Equity Premium Puzzle: 100 Years of Bad Luck : The Equity Premium Puzzle in Norway

Abstract

This master thesis investigates the Norwegian equity premium puzzle for the period 1900-2008. I give a detailed overview over the stock, bill and bond market and follow the consumption based asset pricing model to relate the equity premium to the volatility in consumption innovation and the coefficient on relative risk aversion (RRA). I find that the Norwegian data implies a lower coefficient on RRA compared to what is calibrated for other countries. Mehra and Prescott (1985), Abel (1999) and Campbell and Cochrane (1999) implicitly assume perfectly correlated stock returns and consumption innovation in their assessment of the equity premium puzzle. Following this approach the implied RRA from the Hansen-Jagannathan bound equation is 6 for the whole period and 15 for the post WWII period. If the calculations are performed with the observed correlations between consumption innovation and stock returns, the implied RRA jumps to 37 for the whole period and 85 for the post WWII period. The implied RRA of 6 does not constitute an equity premium puzzle for Norway, however it implies a time preference parameter of 50%, and hence a risk-free rate puzzle. The relatively low RRA parameter for Norway arises from a low Sharpe ratio for excess returns (synonymously a relatively more volatile stochastic discount factor) and more volatile consumption innovation compared to other countries. The equity premium puzzle is definitely smaller in Norway than other counties. The observed high equity premium of 7.33% can perhaps be explained by theories from the field of behavioral economics, heterogeneous agents and the use of geometric returns as opposed to arithmetic returns, all of which are lightly discussed in this thesis. In accordance with the Common Stock Theory put forward by E. Smith (1924), I find that stocks outperformed bonds, providing both a higher return and a lower standard deviation in the period 1900-1970.