Security Selection, Earnings Management, and Pro Forma Forecasting Principles

Security Selection Based on Earnings Surprises

  • Conceptual Framework and Efficient Market Hypothesis (EMH):

    • The relationship between earnings forecasts and stock price movements provides a critical test of the Efficient Market Hypothesis (EMH).

    • Before Earnings Announcements:

      • Investment decisions are based on the Expected Earnings Surprise.

      • The central comparison is between the investor's EPS (Earnings Per Share) forecast and the market analyst consensus.

      • Trading Rule:

        • If Your Forecast > Market Consensus: This represents a positive expected surprise; one would typically buy or hold a long position.

        • If Your Forecast < Market Consensus: This represents a negative expected surprise; one would typically sell or short.

    • After Earnings Announcements:

      • The focus shifts to the Actual Earnings Surprise.

      • The comparison is between the actual reported EPS and the market consensus prior to the announcement.

      • Trading Outcome:

        • Actual > Market Consensus: Positive surprise.

        • Actual < Market Consensus: Negative surprise.

Empirical Evidence: Cumulative Abnormal Returns (CAR)

  • Figure 8.5: Response to Earnings Announcements:

    • The data tracks the Cumulative average excess return (%) against the days from the earnings announcement (ranging from -20 to +90 days).

    • The graph categorizes firms into deciles (1-10) based on the magnitude of the earnings surprise.

    • The highest decile (10) shows a distinct upward drift in excess returns following the announcement (t=0t=0), while the lowest decile (1) shows a sharp and continuing downward trend.

    • This phenomenon is often referred to as the post-earnings-announcement drift (PEAD), which challenges the semi-strong form of EMH, as the market seems to take time to fully incorporate the surprise into the stock price.

The Economic Implications of Corporate Financial Reporting

  • CFO Survey Results (Graham, Harvey, and Rajgopal, 2005):

    • Key Metric: CFOs overwhelmingly believe that outsiders (investors and analysts) prioritize earnings over cash flows as the primary metric for valuation.

    • Primary Benchmarks:

      • Quarterly earnings from the same quarter in the previous year.

      • The analyst consensus estimate.

    • Strategic Trade-offs: There is a constant tension between the short-term pressure to "deliver earnings" and long-term value-maximizing investment decisions.

    • Market Credibility: Meeting or exceeding benchmarks builds credibility. Consistently hitting targets is perceived to maintain or increase the firm's stock price.

    • The "Hidden Problems" Narrative:

      • Small EPS misses (even by one or two cents) can trigger severe market reactions.

      • The market assumes most firms can "find the money" to hit a target. Failure to do so suggests hidden fundamental problems.

      • If a firm previously provided guidance and then misses, it suggests poor management and an inability to predict its own future performance.

    • Accounting Quality: Managers are willing to make economic sacrifices (e.g., cutting R&D or maintenance) to hit targets rather than using within-GAAP accounting adjustments, largely due to the post-Enron stigma regarding accounting fraud.

The Preference for Earnings Smoothing

  • Smooth vs. Volatile Earnings:

    • CFOs believe volatile earnings are riskier than smooth earnings, even when cash flows are held constant.

    • Smooth earnings make the tasks of analysts easier, increasing predictability.

    • Risk Premium: Less predictable earnings (missed targets or high volatility) command a higher risk premium in the market.

    • Sacrificing Value: An astounding 78%78\% of surveyed executives admitted they would sacrifice economic value to ensure smooth earnings.

  • Equilibrium of Overreaction:

    • Managers perceive the market's severe reaction to surprises as a "reality."

    • Therefore, sacrificing long-term value to avoid short-term turmoil (equity/debt market volatility) is often viewed as the "lesser evil" or the optimal choice in the current financial environment.

    • Dividend Policy Parallel: This mirrors findings by Brav et al. (2005), who noted that strong market reactions force executives to maintain dividends at all costs, even if it means bypassing positive Net Present Value (NPV) investments.

Case Study: Cleveland-Cliffs CEO Lourenco Goncalves

  • Context (October 19, 2018):

    • CEO Lourenco Goncalves famously berated analysts following a stock price drop.

    • The Incident: The stock price fell nearly 6%6\% on a Friday despite the company reporting increased profit and improved revenue.

    • The Cause: Excluding specific items, the reported EPS was $0.64\$0.64, which fell short of the analyst consensus estimate of $0.66\$0.66 (per FactSet).

    • Lesson: Market reactions are frequently driven by the performance relative to the consensus estimate rather than absolute year-over-year gains in profit or revenue.

Methods for Forecasting Sales and EPS

  • Equity Analysts:

    • Provide stock recommendations (Buy/Hold/Sell).

    • Establish target prices.

