| dc.description.abstract | This thesis examines the Morningstar rating system as a predictor of mutual fund
performance for mutual funds investing primarily in Norwegian equity. The predictive
abilities are examined using a data set free of survivorship bias, which contains data
on 136 mutual funds from January 2002 to December 2019. Mutual fund performance
is evaluated using estimates of alpha from three different factor models of performance
measurement: the Capital Asset Pricing Model (CAPM), Fama and French 3-factor model,
and Carhart 4-factor model. To comprehensively test for predictive abilities in the rating
system, both random effects panel data regressions and strategies of buying historically
top-rated ("winners") versus low-rated ("losers") funds are employed. The results indicate
findings that are robust across different performance measures and styles of funds. First,
for the period before the financial crisis, from January 2002 to March 2008, low ratings
from Morningstar generally indicate relatively poor future performance. Second, for the
period during and after the financial crisis, from April 2008 to December 2019, low ratings
from Morningstar, on the contrary, indicate relatively high future performance. Hence, in
the period before the financial crisis, investors could invest in past winners to generate
higher returns than an investment in past losers would have yielded. Contrary, during and
after the financial crisis, the loser strategies outperformed the winner strategies. Third,
there is little statistical evidence that Morningstar’s highest-rated funds outperform the
next-to-highest, median, and next-to-lowest rated funds in both periods. | en_US |
