Hungary or Columbia

Technical Analysis of Columbia Financial Market Advocates of market efficiency generally believe that it is impossible for technical analysts to predict recurring-price pattern and technicians could not beat buy-and-hold strategy using technical trading rules. However, technical traders believe that there is some form of market efficiencies and technical traders can use price patterns strategy to beat buy-and-hold strategy. Objective of this paper is to investigate Columbia Stock Market and attempt to find technical trading rules to predict changes in the Columbia Stock Market Index.

The paper identifies trading strategy to be used to beat buy-and-hold strategy to enhance efficiency of Columbia stock market. Within the past few decades, the stock markets in the emerging economies have experienced rapid growth, which have attracted investors across the globe. With the increase in the price movements in the emerging economies, greater importance has been given to market efficiency.

While technical trading rules have been implemented to test market efficiencies in the emerging countries such as BRIC (Brazil, Russia, India and China), there is a paucity of research that confirms that technical trading rules could be applied in the Columbia stock market. This study evaluates different technical trading rules on Columbia stock market and determines whether changes in the Columbia market index could be predicted using technical analysis.

The remainder of the paper is structured as follows: Section II of the paper discusses literature review, while section III discusses the data and methodology, while section IV provides the empirical results on technical trading rules. Section V provides buy-and-hold strategy and section VI provides the conclusions. II. Literature Review Dominant theme in financial market since the last few decades is the concept of efficient market systems.

Fama (1970) defines efficient market system as the market, which its security price always reflects available information and any new information simultaneously and quickly reflect in prices. Moreover, news of any company always unpredictably arrive randomly leading to price changes unpredictably or follow a random walk. According to Fama perspectives, there are three forms of EMH (Efficient Market Hypothesis): a) the weak form, b) the semi-strong form, and c) the strong form.

Supporters of the weak-form market efficiency show that investors will not be able to drive up profits above buy -- and hold strategy using trading rules, which depend solely on past market information such as volume or price, and the implying technical trading, will be useless. More than three decades of research shows that financial economists, academic researchers and practitioners have not yet reached a consensus whether technical trading rules could yield profitable trading results. However, overwhelming number of economists and financial analysts support the "weak-form" efficient market hypothesis.

While the earlier research strongly supported the random walk hypothesis, however, the semi-strong form of Efficient Market Hypothesis has been the basis of most empirical research. Early results from the literatures reveal that profitability derived from the technical trading was overwhelmingly negative. For example, Fama and Blume (1966), Alexander (1964), Larson (1960), Granger and Morgenstern (1963), Van Horn and Parker (1967), Jensen and Benington (1970), Mandelbrot (1963), Fama (1965), and Osborne (1962) all supported the weak form of market efficiency. Since the beginning of the1990s, financial analysts have used technical trading rules to evaluate financial market performances.

In 1990s, technical trading rules have enjoyed a renaissance in both academic circles and Wall Street. Several papers deliver evidences that simple predicting trading rules are very effective in predicting stock returns. Contrary to fundamentalists who use balance sheet or income statements of a company to predict stock returns, technical analysis is based on the assumption that past volume, prices and many other indicators could be used to detect future price movements.

The art of technical analysis is to identify changes in prices and maintain investment postures using the trade signals. (Pring, 1991). Murphy (1999) points out that technical analysis is the study of market actions such as volume and price to forecast future price actions. To test power of technical trading rules, stock technicians use the WFEMH (Weak-Form Efficient Market Hypotheses) using past returns to determine random walk testing. Following the test, technicians use various trading rules to predict profits using buy-and-hold strategy.

While many studies have investigated whether technical rules could be used to provide a superior investing performance, however, the most comprehensive recent study of Brock, Lakonishok, and LeBaron (1992) (BLL) using 90 years of daily stock prices discovered 26 technical rules that had been applied to Dow Jones Industrial Average and significantly outperformed a holding cash benchmark. BBL analyzed moving and Dow John Index from 1897 to 1985 using various short and long run moving average of prices in order to generate buy and sell signals.

