📖 有效市场假说（EMH）三种形式

Equity Investments

I. Lesson Focus

This lesson examines the Efficient Market Hypothesis (EMH) and its three forms—weak, semi-strong, and strong. Candidates must be able to define each form, identify the information set incorporated in prices, describe appropriate empirical tests, interpret market anomalies, and determine the investment implications for active versus passive management.

II. The Problem

Suppose you are a portfolio manager who spends considerable time and resources studying company financial statements, following news, and analyzing price charts, yet your portfolio consistently fails to outperform a simple buy-and-hold strategy in the CSI 300 Index. Does this mean your research efforts are wasted? The Efficient Market Hypothesis (EMH) directly addresses whether security prices fully and instantaneously reflect all available information, making it impossible for active managers to consistently earn abnormal returns. This lesson systematically explains the theoretical foundation of EMH, the precise definitions and tests of its three forms, real-world anomalies that challenge each form, and the practical consequences for investment strategy. Mastery allows candidates to correctly classify which form of EMH is violated by a given piece of evidence or anomaly.

III. Theoretical Foundation of the Efficient Market Hypothesis

The Efficient Market Hypothesis was formally proposed by Eugene Fama in 1970. Its central proposition is that in an efficient market, security prices adjust rapidly and in an unbiased manner to new information. Consequently, at any moment, the current price equals the intrinsic value, and investors cannot systematically earn abnormal (risk-adjusted) returns through analysis.

For EMH to hold, three conditions must be met: 1. Information is available to all investors at zero or very low cost. 2. Transaction costs are negligible. 3. A large number of rational, profit-maximizing investors react instantaneously to new information.

When these conditions are satisfied, prices follow a random walk and active management cannot generate positive alpha on a consistent, after-cost basis.

IV. The Three Forms of EMH and Their Tests

EMH is categorized into three progressively stronger forms based on the breadth of the information set reflected in prices.

1. Weak-Form Efficiency

Definition: Security prices fully reflect all historical price and volume information.

Implication: Technical analysis is useless. Past price patterns cannot be used to forecast future prices.

Empirical Tests: - Serial correlation (autocorrelation) tests: coefficients near zero support weak-form efficiency. - Runs test: checks whether successive price changes are random. - Filter-rule tests: examines whether mechanical trading rules based on past prices generate abnormal returns.

Investment Implication: Strategies such as moving-average crossovers, trend lines, and chart patterns should not produce persistent abnormal returns.

2. Semi-Strong Form Efficiency

Definition: Security prices fully reflect all publicly available information, including historical prices, financial statements, news releases, analyst reports, and macroeconomic data.

Implication: Fundamental analysis adds no value. Investors cannot earn abnormal returns by analyzing publicly released information.

Empirical Tests: - Event studies: measure abnormal returns around public announcements. - Key events include earnings announcements, stock splits, mergers, and index reconstitutions. - Compute abnormal return AR_t = Actual return – Expected return (often from market model: R_t = α + βR_m + ε). - Cumulative abnormal return CAR(t1,t2) = Σ AR_t. - If CAR adjusts fully on the announcement date and shows no statistically significant drift afterward, the market is semi-strong efficient.

Investment Implication: Active fundamental stock picking is unlikely to outperform passive index funds on a consistent, risk-adjusted, after-fee basis.

3. Strong-Form Efficiency

Definition: Security prices reflect all information, both public and private (including insider information).

Implication: Even corporate insiders cannot earn abnormal returns. Insider trading should not be profitable.

Empirical Tests: - Examine the performance of corporate insiders, mutual-fund managers, and hedge-fund managers. - Most studies show that, on average and after fees, actively managed funds do not outperform their benchmarks on a risk-adjusted basis. - Direct tests of illegal insider trading profitability (often documented by regulatory enforcement actions).

Real-World Status: Strong-form efficiency is almost universally rejected because insider trading has been shown to generate profits, as evidenced by numerous SEC enforcement cases.

V. Market Anomalies and Challenges to EMH

Numerous empirical regularities appear to contradict EMH and have given rise to behavioral finance, which argues that investors exhibit systematic cognitive biases (overconfidence, herding, loss aversion, etc.).

Common anomalies include: - Calendar anomalies: January effect, weekend effect—challenge weak and semi-strong forms. - Size effect: Small-capitalization stocks historically outperform large-cap stocks. - Value effect: High book-to-market stocks outperform growth stocks (incorporated in Fama-French factors). - Momentum: Stocks with strong past 6–12 month performance continue to outperform—challenges weak form. - Long-term reversal: Stocks that performed poorly over 3–5 years subsequently rebound. - Post-earnings announcement drift (PEAD): Stocks with positive earnings surprises continue to rise for weeks after the announcement—direct challenge to semi-strong form. - Insider trading profits: Direct refutation of strong-form efficiency.

These anomalies suggest that prices may deviate from intrinsic value for prolonged periods, yet many academics still view EMH as a useful benchmark rather than a literal description of reality.

