City Research Online

Abstract

In this study 118 unit trusts are examined for the period 1966-1975, the two sub-periods: 1966/70 and 1971/75 in the light of Capital Market Theory;

The Efficient Market Hypothesis states that returns are generated by a "fair game" process, which amplified by the Capital Asset Pricing Model, implies a risk-free return and a risk-premium proportional to the covariance of the portfolio (or asset) to the market portfolio;

Since 1945, the proportion of equity, preference shares and ordinary shares and commercial loans held by investing institutions has grown remarkably. In testing unit trust portfolios, the actual behaviour in the market against a benchmark, alternative is examined. The theoretical alternative assumes no investment judgement was used to select portfolios.

The initial tests using correlation analysis fail to indicate any evidence of consistent continuity of performance. There was evidence of state dependent behaviour where unit trusts would rank similarly given consistent market conditions and dis-similarly given changed market conditions;

When introducing the risk-adjusted benchmark, the performance statistics indicated that unit trust managers were unable to generate consistent above-average results, though there was some evidence that both good and bad performance tended to persist between the two sub-periods 1966/70 and 1971/75;

It was noted that the trusts underwent considerable changes in their market risk exposure, as measured by beta, the latter sub-period seeing a large reduction in risk;

The individual categories of trusts had different performance results with the income group achieving greater returns for a given level of risk, while the specialized group achieved lower returns for given risk;

In considering the factors that may affect the performance of the unit trust portfolios, on the whole while some tests gave results which were statistically significant, the conclusion is that most of the supposed effects on performance, if they do operate are hardly significant enough individually to explain the great differences in outturn;

A multi-variate analysis of the factors relating to performance found that a strong management factor was responsible for the differences in return for the period analysed. However, the large qualitative element of the analysis precludes its future predictive value;

A discriminant analysis of the unit trusts found that 50 percent, using whatever combination of independent variables, were correctly classified. This result indicates that it appears many
unit trusts do not conform to the qualitative labels which are given in their own literature and reported in the press or the Unit Trust Yearbook.

Publication Type:	Thesis (Doctoral)
Subjects:	H Social Sciences > HG Finance
Departments:	Bayes Business School > Bayes Business School Doctoral Theses Bayes Business School > Finance Doctoral Theses

Export

Downloads