Criticism against Efficient Market Hypothesis

Get Complete Project Material File(s) Now! »

THEORETICAL FRAMEWORK

Technical analysis

The concept of technical analysis is over 100 years old, examples of this is the Dow Theory (Brock et al., 1992) and the Japanese Candlestick Charts, which was first used when speculating in rice prices (Nison, 1994).
Technical analysis refers to many different techniques with the purpose of attempting to forecast future prices from past prices, or patterns. Both Brock et al. (1992) and Achelis (2001) state that past patterns have a tendency to repeat themselves. A basic assumption concerning technical analysis is; that since stock prices are based on supply and de-mand, which means that the major interest is in the market and its behavior. So instead of spending time trying to predict demand or looking at production costs the technical analyst can focus on finding trends and patterns (Kirkpatrick and Dahlquist, 2006)
The other type of analysis, which is commonly known as the opposite of technical anal-ysis, is the fundamental analysis. While the technical analysis is primarily used for short-term investments, since it’s helpful predicting trends, the fundamental analysis is more used for long-term investments. This method focuses on finding mispriced stocks and predicting the value and the potential profitability of a security or a company based on earnings, P/E-ratios and other financial ratios (Bodie et al. 2011).
Even though technical analysis has been criticized for its potential to predict the future prices of the market (Malkiel, 2005), there are studies suggesting that not only could one predict future stock price movements using technical analysis, but also gain excess return compared to a simple buy and hold strategy. Metghalchi et al. (2008) and Chong & Ng (2008) are two examples of studies that emphasises positive performance of tech-nical analysis.

Efficient Market Hypothesis

To understand the significance of our thesis and the tests we will conduct, it is im-portant to describe the efficient market hypothesis (from here on referred to as EMH), developed by Eugene Fama (1970).
The EMH asserts that markets fully reflect all available information in security prices and thereby are efficient. When the markets are efficient and information is fully re-flected, securities will neither be over- nor undervalued, which means that one cannot expect any higher return than the one associated with the risk one are willing to take (Torssell, Nilsson, 2000).
The EMH consists of three different forms, or degrees, of efficiency. The Weak form of efficiency implies that prices fully reflect all information contained in historical prices, which would mean that the usefulness of technical analysis is rejected; Semi-Strong form implies that stock prices reflects both historical prices and all publicly available in-formation relevant to the market; The Strong form of EMH implies that all information known by any market participant is fully reflected in the prices (Malkiel, 1989).
Since EMH states that security prices fully reflect all available information (Fama, 1991), an important conclusion is that, even though there are some positive correlation in day-to-day price changes, it should not be possible to gain excess return through trad-ing systems since it would cause a too large amount of transactions which implies that even a small commission fee would hinder such an attempt (Fama 1970).
Malkiel (2005) claim that the robustness concerning the, alleged, predictable patterns is questionable. Malkiel (2005) also mention that the failure of most professionals should be evidence enough for market efficiency:
‘The strongest evidence suggesting that markets are generally quite efficient is that professional investors do not beat the market’ (Malkiel, 2005, p.2)
Malkiel’s (2003) conclusion is that, even if there are occational mispricing and anoma-lies in the stock market, those will most probably not last in the long-term due to the ef-ficiency of the market, in the end the belief of the stock markets efficiency will not be abandoned. Lee & Yen (2008) draws a similar conclusion – beliving in the concept of efficiency and its continuing, important, role in markets.

Criticism against Efficient Market Hypothesis

In the late 1970’s and through the 1980’s the findings concerning EMH was mixed, though studies from authors such as Watts (1978) and Chiras & Manaster (1978) argues that there are signs of inefficiency. On the other hand, Jensen (1978) believe that there is no reason for abandoning the idea of efficiency. During the 1990’s and the beginning of the 21st century the challenger of EMH is the idea of behavioral finance (Lee & Yen, 2008).
Among others; Brock et al. (1992), Chong & Ng (2008), Gunasekarage & Power (2001) and Metghalchi, et al. (2008) finds evidence of some predictive powers when using technical analysis, these proofs makes the EMH questionable since, as mentioned earli-er, it should not be possible to predict future price movements.
However, the EMH cannot be rejected until it is proven that identified trends can be used to earn risk-adjusted excess return (Torssell and Nilsson, 2000).

 High frequency trading

High Frequency Trading (HFT) is symbolized by a large amount of computer driven trades, that lasts from less than a minute up to a couple of hours, every day. The HFT trading holds a lower average return for each trade made. Due to volatile markets and overnight rate, High Frequent money managers usually reduce or eliminate their over-night holdings. These computer driven trades are taking advantage of occasional mis-pricing in the market, which pushes the markets towards efficiency (Aldridge, 2010).
A high frequency trader is trying to go undetected when trading, by using a highly de-veloped computer system, but there is still a risk for errors and that’s why supervision from a human is important (Aldridge, 2010).
HFT can be done from all over the world and a large number of firms using it are locat-ed in cities with large stock exchanges, for example Chicago, New York, London and Singapore. This gives them the advantage to develop fast trading strategies (Aldridge, 2010). According to Aldridge (2010), the competitors in this business are mainly hedge funds, investment banks and independent proprietary trading firms. The use of a strate-gy based on High Frequency Trading is associated with high costs and advanced tech-nonolgy, therefore we do not consider it an option for the individual investor.

