Equity-Focused Funds

Capital Asset Pricing Model Regression

As with the emerging markets regressions above, the regressions for the Nordic markets will also put focus on the equity focused funds in this section. Of the 974 funds with geographical focus area in the Nordic markets, a little under half of the are equity focused funds. More precisely there are 479 funds in the data analysed in this section. Regressing this data with methodology as explained yields the following results.

83

Table 5.18: CAPM regression, equity focused funds, Nordic markets

Where the previous regression including all asset classes showed all three coefficient estimates were significant, this regression has an insignificant coefficient estimate for the fee variable. This means that the regression does not find any significant impact of fees on returns for equity focused funds in the Nordic markets. The other two variables have significant coefficient estimates like in the

previous regressions. An interesting thing to note is that the intercept is much higher in this model, which is of course due to the equity focused funds having a higher return on average than all the funds on average. This is supported by the tables shown in appendix 1, where equity focused funds clearly have the highest mean return.

The regression reports a R-squared value of 0.3108, meaning it is almost exactly the same as for the regression including all asset classes. Surprisingly the model is not a better fit for equity focused funds than what it is for all the different asset classes, even though the methodology is constructed to explain equity returns.

As with the other regressions, a test for problems with heteroscedasticity will be performed by looking at the residual plots. The residual plots are all shown in appendix 12. The plot with predicted y-values on the x-axis are shown in figure 5.12 below as well.

84

Figure 5.12: residual vs predicted Y-values plot for CAPM regression, equity focused funds, Nordic markets

The graphs both in appendix 12 and the one above looks very similar to the ones presented with the previous regression, except there are less observations. This is of course not a coincidence, and since some of the coefficient estimates are very similar and many of the funds are included in both, this also should not come as a surprise. The plot above shows the same three outliers. This picture is almost the same for the other two plots in the appendix. As with the previous regression this shows a very stable variance in the residuals, which is a clear sign of homoscedasticity and hence there is no reason to question the reliability of that assumption.

Fama-French Three-Factor Model Regression

As has been done earlier, the model used above will be extended to include the extra two factors in Fama-French's Three-Factor Model, Small-Minus-Big and High-Minus-Low. This will add further validity to the conclusions drawn, either by possibly changing the results or by simply confirming the estimates from the previous regressions. The Fama-French Three-Factor Model, including the fee variable, produces the following results shown in table 5.19 below.

85

Table 5.19: Fama French 3-factor regression, equity focused funds, Nordic markets

The results are not surprising. The coefficient estimates for the intercept and the market return have not changed much and they are still significant at 99% confidence level, as the p-value is below 0.01. Both of the new factors are highly significant at 99% confidence level as well, but they do not change the conclusion about the fee variable, as there seems to be no significant relationship between fees and returns for equity focused funds with the Nordic markets as their geographical focus area. What can be said, is that even though no significant relationship exists, there is no reason for concluding that higher fees should lead to lower returns.

The regression returns a R-squared value of 0.3310, which means it has increased around two percentage points by adding the two additional factors. This is not surprising as they are highly significant and does not change the conclusions already drawn about the variables included in the previous regressions. This means the new variables are able to explain some of the variance not already explained in the previous regression.

The test for heteroscedasticity is performed using the residual plots shown in appendix 13. The residual plot with predicted y-values on the x-axis are also printed in figure 5.13 below to highlight differences or lack of differences compared to the previous regression.

86

Figure 5.13: residual vs predicted Y-values plot for Fama-French regression, equity focused funds, Nordic markets

The plot above and the additional plots in the appendix shows no to little signs of heteroscedasticity.

The only plot which shows some tendencies of heteroscedasticity are the residual plot with the fee variable on the x-axis. The tendency is not strong and hence the assumption about homoscedasticity in the error terms are assumed to hold in this regression.

Interestingly, and maybe a bit surprising, the error plot above looks almost exactly the same as the one in the previous regression, in the sense that the two additional factors have failed to explain the variance in the outliers. This means that even though the R-squared has increased by two

percentage points, the residuals looks almost exactly the same.