FAT REVISION GUIDE TOPIC 1 - INTRODUCTION Theory set of hypothetical, conceptual, & pragmatic principles forming the general framework of reference for a field of inquiry Understand economic forces and incentives, and predict likely outcomes, of demand, production, and supply of accounting information, and its impact when an economic event occurs FAT is used to explain, predict, and prescribe accounting practice and the reactions to those practices FAT explains when accounting information is decision- useful Accounting information is a key ingredient in efficient capital markets Normative theories are prescriptive theories (best valuation model, best measure of profit) Efficient Market Hypothesis gives info on how we should account Arrow Theorem no true satisfying accounting concept (conflicting interests) Decision- usefulness characteristic of information that is useful to rational investors IA occurs when one party is at an informational disadvantage to the other party Adverse Selection valuation objective, selection (pre- transaction) problem (management has more inside information than outside investors) Role of information is to reduce adverse selection and thus improve operation of capital markets Best measure of net income to control adverse selection best measure to motivate managerial performance (rewards, etc.) o Investors prefer current value method o Managers prefer HCA (conservatism better reflects manager performance, i.e. change in fair value is beyond control) Global Financial Crisis o Subprime mortgage brokers who sold the loans to banks did not care anymore about selling bad loans moral hazard o When defaults rose and house prices fell, banks assets are now junk adverse selection for investors TOPIC 2 IDEAL ACCOUNTING Ideal Accounting System as a benchmark to assess existing accounting system (HCA, FVA) o Perfectly relevant o Perfectly reliable What is theoretically ideal? o Known future cash flows (relevant) o Unbiased, precise info (reliable)! Thus, we can fully resolve adverse selection and value a firm/asset! Precise no random error! Unbiased no systematic error A. IDEAL CONDITIONS CERTAINTY Assumptions o Future Cash Flows certain & known o Cost of Capital given PV basis of accounting o PV of FCF " firm value o No information asymmetry Income recognised as changes in PV occur o True NI = changes in PV for a period o In reality, we will never know True NI! Probabilities not objective Financial Statements: B/S show PV of total assets of the firm o B/S is the only FS needed! Net Income 100% predictable o B/S satisfies the valuation objective! Resolves adverse selection o PV also value- in- use Expected NI = Accretion of discount = r PV year 0 r " cost of capital Expected Net Inc. = Realised Net Inc. o Why? Because of perfect certainty of CFs and only 1 state (No A.E.) B/S at year 0 o Assets = PVFCF 0 o Equity = PVFCF 0 B/S at year 1 o Assets = CF 1 + PVFCF for (t- 1) yrs o Equity = PVFCF 0 + NI 1 Notes o Book Value = PV = Market Value = FV o Relevance info about FCF of the firm o Reliability precise and unbiased o B/S is 100% relevant & reliable because PV is precise, unbiased, and reflects expected future cash flows! ONLY IN IDEAL CONDITIONS 1
Why is there a trade- off in real life? Because of lack of perfect certainty B. IDEAL CONDITIONS UNCERTAINTY Assumptions o Future Cash Flows certain & known o Cost of Capital given o States of Nature known & observable o Probabilities are objective and known! Objective long run true prob. No information asymmetry 100% reliable and relevant o Information about FCF o FCF are precise and unbiased In reality, probabilities are not objective (i.e. true probability of states not known) Assumption of RISK NEUTRAL INVESTORS Financial Statements: B/S at year 0!! o Assets = Exp(PVFCF 0 ) o Debt = PV(interest) 1 + + PV(Principal + Interest) n discounted at yield % o Equity = residual o Exp(PVFCF 0 ) =!"!""#!"!"# p!""# + p!!!!!"#!!!! o Expected Net Inc. = r Opening Equity (Accretion of discount) B/S at year 1 o Assets = CF 1,g/b + Exp(PVFCF t- 1 ) o Equity = Exp(PVFCF 0 ) + RNI - Dividends o RNI= Expected NI + Abnormal Earnings o Amort. = Exp(PVFCF 0 ) Exp(PVFCF t- 1 ) Total Asset Value = PV = Market Value = ensured by arbitrage of risk- neutral investors Expected Net Inc. Realised Net Inc. I/S has no value in ideal states because the B/S can solely explain the expected FCFs I/S will not alter the user s expectations because the obj. probability of the states will never change (perfect certainty of inputs) o Unexpected state probabilities are realised and don t affect future states (probabilities a random walk) o Thus, unexpected earnings do not persist (unexpected earnings random) Dividend policy does not matter under ideal conditions if investors and firm both earn interest