Kernel methods for smoothing of options- implied volatility
(2011) NEK791 20112Department of Economics
- Abstract (Swedish)
- In today’s financial markets, the estimation of implied volatility of options plays an important part. Through the options prices, the implied volatility shows the size of market movement that is anticipated by the market’s participants. It is used as a benchmark for pricing all kind of assets and an erroneous number can be very costly. There are several methods for estimating and establishing patterns for the implied volatility. In latter years, nonparametric methods have grown in popularity, not least the kernel regression methods.
This thesis compares two different kernel regression methods, the Nadaraya-Watson- and the Local linear regressions for smoothing the implied volatility of options on OMX, an exchange where options are... (More) - In today’s financial markets, the estimation of implied volatility of options plays an important part. Through the options prices, the implied volatility shows the size of market movement that is anticipated by the market’s participants. It is used as a benchmark for pricing all kind of assets and an erroneous number can be very costly. There are several methods for estimating and establishing patterns for the implied volatility. In latter years, nonparametric methods have grown in popularity, not least the kernel regression methods.
This thesis compares two different kernel regression methods, the Nadaraya-Watson- and the Local linear regressions for smoothing the implied volatility of options on OMX, an exchange where options are sparsely traded. For each kernel regression method two different bandwidth selectors are used. Using market data from the latter years of market turbulence, data from a group of days with low market volatility is examined along with a group of days with high market volatility. Before carrying out the regressions freedom of arbitrage is ensured.
The findings are that the different regression methods show similar results for either of rhte bandwidth selector in both kind of markets (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/2158342
- author
- Olsson, Ola LU
- supervisor
- organization
- course
- NEK791 20112
- year
- 2011
- type
- H1 - Master's Degree (One Year)
- subject
- keywords
- Implied volatility, Kernel, Smoothing, Regression, Nadaraya-Watson, Local linear, Bandwidth, Akaike, Cross-Validation
- language
- English
- id
- 2158342
- date added to LUP
- 2011-10-03 11:21:28
- date last changed
- 2011-10-03 11:21:28
@misc{2158342, abstract = {{In today’s financial markets, the estimation of implied volatility of options plays an important part. Through the options prices, the implied volatility shows the size of market movement that is anticipated by the market’s participants. It is used as a benchmark for pricing all kind of assets and an erroneous number can be very costly. There are several methods for estimating and establishing patterns for the implied volatility. In latter years, nonparametric methods have grown in popularity, not least the kernel regression methods. This thesis compares two different kernel regression methods, the Nadaraya-Watson- and the Local linear regressions for smoothing the implied volatility of options on OMX, an exchange where options are sparsely traded. For each kernel regression method two different bandwidth selectors are used. Using market data from the latter years of market turbulence, data from a group of days with low market volatility is examined along with a group of days with high market volatility. Before carrying out the regressions freedom of arbitrage is ensured. The findings are that the different regression methods show similar results for either of rhte bandwidth selector in both kind of markets}}, author = {{Olsson, Ola}}, language = {{eng}}, note = {{Student Paper}}, title = {{Kernel methods for smoothing of options- implied volatility}}, year = {{2011}}, }