Lect06EffMark

							Lecture 6: Efficient Markets and
       Excess Volatility
The Efficient Markets Hypothesis
•   History of the Hypothesis
•   Reasons to think markets are efficient
•   Reasons to doubt markets are efficient
•   Technical analysis
•   Empirical evidence in literature
•   Homework assignment and regressions
    Earliest Known Statement
“When shares become publicly known in an
 open market, the value which they acquire
 there may be regarded as the judgement of
 the best intelligence concerning them.”
     - George Gibson, The Stock Exchanges
 of London Paris and New York, G. P.
 Putnman & Sons, New York, 1889
       Intuition of Efficiency
• Reuter’s pigeons and the telegraph
• Beepers & the internet
• Must be hard to get rich
      Textbook Version Today
As one of the six most important ideas in
  finance:
“Security prices accurately reflect available
  information, and respond rapidly to new
  information as soon as it becomes
  available” Richard Brealey & Stewart
  Myers, Principles of Corporate Finance,
  1996
         Harry Roberts, 1967
• Weak form efficiency: prices incorporate
  information about past prices
• Semi-strong form: incorporate all publicly
  available information
• Strong form: all information, including
  inside information
      Price as PDV of Expected
              Dividends
• If earnings equal dividends and if dividends grow
  at long-run rate g, then by growing consol model
  P=E/(r-g), P/E=1/(r-g). (Gordon Model)
• So, efficient markets theory purports to explain
  why P/E varies across stocks
• PEG ratio is popular indicator = g’/(P/E), where
  g’ is short-run growth rate; popular rule of thumb:
  buy if PEG<0.5
• PEG rule of thumb makes sense only if g’ bears a
  certain relation to g; not a sensible rule.
• Efficient markets denies that any rule works
Reasons to Think Markets Ought
        to Be Efficient
• Marginal investor determines prices
• Smart money dominates trading
• Survival of fittest
Reasons to Doubt these Reasons
• Marginal investor: wealth matters
• Smart money: matter of degree. Limits to
  arbitrage theory
• Survival of fittest: life cycle renews
        Psychological Factors
• Gambling behavior
• Overconfidence
• Slowness to make money, futility of career
  trying to prove others of one’s ability
• Siegel and Peter Lynch
 Popular Doubters of Efficiency
• Peter Lynch: Elementary school children
  beat professionals
• Beardstown Ladies
• Robert Kiyosaki Rich Dad, Poor Dad
• Motley Fool
         Raskob on the Market
“Suppose a man marries at the age of twenty-three
  and begins a regular saving of fifteen dollars a
  month – almost anyone who is employed can do
  that if he tries. If he invests in good common
  stocks and allows the dividends to accumulate, he
  will have at the end of twenty years at least eighty
  thousand dollars. . .I am firm in my belief that
  anyone not only can be rich but ought to be rich.”
  John J. Raskob, Ladies Home Journal, 1929
        Raskob’s Calculation
Annuity formula (converted to terminal value)
 shows that Raskob assumed 26% per year
 returns:

                   15(10196) 240
                       .             15
          80,000                
                      .0196        .0196
          Technical Analysis
• Robert D. Edwards & John Magee,
  Technical Analysis of Stock Trends, 1948.
• Hand drawing of charts, judgmental
  interpretation of patterns
• Difficult to test success of technical analysis
• Harry Mamaysky, SOM finds some success
  in their methods.
     Head & Shoulders Pattern
• Initial advance attracts traders, upward
  momentum. Smart money begins to distribute
  stock, trying not to kill demand.
• Eventually downturn, but smart money comes in
  to support demand, manipulation. (left shoulder)
• Upward momentum resumes, ends when smart
  money has distributed all shares; market drops.
• New traders try to exploit well-known tendency to
  rally. New weak rally, right shoulder, then a
  breakout. (Edwards & Magee)
    Random Walk Hypothesis
• Karl Pearson, Nature, 72:294, July 27,
  1905. Aug 10, 1905, walk of drunk
• Burton Malkiel, A Random Walk Down
  Wall Street, 1973.
 Random Walk & AR-1 Models
• Random Walk: xt=xt-1+t
• First-order autoregressive (AR-1) Model:
  xt=100+(xt-1-100)+t. Mean reverting (to
  100), 0< <1.
• Random walk as approximate implication of
  unpredictability of returns
• Similarity of both random walk and AR-1 to
  actual stock prices
Random Walk & AR-1(=.95)
    115


    110


    105
                                                           Random Walk
x




                                                           AR-1
    100


    95


    90
          1

              6

                  11

                       16

                             21

                                  26

                                       31

                                            36

                                                 41

                                                      46
                            Time Period
          Obvious Examples of
              Inefficiency
•   Jeremy Siegel – Nifty-fifty did well
•   Rebalancing
•   Most closed out
•   Polaroid and Edwin Land
                Tulipmania
•   Holland, 1630s.
•   Peter Garber, Famous First Bubbles
•   Mosaic virus, random-walk look
•   Free press began in Holland then.
           Dot Com Bubble
• Toys.com: Had disadvantage relative to
  bricks & mortar retailers starting web sites
• Lastminute.com: travel agency, sales in
  fourth quarter of 1999 were $650,000,
  market value in IPO ins March 2000 was $1
  billion.
   Problem Set #3: Forecast the
            Market
• Step 1: Get stock price data on spreadsheet,
  as from yahoo.com.
• Step 2: Create new column showing
  percentage price changes
• Step 3: Create new Column(s) containing
  forecasting variables
• Step 4: Test for significance and interpret
  results.
 Significance Test in Regression
• Use the R2 which is the fraction of the
  variance of the dependent variable that is
  explained by the regression.
• Compute F statistic (k, n-k-1 degrees of
  freedom, and check that it is above critical
  value for significance at 5% level.
• Issues of data mining, etc.
                 F Statistic
• F statistic with k, n-k-1 degrees of freedom,
  where k = number of independent
  (forecasting) variables and n = number of
  observations:

                           R2 / k
                F
                   1  R2  / (n  k  1)
      Regression Output - Excel
•   Intercept, X Variable, X Variable
•   T statistic, P value
•   F statistic, P value
•   R squared