The authors tested 50, 150,200 days long moving averages and short moving averages of 1,2,5 days. Their findings showed that stock market is not efficient when using the weak form efficient market hypothesis and presented findings to show that technical rules have predictive power. The researcher such as Bessembinder and Chan et al. (2006) also used BBL's moving average to investigate whether it is possible to predict stock market indices using some simple form of technical analysis. Bessembinder et al.

(1995) conclude that BBL technical rules are very effective in predicting stock price movement in South Korea, Japan, Hong Kong, Taiwan and Thailand. Moreover, Ergul, Holmes and Priestley (1997) use the prices of 63 stocks traded daily on the Istanbul Stock Exchange to evaluate the technical trading rules. The authors conclude that technical analysis can be used to predict stock returns, which cannot be predicted with previous analysis.

Pruitt and White (1998) also use the University of Chicago's Research in Security Prices (CRSP) daily data tapes between 1976 and 1985, and their findings reveal that technical trading rules are very effective in outperforming a simple buy-and-hold strategy after accounting for transaction costs. Bessembinder and Chan (1998) also confirm the basic BLL results by pointing out that the BLL findings lie parallel with the concepts of market efficiency even after accounting for the transaction costs. Ratner and Leal (1999) also support the predictive power of stock return of technical trading rules. (Gencay 1998a, 1998b).

Kwon and Kish (2002) in their own case apply three popular technical trading rules to NYSE index of between 1962 and 1996. The authors conclude that the "technical trading rules have the potential to capture profit opportunities over various models when compared to buy and hold strategy." (Metghalchi. Glasure, and Garza-Gomez, 2011. p.4). However, Ready (2002) points out in his recent study that the apparent success of the BLL moving average rules is due to a spurious result of data snooping and the issue should not persist in the future.

Researchers have also applied the technical trading rules to foreign exchange markets. Taylor and Allen (1992) point out that technical advice could be self-fulfilling for foreign exchange markets. Mengoli (2004) also collects data of the listed securities from the Italian Stock exchanges over the period of 1950- 1995 and conclude that technical trading rules are very effective to determine market profitability.

Neely, and Weller, (2011) argues that 30% of the traders in the foreign exchanges markets in the United States are being implemented by the technical trading rules and there have been continuous used of technical trading rules for foreign exchange market. However, some studies do not support technical trading rules. Szakmary, Davidson, and Schwarz (1999) apply technical trading rules on stocks listed in Nasdaq index and conclude that individual stock perform poorly using technical trading rules strategies, however, the Nasdaq index earn satisfactory abnormal returns.

The author believes that abnormal returns can still be insignificant since high level of transactional costs is generally associated to NASDAQ trading. Sullivan, Timmermann and White (1999) find no evident of profitability after applying the technical strategies to the S & P. And Dow Jones Industrial Average. Coutts and Cheung (2000) apply technical strategies on the Hang Seng returns over a period of 1985 -- 1997 and conclude that both the trading breakout rules and moving average fail to provide net transaction costs and positive abnormal returns.

III Data and Methodology The study collects data from Datastream's daily index prices of the Columbia stock market from October 12, 2001 to October 12, 2011, and computes daily returns based on the changes in the stock index level. To determine daily interest rate, the study uses DataStream's daily Columbian interbank one month and Columbia Treasury three-month rates. Technical analysis is based on the assumption that prices move in trends and determined by the traders' changing attitude following the political, economics, and psychological forces.

In other words, technical analysis uses the past behavior of market data to assist in making trading decision in financial markets. The historical price data generally assist in making these decisions. As being put forwards by Pring (1991), the technical analysis is a reflection that prices move in trends and generally being determined by changing attitude of investors. Typically, technical approach is based on the idea historical data can be used to forecast future price movement based on the monetary, economic, political and psychological forces.

Pring (1991) further argues that the goal of technical analysis is to identify changes in trends and maintain investment posture until there is a reversal in trends. Major Trend-Determining Techniques is based on buy (sell) signals emitted when the price of index moves below or above the local minimum or maximum. This study uses both short-term moving and long-term moving average.

