Class #15 Accounting Trading Strategies Do Investors Understand Accounting? 15.535 - Class #15 1Road Map: Where do things fit? • Risk Analysis: –CAPM – 3 Factor Model: Size and B/M Matter – Combine with Cash Flow Analysis • Where Now? – Recall discussion in first class about market efficiency edbate • Application of Fundamental Analysis … Can we use financial accounting numbers to identify mis-priced stocks? 15.535 - Class #15 2Does the market set stock prices correctly all the time? • EMPHASIZE: Mkts are very competitive! • But .. Evidence that markets may not be perfectly efficient Æ Possible (risky) arbitrage opportunities. • Question: Can we use current (historical) financial accounting information and fundamental analysis to “pick” which stocks will do better/worse in the upcoming months/years? – Answer: There is growing evidence that this appears to be possible! 15.535 - Class #15 3What is the correct benchmark for “Beating the Market”? • A high stock return (relative to other stocks) does not immediately imply you are getting a “free lunch” or an arbitrage opportunity exists! • Asset pricing models: There is a trade-off between risk and return. – Higher risk stocks should have higher returns. • What is the expected return on stock? … It depends on the stock’s systematic risk! • Simple case – CAPM: E( R ) = Rf + E*(Rm-Rf) – Expected return is increasing in systematic risk! 15.535 - Class #15 4Abnormal Stock Returns: Getting the benchmark correct • Abnormal stock performance must be calculated relative to the stock return predicted by CAPM (or other model): • D = Abnormal return = Actual return – { Rf + E*(Rm-Rf) } – Abnormal return is known as the “alpha”. – A positive (negative) alpha means that the stock provided a higher (lower) return than predicted for a given level of systematic risk. – Strategy: Attempt to go “long” in stocks that will have future positive alphas and “short” in stocks that will have negative alphas. 15.535 - Class #15 5How might we predict alpha’s? • Keys to fundamental analysis: – abnormal returns. – Next, find a technique for “picking” stocks that will First, get the benchmark correct for determining systematically overperform (underperform) the benchmark. – Involves historical analysis … Hope that past strategies will work in future. Why might a successful strategy from the past disappear? • What is an alternative interpretation that we can systematically identify firms with high/low alphas? 15.535 - Class #15 6Valuation & “Beating the Market” • Valuation implicitly assumes market “inefficiency” … basis for active management. • Quantitative Models: – There should be a direct link between current stock price and: • Current earnings (proxy for future earnings and CF’s) • Current book value of equity (proxy for liquidating value) – If this link is absent, then there may be an (risky) arbitrage opportunity. (Go back to DCF!) – “Value” Models: Find companies that are “cheap” relative to others in terms of fundamentals derived from income statements and balance sheets. 15.535 - Class #15 7Value Strategies • Quantitative Models – “Value Strategies”: – Assumption: Investors do not understand fundamentals today, but stock prices will adjust once investors see realized accting performance. – Examples: Strategies based on Book-to-price (B/M), Earnings-to-Price (E/P) – What does high B/M mean? Low B/M? – What does high E/P mean? Low E/P? – Key assumption …. earnings and book equity are comparable across firms. Get the accounting right! 15.535 - Class #15 8Misunderstanding information events and “Drift Strategies” • Drift Models: – Does the market immediately (and correctly) react to an information release that affects company risk or future cash flows? – Post earnings announcement drift • Company announces higher than expected earnings (lower than expected earnings) …. Stock price increases (decreases) on announcement • But stock price continues to go up (down) in the subsequent weeks and months. • Appears to be a profitable trading strategy 15.535 - Class #15 9Other “Drift” Strategies • In general, there appears to be “momentum” in stock returns: – GENERAL MOMENTUM: Firm with highest (lowest) stock returns over past 6-12 months are likely to experience high (low) stock returns in next 6-12 months. – Post earnings announcement drift is one example of underreaction to earnings news Æ momentum. – Other examples: Underreaction to bond-rating changes (which are positively autocorrelated). – Evidence that momentum may be attributable to industry effects vs firm-specific effects. 15.535 - Class #15 10Quality of Earnings Trading Strategy (Accruals Anomaly) • Managers often have incentives to “fool the market” by using their financial reporting discretion to report high earnings. (Why?) • Example: Accruals – Aggressively book sales before they are justified. – Underreport expenses (or defer current expenses). • Key Question: Do investors and analysts understand that managers may be misreporting earnings? Do they know how to back out the accounting distortions? 15.535 - Class #15 11Some Red Flags • Managers that use different accounting methods/estimates compared to other firms in the industry. • Unexplained changes in accounting methods/estimates. • Large gap between reported income and cash flow from operations. • Unusual transactions that boost earnings. • Significant related party transactions. 15.535 - Class #15 12The Quality of Earnings Ratio ( Earnings )  (Cash _ From _ Operations ) Average _ Total _ Assets • Annual ratio low => ‘Low Quality’ • Annual ratio high => ‘High Quality’ 15.535 - Class #15 13“Widely-Accepted” Evidence on Fundamental Trading Strategies Strategy Claimed Direction of Effect E/P ratio High E/P leads to high future abnormal stock returns B/M ratio High B/M leads to high future abnormal stock returns CF/P ratio High CF/P leads to high future abnormal stock returns var(CF)/P ratio High var(CF)/P leads to low future abnormal stock returns V/P ratio Use predicted firm value from abnormal earnings model and compare to stock price: High V/P leads to low returns. Short term High stock return this month leads to low stock return next month.