Financial Volatility Forecasting by Nonlinear Support Vector Machine Heterogeneous Autoregressive Model: Evidence from Nikkei 225 Stock Index

Khan, Md. Ashraful Islam

doi:10.5539/ijef.v3n4p138

Cited by 17 publications

(10 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The HAR-RV model was developed in accordance with the heterogeneous market hypothesis proposed by Muller et al (1997) and the long memory character of realised volatility by Andersen et al (2003). Empirical studies have shown that the HAR model has high forecasting performance for future volatility, especially for out-of-sample data with different time horizons (Corsi 2003;Khan 2011).…”

Section: Introductionmentioning

confidence: 99%

Forecasting of Realised Volatility with the Random Forests Algorithm

Luong

Dokuchaev

2018

JRFM

View full text Add to dashboard Cite

The paper addresses the forecasting of realised volatility for financial time series using the heterogeneous autoregressive model (HAR) and machine learning techniques. We consider an extended version of the existing HAR model with included purified implied volatility. For this extended model, we apply the random forests algorithm for the forecasting of the direction and the magnitude of the realised volatility. In experiments with historical high frequency data, we demonstrate improvements of forecast accuracy for the proposed model.

show abstract

Section: Introductionmentioning

confidence: 99%

Forecasting of Realised Volatility with the Random Forests Algorithm

Luong

Dokuchaev

2018

JRFM

View full text Add to dashboard Cite

show abstract

“…However, despite the large amount of papers that refer to the application of ANNs for financial time series forecasting (e.g. White (1988), Khan (2011)), only few works focus specifically on their application on forecasting conditional volatility. The majority of these studies foresees the combination of a GARCH model with a NN architecture, see for example Donaldson and Kamstra (1997) and Hu and Tsoukalas (1999), with few exceptions, e.g.…”

Section: Introductionmentioning

confidence: 99%

Cholesky–ANN models for predicting multivariate realized volatility

Bucci

2020

Journal of Forecasting

View full text Add to dashboard Cite

Accurately forecasting multivariate volatility plays a crucial role for the financial industry. The Cholesky-Artificial Neural Networks specification here presented provides a twofold advantage for this topic. On the one hand, the use of the Cholesky decomposition ensures positive definite forecasts. On the other hand, the implementation of artificial neural networks allows to specify nonlinear relations without any particular distributional assumption. Out-of-sample comparisons reveal that Artificial neural networks are not able to strongly outperform the competing models. However, long-memory detecting networks, like Nonlinear Autoregressive model process with eXogenous input and long shortterm memory, show improved forecast accuracy respect to existing econometric models.

show abstract

“…This nonlinear and complex nature of stock market makes SPP more difficult [7,8]. The prediction of future stock prices is defined by factors like data intensity, noise, non-stationary, random nature, uncertainties and invisible interlinks [9][10][11]. Generally, there are two classifications of SPP which includes the SPP movement and predicting the value of the stock price.…”

Section: Introductionmentioning

confidence: 99%

Social Spider Optimization with Tumbling Effect Based Data Classification Model for Stock Price Prediction

Saravanan¹,

Sujatha²,

Kadiravan³

et al. 2019

IJITEE

View full text Add to dashboard Cite

In last decade, data classification become more famous which aims to classify the data to a fixed number of classes. The data classification problem is treated as an NP hard problem and different optimization models are presented to resolve it. This paper introduces a social spider optimization (SSO) algorithm with tumbling effect called SSO-T algorithm to solve the data classification problem. First, the SSO algorithm is derived to solve the classification process which is considered as a NP hard problem. Next, to further enhance the exploration process of SSO algorithm, it is modified by the inclusion of tumbling effect, called SSO-T algorithm. For validating the results of the SSO and SSO-T algorithm, a real time problem of stock price prediction (SSP) is employed. For experimentation, the results are validated by testing the SSO and SSO-T algorithms against four datasets such as Dow Jones Index (DJI) dataset, three own datasets gathered from Yahoo finance on the basis of daily, weekly and yearly. The empirical result states that the proposed algorithms perform well and it is noted that the classification performance of the SSO algorithm is increased by the inclusion of tumbling effect.

show abstract

Financial Volatility Forecasting by Nonlinear Support Vector Machine Heterogeneous Autoregressive Model: Evidence from Nikkei 225 Stock Index

Cited by 17 publications

References 39 publications

Forecasting of Realised Volatility with the Random Forests Algorithm

Forecasting of Realised Volatility with the Random Forests Algorithm

Cholesky–ANN models for predicting multivariate realized volatility

Social Spider Optimization with Tumbling Effect Based Data Classification Model for Stock Price Prediction

Contact Info

Product

Resources

About