2010年4月6日学术报告通知.pdf
2010年4月6日学术报告通知.pdf
1. The objectives of accounting research 1 1. The objectives of accounting research a. The setting of accounting policy to maximize welfare by improving resource allocation, which subsumes market efficiency and information useful for contracting.(both within and outside the organization 1 1. The objectives of accounting research a. The setting of accounting policy to maximize welfare by improving resource allocation, which subsumes market efficiency and information useful for contracting.(both within and outside the organization b. Accounting innovation and entrepreneurship : Research should not necessarily conform to existing paradigms. 1 2. Means 2 2. Means a. Cannot be based solely on descriptive (experimental or archival) research 2 2. Means a. Cannot be based solely on descriptive (experimental or archival) research • Policy formulation must be based on normative research. 2 2. Means a. Cannot be based solely on descriptive (experimental or archival) research • Policy formulation must be based on normative research. • In formulating normative innovations, assumptions required in the modeling may need to be tested. 2 2. Means a. Cannot be based solely on descriptive (experimental or archival) research • Policy formulation must be based on normative research. • In formulating normative innovations, assumptions required in the modeling may need to be tested. • This is parallel to scientific inquiry, such as in the field of medicine: develop a theory of the antecedents of a certain disease, and of a potential cure, experimentally or empirically test the assumptions made by the theory, and then subject the implications to testing if feasible. 2 2. Means 3 2. Means • This calls for integrated research programs that combine the normative and the descriptive. • I am not addressing here research for personal consumption, which does not require publication or dissemination. The latter would be akin to art enjoyment. 3 3. Research in accounting today Managers Markets Wise Foolish 4 3. Research in accounting today Managers Markets Wise Foolish Wise 4 3. Research in accounting today Managers Markets Wise Foolish • Analytical equilibrium Wise modeling • Tests of models with rational parties: • signaling, optimal compensation, optimal performance measures, etc 4 3. Research in accounting today Managers Markets Wise • Analytical equilibrium Wise modeling • Tests of models with rational parties: • signaling, optimal compensation, optimal performance measures, etc Foolish • Tests of market efficiency with respect to accounting numbers • The market sees through attempts to manipulate accounting numbers • Value relevance: Association tests 4 3. Research in accounting today Managers Markets Wise • Analytical equilibrium Wise modeling • Tests of models with rational parties: • signaling, optimal compensation, optimal performance measures, etc Foolish • Tests of market efficiency with respect to accounting numbers • The market sees through attempts to manipulate accounting numbers • Value relevance: Association tests Foolish 4 3. Research in accounting today Managers Markets Wise • Analytical equilibrium Wise modeling • Tests of models with rational parties: • signaling, optimal compensation, optimal performance measures, etc Foolish • Tests of market efficiency with respect to accounting numbers • The market sees through attempts to manipulate accounting numbers • Value relevance: Association tests • Pernicious earnings Foolish management, pernicious smoothing • accounting-market anomalies: accruals anomaly, price-earnings ratios, Price-CFO ratios, etc 4 3. Research in accounting today Managers Markets Wise • Analytical equilibrium Wise modeling • Tests of models with rational parties: • signaling, optimal compensation, optimal performance measures, etc • Pernicious earnings Foolish management, pernicious smoothing • accounting-market anomalies: accruals anomaly, price-earnings ratios, Price-CFO ratios, etc Foolish • Tests of market efficiency with respect to accounting numbers • The market sees through attempts to manipulate accounting numbers • Value relevance: Association tests • Difficult to test: Management does not attempt to fool the market, when in fact the market could be fooled 4 3. Research in accounting today Managers Markets Wise • Analytical equilibrium Wise modeling • Tests of models with rational parties: • signaling, optimal compensation, optimal performance measures, etc • Pernicious earnings Foolish management, pernicious smoothing • accounting-market anomalies: accruals anomaly, price-earnings ratios, Price-CFO ratios, etc Foolish • Tests of market efficiency with respect to accounting numbers • The market sees through attempts to manipulate accounting numbers • Value relevance: Association tests • Difficult to test: Management does not attempt to fool the market, when in fact the market could be fooled 4 3. Research in accounting today 5 3. Research in accounting today • The only sustainable long-term equilibrium is the wise-wise cell above 5 3. Research in accounting today • The only sustainable long-term equilibrium is the wise-wise cell above • Any empirical tests documenting findings in other cells must be cautiously evaluated. (i) For example, in the managers-wise-marketfoolish cell, whether the risk adjustments were properly made should be carefully examined when documenting anomalies. 