on financial assets at same rate Certainty vs. Uncertainty o Working with Expected PVs o Expected CFs Realised CFs! Expected FCFs for period t = (p x CF t ) good + (p x CF t ) bad! Realised FCFs for period t = CF t Abnormal Earnings = Realised Expected CF Realised NI = Expected NI +/- Abnormal Earn. Although Expected NI Realised NI, no new information in income as abnormal earnings do not persist Expected Net Income after state realisation = r Opening Equity Expected Net Income before state realisation = (r Open Equity) good + (r Open Equity) bad Again, B/S shows PVs and I/S shows changes in PVs F/S still reliable but more volatile (Total Asset Value at year 1 depends on the realised state) Uncertainty per se is not a problem C. LACK OF IDEAL CONDITIONS Ideal Systems cannot be set up o Future Cash Flows unknown o Cost of Capital unknown o States of Nature unknown o Probabilities are subjective/unknown All of the above are subject to estimation not perfectly reliable Once objective probabilities are lost, ideal accounting is NOT achievable Also, markets are incomplete o Markets may not exist for all assets and liabilities Implications o Trade- off bet. Relevance & reliability o HCA vs FVA o FVA reliability traded- off o Too much estimation junk (noise) RRA application of non- ideal PV accounting Basic RRA problem firm does not operate under ideal conditions! When oil reserves are discovered, unrealised gains are recognised (but many estimations) As opposed to HCA, revenue is recognised only when oil is sold/realised 2
Relevant, but not reliable (sometimes to the point where relevance is destroyed) However, still not completely relevant it uses period- end prices, interest rate is specified, etc. so it reduces the extent to which the PV predicts FCFs and risk, but increases reliability Disclosure RRA is not reliable, it should not solely be relied upon in performance eval., and that substantial changes to estimates may be made in the future RRA is still useful nonetheless (being more relevant than HCA) RRA mixed results on whether it produces a strong or weak market reaction why? o Low reliability o RRA is pre- empted by more timely sources of reserves information (announcements, forecasts) HCA income statement and matching principle focus (information perspective) o Attempt to measure economic value added (via matching principle) FVA balance sheet focus (measurement perspective) o Attempt to measure firm value o But fails to take matching principle into account Other notes: o PV of Debt at n years = PV(interest) 1 + + PV(Principal + Interest) n! Discounted at yield rate! Interest at coupon rate If certainty, Income Statement Net Income (Accretion of Discount) Or alternatively, Income Statement Gross Cash Flow (w/o dividends) Add: Amortisation of Debt Interest Income Less: Amortisation of Asset Interest Expense Realised Net Income Balance Sheet: Balance Sheet as at Year N Cash(+/- Interest) Debt (Amort.) Assets () () Equity (RNI- Div.) o Interest accrues on Cash past yr 1 at the interest rate o Cash t = (Cash t- 1 (1 + r)) Debt Int. Div. o Equity t = Opening Equity + RNI Dividends o Debt = PV(interest) 2 + + PV(Principal + Interest) n o Interest expense not shown (already accounted for as amortisation of debt) PVFCF = Market Value in ideal conditions because it is the max (min) price at which the buyer (seller) is willing to buy (sell) the asset such that they earn interest at the ideal interest rate (equilibrium) Expected NI = weighted average of ex- post net incomes (W.AVE of GN and BN realised net incomes) = [p good (RNI good )] + [p bad (RNI bad )] If uncertainty, Income Statement Expected Net Income (Accretion) Add: Abnormal Earnings Actual Cash Flow Expected Cash Flow Realised Net Income 3
TOPIC 3 DECISION USEFUL INFORMATION Understand first how rational investors make decisions to be able to generate decision- useful reports in non- ideal conditions Information perspective accountants role is just to supply useful information; accountants must know investors decision needs SINGLE PERSON DECISION THEORY Captures average investor behaviour in making decisions in uncertainty Information is decision- useful when it affects the investors decision, i.e. revise their prior subjective beliefs upon receipt of this information about future outcomes Assume RATIONAL & RISK AVERSE investors o Risk averse declining marginal utility of wealth (concave function) o Losses have more impact than gains o Trade- off between risks & returns The investment decision (risk averse): o Utility for each state for each investment alternative = payoff o EU = SUM (prior prob. x utility) o Choose alternative with highest EU Alternatively, investor could reduce information asymmetry via financial reports (cheap information source) o But financial