Sammendrag
This research studies the asymmetric volatility in the Norwegian equity market. Using trading volume, market value, book-to-market value, and average return as portfolio indicator at t-1 in addition to 5 market indices I consider the asymmetric relationship between asset returns and changes in volatility. Portfolio returns are characterised by negative and positive serial correlation induced by non-synchronous price adjustment among the component stocks. The GARCH family of stochastic has shown to represent well the stochastic volatility of stock returns. Using an Exponential-GARCH model for the variance of above categorised firm portfolios and market indices, conditioned on the auto-correlation structure in the conditional mean, striking differences within and above portfolios categories appear. Trading volume and market volume firm portfolios sort the serial correlation nicely from significant positive to significant negative coefficients. That is, highly traded and high market value portfolios show slow price adjustments while the lowly traded and low market value firms overreacts and reverse the price adjustment process. Moreover, lowly traded, small market value, and high book-to-market firms show a smaller influence on the conditional variance than highly traded, high market value, and low book-to-market value firms. In contrast, the past conditional variance exerts a greater influence over the current conditional variance in the case of lowly traded, low market value, and high book-to-market value firm portfolios. The combination of these two features of the conditional variance suggest that although shocks to the volatility of lowly traded, low market value, high book-to-market value firm portfolios have less impact than shocks to the volatility of highly traded, high market value, and low book-to-market value firm portfolios, they are much more persistent. The GED-parameters of the EGARCH model provide an asymmetric relationship between returns and changes in volatility. In combination the GED-parameters suggest positive responses to shock, negative assymetric volatility and thick tales. That is, all major episodes of high volatility are associated with market drops. Moreover, the conditional volatility parameters indicate substantial persistence of shocks to the conditional volatility. Finally, the latest years has made the asymmetry more severe for highly traded and/or high market value firm portfolios. That is, high volatility and markets drops are more extreme for highly traded and/or high market value firms than lowly traded and low market value firms. This effect has continued to thicker tales for the value weighted indices, while equal weighted indices show thinner tales.
Vis fullstendig beskrivelse
