Risk and Return : A test of CAPM on the Swedish financials and industrials stock market before and after the financial crisis of 2008

Abstract: In hindsight one can often learn a valuable lesson or two. Looking back at the financial crisis of 2008, perhaps more than two lessons can be learned. The grave chock that occurred on the financial markets at the end of 2008 left the whole world questioning what had caused it. Theories that had previously been accepted as general truth were subjected to critique in the light post-crisis. The Capital Asset Pricing Theory or CAPM is one of these theories. It seeks to explain the relation between risk and returns. There are many studies critiquing CAPM that have been published after the crisis. Most studies that we found (post- and pre-crisis) are conducted based on portfolios and not individual stocks. Also we could not find a single study that compared if CAPM performed differently on stocks between different sectors. Based on our ontological and epistemological approaches of positivism and objectivism, we chose a deductive research approach and performed a quantitative data collection. This study was conducted on the Large Cap index listed on the Stockholm Stock Exchange (SSE). We chose 12 companies from the Financials sector and 12 companies from the Industrials sector, in total 24 companies. The period which we tested, were the years 2005-2012, 8 years in total. Comparison was made between the two sectors and between two different periods. The periods were Period 1 2005-2008 (pre-crisis) and 2009-2012 (post-crisis) We found that the original CAPM model, Sharpe-Lintner, did not explain the assumed relation between risk and return in the model, with Beta as the only risk-factor. In the second period (post crisis) CAPM did not have a single regression that showed that there was a relation between risk (Beta) and return (excess return). Pre crisis the majority (88%) of the regressions to test CAPM showed that there was no relation between the variables (risk and return). The Industrial sector had more positive correlations coefficients than the financial sector, but they were still to few to support the models postulated risk-return relationship.