L10 - EMH

Question 1

What is the Efficient Market Hypothesis?

Answer 1

• The Efficient Market Hypothesis (EMH) asserts that prices in money and capital markets are efficient.
• It is an empirical hypothesis represented by various equivalent statements:
  
  • Current market prices reflect all available information
  • It is impossible to earn excess returns by using information available to the market to guide your trading
  • Information is rapidly and accurately disseminated
  • It is not possible to predict price changes by using current information

Question 2

What is the Fair Game Model of EMH?

Answer 2

Question 3

What is the Martingale property?

Answer 3

• E(p_t/omega(t)) = p_t as we know what the value of p is today

Question 4

What is the other form of the Martingale property/model?

Answer 4

• Xt+1 is also called the martingale difference sequences
  
  • Omega contains all the information about the past prices

Question 5

Why does a Random Walk have the Martingale property?

Answer 5

• Expected return in week 1 is 100.25
  
  • mu in the martingale process is the expected return based on the coin toss
    
    • =0.25%
  • Martingale process states that Pt+1 = Pt (1+ mu)
    
    • This gives us 100.25 from the first to second period so thus a random walk has a martingale property
• Say if we played this game over 5 years and plotted the returns
  
  • graph is indistinguishable from the returns on the S&P500 over the past 5 years (prices cannot be predicted because the stock market is a random walk)

Question 6

What does the random walk theory have to do with stock prices that deviate from predictions?

Answer 6

Question 7

What is the different levels of efficiency?

Answer 7

• We can define three levels of market efficiency, which are distinguished by the degree of information reflected in security prices:
  
  • Weak Form: all current and past information on prices and returns, or market trading data in general.
  • Semi-strong form: all current and past public information
  • Strong form: all current and past information, public or private.

Question 8

Implication 1 of EMH?

Answer 8

• Information search is pointless
• Why spend valuable resources looking for inefficiency if the market is efficient?
  
  • This is not rational But if people are not rational, prices will be inefficient.
• This is the EMH paradox (resolved by Grossman-Stiglitz argument).
• Efficiency is not an “all-or-nothing” matter.
• Investors spend resources until the marginal cost of gathering and processing information is just equal to the marginal benefit.
  
  • 10bps is nothing for a $500 Robinhood investor but worth billions to a mutual fund with billions AUM

Question 9

What is Fundamental Analysis?

Answer 9

• Value a firm using its attributes and expected future condition
• s Find the stock price that reflects fundamental value, leading to buy or sell decisions.
  
  • Estimation of future earnings, dividends, interest rates
  • Analysis of financial statements
  • Risk analysis
• Problem: if the market price already embodies fundamental information, why do fundamental analysis?
• This is the EMH paradox again
• For fundamental analysts, the challenge is to perform an analysis that is better than that of anyone else. (Keynesian animal spirit analyst are trying to guess what the others are trying to guess about their own predictions of the market)
  
  • Find investment opportunities that end up performing better than other people expec

Question 10

What is implication 2 of EMH?

Answer 10

• Active or Passive Portfolio Management
• The EMH implies that professional portfolio managers have no informational advantages over other investors
• Buy and hold should outperform active investment because of lower transactions costs
• Ordinary investors can reduce transactions costs by investing in mutual funds
• Passively-managed mutual funds track an index such as S&P500 of FTS100.
  
  • Active investment does similar/ worse than tracking an index

Question 11

When can active portfolio management be beneficial?

Answer 11

• An investor’s optimal portfolio may vary with their age, tax bracket, degree of risk aversion, etc.
• Professional portfolio managers may not have special information but they can design portfolios for specific purposes:
  
  • Tax efficiency
    
    • high earning find it advantageous to buy tax-exempt municipal bonds despite their relatively low tax yields
      
      • Generally, tilt portfolio towards in the direction of capital gains opposed to interest income, as it is tasked less heavily and because the option to defer the realisation of capital gain income income
  • Particular levels of systematic risk for example by holding specific positions in treasury bills
    
    • Toyota investors should invest additional amounts of stock into automobiles, seeing as most of their compensation already depends on the companies well being
    • Older investors who are living off savings might choose to avoid long-term bonds whose market values fluctuate dramatically with changes in interest rates. –> need to conserve their principal
    • Younger investors may be more inclined towards long term inflation-indexed bonds, stable flow of real income that is locked in over a long period with these bonds can be more important than the preservation of principal to those with long life expectancy
• “Beating the market” need not be the goal of portfolio management

Question 12

Implication 3 of EMH?

