Investment risk forecasting model using extreme value theory approach combined with machine learning

被引:0
作者
Melina, Melina [1 ]
Sukono [2 ]
Napitupulu, Herlina [2 ]
Mohamed, Norizan [3 ]
机构
[1] Univ Padjadjaran, Fac Math & Nat Sci, Doctoral Program Math, Sumedang 45363, Indonesia
[2] Univ Padjadjaran, Fac Math & Nat Sci, Dept Math, Sumedang 45363, Indonesia
[3] Univ Malaysia Terengganu, Fac Ocean Engn Technol & Informat, Kuala Terengganu 21030, Malaysia
来源
AIMS MATHEMATICS | 2024年 / 9卷 / 11期
关键词
backtesting; extreme value theory; GRU; LSTM; machine learning; multivariate; non-linear; RNN; VaR; VALUE-AT-RISK; MARKET; TIME;
D O I
10.3934/math.20241590
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
Investment risk forecasting is challenging when the stock market is characterized by non-linearity and extremes. Under these conditions, VaR estimation based on the assumption of distribution normality becomes less accurate. Combining extreme value theory (EVT) with machine learning (ML) produces a model that detects and learns heavy tail patterns in data distributions containing extreme values while being effective in non-linear systems. We aimed to develop an investment risk forecasting model in the capital market with non-linear and extreme characteristics using the VaR method of the EVT approach combined with ML (VaR GPD-ML(alpha) ). The combination of methods used is a multivariate time series forecasting model with RNN, LSTM, and GRU algorithms to obtain ML-based returns. The EVT method of the POT approach was used to model extremes. The VaR method was used for investment risk estimation. The backtesting method was used to validate the model. Our results showed that determining the threshold based on the normal distribution will identify extreme values with the ideal number, minimum bias, and distribution of extreme data following GPD. The VaR GPD-ML(alpha) model was valid in all samples based on backtesting at alpha = 0.95 and alpha = 0.99. Generally, this model produces a greater estimated value of investment risk than the VaR GPD(alpha) model at the 95% confidence level.
引用
收藏
页码:33314 / 33352
页数:39
相关论文
共 75 条
[1]   Modeling of extended osprey optimization algorithm with Bayesian neural network: An application on Fintech to predict financial crisis [J].
Abdullayev, Ilyos ;
Akhmetshin, Elvir ;
Kosorukova, Irina ;
Klochko, Elena ;
Cho, Woong ;
Joshi, Gyanendra Prasad .
AIMS MATHEMATICS, 2024, 9 (07) :17555-17577
[2]  
Aggarwal C., 2018, Neural Networks and Deep Learning, V1st
[3]   Poly-linear regression with augmented long short term memory neural network: Predicting time series data [J].
Ahmed, Supriyo ;
Chakrabortty, Ripon K. ;
Essam, Daryl L. ;
Ding, Weiping .
INFORMATION SCIENCES, 2022, 606 :573-600
[4]   Forecasting oil, coal, and natural gas prices in the pre-and post-COVID scenarios: Contextual evidence from India using time series forecasting tools [J].
Alam, Md Shabbir ;
Murshed, Muntasir ;
Manigandan, Palanisamy ;
Pachiyappan, Duraisamy ;
Abduvaxitovna, Shamansurova Zilola .
RESOURCES POLICY, 2023, 81
[5]  
[Anonymous], 2000, J Empir Financ, DOI DOI 10.1016/S0927-5398(00)00012-8
[6]   Forecasting of COVID-19 using deep layer Recurrent Neural Networks (RNNs) with Gated Recurrent Units (GRUs) and Long Short-Term Memory (LSTM) cells [J].
ArunKumar, K. E. ;
Kalaga, Dinesh, V ;
Kumar, Ch Mohan Sai ;
Kawaji, Masahiro ;
Brenza, Timothy M. .
CHAOS SOLITONS & FRACTALS, 2021, 146
[7]  
Bali T.G., 2003, Journal of Empirical Finance, V10, P455, DOI [10.1016/S0927-5398(02)00070-1, DOI 10.1016/S0927-5398(02)00070-1]
[8]   RESIDUAL LIFE TIME AT GREAT AGE [J].
BALKEMA, AA ;
DEHAAN, L .
ANNALS OF PROBABILITY, 1974, 2 (05) :792-804
[9]   Extreme risk spillovers across financial markets under different crises [J].
Cao, Yufei .
ECONOMIC MODELLING, 2022, 116
[10]   Extreme risk spillover of the oil, exchange rate to Chinese stock market: Evidence from implied volatility indexes [J].
Chen, Lin ;
Wen, Fenghua ;
Li, Wanyang ;
Yin, Hua ;
Zhao, Lili .
ENERGY ECONOMICS, 2022, 107