The short-term moving average is 9,and 12 and 26 and long-term moving average is 50, 100, and 200 day's local minimum or maximum and the strategy to employed in this study is as follows: We are in the market when a buy signal is emitted and when the price level has broken out above the local maximum, We out of the market and exit the long when a sell signal is emitted and the price level has moved below the local minimum.

Based on the technical trading rules, the study can be "either in the market (buy days) or out of the market (sell days)." (Metghalchi. et al., nd. p.4).The methodology reveals that the study follows the break out rules, which assists in observing the price movement in other to make a trading decision.

If we are out of the market and our closing price is above the local maximum, then trader should in the market the following day by buying the index at today's closing price (Next must be the day to buy). Thus, the next day's return is going to be difference between the logarithms of today's closing price and the log of the closing price of next day.

Once we are in the market, the position is a buy day until the closing price is below the minimum of the local price. In that day, the next action is to sell the price index at the closing price and be out of the market the next day, which is the sell day. Meanwhile, we continue staying out of the market as long as price is not above the local maximum.

Thus, the study defines mean sell, X(s), mean buy, X (b), and returns as revealed in the following equations: The equation 1 defines the moving average. MAt (M)= 1 M-1 (1) M ? Pt-i i=0 1 X (b) = N (b) ?Rb (2) In the equation 2, when the system generates sell signal we are out of the market however, we are in the market when the system generate the buy signal. Thu, some days we are in the market and some days we are out of the market.

In the equation 2, we define the mean of buying days and average returns of the days in the market. 1 X(s) = N(s) ?Rs (3) The equation 3 defines the average returns of sell days. As being revealed in the equation 1 and 2, the N (b) are total number of buy days and N(s) are total number of buy sell days. Moreover, Rb are daily returns of buy days and Rs are daily returns of sell days.

The study "tests whether the returns of any trading breakout rules are greater than a buy and hold strategy and whether the mean buy is different than the mean sell." (Metghalchi, et al., nd, p. 5) specifically based on the following hypothesis: H0: "X (b) - X (h) =0, X(s)-X (h) = 0, X (b) -- X(s) =0" HA: "X (b) -- X (h) ? 0, X(s) -- X (h) ? 0, X (b) -- X(s) ? 0" "Where X (h) is the mean return for the buy-and-hold strategy.

The test statistic for the mean buy returns over the mean buy -- and hold strategy is": (Metghalchi, et al., nd, p. 5) Equation 4 is the test statistics of two mean that is comparing means to reveal whether the mean of buy and mean of sell different from each other. X (b)- X (h) t= ( 4) Var (b) / Nb+ Var (h) / N Where Var (b) is the variance of buy and Var (h) is buy-and-hold returns.

The formula presented above is used to test the mean sell returns and the mean buy-and-hold strategy. The formula is also used to test the mean of buy returns over the total mean of sell returns, and this is done by replacing appropriate variables in the t-statistic formula. Moreover, the study uses the Exponential Moving Average (MVA) and MACD (moving average convergence/divergence) for our methodology.

The MACD is employed to spot changes direction, strength, duration and momentum of the stock prices; however, EVA is similar to the simple moving average except that more weight is given to the latest data. IV - Empirical Results The study presents the average returns of the daily interest rates of the Columbia money market from October 12, 2001 to October 12, 2011 and the interest rate return is 0.000177571, however, the average return for buy-and-hold for the entire period is 0.001070314 (.0107% per day) with the Standard Deviation of buy-and-hold is 0.0140.

Thus, the t-value of the buy and hold for the entire period (2608 observations) is equal to 390 (0.00107 divided by 0.0140 / ?2608). The annual average for the entire period is 14.07%. The table 1 presents the summary result of the standard moving average based on the trading rules. The rules report the mean returns of sell days, mean returns of buy days as well as standard deviation of returns of sell, and buy days.

Based on the t-statistics in equation 4, the results test the different of mean sell and mean buy from "from the unconditional 1-day mean, and buy-sell from zero." The study presents the results of the breakout rules in table 1, and the rules (test) for the breakout rules are the maximum of 9, 12, 18, and 26. In the study, the Test (9-26) means we are going to be in the market if the prices are above Maximum of 26 days and stay in the market till the stock prices are below minimum of 9 days.