Answer 12

• If asset prices are mispriced, then incorrect signals are sent to the market, to investors and to firms.
• Cash may flow into over-valued activities:
• For example, the dot.com bubble of the late 1990s led to over-investment in firms in IT industries.
• Many such firms subsequently lost money or failed altogether.

Question 13

What are the three major issues in evaluating the EMH?

Answer 13

• Academics and market analysts tend to disagree about market efficiency We can think of three major areas of disagreement:
  
  • A. The magnitude issue
  • B. The selection bias issue
  • C. The lucky event issue

Question 14

What is the magnitude issue?

Answer 14

• Small opportunities will exist to make abnormal profits, even if prices are very close to fundamental value.
  
  • 0.001 x 5 billion = 5 million
• Professional analysis may be worthwhile even for a very small % increase in portfolio performance.
• This may not be detectable by statistical analysis but may still represent a significant return
• The market is not “all-or-nothing” efficient,

Question 15

What is the selection bias issue?

Answer 15

• Tests of the EMH based on published investment strategies (“tips”) will likely show that these strategies are worthless. –> publically available so everyone arbitrages the strategy away
• Truly profitable strategies or “tips” would not be publicised.
  
  • Renaissance highly profitable closed fund ‘ medallion’ has been profitable for all but one year since its inception (In 1989), –> but not many people know what quantitative strategies they use
  • Last year it surged to 76%
• Published techniques may not work but effective private techniques may exist

Question 16

What is the lucky event issue?

Answer 16

• A large investment surplus need not imply either special skill or market inefficiency.
• Successful professional portfolio management receives high publicity.
• Unsuccessful management receives little publicity
• Both Success and lack of success may be due to chance.
• Abnormal performance is only evidence of inefficiency if it is persistent over time.

Question 17

What evidence do we look at to test the weak-form efficiency?

Answer 17

• Returns over the Short Horizon
  
  • Momentum: Good or bad recent performance continues over short to intermediate time horizons (esp. portfolios)
    
    • Jegadeesh and Titman (1993) found a momentum effecting which good or bad recent performance of particular stocks continue over time,
      
      • They conclude that while the performance of individual stocks is highly unpredictable, portfolios if the best-performing stocks in the recent past appear to outperform other stocks with enough reliability to offer profit opportunities
• Returns over Long Horizons
  
  • Episodes of overshooting followed by correction
  • –Long-run mean reversion (DeBondt and Thaler, 1985): Positive (negative) returns tend to be followed by negative (positive) returns(reversal effect)
    • looking at the 35 top winners and losers in one period, the loser tended to outperform the winners on an average of 25% in the following three periods
• Use of serial correlation

Question 18

What was the correlation test (short-term predictability) for weak-form efficiency?

Answer 18

• Tests of short-term predictability examine whether return in the prior period (usually a day or days) can predict today’s return.
• Correlation tests are tests of a linear relationship between today’s returns and past returns.
  
  • rt = a + br_t-1-T + e_t
• where a measures the expected return, uncorrelated to previous return, usually positive. The term b measures the relationship between the previous return and today’s return
  
  • If be is statistically significant, it means we can use past returns to predict future returns

Plotting successive trading day returns of 4 companies found no evidence of serial correlation –> Consistent with EMH

The second picture shows daily correlations coefficient(usually negative but close to zero) and not significant –> again supports EMH

• But found that the positive returns of today are negatively correlated with returns in two years time???

Question 19

What evidence do we have against weak-form and semi-strong form efficiency in relation to broad market returns?

Answer 19

• Fama and French, 1988:
  
  • Aggregate returns are higher with higher dividend ratios
• Campbell and Shiller 1988:
  
  • Earnings yield can predict market returns
• Keim and Stambaugh (1986):
  
  • Bond spreads can predict market returns

One could argue that these imply that abnormal stock returns can be predicted. More probably however these variables are proxying for variations in the market risk premium

Question 20

Evidence against of Semi-strong form efficiency?

