase in the size of analyst coverage from the 25th to the 75th percentile indicates a drop in the level of DAs equal to 14% (26%) of the sample mean (median) 18

19 The Effect of Analyst Coverage on Earnings Management 2SLS tests S&P 500 index dummy as instrument The coefficient on analyst coverage is negative and significant The estimates from the 2SLS tests are larger and more significant than those of the OLS regression 19

20 Covered vs. Uncovered Firms Do uncovered firms engage in more earnings management than covered firms? Use dummy for analyst coverage as an explanatory variable Covered firms have a much lower level of discretionary accruals than uncovered firms the discretionary accrual level of a firm with analyst coverage is lower than that of an uncovered firm by about 20% (37%) of the sample mean (median) 20

21 Characteristics of Analysts Do different types of analysts have different effects on earnings management? Analysts who are better at processing information and analysts with more experience are better forecasters than other analysts Are they also better monitors? i.e. firm s earnings management behavior decreases with the number of superior analysts following the firm? 1. Affiliation Research shows that analysts from more prestigious brokerage houses make more accurate forecasts and their revisions have a bigger impact on the market: Top brokerage houses are able to hire better talent than non-top brokerage houses Analysts with top brokerage houses have better resources to conduct their research, such as more extensive databases and better research support staff Top brokerage houses have better distribution channels 2. Experience - Experienced analysts are better at incorporating past information and make more accurate forecasts: General experience - the number of years an analyst has worked as an analyst Firm-specific experience - the number of years an analyst has followed a given company An average firm in the sample is followed by analysts with 6.55 years of experience as analyst and 3.05 years of experience with the firm 21

22 Characteristics of Analysts Firms with more analysts from top brokerage houses have a lower level of discretional accruals An experienced analyst from a top broker has a much stronger impact on the firm he or she covers than an inexperienced analyst from a non-top broker Analysts firm-specific experience has a stronger effect on earnings management than general experience analysts forecasting ability benefits more from firmspecific experience than from general experience 22

23 Conclusion We explored how to deter opportunistic earnings management if and when traditional governance devices do not seem to work well We studied the role of information intermediaries in corporate governance within the context of the effect of analyst coverage on earnings management: A higher level of analyst coverage is related to less earnings management The research suggests that a high level of analyst coverage creates a better information environment for firms and leads to less asymmetric information The results suggest that the role of information intermediaries in corporate governance is potentially significant 23

24 Credit Rating Agencies Credit rating agencies (CRAs) provide opinions on the creditworthiness of entities and their financial obligations provide the public, and government regulators with an estimate of the default risk associated with particular bond issues The credit rating process involves analysis of both business risk and financial risk Ratings are broadly divided into Investment Grade and Speculative Grade This distinction is important because some investors are prohibited from holding noninvestment grade bonds 24

25 The Role of CRAs Valuation role by disseminating information to market participants 1. ratings timeliness 2. information usefulness ratings accuracy and the extent to which CRAs provide additional relevant information Facilitate contracting ratings as credit quality benchmarks in private contracting (loan agreements, bond covenants etc.) and in institutional investors in-house investment rules (only high quality instruments are allowed) 1. Ratings stability change only when fundamental credit risk changes: through-the-business-cycle rating approach 2. Rating conservatism - greater level of verification is required for credit rating upgrades than for downgrades reduces the risk that a business will be portrayed as being financially stronger than it really is Conflicting incentives for the large CRAs 25

26 Why Are Credit Rating Agencies Different? During the early 1970s, the SEC decided that instead of regulating the credit rating industry, it would begin relying on the ratings of a handful of major credit rating agencies in making certain regulatory determinations and created the NRSRO concept. These agencies are Nationally Recognized Statistical Rating Organizations (NRSROs) whose ratings are fully recognized by the SEC, Federal and state banking and insurance regulators, and other U.S. oversight bodies There are three main NRSROs: Moody s, Standard & Poor s and Fitch Regulatory Licenses: A good rating entitles the issuer to certain advantages related to regulation Creates oligopolistic pressure and generate economic rents for CRAs that persist even when they perform poorly and otherwise would lose reputational capital => it is costly 26

27 Why Are Credit Rating Agencies Different? Liability - lack of exposure to civil and criminal liability Unlike equity analysts (who make buy, hold, or sell recommendations) credit rating agencies say that they are not providing investment advice and they are merely financial journalists publishing opinions Profitability As of early September 2005, Moody s market capitalization was more than $15 billion supported by assets of just $1.4 billion The value of Moody s shares has increased by more than 300 % during the past five years whereas most banks shares have declined in value 27