When the prices are below the local minimum of 9 days, we close the position and move out of the market and until the prices are above the maximum of 26-day. For each test rule, the study reports mean returns for buy days and mean return for sell days as well as the standard deviation (SD) of returns on both buy and sell days with the total number of buy and sell days.

As being revealed in equation 4, the t-statistics reveals the number in the parentheses, which tests the difference between the mean buy and mean sell. The first row in the table 1 reveals results based on the breakout rules (9, 9). We are in the market (buy days) when price level is above the local maximum of 9 days, however, we are out of the market if the price is less than local minimum of 9 days.

Thus, we will be in the market as far as the price is above the minimum of the last 9 days. However, immediately the prices are below the minimum of the last 9 days, we exit the market meaning that we will be out of the market (sell days). When the price is above the maximum of 9 days, we immediately re-enter the market and stay in the market unless the price is below the minimum of the past 9 days. In the last row of Table 1 are the results of trading rule (26, 9).

The rule is that we will remain in the market if the price above local maximum of 26 days (sell days), however, if the price level is below the local minimum of 9 days. In other words, we will remain in the market, when the price greater than the minimum of the last 9 days.

However, when the prices are below the minimum of 9 days the next action is to move out of the market (sell days) and we will remain out of the market until the price is less than maximum of the past 26 days. Moreover, when the price is above the maximum of the past 26 days, we move back to the market and remain in the market until the price above the minimum of the past 9 days.

Table 1: Statistical Results for Standard Moving Average and MACD Rules The results present daily data from October 12, 2001 to October 12, 2011. The moving average of short and long percentage different is to generate signal. The Nb and Ns are the buy and sell signal that have been reported in each period. Moreover, the SDb and SDs denotes the standard deviation of both buy and sell signals, respectively.

More importantly, the numbers in the parentheses provides the t-statistics testing of the difference of the mean buy and mean sell "from the unconditional 1-day mean, and buy-sell from zero." Numbers being marked with asterisks are the significant of 5% level for a two-tailed test.

Rules Mean Buy Mean Sell Buy-Sell SDb SDs Nb Ns MACD & MA9 (9, 9) 0,00460223 -0,001293 0,00589534 0,010426033 0,015409396 (8.27)* (-4,95)* (11,57)* MACD & MA9 (18, 18) 0,0035 -0,0009 0,0044 0,010515827 0,015884711 (5,865)* (-3.9425)* (8.450)* RSI>(50) 0,002148 - 0.00084 0,00298 0,011414827 0,01749 (2.75)* (-2.99)* (4.665)* Our results in Table 1 are amazing revealing that all the buy-sell difference are positive and the t-statistics are very significant thereby rejecting the null hypothesis. All the buy returns also have the significant value of rejecting the null hypothesis. The same is for the mean sell returns; all our mean sell returns significantly reject the null hypothesis.

The standard deviation of sell days and buy days are similar with those of sell days. Revealing that the volatility is the same for both buy and sell days period. The results in Table 1 show that buy -- sell difference are positive in Column 4 of the Table 1 and the t-statistics of the difference are positive showing that t-statistics is significant thereby rejecting the null hypothesis of equality with zero.

In the Columbia financial market, technical analysis, and the trading breakout rules have predictive powers, and based on predictive power, the study designs trading strategies to surpass the buy-and-hold strategy. V Trading Strategies The study provides a degree by which traders could use the technical trading rules to earn trading profits, which could be used to beat buy-and-hold strategy.

Based on the fact that the mean buy is greater that the mean sell and the profitability of the technical analysis is based on the trading strategies and the position that traders should take when the trading rules emit sell signals. Typically, the traders are not to invest when the system emits sell signal, the trader will earn zero profits based on the mean return of (Nb/N)*X (b) + (Ns/N) *0. To be beat buy-hold-strategy, the study suggests a simple strategy by which the traders will be in.