READ  Game-theoretic model for control effects of embeddedness on trust in social networks

Technical indicators

Moving average

MA is the most commonly used instrument for technical analysis, the idea with MA is to identify a trend that could be used when trying to forecast future movement of a secu-rity. The method builds on a moving average of historical prices and is thereby a smoothing effect of the actual and more volatile security price curve. There are several different types of moving averages where the weights for historical prices are treated differently, mainly the difference between those are the weight that they put at the most recent data (Brock et al. 1992). The MA rule is a type of timing strategy and the per-formance of the model is often compared to a buy and hold strategy (Torssell, J. & Nils-son, 2000).
The calculation of MA reposes on an average of historical closing prices for the securi-ty. The length of the measurement periods vary depending on investors preferences for desired observation length, a long-term trend is often observed when measuring time se-ries between 100 and 200 days while popular time series for short-term trend often lies within an intraday average up to a 25 day time window. Time series in-between the long-term and short-term windows mentioned above are to be seen as intermediate lengths (Zhu, Y & Zhou, G. 2009)
According to the basic rules for moving averages, the idea is to look for crossovers, in most case one use dual or triple MA’s where different time series lengths are chosen. When those MA lines cross each other, signals are generated. The buy signal is ob-served when the short period average crosses above the longer period, and the sell sig-nal is generated when the short period crosses below the long period average (Torssell, J. & Nilsson, 2000). The MA rules are meant to exploit patterns in the security returns, this by giving a buy signal at the beginning of an upward trend, then investors are ex-pected to hold the security until a downward trend is identified where the MA rule gen-erates a sell signal.

Simple moving average

A simple moving average is constructed to assign all data points in the observation with same weights. It uses historical price data and creates an average by adding the desired length and then divides by the number of time periods.
!! ! »#$% = Simple moving average!
Where n denote our chosen number of time periods while price is the closing price for the security (Achelis, 2001).

Exponential moving average

In contrast to a simple moving average that assign equal weight to all observations, the exponential moving average is placing larger weight to the most recent data and expo-nentially decrease the weight of earlier observations. EMA is included in the study since it is used when calculating the values of Moving Average Convergence Divergence.
The weighting or smoothing constant assigns a number between 1 and 0 depending on the size of the sample; the value the weighting factor is given decides the importance of older observations. The larger the value the weighting factor is assigned, the more weight is put on most recent prices.
The smoothing constant is given by dividing two with n plus one, letting n denote num-ber of time periods (Achelis, 2001). ( ! − !)*W+ EMAy = Exponential moving average Once the smoothing constant is given the exponential moving average could be calcu-lated. Given the above equation, Closep signify the closing price, EMAy is the exponen-tial moving average of yesterday and W denote the smoothing constant (Achelis, 2001). Also for the first observation of the exponential moving average, let EMAy denote a simple moving average holding the chosen parameter length.

Moving Average Convergence Divergence

The MACD (Moving Average Convergence Divergence) indicator was developed by Gerald Appel, it is one of the most commonly used momentum indicators due to its simplicity to use. The model is basically built on moving average and measures a long-term MA against a more short-term one, the most common length is the 26-day EMA against a 12-day EMA. It shows how the moving average converges or diverges from the actual price trend. (Murphy, 1999)

1 Introduction and Background
1.1 Problem Statement
1.2 Purpose
1.3 Previous research
1.4 Limitations of the study
2 THEORETICAL FRAMEWORK
2.1 Technical analysis
2.2 Efficient Market Hypothesis
2.3 Criticism against Efficient Market Hypothesis
2.4 High frequency trading
2.5 Technical indicators
2.5.1 Moving average
2.5.2 Simple moving average
2.5.3 Exponential moving average
2.5.4 Moving Average Convergence Divergence
2.5.5 Relative Strength Index
3 Methodology
3.1 Deductive approach
3.2 Collection of data
3.3 Quantitative method
3.4 Stock selection
3.5 Observation period
3.6 Selection of technical indicators
3.7 Assortment of parameters
3.8 Basic calculation procedures
3.9 Risk adjustment and statistics
3.9.1 Sharpe ratio
3.9.2 Jensen’s Alpha
3.9.3 Statistics
3.10 Transaction costs
3.11 Out of sample test
3.12 The Backtesting of data in practice
3.13 Limitations of the methodology
4 Empirical Results
4.1 Result from out of sample prediction
5 Analysis and discussion
6 Conclusion
6.1 Proposal for further research
6.1.1 Different parameters and strategies
6.1.2 Different Swedish stock lists
References
Appendix
GET THE COMPLETE PROJECT

Related Posts