Technical Analysis

FINANCIAL MARKETS (AF355) LECTURE 6

RETURNS TO TECHNICAL ANALYSIS

ISSUES DISCUSSED

Returns to technical analysisMarket overreactions

Moving average oscillators and trading range break-outs Reconciliation of test results

Dr. Pascal Frantz: Lecture 6, AF355 2

TESTS OF RETURNS TO TECHNICAL ANALYSISDoes The Stock Market Overreact? (1)

De Bondt and Thaler [BT, (1985)]

Hypothesis

Investors tend to overreact to unexpected and dramatic news events




Test

• Based on their market-adjusted cumulative abnormal returns, stocks are allocated to portfolios:– The top 35 (or the top 50 or top decile) are assigned to the

winner portfolio and the bottom 35 (or the bottom 50 or bottom decile) are assigned to the loser portfolio;

• The performances of the extreme portfolios are measured over 3 years;

• BT derive the differences in market-adjusted cumulated average returns between the loser and winner portfolios over time.







Results

• The results of the test are consistent with the overreaction hypothesis:

Over the last 50 years, loser portfolios of 35 stocks outperform winner portfolios of 35 stocks by about 23% in the 3-year period following the formation of the portfolios;

• Most of the trading gains come from the loser portfolios (+19.6%);

• Most of the market correction occurs in years 2 and 3: the difference in performance between extreme portfolios is not statistically significant after 1 year;

Story not dissimilar to Lakonishok, Shleifer, and Vishny (1994);


TESTS OF RETURNS TO TECHNICAL ANALYSISSimple Trading Rules and the Stochastic Properties of Stock Returns

Brock, Lakonishok, and LeBaron [BLL, (1992)]

Simple Trading Rules: Momentum Strategies (1)

Fama and Blume (1966) filter rule

“Using price history, buy a stock if the price rises x%, hold it until the security falls x%, then sell and go short. Maintain this short position until the price rises x%, then cover the short position and establish a long position”

Momentum strategy!


TESTS OF RETURNS TO TECHNICAL ANALYSIS Fama and Blume (1966) filter rule






Moving average oscillator

• According to the moving average rule, "buy" and "sell" signals are generated by 2 moving averages of the level of the market index:– A long-period average;– A short-period average;

• This strategy calls for buying when the short-period moving average rises above the long-period moving average and for selling when the short-period moving average falls below the long-period moving average;




Simple Trading Rules: Momentum Strategies (2 - cont)


• The moving average decision rule is often modified by introducing a band around the moving average in order to reduce the number of buy and sell signals by eliminating "whiplash" signals when the short and the long period moving averages are close;






• Two main variants:– The variable length moving average (VMA) method;

– The fixed length moving average (FMA) method;





Variable length moving average oscillator

• The VMA method calls for:– Staying long as long as the short moving average stays

above the long moving average (adjusted for the band width);

– Staying short as long as the short moving average stays below the long moving average (adjusted for the band width);





Fixed length moving average oscillator

• The FMA method calls for:– Going long for a fixed number of days following the

moment when the short moving average cuts (from below) the long moving average (adjusted for the band width);

– Going short for a fixed number of days Following the Moment when the Short Moving Average Cuts (from above) the long moving average (adjusted for the band width);





Trading range break-outs • According to the trading range break-outs rule:

– A "buy" signal is generated when the price penetrates the “resistance level” defined as the local maximum price;

– A “sell” signal is generated when the price penetrates the "support level" defined as the local minimum price;

In essence, technical analysts recommend buying when the price rises above its last peak and selling when the price sinks below its last trough;





Trading range break-outs

The underlying rationale is that the price has difficulties penetrating the support level because many investors are willing to buy at the minimum price and that the price has difficulties breaking through the resistance level because many investors are willing to sell at the maximum price;





Results

Evidence of positive returns (validation of trading rules gross of trading costs);

But does not come cheaply (high trading costs)!

Buy signals consistently generate higher returns than sell signals;


RECONCILIATION OF TEST RESULTS (1)The Paradox

• De Bondt and Thaler: Overreaction;

Contrarian strategy: buying losers and short-selling winners;

• Brock, Lakonishok, and leBaron: Trend extrapolation (moving average oscillators);

Mainstream (momentum) strategy: buying winners and selling losers;


RECONCILIATION OF TEST RESULTS (2) Jegadeesh and Titman (1993) Poterba and Summers (1987)

• Jegadeesh and Titman (1993) obtain excess returns from buying winners and selling losers over periods of 3-12 months (momentum strategies). Subsequent excess returns are negative;

• Apparently contradictory investment rules may be reconciled by evidence from Poterba and Summers (1987): stock returns are positively autocorrelated over short horizons but negatively over long horizons;


RECONCILIATION OF TEST RESULTS (3)Positive Feedback Investment Strategies and

Destabilizing Rational Speculation

Delong, Shleifer, Summers, and Waldman (1990)

Summary

• Feedback traders buying winners cause prices to overreact leading eventually to a sell-off by fundamentalists!


RECONCILIATION OF TEST RESULTS (3 -cont)Positive Feedback Investment Strategies and



Model

• Stylised model with 2 types of agents:

– Fundamentalists, getting signals about intrinsic values;

– Chartists, learning indirectly about intrinsic values from prices;


RECONCILIATION OF TEST RESULTS (3 -cont)Positive Feedback Investment Strategies and



Dynamics

• A good signal received by fundamentalists causes prices to increase;

• When chartists observe the rise in prices, some chartists start to buy, causing further price increases, causing more chartists to buy;

• Eventually, share prices are so much above intrinsic values that fundamentalists sell;