Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach
碩士 === 國立臺灣海洋大學 === 應用經濟研究所 === 98 === Forecasting the stock return volatility needs an appropriate approach and has been an important issue in finance. Traditional volatility models, based on the assumptions of linearity and stationarity, can only use the returns series calculated by the non-stat...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2010
|
Online Access: | http://ndltd.ncl.edu.tw/handle/69583906980567783470 |
id |
ndltd-TW-098NTOU5452007 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-098NTOU54520072015-10-13T19:35:32Z http://ndltd.ncl.edu.tw/handle/69583906980567783470 Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach 股票報酬波動率預測:應用Hilbert-Huang Transformation方法 Sheng-Wen Wang 王聖文 碩士 國立臺灣海洋大學 應用經濟研究所 98 Forecasting the stock return volatility needs an appropriate approach and has been an important issue in finance. Traditional volatility models, based on the assumptions of linearity and stationarity, can only use the returns series calculated by the non-stationary stock price with some underlying correlated and periodic signals. Therefore, forecasting volatility using traditional models may be less accuracy due to the interference among the underlying signals. To overcome the mentioned problem, we first applied the Hilbert-Huang Transformation (HHT) by Huang et al. (1998) to decompose the stock price index of Taiwan Stock Exchange and Standard and Poor’s 500 during 1999-2009 into several intrinsic mode functions (IMFs) with different frequencies from fine-to-coarse and a trend, and then sequentially removed some specific IMFs which may result in interference among underlying signals. After removing the IMFs of interference from the original stock price, the filtered returns can be calculated. Consequently, we find that applying filtered returns to different volatility models improves the forecasting performance. Moreover, the amplitude in physics is defined as the magnitude of departure of the signal from the average position, which can be analogous to the volatility in finance. We then apply Hilbert transform to each IMF from EMD procedure on stock returns and obtain Hilbert Spectrums with time-varying instantaneous frequencies and amplitudes which can proxy stock returns volatility. The results showed that the amplitude approach does not improve and has similar forecasting performance with traditional volatility measures. However, amplitudes can still be served as an alternative proxy for stock returns volatility and worthy of further research. Fu-Sung Chiang Yi-Hao Lai 江福松 賴奕豪 2010 學位論文 ; thesis 99 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立臺灣海洋大學 === 應用經濟研究所 === 98 === Forecasting the stock return volatility needs an appropriate approach and has been an important issue in finance. Traditional volatility models, based on the assumptions of linearity and stationarity, can only use the returns series calculated by the non-stationary stock price with some underlying correlated and periodic signals. Therefore, forecasting volatility using traditional models may be less accuracy due to the interference among the underlying signals. To overcome the mentioned problem, we first applied the Hilbert-Huang Transformation (HHT) by Huang et al. (1998) to decompose the stock price index of Taiwan Stock Exchange and Standard and Poor’s 500 during 1999-2009 into several intrinsic mode functions (IMFs) with different frequencies from fine-to-coarse and a trend, and then sequentially removed some specific IMFs which may result in interference among underlying signals. After removing the IMFs of interference from the original stock price, the filtered returns can be calculated. Consequently, we find that applying filtered returns to different volatility models improves the forecasting performance. Moreover, the amplitude in physics is defined as the magnitude of departure of the signal from the average position, which can be analogous to the volatility in finance. We then apply Hilbert transform to each IMF from EMD procedure on stock returns and obtain Hilbert Spectrums with time-varying instantaneous frequencies and amplitudes which can proxy stock returns volatility. The results showed that the amplitude approach does not improve and has similar forecasting performance with traditional volatility measures. However, amplitudes can still be served as an alternative proxy for stock returns volatility and worthy of further research.
|
author2 |
Fu-Sung Chiang |
author_facet |
Fu-Sung Chiang Sheng-Wen Wang 王聖文 |
author |
Sheng-Wen Wang 王聖文 |
spellingShingle |
Sheng-Wen Wang 王聖文 Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach |
author_sort |
Sheng-Wen Wang |
title |
Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach |
title_short |
Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach |
title_full |
Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach |
title_fullStr |
Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach |
title_full_unstemmed |
Forecasting Stock Return Volatility: A Hilbert-Huang Transformation Approach |
title_sort |
forecasting stock return volatility: a hilbert-huang transformation approach |
publishDate |
2010 |
url |
http://ndltd.ncl.edu.tw/handle/69583906980567783470 |
work_keys_str_mv |
AT shengwenwang forecastingstockreturnvolatilityahilberthuangtransformationapproach AT wángshèngwén forecastingstockreturnvolatilityahilberthuangtransformationapproach AT shengwenwang gǔpiàobàochóubōdònglǜyùcèyīngyònghilberthuangtransformationfāngfǎ AT wángshèngwén gǔpiàobàochóubōdònglǜyùcèyīngyònghilberthuangtransformationfāngfǎ |
_version_ |
1718042290986942464 |