Mutual Fund Specifics - Dissertation Example

Introduction A “mutual fund” is a scheme for investment which is professionally run by pooling money from many investors to buy securities. The mutual funds firm management undertakes market analysis that enables the fund to make a decision on asset allocation to invest depending on the expected return and risk of the assets. The decision that the managers of the firms take has in the past raised questions on how their analysis are able to beat the market consistently. To out perform the market returns, the manager is expected to demonstrate the ability to make correct market timing and market selection. Market selection is when the manager is able to select among the assets traded in the market the lowly priced asset and sell it at a higher price in the future due to its rise in returns. Market timing funds in most cases moves towards highly concentrated industry, fund which are large and align to small-cap stocks. The decisions that some managers do make sometime do outperform the market while at times they underperform. This has raised the debate whether the manager’s performance is guided by luck or skill in the manner under which they arrive at decision making. This paper undertakes to investigate how a number of mutual funds analyses have faired in their performance in the past years from a given data of selected fund firms. Finance literature has two contrasting strands on how optimal asset allocation is arrived at. On one hand, the argument has being that aggregate returns on the stock market are predictable and thus, investors are able to reach at optimal asset allocation based on the predictability strand. In contrast, argument has being that there is minimal evidence that investors utilize the predictability of aggregate stock market returns in their asset allocation. Investors in the past have being interested in funds that have large annual returns like the case of Fidelity Magellan mutual fund which outperformed S&P 500 index for 13 years in its 11 indexes from 1977 to 1989 under the stewardship of Peter Lynch. However, a number of funds making outstanding profits have being collapsing and investors are in the present days interested in other dynamics of fund performance. The problem has being the difficulty an investor faces in choosing the right manager to out perform the market and maintain. This paper undertakes a research that focuses on performance of some mutual funds by market timing and security selection. Market timing means that the manager has the ability to predict price changes of securities and thus, they invest or withdraw funds in a timely manner from an investment. Security selection on the hand means that the manager has the ability to identify and select lowly priced securities that will provide returns in the future. 2. Literature Review Literature that has dwelled on evaluating performance of mutual funds has being very successful in the foundation of modern days theory on portfolio and how assets are valued (Guerard, 2009). The investors understanding on how to compile a portfolio by taking care of risk and returns has being contributed by Markowitz (1959) and Sharpe (1964). An investor will select a portfolio currently that is able to produce returns later. Sharpe (1964) in analyzing 34 open-ended mutual funds found out that the capital market efficiency is usually high. Also, managers are more interested in evaluating risk and engaging in diversification instead of evaluating on mispriced securities. According to Sharpe (1964), an investor is able to achieve any return on assets along the capital market line if he/she undertakes primary diversification at equilibrium because capital asset will have adjusted. This is because investors avoids risks in selecting among portfolios and are only concerned by mean and variance of their investment. The expected return can be maximized by undertaking additional risk on the holdings. Thus, in the market there will be two prices of interest rate and the risk price and for additional return per unit is as a result of additional risk per unit. A method referred as a reward to volatility was developed by Treynor (1965) to measure portfolio performance. According to contributions of Treynor (1965), manager’s performance risk adjusted is derived by dividing the portfolio return less risk free rate by portfolio beta (portfolio return – risk free rate)/ portfolio beta. This contribution was followed by Sharpe (1966) proposal that the manager is supposed to be penalized for failure to diversify fully by dividing the portfolio return less risk free by portfolio volatility ((portfolio return – risk free rate)/ portfolio volatility. The two ratios have being generally labeled as Treynor ratio and Sharpe ratio respectively. Treynor and Mazuy (1966) developed a model that is able to test mutual funds manager’s ability of timing the market and security selection. The finding of their research was that no mutual fund performance is able to beat the market always. The research was based on 57 mutual funds in 1953 to 1962 which revealed that an investor move to mutual funds is pegged on market fluctuation. Treynor and Mazuy suggested that an increase in the rate of return in mutual funds is more likely to be influenced by the manager’s ability to identify lowly priced securities rather than out guess the market returns. Later on, Jensen (1968) carried out a research on mutual funds based on Capital Asset