reports still do not reflect the true state of nature (imperfect relevance and reliability) o Report may reveal bad news for a high- performing firm and good news for a low- performing firm Bayes Theorem explains how an investor will change expectations (conditional probability based on given information) We assume that investors are BAYESIAN o We can assume the probabilities of good/bad news given a high/low performing firm (e.g. P(GN H)) o Because this is what we can observe via historical trends o But we want to find P(m 1 θ) Bayes Theorem: Suppose good news is received We want to know the probability that the firm is in a high- performance state given that we received good news P m! θ = P m! P(θ m! ) P m! P θ m! + (P m! P θ m! ) m! = State that we want to know about (e.g. H) θ = Evidence from the actual, realised event which we will use to make inferences (Could be GN or BN, depending on the actual condition realised by the investment) We now get the posterior state probabilities, Note: P(m! θ) = 1 - P(m! θ) Recalculate Expected Utilities for each state for each investment alternative, for example: o EU = SUM (posterior prob. x utility) Note: the rarer the state of nature which we are testing, the greater the chance of false positives The Information System: o Two states: High and Low o Main diagonal probabilities! When it is high, F/S are reliable and relevant " information system quality is high The higher the main diagonal probabilities (MDPs), the better the investor can predict the state of nature (i.e. F/S informativeness) MDPs of 1 = perfect reliability & relevance o Not achievable How are the Information System probabilities known? MDPs if o GAAP quality (if F/S audited by high- quality auditors, if F/S standards are high- quality) o Analysts revisions (if analysts revise their views, high quality F/S) o Sampling (historical trends) Prior & posterior probabilities are subjective But if F/S is informative (i.e. MDPs are high) then posteriors are better than priors F/S is useful to the extent that GNs or BNs will persist into the future Conservatism increases P(BN L) o Conservatism, if the firm value is low, will force firm to revalue assets down o Thus, if reports show (L), under conservatism, probability that it is indeed (BN) is higher 4
Why is F/S informative? o Current period earnings information can predict future period earnings! Current earnings correlated with Future earnings! Higher correlation, higher MDPs o These help form expectations about future CFs/dividends! GN in current earnings will likely persist in the future! Thus, s/h should bid up prices o These help provide share value, i.e. PV of future dividends (Dividend Model) The information system (F/S informativeness) will increase as relevance increases but only to the extent that reliability will not suffer Professional Bodies: IASB adopts entity view reporting to all capital providers not just shareholders (reports made in the entity s POV) o Assumes investors are risk averse SPDT! Future- oriented (implies that investors want future- oriented information, i.e. expected returns and risk) Move towards increasing use of accruals to anticipate unrealised losses which predict future cash flow reductions (conservative) Earnings number can contain info to help predict FCFs (e.g. accrual accounting) DeChow (1994) Accruals subject to management discretion, opportunistic manipulation of earnings The recording/preparation of accruals will determine the quality of the F/S Fundamental principles of accrual accounting (which make earnings informative) o Revenue Recognition Accounting o Matching Principle! Resolves timing & matching problems which cause cash flows to be a noisy measure of firm performance! The accrual earnings number gives a more precise measure of the true economic value added However, note the trade- off of relevance and reliability (because of managerial estimates) Lower reliability affects F/S informativeness and in turn the updating of the priors 5
TOPIC 4 EFFICIENT SECURITIES MARKETS Assumes the whole market is efficient (single person decision theory " multiple persons) Why worry about efficiency? o Level of efficiency has impacts on the principles as to what info should be supplied o If the market is fully efficient, the market reaction is a good way to measure whether F/S information is useful (measurement of decision- usefulness) Two types of information: o Public (ABS Statistics, financial reports) o Insider (firm has info on its underlying performance but not public) Role of financial reporting to increase public information to get closer to true firm value EFFICIENCY SECURITIES MARKETS (ESM/EMH) Semi- strong Efficient Market Hypothesis One where prices of securities traded on that market, at all times, fully reflect all publicly