A portfolio selection problem arising from real-estate investments

Abstract

Investments have been used as a medium to mitigate the effects of inflation for many years and it is expected that they will be used for many years to come. Not only do investments allow for an opportunity to increase the net worth of a sum of money, they also allow for a
source of passive income. With every investment, however, there is an associated risk but, higher risks are typically associated with a more enticing reward. From the broad scope of potential investment opportunities, the primary focus of this paper is to consider investments within the realm of real estate. Although property investments have the potential of generating a satisfactory return and
may generally be considered a safe investment, as with any investment, poor decision making may still results in the loss of capital. A multi-period portfolio selection model may assist a potential investor in determining an optimal investment plan. The multi-period  portfolio selection model considers the future value and rental incomes of a set of potential properties, in determining the optimal investment plan over a given time horizon, aimed at maximising the expected net present value of the portfolio. The quality of the investment plan, however, is dependent on the accuracy of the predicted future value and rental income of the properties under consideration. Therefore, in order to determine these future values accurately, a time series forecasting model is proposed. The forecasting model predicts the values for property values and rental income over a selected time horizon, which serves as input to the multiperiod portfolio selection model. In short, the goal of this paper is to apply suitable time series forecasting methods in order to generate predicted values with an acceptable accuracy, to serve as input for a multi-period portfolio selection model to determine an optimal investment plan as output.