5 (Continued) 6 (Continued) (ii) In the case of pernicious earnings management, what market frictions and /or institutional deficiencies are responsible? 6 (Continued) (ii) In the case of pernicious earnings management, what market frictions and /or institutional deficiencies are responsible? (iii) And, more importantly, what feasible policies can researchers offer to rectify the deficiencies? 6 (Continued) (ii) In the case of pernicious earnings management, what market frictions and /or institutional deficiencies are responsible? (iii) And, more importantly, what feasible policies can researchers offer to rectify the deficiencies? (iv) How could assumptions made in the process of formulating such policies be tested and validated? foolishness? What is the remedy? 6 (Continued) 7 (Continued) (v) In the managers-foolish-market-wise cell any findings should be explained in terms of the factors that impede management from learning that the market is wise, that it sees through their manipulations. 7 (Continued) (v) In the managers-foolish-market-wise cell any findings should be explained in terms of the factors that impede management from learning that the market is wise, that it sees through their manipulations. (vi) What is the equilibrium producing this phenomenon? Is it job horizon? Is it simply stupidity? 7 (Continued) (v) In the managers-foolish-market-wise cell any findings should be explained in terms of the factors that impede management from learning that the market is wise, that it sees through their manipulations. (vi) What is the equilibrium producing this phenomenon? Is it job horizon? Is it simply stupidity? (vii) And again, with respect to the managers-wisemarket-foolish cell, what market frictions or institutional deficiencies sustain long-term market 7 4. More in-depth: studies of value relevance 8 4. More in-depth: studies of value relevance a. Can accounting system A be compared with accounting system B based on value relevance (association) studies? 8 4. More in-depth: studies of value relevance a. Can accounting system A be compared with accounting system B based on value relevance (association) studies? • Not unless the two equilibria are evaluated: accounting system A associated with equilibrium price P(A), and accounting system B associated with equilibrium price P(B). 8 4. More in-depth: studies of value relevance 9 4. More in-depth: studies of value relevance b. Are prices or returns the best benchmark against which to evaluate information content of accounting numbers? 9 4. More in-depth: studies of value relevance b. Are prices or returns the best benchmark against which to evaluate information content of accounting numbers? i) Prices may be a poor benchmark because of noise generated by liquidity and noise traders 9 4. More in-depth: studies of value relevance b. Are prices or returns the best benchmark against which to evaluate information content of accounting numbers? i) Prices may be a poor benchmark because of noise generated by liquidity and noise traders ii) Consider the circularity. If accounting alone informs the markets, tests are unnecessary. See the following Venn diagram. 9 Circularity : Venn Diagram 10 Circularity : Venn Diagram Inside Information 10 Circularity : Venn Diagram Inside Information Non-accounting 10 Circularity : Venn Diagram Inside Information Non-accounting Accounting 10 Circularity : Venn Diagram Inside Information Accounting Non-accounting Noise 10 4b. Price or returns as the benchmark for informativeness 11 4b. Price or returns as the benchmark for informativeness iii) If additional public and/or private sources inform the market, the association amounts to testing whether accounting numbers reflect non-accounting information otherwise impounded in prices. 11 4b. Price or returns as the benchmark for informativeness iii) If additional public and/or private sources inform the market, the association amounts to testing whether accounting numbers reflect non-accounting information otherwise impounded in prices. • In this case, whatever the association, the accounting information would be redundant under this maintained hypothesis: prices better reflect values -> there is no demand for accounting. 11 4b. Price or returns as the benchmark for informativeness 12 4b. Price or returns as the benchmark for informativeness iv) Nonetheless, non-accounting information may not duplicate the accounting information but help assess its reliability. 