available information on those securities o Unbiased & Instantaneous reaction by the market on all publicly known info o No arbitrage because of instantaneity (i.e. no abnormal profits from trading on public information) ESM is a relative concept (relative to public information) ESM prices co not necessarily reflect its real (fundamental) underlying value Only new/unexpected information will cause rapid revision of prior beliefs (single person decision theory) Price fluctuations follow a random walk because expected information (seasonality of business, retirement of CEO) is already impounded in price How do prices reflect all publicly known info? Random errors may affect an individual investor s estimates, but Differences in individual forecasting ability is cancelled out when the consensus is formed Thus, on average, market estimates of prices are unbiased (uses all publicly available information) Also, individual decisions are independent Thus, share price fully reflects these publicly available information Efficient Market Price reflects all public info True Fundamental Price reflects all info Thus, for a rational, risk- averse investor: Investing is a fair game there is no arbitrage opportunity for abnormal profits because of instantaneity o Efficient markets instantaneously fully reflect all public information into price to compensate for systematic (market) risk Market prices follow a random walk, i.e. changes in prices are not correlated over time (change today change tomorrow, etc.) o If prices don t follow a random walk, then market is inefficient (slow) Naïve investors are price protected the market price already fully reflects/impounds all public information instantaneously and unbiasedly via sophisticated investors Prices are hard to predict over the short term (random walk); but Shiller found that it is easier to predict over the long term o Share prices fluctuate more than dividends (shares more volatile) o Price/Dividend ratio fall when high, and rise when low (more predictable) Implications of ESM for Financial Reporting Investors are pricing the firm s expected future cash flows using publicly available info Thus only accounting info that affects expected future cash flows will be priced o Cash flow/dividend implications 1. Assuming full disclosure and no cash flow implications, accounting policies do not affect share prices (e.g. depreciation method) 2. Full disclosure enhances efficiency the more information is made public, (1) the more information will be used by investors in revising priors, and (2) the less the impact of inside information o Thus, if markets are efficient, more info will be impounded on to prices o The close the price will reflect its true fundamental value (less inside info) o Only to the extent that benefits exceed the costs of full disclosure 3. Naïve investors don t need to analyse financial reports as they re price protected 6
4. Accountants are in competition o Public info can come from any source Information Asymmetry (IA) ESM still have information asymmetry insider information not in price IA encourages information search o More information = reflective price IA gives rise to adverse selection o Insider trading, resource misallocation Consequence of Adverse Selection: Prices don t reflect the true fundamental value (misallocation of scarce resources) o At the cross section of shares, some will be overvalued, some will be undervalued, relative to the share s fundamental value All share prices suffer o Cost of capital increases o Estimation and Information Risk for all securities (investors don t know which firm is the cherry or the lemon) o Thus investors reduce offer prices for all shares Thin/Incomplete Markets o Reduced offer price increases the lemons and reduces the cherries in the market o Investors realise that the market is full of lemons and will withdraw o Lower level of capital investment (thinner markets) result Incentivising Market Efficiency Create incentives to fully reveal all info to alleviate adverse selection, estimation, and information risk which lowers cost of capital, and to reduce investor concern on inside info o Financial Reporting full disclosure (e.g. inclusion of MD&A in annual reports may increase market price reactions to annual reports) o Share price will be closer to its true fundamental value Create penalties to avoid adverse selection with insider trading laws THE INFORMATION PERSPECTIVE Assumes markets are efficient o Share prices fully reflect public info o Public info can come from any source Assumes investors are Bayesian o What is useful is information that can lead to revision of prior beliefs Investors responsibility to measure/predict the value and performance of the firm in making decisions Accountant s role is to provide useful