    • Prepare Pro forma financial statements to forecast quarterly revenues and EPS.

    • Market Coverage: Small stocks typically have an average of 1010 analysts; large stocks average 1414 analysts.

  • Estimize: A platform utilized to aggregate crowdsourced earnings and revenue estimates, often providing a different "edge" compared to traditional Wall Street consensus.

Constructing Pro Forma Financial Statements (Quarterly)

  • Core Methodology: Based on Valuation by Titman and Martin. It utilizes both annual data (10-K, filed within 60 days of fiscal year-end) and quarterly data (10-Q, filed within 45 days of quarter-end).

  • Step-by-Step Pro Forma Income Statement Logic:

    1. Sales: Calculate as Prior Sales (same quarter)×(1+gS)\text{Prior Sales (same quarter)} \times (1 + g_S), where gSg_S is the sales growth rate.

    2. Cost of Goods Sold (COGS): Calculated as Sales×(1−Gross Profit Margin)\text{Sales} \times (1 - \text{Gross Profit Margin}).

    3. Gross Profit: Calculated as Sales×Gross Profit Margin\text{Sales} \times \text{Gross Profit Margin}.

    4. SG&A (Selling, General, and Administrative): Forecasted using the \text{Sales} \times \text{SG&A to Sales ratio}.

    5. Research and Development (R&D): Forecasted using the \text{Sales} \times \text{R&D to Sales ratio}.

    6. Depreciation Expense: Calculated as Prior Depreciation+CAPEXUseful Life\text{Prior Depreciation} + \frac{\text{CAPEX}}{\text{Useful Life}}.

    7. Interest/Dividend Income: Calculated as Investments (at start of period)×Return Rate\text{Investments (at start of period)} \times \text{Return Rate}.

    8. Interest Expense: Calculated as Debt (at start of period)×Interest Rate\text{Debt (at start of period)} \times \text{Interest Rate}.

    9. Provision for Income Tax: Calculated based on the effective tax rate applied to Income Before Taxes.

    10. Earnings Per Share (EPS): Calculated as Net IncomeDiluted Shares Outstanding (from last quarter)\frac{\text{Net Income}}{\text{Diluted Shares Outstanding (from last quarter)}}.

Case Example: Better Buys Inc. (2014-2018)

  • Forecast Assumptions:

    • Revenue Growth: Constant at 6.0%6.0\%.

    • COGS: Percentage of sales at 64.8%64.8\%.

    • Fixed Cash Operating Expenses: $2,000,000\$2,000,000.

    • Variable Cash Operating Expenses: 16.0%16.0\% of sales.

    • Depreciation: 10.0%10.0\% of fixed assets from period t−1t-1.

    • Nonoperating Income: 10.0%10.0\% of investments from period t−1t-1 ($300,000\$300,000 constant in example).

    • Interest Expense: 7%7\% of total liabilities from period t−1t-1.

    • Tax Rate: 40.0%40.0\%.

    • Dividend Policy: Fixed at $200,000\$200,000 per period.

Mathematical Foundations of Forecasting

  • Forecasting Revenue Growth (gSg_S):

    • The objective is to minimize the Mean Squared Error (MSE), where Error=Actual−Predicted\text{Error} = \text{Actual} - \text{Predicted}.

    • The best forecast is considered the conditional expectation (mean).

    • Expected Revenue Growth formula:         Expectation (mean)=E[g]=∑i=1npigi\text{Expectation (mean)} = E[g] = \sum_{i=1}^{n} p_i g_i         Where:

      • gig_i = possible growth outcome.

      • pip_i = probability of that outcome.

    • Revenue Prediction:         Expected Revenuet=Revenuet−1×(1+E[g])\text{Expected Revenue}_t = \text{Revenue}_{t-1} \times (1 + E[g])

  • Historical Estimation Example:

    • To forecast Q2 2023, evaluate prior Q2 data (e.g., Q2 2019 through Q2 2022).

    • Growth Calculation (gg):         gt−1=revt−1revt−2−1g_{t-1} = \frac{\text{rev}_{t-1}}{\text{rev}_{t-2}} - 1

    • Assign probabilities (pp) to these historical growth rates based on market beliefs to derive the expected growth (E[g]E[g]).

  • Forecasting Interest and Net Income Components:

    • Interest Income (Q4 2023): Annual Return×InvestmentsQ320234\text{Annual Return} \times \frac{\text{Investments}_{Q3 2023}}{4}

    • Interest Expense (Q4 2023): Interest Rate×DebtQ320234\text{Interest Rate} \times \frac{\text{Debt}_{Q3 2023}}{4}

    • Net Income (NI): Net Operating Income+Net Other Income (Expense)\text{Net Operating Income} + \text{Net Other Income (Expense)}