Worked Cases

Case 1: Weak-Form Test — Serial Correlation

An analyst collects 250 daily returns for a stock and calculates a first-order autocorrelation coefficient ρ = 0.12 with a t-statistic of 1.89 (critical value ≈ 1.96 at 5 % significance). Does this support weak-form efficiency?

Solution:
The t-statistic of 1.89 is below the critical value of 1.96; therefore, we cannot reject the null hypothesis that ρ = 0. The return series is consistent with a random walk. Technical analysis based on past prices is unlikely to generate persistent abnormal returns, supporting weak-form efficiency.

Case 2: Semi-Strong Form — Event Study

On 10 April 2023, Company A releases quarterly earnings that beat expectations. Cumulative abnormal returns are: CAR(−5,0) = +0.8 %, AR(0) = +4.2 %, and CAR(1,20) = +0.3 % (statistically insignificant). Does this support semi-strong efficiency?

Solution:
The price adjusts sharply on the announcement date itself, with virtually no subsequent drift. Because publicly released earnings information is incorporated instantaneously and unbiasedly, the evidence supports semi-strong-form efficiency. Investors cannot profit by trading on the announcement after it becomes public.

Case 3: Strong-Form Efficiency and Insider Trading

A company executive purchases shares 30 days before a major positive announcement that drives the stock price up 18 %. The executive realizes a profit of USD 120,000, later investigated by the SEC. What does this imply for the three forms of EMH?

Solution:
Profitable trading on material non-public information directly violates strong-form efficiency, as insider information was not already reflected in price. However, if the price adjusts fully and immediately upon public release, semi-strong efficiency may still hold. This case is typical real-world evidence against strong-form EMH.

Traps

Key Formulas

• Abnormal return: AR_t = R_t – E(R_t)
  (E(R_t) often estimated via market model: R_t = α + βR_m + ε)
• Cumulative abnormal return: CAR(t1,t2) = Σ AR_t from t1 to t2
• Jensen’s alpha: α = R_p – [R_f + β(R_m – R_f)]
• Random walk (weak form): P_t = P_{t−1} + ε_t (where ε_t is unpredictable)
• Semi-strong test criterion: CAR should jump on announcement and then remain statistically flat
• Strong-form implication: Even insiders earn zero abnormal return

Practice Questions

Q1. According to the weak-form efficient market hypothesis, which of the following strategies should not consistently generate abnormal returns?
A. Low price-to-earnings stock selection
B. Moving-average crossover rules
C. Trading on material non-public information
D. Passive index replication

Q2. A study finds that stocks continue to earn statistically significant positive cumulative abnormal returns from day 5 to day 30 after an earnings announcement. This finding most directly challenges which form of EMH?
A. Weak form
B. Semi-strong form
C. Strong form
D. None of the above

Q3. Which of the following pieces of evidence would best support strong-form market efficiency?
A. Technical trading rules generate no abnormal returns
B. The average mutual fund underperforms its benchmark after fees
C. Corporate insiders are unable to earn abnormal returns
D. Small-cap stocks outperform large-cap stocks over long periods

Q4. In an event study testing semi-strong efficiency, which statement is most accurate?
A. Only the 20-day post-announcement CAR needs to be examined
B. Significant pre-announcement CAR supports semi-strong efficiency
C. Semi-strong efficiency requires that prices adjust fully and immediately upon public release, with no subsequent drift
D. Event studies are primarily used to test weak-form efficiency

Q5. A stock’s monthly returns over the past 60 months have a first-order autocorrelation of 0.03 (t-statistic = 0.45). This result most strongly supports:
A. Weak-form efficiency
B. Semi-strong-form efficiency
C. Strong-form efficiency
D. Behavioral finance explanations

Q6. Which of the following anomalies most directly challenges semi-strong-form efficiency?
A. The January effect
B. Momentum effect
C. Post-earnings announcement drift
D. Long-term reversal effect

Q7. If a market is semi-strong efficient but not strong-form efficient, which statement is correct?
A. Both technical and fundamental analysis are useless
B. Fundamental analysis is useless, but insiders may profit from private information
C. Technical analysis works, but fundamental analysis does not
D. Even insiders cannot earn abnormal returns

Q8. The Fama-French three-factor model adds SMB (size) and HML (value) factors primarily to explain which category of EMH challenges?
A. Calendar anomalies
B. Size and value effects
C. Insider trading profits
D. Technical chart patterns

Answers

Takeaways

• The three forms of EMH expand the information set sequentially: historical prices → all public information → all information including private data.
• Weak form rejects technical analysis; semi-strong rejects fundamental analysis; strong form rejects profitable insider trading.
• In event studies, the absence of post-announcement CAR drift is the key criterion supporting semi-strong efficiency.
• Empirical evidence most strongly supports semi-strong efficiency in developed markets; strong-form efficiency is almost universally rejected.
• Size, value, momentum, and PEAD anomalies have fueled behavioral finance, yet EMH remains the central benchmark for evaluating active-management performance.
• Exam questions frequently require candidates to map a specific anomaly or research finding to the precise form of EMH it challenges.