Pricing Model and came out with an absolute measure that reiterated that no mutual funds are able to give abnormal returns if the transaction costs are considered in the computation. He analyzed 115 funds in a period that run from 1955- 64 and discovered that on average, mutual funds less expenses cannot be able to forecast security prices that will enable them to beat the market benchmark. The analysis also, revealed that there is little evidence that any fund is able to perform significantly better than any other if it is picked randomly. This model assumes that risk is stationary over time and thus, it does not put in to consideration the presence of market timing by fund managers (Anderson and Ahmed, 2005). There however, in some literature it is noted there is potential bias that can occur if there is the presence of market timing. Grinbaltt and Titman (1989) conducted a research that was able to show that market timing success results in systematic risk that tend to be biased upwards and the constant risk is biased downwards. According to Treynor and Mazuy (1966), it is not appropriate for fund managers to go by a linear model approach when they are predicting market condition changes. A quadratic function is the one which is supposed to be used by managers holding large or small portfolios of risky assets and the market prediction moves upward or downward. Lehmann and Modest (1987) argument was that market timers are supposed to generate money whenever the market rises or it goes down because the shared market return is large. The argument is that market timers are able to increase or decrease holdings of a market portfolio when the returns in it are high or low due to their forecasting ability. Treynor and Mazur measure though, has some shortcoming because it cannot be able to evaluate in separation the impact of security selection and ability to time funds performance. A test that performed by Chua and Woodward (1986), in Canada, United States and Britain showed that the performance of mutual funds that were based on market timing registered a poor performance (Knight and Satchell, 2002). According to Bello and Janjigian (1997), in their research on 633 mutual funds between 1984 and 1994 they noted there is positive selectivity and market timing ability (Jelicic, 2010). Also, Daniel, Grinblatt, Titman and Warner (1997) suggest that in aggressive growing mutual funds there is an element of selectivity ability but they lack the market timing ability (Smith and Shawky, 2011). 3. Research hypothesis The research has two tailed hypothesis of a null hypothesis and an alternative that helps in arriving on whether mutual funds do deliver alpha. The null hypothesis is that abnormal performance does not exist while the alternative hypothesis is that abnormal performance exists in mutual funds. To test the hypothesis the significance level has being set at 5%. In rejecting or failing to reject the null hypothesis, statistical errors of error 1 and error 2 are to be observed. Error 1 in the context of the null hypothesis of our case is coming up with a finding that shows that abnormal performance exists in mutual funds when in reality it may be untrue. Type 2 error will be coming up with finding that shows that abnormal performance in mutual funds do not happen when in reality they do happen. 4. Research methodology The data that is utilized in making the finding consists of monthly returns of 38 listed mutual funds whose running period dates from 31st January 1990 to 31st December 2008. A one month London Treasury bill has being used as the risk free rate benchmark and monthly weighted returns of all London stocks index for market returns. The performance test of the funds managers is to be tested by investigating their ability in market timing and selectivity. To reveal their capacity in the two performance elements, Jensen (1968) model and Treynor and Mazuy (1966) model are to be used in the performance analysis. Jensen (1968) model computation is as follows E(Zp) = alpha + zf + Beta* [E(Zm) – zf] Whereby; E(Zp) is the expected return, zf is the risk free rate and [E(Zm) – zf] is the excess return of the market portfolio. The intercept which is alpha is supposed to be equal to zero in Jensen’s model. If the alpha value is greater than zero, it means that the portfolio manager possesses selective ability while a lesser alpha than zero implies an inferior manager. Jensen study divided performance of funds in to two different dimensions as follows. (1) Performance as a result of manager’s ability to increase the returns due to successful market prediction of future prices of securities. (2) Performance as a result manager’s ability to minimize the insurable risk through efficient diversification. Treynor and Mazuy (1966) model of assessing selectivity and timing ability is computed in a quadratic equation as following; rpt = ap + birmt2 + b2rmt2 + ept Where; rpt is the excess return earned on portfolio p in excess of the risk-free-rate during period t covered, ap is the estimated selectivity performance, bi is the portfolio’s estimate of systematic risk, rmt is the excess return of the market portfolio over the risk free rate during period t, b2 is the estimated indicator of the market timing performance and ept is the residual excess return on portfolio p during period t. Under this model, when ap is a positive value it is an indicative of selectivity ability while a positive b2 is an indication of market timing