Determinants of investment decisions for pension funds in Kenya

Abstract

The aim of this study was to find out the determinants of investment decisions among pension Funds in Kenya. The study used both primary and secondary data to establish the determinants. The primary data were collected using the Likert scale in questionnaires sent to the various pension funds. The secondary data on annual income of the funds and the annual values of assets were collected from the databases of the pension funds. The analysis using means and regression was done to determine the determinants of how the pension funds in Kenya choose to invest members’ funds. Results show that generally expected return; the risk-taking capacity; risk level in the desired investment; nature of risk in the global investment markets and investment portfolio desired were the most influential factors that determined investment decisions across all the firms. The least influential results across the pension schemes were consistency in returns, decision-making preferences of the decision makers, benchmarking with other pension funds, social responsibility issues and the nature of the fund owners. The correlation among the dependent variables namely average return and independent variables namely risk, expected return and investor characteristic variable was generally low indicating low level of interrelations among them. The factors were concluded to be independent of each other. The highest level of positive correlation was between the dependent variable and risk meaning that the higher the risk, the higher the return realized. The highest negative correlation was between the characteristics of the investors and risk meaning that the lower the importance attached to investor characteristics, the higher the risk. The regression of average return against risk variable, expected return variable, and investor characteristics variable was significant according to the F-Test, though the variation in the realized return was not strongly explained by the variables identified. This means that, though the variables identified were important to the realized return, there were some variables that were not captured by the model.