Kauppakorkeakoulun julkaisuportaali
Aalto-yliopiston kauppakorkeakoulun gradujen tiedot nyt Aaltodocissa: Aaltodoc-julkaisuarkisto
Kauppakorkeakoulu | Rahoituksen laitos | Rahoitus | 2013
Tutkielman numero: 13138
Modeling Intraday Implied Volatility: Evidence From EURO STOXX 50
Tekijä: Tikanoja, Juuso
Otsikko: Modeling Intraday Implied Volatility: Evidence From EURO STOXX 50
Vuosi: 2013  Kieli: fin
Laitos: Rahoituksen laitos
Aine: Rahoitus
Asiasanat: rahoitus; financing; pörssit; stock exchanges; osakemarkkinat; stock markets; optiot; option; kurssivaihtelut; volatility
Sivumäärä: 72
Kokoteksti:
» hse_ethesis_13138.pdf pdf  koko: 2 MB (1061930)
Avainsanat: Implied volatility; Modeling; Intraday; Patterns; Options; EURO STOXX 50; VSTOXX
Tiivistelmä:
MODELING INTRADAY IMPLIED VOLATILITY: EVIDENCE FROM EURO STOXX 50

PURPOSE OF THE STUDY The objective of this thesis is to study intraday implied volatility with high-frequency observations. Specifically, I study if systematic intradaily and weekly patterns exist in implied volatility of EURO STOXX 50. Furthermore, I study if the implied volatility can be modeled using the possible patterns and time series econometric methods. Additionally, I study if the modeling can provide abnormal economic profits.

DATA AND METHODOLOGY

The dataset includes over 110 000 observations from VSTOXX, the implied volatility index of EURO STOXX 50 from January 2012 to the end of November 2012. Additionally, I acquire similar time-series of several financial instruments. To test the hypotheses, I study patterns with means and variances of VSTOXX. In addition, I model the data with ARMA family models and with several variables. Furthermore, I try to take advantage of possible patterns in intradaily/weekly implied volatility and use dummy variables in the modeling.

RESULTS

The empirical results of this thesis show that both intradaily and weekly patterns exist in implied volatility. Three possible reasons for the patterns were found: systematic patterns because of the VSTOXX formula misspecification, patterns driven by stock markets and patterns driven by traders. Modeling of implied volatility is found to be possible. ARMA models perform the best with the dataset. Several economic/financial variables are significant. Also, the weekday/time dummies were significant. However, when measuring the directional accuracy of the models, it seems that only ARMA model can forecast the direction and adding external variables is not useful. Even though proper trading simulation was not conducted, it seems that abnormal profits would not be created with models
Verkkojulkaisut ovat tekijänoikeuden alaista aineistoa. Teokset ovat vapaasti luettavissa ja tulostettavissa henkilökohtaista käyttöä varten. Aineiston käyttö kaupallisiin tarkoituksiin on kielletty.