Optimum portfolio visualiser for risky assets using mean-variance model / Ruqayyah Apandi, Nur Suhailah Mohd Bakhtir and Nur Fatini Ramli

This research focuses in minimising the risk using mean risk model that
was first introduced by Markowitz (1952) for solving portfolio selection
problem. Thus, a variance is used as a risk measure in this project. The
scenario returns were obtained based on the historical monthly returns
from FBMKLCI. The mean-variance model and data set are being implemented
in Microsoft Excel and there are different level of target returns
which the optimal portfolios arc evaluated. Hence, the purpose of this
study is to optimise portfolio of risky assets under different level of target
return using mean-variance model. Next, to validate in-sample portfolios
obtained using the out-of-sample analysis. The in-sample result shows
that diversification allows us to reduce the risk of the portfolio without
sacrificing potential returns and it also shows that the lower the target
return, the lower the risk and the higher the target return, the higher
the risk. Based on the out-of-sample analysis, when the expected realised
return is low, it will give a low realised risk, when the expected realised
return is medium, the realised risk will also be medium and when the
expected realised return is high, the realised return is also high. Consequently,
to develop user interface as an optimal portfolio visualiser. The
user interface design is used to visualise the composition of portfolios and
realised returns in graphicas view to help the user quickly absorb and
interpret the presented result after they have entered the specific target
return. Generally, based on the results that we obtained, we can conclude
that mean-variance is applicable and widely used, as the method is easy
to be calculated, but only favorable at low target return. If we were to design
this study again, there are several changes that we would make. Most
importantly we would go for a longer time period in order to create more
scenario returns, include other types of data set, not only from FBMKLCI
and to include more methodological work on how to robustly capture the
impact and outcomes of different kind of risk measure in optimisation
portfolio such as value-at-risk and also conditional value-at-risk.