Intelligent Investment Approaches for Mutual Funds: An Evolutionary Model

Introduction

Recent times have witnessed a general concept that massive volumes of capital are being sold by the Stock Markets throughout the world. Every country’s economy happens to be associated and influenced severely by the behavior of the Stock Markets. Additionally, in the last few years or precisely more than a decade or two, the stock markets has become an easily available investment option for professional investors and also for common investor as well. As a result they did not remain related only to macroeconomic factors, but they have begun influencing day to day life directly. Therefore they have assumed significance and direct social impacts. Stock prices had always been a significant parameter for development of the companies, the constituting sector as well as the economic stability and growth of the country as a whole. One common characteristic of all Stock Markets is that they are uncertain, and they are correlated with short and long term future status. This characteristic, although an unwanted one from the investor’s point of view, but at the same time is also unavoidable one whenever he or she takes the Stock Market into consideration as the asset building option. The best option that one can try is to attempt to reduce this uncertainty. Stock Market prediction is one of the tool under this condition. Mutual Fund (MF) is a platform which collects money from various investors and the fund manager invests the money in various stocks and bonds. These bonds and stocks on which the investment is made constitute the portfolio of the fund. A mutual fund is managed by a company called fund house which hires a team of experts who executes the crucial investment of the investors’ money into selected stocks so as to provide maximum returns to the investors. The earning from the mutual fund is dispersed to the investors after paying the operating costs of the fund house. Every mutual fund comes with a certain objective and the investments made are also as per the offer documents. Investors also known as unit holder invest their money in a mutual fund which in turn is invested across a range of sectors thus minimizing the risk. Usually a fund house comes with a number of mutual funds each with certain objectives. Stock prices determine the development of banking operations viz. deposit and loan interest rates which in turn determines the mobility of money in the market and a whole lot of other economic factors of day to day life. Consequently, all through last few decades, the task of stock price speculation has assumed paramount importance in the capital market. Mutual Funds and other financial houses, who work on equity linked investments always, have the tremendously heavy task of speculating the stock prices so as to provide proper returns to the investors. If the same is not done accurately, either the investment will cease or the company will incur loss giving returns. Therefore, an accurate prediction is always essential so as to sustain in the market. Thanks to the field of Business Intelligence which offers varied metaheuristic approaches for accomplishing the above using the statistical knowledge base and input from external factors. The models involve extracting the data from statistics of past years, modeling them appropriately, adjusting them with external factors and finally speculating the trajectory of the prices for the time to come. The proposed chapter endeavors to model the past history of certain sectors of stocks so as to give a realistic prediction of the same in the time to come. The neuro fuzzy inference system is quite an established tool for forecast of data based on available data. Consequently, the authors have resorted to the same so as to predict the prices of stocks. There may be two different views of investment (return) maximization: viz. Long term investments and Short Term Investments. Analogously we also have two types of mutual funds viz. close ended and open ended that fetches returns likewise. The current chapter we confine our strategy to open ended mutual funds. We execute the same using historic patterns of the statistical data over the last 4 years, starting from the year 2014. In the next portion a history of the Mutual Fund, Net Asset Value and different types of mutual fund schemes are given. In the related work section the literature survey on the use of machine learning in portfolio management is given. Finally the data set preparation for the work and the proposed work is discussed with result and discussion.