28 Why Are Credit Rating Agencies Different? Conflicts of Interest: Fee Structure: approximately 90 % of rating agency revenues come from issuers who pay for ratings might encourage the CRAs to issue more favourable ratings and to be less diligent in probing for negative information Credit rating agency board members serve in various capacities for companies that the credit rating agencies rate Ancillary services: rating a company and, at the same time, consulting them how each transaction will affect their credit ratings or how to improve it leads to conflicts of interests Unsolicited ratings issuers are rated without request: Less self-selection process in which only low-credit-risk issuers are rated The conflict involves the pressure when the agency threatens the issuers with unfavorable ratings to persuade them to pay fees for the ratings Unlike other financial intermediaries, credit rating agencies have not been pressured to eliminate such conflicts 28

29 Have Rating Agencies Become More or Less Conservative? The identified conflicts of interest may have led the agencies to relax their standards: too generous ratings relative to the default risk of the securities Inflated ratings could be due to the demand from the buy side investors: increase institutional investors flexibility in investing reduce the amount of capital institutions have to maintain against their investments increase their perceived risk-adjusted profitability in the eyes of less-sophisticated ultimate investors by making it appear that a AAA rated investment is earning a AA-rated return Given that many financial institutions made capital allocation decisions based on these ratings (and ultimately failed) rating agencies have even been accused of causing the crisis Study the changes in the standards applied by rating agencies over time, and the consequences of these changes for corporate behavior and debt pricing 29

30 Have Rating Agencies Become More Conservative? Monthly data on debt ratings issued by Standard & Poor s over S&P s ratings fall into 21 categories. The lower the rating, the higher the expected default risk Firms rated BBB and above = investment grade Firms rated below BBB = noninvestment grade or junk-rated firms 30

31 Distribution of Ratings The fraction of firms with AAA, AA, or A ratings has declined The fraction of firms with BBB and BB ratings has increased The credit quality of U.S. corporate debt issuers has worsened over time Or have rating agencies become more conservative? 31

32 Estimate a ratings model with potential explanatory variables: EBITDA to interest payments (IntCov) profitability = EBITDA divided by sales (Profit) long- and short-term debt divided by total assets (Book Lev) the log of the book value of assets (Size) long- and short-term debt divided by EBITDA (Debt/EBITDA) volatility of profitability (Vol) using current and the four previous years data cash and marketable securities divided by total assets (Cash/Assets) convertible debt divided by total assets (ConvDe/Assets) rental payments divided by total assets (Rent/Assets) tangibility, measured as net property, plant, and equipment divided by total assets (PPE/Assets) capital expenditures divided by total assets (CAPEX/Assets) firm s beta (Beta) firm s idiosyncratic risk 32

33 Estimate a ratings model with potential explanatory variables: All the explanatory variables are statistically significant and have the expected sign, with the exception of the cash ratio. Firms have worse credit ratings when they: have more debt of various kinds pay higher rents have lower interest coverage are less profitable have more volatile profits are smaller have fewer tangible assets have lower capital expenditures Have higher beta and idiosyncratic risks Cash holdings => firms with higher default risk in the long run are more likely to save cash 33

34 Estimate Ratings Model (continued) Study year dummies: measure the increase in the credit rating variable (the decline in ratings quality) with respect to first sample year of 1985 All the year dummies are positive and statistically significant and they increase over time! Tougher credit rating standards over the sample period: AAA firm in 1985 would be rated AA by 2009 BBB-rated firm in 1985 would have lost its investment grade rating by 2009 = Change in rating of 3.12 notches Ratings are often also based on qualitative criteria not observed here: include firm fixed effects => assumes that any unobserved firm specific factors are constant (and controlled) over the sample period No significant change in year dummy variables 34

35 The Evolution of Default Rates have rating agencies become more conservative? 1. Rating agencies might have simply adjusted their criteria over time to respond to changes in the macroeconomic environment (default risk has increased over time) 2. The more stringent rating criteria are not reflected in increased default probabilities => firms obtain ratings that are worse than they merit based on historical standards Examine five-year default rates by rating categories from Moody s 35

36 Default Rates for the Rating Categories If the ratings conservatism was warranted, we would expect to observe an upward trend in the rates of defaults There appears to be a decline in default rates for investment grade categories and for the stronger noninvestment grade categories We do not see a decline in defaults for the worst credits (Caa C) 36

37 The Estimation of Default Rates Determine whether the observed decline in default rates is statistically significant: estimate a time-series regression of five-year default rates against a linear time trend variable The linear trend shows declines in default rates for all but the worst rating categories The given rating category does not imply the same default risk over time, => the increased conservatism in ratings is not fully warranted. Linear Trend takes the value of 0 in 1985, 1 in 1986, 2 in 1987, 3 in 1988, etc. 37

38 The Estimation of Recovery Rates What if the decline in default rates has been accompanied by a decline in recovery rates? investor losses across rating categories have remained constant over time Recovery rates do not exhibit a trend Increased conservatism in debt ratings is not fully warranted 38