Answer 20

• P/E Effect (adjustment for risk?)
  
  • Even if the portfolio is adjusted for risk (Basu, 1983) found that a portfolio of P/E ratios tend to provide higher returns than high P/E portfolios
    
    • NOt very plausible to earn superior returns through this as it is such a simple procedure, –> CAPM beta hasn’t fully adjusted for all risk and P/E will act as a descriptor of risk and will be associated with abnormal return if the CAPM is used as a benchmark
• Small Firm Effect (January Effect)
  
  • (Banz, 1981) –> The NYSE stocks split into 10 portfolios ranked by size, between 1926-2015 the small firms consistently outperformed, with the average annual return between the smallest and largest portfolio being 7.65%
• Neglected Firm Effect, Liquidity Effects
  
  • Small firms perform better than large ones because not a lot is looking at them or a lot unknown about them (more risk more returns)
    
    • Yet high trading costs on small stocks can easily wipe out apparent normal returns
• Book-to-Market Ratios (Fama-French)
• Post-Earnings Announcement Price Drift

Question 21

What is post earnings announcement drift?

Answer 21

• When a firm announces earnings will be much larger than e
• xpected, will this news be reflected in share price the same day or over the next week? Studies that have been devoted to the effect of an announcement on share price are known as “event studies”.
• They aim to determine what information is reflected in price and, if its impact is unclear, to determine whether the announcement is good or bad news.

The methodology of event studies is fairly simple

• 1. Collect a sample of firms that had a surprise announcement (the event).
• 2. Determine the precise day of the announcement and designate this day as zero.
• 3. Define the period to be studied. (people may know about it before the announcement - include up to 90 days before the event)
• 4. For each of the firms in the sample, compute the return on each of the days being studied.
• 5. Compute the “abnormal” return for each of the days being studied for each firm in the sample
• . 6. Compute for each day in the event period the average abnormal return for all the firms in the sample

We find that there is a drift towards the true price prior to the announcement ( failure of EMH semi-strong form) –> information isn’t know by everyone (asymmetric information) –> should see just a jump on announcement day like in figure 7

Question 22

Evidence of Strong-form efficiency and insider trading?

Answer 22

• The ability of insiders to trade profitability in their own stock has been documented in studies by Jaffe, Seyhun, Givoly, and Palmon
  
  • When insiders are buying/selling their stong it is positive/negative correlated to the price
  • Tricky to do as it requires them to collect information that’s not easy to get access to
• SEC requires all insiders to register their trading activity

Question 23

How do we interpret the anomalies in the empirical evidence for EMH?

Answer 23

• The most puzzling anomalies are price-earnings, small-firm, market-to-book, momentum, and long-term reversal.
• Fama and French argue that these effects can be explained by risk premiums.
  
  • Firms that outperform how they were expected to do –> people held them even though there was an added risk they were doing poorly in the markets general view
• Lakonishok et al. (1996) argue that these effects are evidence of inefficient markets.
  
  • analyst extrapolate past performance too far into the future and therefore overprice firm with recent good performance and underprice firms with recent poor performance
    
    • When market participants realise their errors, prices reverse (consistent with reversal effect, small firms and book-to-market effect)
    • Porta (1996) –> Analysts are overly pessimistic about firms with low growth prospects and overly optimistic about firms with high growth prospects. When these too extreme expectations are corrected the low growth firms tend to out perform high-expected growth firms
• Anomalies or Data mining?
  
  • Are these actually inefficiencies or unexplained puzzles in the financial markets
    
    • is it chance that if you ran tests over and over again to examine stock returns along enough dimensions, simple chance will cause some criteria to appear to predict returns
      
      • small-firm effect published in the 1980s disappeared for much of the rest of the decade
      • After research was published abnormal returns in stocks exhibiting certain ‘market-beating characteristics’ decayed by 35% (Mclean and Pontiff, 2016)
• Bubbles and irrationality
  
  • Human irrationality causes the mispricing itself
  • Game stop rally had nothing to do with fundamentals more of a statement to wall street