info HCA, together with other information, is used Proving the usefulness of Accounting Information If the market is efficient, rational, risk- averse investors will revise their priors upon release of accounting information The investment decision will result in share price movement Event Study tests whether an earnings release is decision useful to capital markets (i.e. if it causes a security price change) Event Window time frame of test o Narrow Window tests causation o Wide Window tests association Unexpected Earnings We need a measure of unexpected earnings to test the usefulness of earnings information o U.E. is new information about the firm o ESM will only react to new information! React only if it will change priors U.E. = Actual Earnings Expected Earnings o Expected Earnings = could be last year s earnings, analysts forecasts o If U.E. = 0 (no reaction, no change), it does not mean that earnings is useless! It means that there is no new information; released earnings is the same as the prior/forecasted earnings, which is already impounded in the share price Abnormal Returns/Market Response Share returns influenced by both firm- specific factors and market- wide factors We want to strip out market- wide factors (isolate firm- specific factors) Firm- specific return measures the price response associated with U.E. (new info) Market Model: A.R. = Realised Return Expected Return o Expected Return = β (R M ) o A.R. = Realised Return - β (R M ) 7
+UE = good news portfolio - UE = bad news portfolio If the information is junk, when we plot the Cumulative Abnormal Returns (CARs) of these portfolios, then we ll see no association Ball & Brown (1968) first to document share price response to reported net income o Computer average abnormal returns, not cumulative abnormal returns o The average abnormal returns for GN shares was strongly positive (and vice versa for BN shares) o Results: clear association of earnings info and share price o Thus, earnings contain info about the underlying firm value & performance which affected prices (consistent with EMH and Rational Decision Theory) HCA- based statements are decision- useful Most information already impounded on price before announcement (market already anticipates the GN/BN in earnings info 12 months prior to announcement) " why? 1. Earnings cause market reaction, thus investors have incentives to figure out earnings early; as time goes investors obtain more accurate estimates of the year s earnings; AR increases evenly 2. If the firm is doing well, the efficient market bids up share price; AR rises over the course of the year; the GN in earnings announcement at year- end is correlated with increasing AR over the year Results show that market is not efficient prices are not random walk (PAD) o Market is slow in reacting to info When accounting information is more/less useful Earnings Response Coefficient measures the change in share price return in response to unexpected earnings A.R. = ERC U.E. ERC is composed of firm characteristics and earnings characteristics ERC Firm Characteristics Higher Beta, Lower ERC o Higher security risk o Risk aversion = dampened demand for higher beta stocks on average Higher D/E, Lower ERC o Higher financial risk (earnings go to debtholders and not shareholders) o The more risky a share is, the less investors care about a $ of U.E. High Growth, Higher ERC o PV of FCF is higher for growth firms ERC Earnings Characteristics Earnings Persistence: Higher earnings persistence, Higher ERC o The more a $ of U.E. earnings will persist in the future, the more investors care about that $ of U.E. ERC > 1 Permanent Earnings o New product line, patents, M&As ERC =1 Transitory Earnings o Sale/Disposal of Assets ERC = 0 Price Irrelevant Earnings o Change in Accounting Policy (assuming no cash differential & full disclosure) Abnormal (abnormal size) & Extraordinary (beyond ordinary business) Items now abolished due to potential for management manipulation in classifying earnings HCA has ERC > 1; vs. FVA with ERC = 1 o HCA recognises revenue when realised (more permanent) o FVA recognises revenue even when it is unrealised (transitory) Earnings Quality: Higher earnings quality, Higher ERC o The more a $ of U.E. reflects the underlying economic fundamentals of the firm, the more investors care about that $ of U.E. Relevance and reliability impact quality Can be represented in the information system (main diagonal probabilities) o Higher quality = higher main diagonals o Higher main diagonals = higher extent to which priors are revised = more decision useful HCA Earnings have better quality than FVA Measuring Earnings Quality: o Magnitude of analysts forecast revision (high revision, high quality) o Regression Regression: ΔWC t = b 0 + b 1 CFO t- 1 + + e 8