39 Implications for Capital Structure Decisions Given that ratings are an important determinant of the cost of debt, we would expect firms that are disadvantaged by this ratings conservatism: to use less debt to opt out of the bond market altogether to decide to hold more cash Measure of ratings conservatism: estimate the ratings model over = the old model use the firms characteristics to predict their debt ratings based on the estimated coefficients of the old model for each year over Rating Conservatism = firm s actual rating its predicted rating 39

40 Implications for Capital Structure Decisions New debt issues as a function of firm s actual rating and the measure of ratings conservatism ratings conservatism affects capital structure decisions: Firms issue less debt when their ratings are worse than predicted increasing the ratings disadvantage by one notch reduces net debt issues by 0.3% of total assets. The increased ratings stringency has had a substantial impact on firms capital structure decisions 40

41 Decision to Access the Bond Market All dummies from 1990 are negative, becoming significant from 1996 onwards; Also becoming more negative over time: Firms are less likely to access the public bond market over time Increased conservatism would disadvantage more those firms whose ratings based on the old model are much better than the ratings based on the new model: The higher this number, the more negatively the firm is affected by conservatism conservatism leads firms to opt out of the public debt market 41

42 Implications for Cash Holdings Firms that suffer more from conservatism may decide to hold more cash to cover their financing needs. Cash levels increase by 0.2 pp. for every notch difference between actual and predicted ratings Conservatism affects firms in three ways: they are less likely to issue debt they are less likely to seek a debt rating they increase their cash holdings 42

43 Implications for Growth and Investments Conservatism leads to a decline in the growth rate of 0.8% for every notch difference between actual and predicted ratings For all investment decisions: capital expenditures to sales cash acquisitions to sales R&D to sales coefficients on the conservatism measures are negative (but only the impact on acquisitions is significant): firms whose ratings are worse than predicted are less willing to make cash acquisitions, possibly because they find that the cost of funding those acquisitions with debt is too high Firms affected more by conservatism experience lower growth rates and spend less money on acquisitions 43

44 Implications for Debt Spreads Debt Spread = Bond Yield - U.S. Government Bond Yield If capital markets take the increased strictness of the ratings into account => debt spreads will be narrow for firms with stricter ratings Estimate the following regression model: Debt spreads increase as ratings worsen spreads increase by basis points for each onenotch decline in rating Firms whose actual rating is one notch worse than predicted have spreads that are pp lower Ratings conservatism impacts debt spreads: An increase in conservatism leads to a considerable tightening of debt spreads 44

45 Summary Debt ratings have become more conservative over the period 1985 to 2009 the rating of the average firm declined by three notches. The findings suggest that this increased conservatism is not warranted: Default rates declined over time for both investment grade and noninvestment grade bonds ratings have become more stringent over time Firms take the increased conservatism into account: They issue less debt and have lower leverage They are less likely to obtain a bond rating over time They increase cash holdings They experience decline in growth and investments in acquisitions Capital markets also take this increased stringency into account firms more affected by conservatism have lower spreads than unaffected firms with the same rating 45

46 Conclusion These findings are in sharp contrast to the research on the ratings of assetbacked securities, which suggests that the ratings have become more inflated over time The conflict of interest argument for explaining inflated ratings for mortgage-backed securities may not apply to corporate bonds Differences in the complexity of instruments can explain why there was ratings inflation in structured products but not in corporate ratings Why rating agencies have become more conservative? It may be the case that pressure on rating agencies increased after some well-known defaults at the start of the century Alternatively, conservatism may just be the outcome of rating agencies learning about the correct model over time 46

47 References: *Yu, F. F. (2008). Analyst coverage and earnings management. Journal of Financial Economics, 88(2), *White, L. J. (2010). Markets: The credit rating agencies. The Journal of Economic Perspectives, 24(2), *BAGHAI, R. P., Servaes, H., & Tamayo, A. (2014). Have rating agencies become more conservative? Implications for capital structure and debt pricing. The Journal of Finance. 47

48 Optional Readings/Presentations: Lin, H. W., & McNichols, M. F. (1998). Underwriting relationships, analysts' earnings forecasts and investment recommendations. Journal of Accounting and Economics, 25(1), Jorion, P., Liu, Z., & Shi, C. (2005). Informational effects of regulation FD: evidence from rating agencies. Journal of Financial Economics, 76(2), Chang, X., Dasgupta, S., & Hilary, G. (2006). Analyst coverage and financing decisions. The Journal of Finance, 61(6), Chen, X., Cheng, Q., & Lo, K. (2010). On the relationship between analyst reports and corporate disclosures: Exploring the roles of information discovery and interpretation. Journal of Accounting and Economics, 49(3),