Forecast quality of the Swedish Volatility Index

Abstract: In this paper, I investigate the forecasting power of implied volatility via a new volatility index for the Swedish stock market (SVIX). By implementing the same methodology as the new VIX index originated from CBOE, I examine the information content of implied volatility and appraise the forecast quality of SVIX using two methods. Firstly, I use option valuation to evaluate the information content of implied volatility. I use four different volatilities and the evidence is clear. Using historical volatility or lagged one day at-the-money implied volatility generates poor results. Evaluating the quality of the Swedish volatility index SVIX and the average between the implied volatility lagged one day of one at-the-money call and one at-the-money put option (AIV), the results are diverted and there is no clear evidence whether to use AIV or the SVIX. Secondly, I evaluate the forecasting performance of the GARCH (1,1) model, SVIX and the AIV. Evidence point in the directions that SVIX and AIV forecasts is of higher quality than the GARCH (1,1) model, which uses historical information to produce volatility forecasts.