12 4b. Price or returns as the benchmark for informativeness iv) Nonetheless, non-accounting information may not duplicate the accounting information but help assess its reliability. • In this case, association may shed light on managerial incentives or skills, but not the desirability of accounting policies. 12 4b. Price or returns as the benchmark for informativeness iv) Nonetheless, non-accounting information may not duplicate the accounting information but help assess its reliability. • In this case, association may shed light on managerial incentives or skills, but not the desirability of accounting policies. • Auditing research addressing the demand for auditing services fits in this category: the impact and usefulness of auditing. 12 4b. Price or returns as the benchmark for informativeness iv) Nonetheless, non-accounting information may not duplicate the accounting information but help assess its reliability. • In this case, association may shed light on managerial incentives or skills, but not the desirability of accounting policies. • Auditing research addressing the demand for auditing services fits in this category: the impact and usefulness of auditing. • Also, information asymmetry can affect the reliability and be manifested in association measures. 12 4b. Price or returns as the benchmark for informativeness 13 4b. Price or returns as the benchmark for informativeness v) Finally, any value relevance tests should employ an appropriate valuation model. There would be no demand for accounting information under information symmetry. 13 4b. Price or returns as the benchmark for informativeness v) Finally, any value relevance tests should employ an appropriate valuation model. There would be no demand for accounting information under information symmetry. vi) Normatively, the role of accounting is to reveal truthfully private information possessed by knowledgeable insiders. The most pressing research question thus would be how to devise arrangements that truthfully elicit private information, and empirically or experimentally test assumptions made in formulating such arrangements. 13 4b. Price or returns as the benchmark for informativeness 14 4b. Price or returns as the benchmark for informativeness viii) This brings into focus the tension between relevance and reliability on the one hand and the role of accounting in informing vs. contracting (stewardship) 14 GAAP AND POLITICS THE CASES OF MARK TO EXIT VALUES Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 1 GAAP OBJECTIVES AND EXTERNALITIES • OVERALL OBJECTIVE: EFFICIENT RESOURCE ALLOCATION. • IMPLIES INFORMATION ON PRIVATE BENEFITS AND COSTS TO STAKEHOLDERS: FOR INVESTORS THIS MEANS HELPING PREDICT RISK AND RETURN • BUT IT ALSO IMPLIES THE PROVISION OF INFORMATION ON EXTERNALITIES (SYSTEMIC RISK) THAT ALLOWS REGULATERS TO ELIMINATE THEIR ADVERSE IMPACTS. Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 2 DERIVED IMPLICATIONS • GAAP should not yield information that, when inputted into users’ decision models, reduces allocative efficiency (the case of exit values). • Information about effects of an enterprise’s decisions on other entities should be disclosed unless it is internalized by the enterprise. Examples: implicit cost of free or cheap government guarantees (guarantees of deposits, debts). Impacts of risky investments by CDS writers on counterparty risk of CDS buyers, etc. Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 3 DERIVED IMPLICATIONS (CONTINUED) • Note that the allocative efficiency objective of GAAP is narrow: standard setters are not concerned with wealth distribution (social welfare). • If allocative efficiency is attained, no conflict between standard setters and policymakers or regulators should be expected. The latter would be concerned only with non-accounting policies and wealth distribution. Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 4 THE CASE OF MARK TO EXIT VALUES • Mark to Market accounting refers to presenting financial assets in balance sheets at “Exit Value” – the amount for which an asset can be sold on the balance sheet date. Exit values from the perspective of external, hypothetical markets • Declines in Exit values (“write-downs”) reduce the capital equity base and sometimes also income. • In today’s illiquid markets declines in exit values can be substantial (close to 600b during the recent financial crises), giving rise to adverse systemic effects. Wednesday, December 02, Joshua Ronen, Stern school of Business 5 2009 Mark to Market: Levels of measurement • Three levels of measurement: 1 Quoted prices in an active market for identical assets or liabilities. 2 Other observable inputs: similar assets’ prices in active markets, prices of identical assets in non-active markets, observed interest rates and yield curves, default rates, etc., inputs corroborated my observable market data. 3 Unobservable inputs that reflect management’s assumptions about market participants’ assumptions used in pricing. Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 6 The problems with Mark to Exit Values • Relies on exit values from the perspective of external, hypothetical markets. • Determination is conditional on assumptions about market participants’ information. • Implies that management’s private information is ignored (reason: lack of reliability) : information asymmetry and a lemons problem. But Transparency requires that the company tell the market what it knows -- and not what it thinks the market knows! Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 7 The problems with Mark to Exit Values (continued) • For complex, hard-to-value securities the information asymmetry is severe, the market for liquidity is incomplete, and credit seizes up. • In the illiquid market, valuations (whether under level 2 or 3) under the watch of conservative auditors are very low, leading to losses and/or equity erosion, rating downgrades, capital insufficiency, and possible insolvency. • Source of systemic risk, especially with a huge volume of credit default swaps outstanding. Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 8 The problems with Mark to Exit Values (continued) • Hypothetical market-based Exit values do not help in predicting enterprise cash flows (object of transparency); they fail to meet the objectives of financial statements. • Market bubbles or busts would be transplanted into the income statement and/or equity. • There were massive complaints by banks and lobbying efforts, but • On April 9, 2009, the FASB clarified the implementation of fair values while maintaining the mark to market rule based on the exit value concept (impact on 2Q09: 7%). Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 9 systemic effects (externalities) of exit values Write-down to artificially low exit values Another $700 billion still expected in the following quarters Adverse impact on CDS issuers (default risk) and on CDS buyers (counterparty risk) Debt-rating down grade Potential insolvency Wednesday, December 02, 2009 Triggered additional capital requirement Joshua Ronen, Stern school of Business 10 Lehman CDS Transactions and Write-Offs 500 450 400 350 Spreads 300 250 200 150 100 Sept. 10 2008: During the fiscal third quarter, the Firm is expected to incur negative gross mark-tomarket adjustments on assets of ($7.8) billion, including gross negative mark-to-market adjustments of ($5.3) billion on residential mortgage-related positions, ($1.7) billion on commercial real estate positions, ($600) million on other assetbacked positions and ($200) million on acquisition finance positions. (This is an earnings pre announcement) LEHMAN CDS Spreads 50 0 Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 11 Morgan Stanley CDS Spreads and Write-Offs 160 140 120 100 80 60 40 20 Nov 7 2007: As a result of these declines in value, Morgan Stanley's revenues for the two months ended October 31, 2007, were reduced by $3.7 billion 0 Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 12 Avoiding the pitfalls • Present at Management’s expected present value of cash flows. Also disclose exit values as a measure of downside risk (if things go wrong). • But Management may not present truthfully (the issue of reliability). • Solution for this lies elsewhere: device mechanisms that induce honesty. Wednesday, December 02, 2009 Joshua Ronen, Stern school of Business 13 Information Risk • Evidence of the relation between voluntary disclosures, financial reporting quality, and cost of capital is inconclusive. • Is information quality diversifiable? Is it priced? 12/8/2009 1 Corporate Governance and Irregularities • Examine the relation between information asymmetry between insiders and outsiders and fraud or accounting irregularities, or • Use the existence of fraud and irregularities to develop a measure of information asymmetry between insiders and outsiders on the board of directors 12/8/2009 2 Information Risk • Issues in measuring cost of capital that relies on earnings forecasts independently of information quality: forecasts accuracy depends on the precision of information provided by the company. 12/8/2009 3 Information Risk • Issues in measuring accounting quality: – Accruals may credibly signal future cash flows – Smoothing may be beneficial – Meeting or beating targets may be truth revealing – Earnings management in general could be either beneficial or pernicious 12/8/2009 4 Information Risk • Suggestion: tests for the effects of information risk using as criterion the ability to predict future cash flows. • Consider the interaction between voluntary or mandatory disclosure and private information acquisition. 12/8/2009 5 Information Environment • False disclosures: track over time whether disclosures were true or false. • Incentive to disclose: distinguish between verifiable (and contractible) and non-verifiable disclosures. 12/8/2009 6 Information Environment • Examine externalities: information transfer effects on competitors, suppliers, customers (manifested in debt, equity, and other derivatives prices), production volume, etc. 12/8/2009 7 Information Environment • Proprietary costs: – Literature finds association between competitive position in product markets and voluntary disclosure strategies – Challenge: measuring and quantifying proprietary costs other than the level of competition as proxied by the level of concentration (look at disclosures in 10ks, on competition, price wars, etc.). 