Tid: 17 juni 2011 kl 14.15-15.00.Sammanträdesrummetrummet 3424, Institutionen för matematik, KTH, Lindstedts väg 25, plan 3. Karta!
Föredragshållare: Daniel Amsköld
Titel: A comparison between different volatility models (Examensarbete - Master thesis)
Abstract The main purpose of this master thesis is to evaluate and compare different volatility models. The evaluation is based on how well the models are imitating the implied volatility of different stock options. Three different times to maturity will be studied. The volatility models are evaluated based on daily deviations from the implied volatility and on daily changes of the modelled volatility. Statistical measurements investigated are Mean Absolute Deviation and R2. The models investigated are historical volatility models, a GARCH model and a model where the implied volatility of an index is scaled with a scaling factor based on historical returns of the asset and the index. The investigation shows that the scaling factor model has a better performance than the other models, but it also shows that it may be better to use the implied volatility of an index, without scaling, instead of the volatility models. For the historical models it is shown that 50 to 75 observations is most appropriate to use to imitate the implied volatility. It is more difficult to evaluate the performance of the GARCH model, since the result of the model is varying. It is also concluded that a few single observations of high absolute returns can result in an overestimation of the modelled GARCH volatility.
|Sidansvarig: Filip Lindskog