| Title | Empirical Analysis of Volatility Forecasting Model Based on Genetic Programming |
| Publication Type | Conference Paper |
| Year of Publication | 2017 |
| Authors | Toriyama, Naoki, Ono, Keiko, Orito, Yukiko |
| Conference Name | Proceedings of the 2017 International Conference on Intelligent Systems, Metaheuristics & Swarm Intelligence |
| Publisher | ACM |
| Conference Location | New York, NY, USA |
| ISBN Number | 978-1-4503-4798-3 |
| Keywords | composability, Genetic Programing, local search, pubcrawl, Realized Volatility, swarm intelligence |
| Abstract | In financial markets, the variance of stock returns plays an important role to reduce a risk, and daily volatility is often used as one of its measurement. We in this paper focus on Realized Volatility (RV), which is one of the most well-known volatility index. Traditionally regression models have been widely used to estimate it, but Genetic Programming (GP) approaches have been proposed recent years. While regression models estimate a suitable equation for forecasting RV, GP approaches estimate a tree (individual) that consists of economic information. Through evolution process, effective economic information can survive, so GP approaches can not only estimate RV values, but also extract effective information. However, GP approaches need computational efforts to avoid premature convergence. In this paper, we proposed a mutation-base GP approach for RV estimation, and analyze which economic information is needed to estimate RV accurately. |
| URL | http://doi.acm.org/10.1145/3059336.3059341 |
| DOI | 10.1145/3059336.3059341 |
| Citation Key | toriyama_empirical_2017 |
Empirical Analysis of Volatility Forecasting Model Based on Genetic Programming