12/8/2009 8 Mandatory Disclosures • The political process of standard-setting: – The narrow objective of the FASB vs. social objectives – Differs across countries – Evolution of standards: rigidity vs. flexibility – How sensitive to the political institutional environment 12/8/2009 9 Mandatory Disclosures • Evidence on lobbying and prediction of lobbying efforts and the resulting equilibrium. 12/8/2009 10 Accounting Choices • Firms might tailor accounting treatments to specific earnings objectives and • Investors may rationally infer or observe earnings management. • This suggests a research opportunity: 12/8/2009 11 Accounting Choices • Managers can record “gains” to prevent covenant violation of imperfect contracts but • Make these “gains” transparent such that equity investors can undo their effect on earnings. 12/8/2009 12 Accounting choices • Are earnings quality and disclosure substitutes or complements? For example, does a firm substitute fundamentally low-quality earnings with additional disclosures? this requires identifying fundamental earnings quality (a property of the underlying business transactions). 12/8/2009 13 Accounting Choices • “complete path” approaches: examine both determinants and consequences of earnings quality in the same research. • Example: external auditors and internal control both affect abnormal accruals, which in turn affect the cost of capital. But is the impact on the latter the same when the source is external auditors rather than controls? 12/8/2009 14 Conservative Accounting • Is there a potential for conservative accounting to assist investors in valuation in addition to the common perception that conservative accounting may assist firms in contracting and governance settings? 12/8/2009 15 Earnings Management and Management Turnover • There is evidence that following restatements higher turnover rates are observed • Are there differences in the types of earnings management (such as those that are detrimental or beneficial to shareholders) that influence turnover decisions? • Example: EM to avoid violating covenants vs. EM to meet analysts’ forecasts. 12/8/2009 16 EM and Boards • In what settings is EM encouraged or deterred? • For example, if a manager meets forecasts by manipulation of earnings does he still get rewarded? Or is he awarded only he achieves earnings without manipulation? 12/8/2009 17 EM and Contracting • Identify situations in which EM is actually beneficial to the contracting parties (e.g., in order to facilitate renegotiation of contracts). • Examine whether EM occurs in situations where the conditions for the revelation principle are satisfied. 12/8/2009 18 Earnings Quality and Debt • Are accounting quality metrics commonly used reasonable predictors of credit quality, or changes in credit quality? • Is high accounting quality associated with boards and monitors that can uncover or prevent agency problems? • Or instead, do boards recognize that managers have substantial discretion and potential for agency conflicts and hence demand that managers commit to high-quality accounting? 12/8/2009 19 Earnings quality and debt • Which attributes of the accounting system are most valuable to lenders? • The role of accounting reports in renegotiation: information about future credit quality and cash flows, and about past investment choices and perquisite consumption. 12/8/2009 20 Using contextual information In Valuation • Incorporate in forecasting profitability industry-specific financial metrics such as same-store sales, load factors for airliners, capacity utilization for manufacturers, etc. • Macro data such as GDP growth, unemployment, inflation expectations, interest rates, shape of the yield curve, commodity prices, currency movements, etc. 12/8/2009 21 Disclosure and Litigation • Focus not only on the relationship between quantity and frequency of disclosures, and litigation costs but also on other attributes: – Positive or negative – Verifiable or non-verifiable – Past and present facts vs. forward-looking information 12/8/2009 22 Corporate governance and financial reporting: Substitutes or Compliments? • Evidence is mixed • Do results depend on the legal environment, e.g. pre-and post- SOX • What penalties do independent and other directors face? (turnover, no-votes by shareholders, fewer memberships in boards • Do compensation committee or audit committee members face penalties upon approving auctions backdating, excess compensation, etc. 12/8/2009 23 Bundled Hard and Soft Information • Do management forecasts and other soft information bundled with earnings announcements confirm or soften the impact of negative surprises? • Relate to firm characteristics and managerial incentives. 12/8/2009 24