ability. This is because the characteristic line is expected to become even steeper as the market portfolio excess returns gets larger. Lack of market timing by the manager is indicated by a negative b2 and if the b2 has an insignificant value, it is an indication of lack of timing or failure of the manager to time the market. Of recent, a research has being conducted by Otten and Bams (2002) in 506 funds from France, Germany, Netherland and United Kingdom from 1991 to 1998. The point of the study is to find out the performance of mutual funds based on Carhart (1997) 4 factor model. The research discovered that the expenditure ratio and age of a fund are indirectly related while the assets of a fund and risk adjusted performance is positively related. Also, the study discovered that UK funds market have a significant positive ? net of expenses and for gross returns. The study makes a conclusion that small stock capitalized mutual funds and portfolios have outperformed the market benchmark in countries such as UK, Netherland and in France. 5. Data collection and analysis Data involved in measuring the performance capability of the mutual funds is from 38 mutual funds returns listed in London and the London Treasury bill return is used as a benchmark of risk free return. The performance of the funds will be analyzed by use of least square regression that is based on Jensen’s alpha model. Jensen alpha; Jensen’s model based on CAPM is used to analyze the alpha. The hypothesis is to be described as following; H0: ? = 0 H1: ? ? 0 From the regressions, there are only 18 alpha values that show they are statistically different from zero at significance level of 5%. This accounts for 47% out of the total 38 mutual funds. The conclusion that can be drawn from this is that there are funs that can be able to outperform the passive benchmark but there are still a majority of mutual funds that cannot be able to out perform the benchmark. This cannot give definite conclusion since the Jensen is based on assumptions of CAPM and it relies on the benchmark given. Assumption tests To validate the model used on the bases of OLS regression, one is supposed to test whether the assumptions in the assumption are met. The assumptions in the regression model are; (1) Normality of term error distribution. (2) Error terms independence (3) Error terms have constant variance which means that the error term is supposed to display a random variable with mean 0 and the same variance. (4) A linear relationship between depended variables and also, in independent variables. Also, to validate the results of Jensen’s alpha, an analysis of assumptions surrounding OLS regressions has being carried out in each fund. From the regression computation the standard error of the estimate represents the accuracy that is expected from the prediction. Treynor and Mazuy In Treynor and Mazuy analyses there are 19 alpha values that are more than zero which represents more than 50% of the 38 mutual funds. This means more than half of the funds are able to time the market. Also, in the regression 13 of the funds have beta value which is more than one which means that less than 50% of the funds are able to out perform the market by security selection ability. 6. Findings Market timing; under the Jansen (1968) model out of the 38 mutual funds analyzed, 17 of them have a positive alpha which indicates 50% of the funds mangers exhibited market timing ability. Under Treynor and Mazuy (1966) model out of the 38 mutual funds analyzed, 22 funds showed market timing ability. Selectivity ability; under the Treynor and Mazuy (1966) model out of the 38 mutual funds analyzed, only 4 of them exhibited market selection ability. 7. Conclusion From the findings of the research, market timing by mutual funds managers appears to be relatively achievable. Managers of the funds as established under this research can be able to time the market by using the market information to make higher profits than the market expected returns. On the other hand, market selectivity seems hard to be attained by managers. The managers of the funds as it is established under this research lack the ability to identify lowly priced securities in the market to make higher profits in the future than the market expected returns. Thus, it can be concluded that managers of funds are able to beat the market by knowing when to withdraw from a certain asset and when to buy a certain asset by using the market information. This means the null hypothesis in this case is rejected. On the hand, there is no skillful ability of funds managers to read from the market information and realize the lowly priced assets that should be bought currently in order to make higher returns in future than the expected market returns. This means on the case of market selectivity ability, the null hypothesis is rejected. Reference Anderson, S and Ahmed, P 2005. Mutual funds: performance of mutual funds. New York. Springer publishers. Guerard, J 2009. Handbook of portfolio construction: contemporary application of Markowitz techniques. New York. Springer publisher. Jelicic, N 2010. Dynamic strategy and performance of German mutual funds: conditional market timing. Hamburg. Diplomica Verlag publishers. Knight, J and Satchell, S 2002. Performance measurement in finance: Treynor-Mazuy measure. Oxford. Butter-Heinemann publishers. Smith, D and Shawky, H 2011. Institutional money management: mutual funds. New Jersey. Wiley & Sons publishers. Read More