Conditional Heteroskedasticity Research Articles

A Stock Index Futures Price Prediction Approach Based on the MULTI-GARCH-LSTM Mixed Model

As a type of financial derivative, the price fluctuation of futures is influenced by a multitude of factors, including macroeconomic conditions, policy changes, and market sentiment. The interaction of these factors makes the future trend become complex and difficult to predict. However, for investors, the ability to accurately predict the future trend of stock index futures price is directly related to the correctness of investment decisions and investment returns. Therefore, predicting the stock index futures market remains a leading and critical issue in the field of finance. To improve the accuracy of predicting stock index futures price, this paper introduces an innovative forecasting method by combining the strengths of Long Short-Term Memory (LSTM) networks and various Generalized Autoregressive Conditional Heteroskedasticity (GARCH)-family models namely, MULTI-GARCH-LSTM. This integrated approach is specifically designed to tackle the challenges posed by the nonstationary and nonlinear characteristics of stock index futures price series. This synergy not only enhances the model’s ability to capture a wide range of market behaviors but also significantly improves the precision of future price predictions, catering to the intricate nature of financial time series data. Initially, we extract insights into the volatility characteristics, such as the aggregation of volatility in futures closing prices, by formulating a model from the GARCH family. Subsequently, the LSTM model decodes the complex nonlinear relationships inherent in the futures price series and incorporates assimilated volatility characteristics to predict future prices. The efficacy of this model is validated by applying it to an authentic dataset of gold futures. The empirical findings demonstrate that the performance of our proposed MULTI-GARCH-LSTM hybrid model consistently surpasses that of the individual models, thereby confirming the model’s effectiveness and superior predictive capability.

Gold Price Prediction Using Skewness and Kurtosis Based Generalized Auto-regressive Conditional Heteroskedasticity Approach with Long Short Term Memory Network

Gold Price Prediction Using Skewness and Kurtosis Based Generalized Auto-regressive Conditional Heteroskedasticity Approach with Long Short Term Memory Network

Heteroscedasticity effects as component to future stock market predictions using RNN-based models.

Heteroscedasticity effects are useful for forecasting future stock return volatility. Stock volatility forecasting provides business insight into the stock market, making it valuable information for investors and traders. Predicting stock volatility is a crucial task and challenging. This study proposes a hybrid model that predicts future stock volatility values by considering the heteroscedasticity element of the stock price. The proposed model is a combination of Generalized Autoregressive Conditional Heteroskedasticity (GARCH) and a well-known Recurrent Neural Network (RNN) algorithm Long Short-Term Memory (LSTM). This proposed model is referred to as GARCH-LSTM model. The proposed model is expected to improve prediction accuracy by considering heteroscedasticity elements. First, the GARCH model is employed to estimate the model parameters. After that, the ARCH effect test is used to test the residuals obtained from the model. Any untrained heteroscedasticity element must be found using this step. The hypothesis of the ARCH test yielded a p-value less than 0.05 indicating there is valuable information remaining in the residual, known as heteroscedasticity element. Next, the dataset with heteroscedasticity is then modelled using an LSTM-based RNN algorithm. Experimental results revealed that hybrid GARCH-LSTM had the lowest MAE (7.961), RMSE (10.466), MAPE (0.516) and HMAE (0.005) values compared with a single LSTM. The accuracy of forecasting was also significantly improved by 15% and 13% with hybrid GARCH-LSTM in comparison to single LSTMs. Furthermore, the results reveal that hybrid GARCH-LSTM fully exploits the heteroscedasticity element, which is not captured by the GARCH model estimation, outperforming GARCH models on their own. This finding from this study confirmed that hybrid GARCH-LSTM models are effective forecasting tools for predicting stock price movements. In addition, the proposed model can assist investors in making informed decisions regarding stock prices since it is capable of closely predicting and imitating the observed pattern and trend of KLSE stock prices.

Decoding Stocks Patterns Using LSTM

Decoding stocks is extensively utilized in the financial sector by numerous organizations. It is volatile in nature, so it’s tough to predict the prices of stock. Numerous methodologies exist for tackling this task, including logistic regression, support vector machines (SVM), autoregressive conditional heteroskedasticity (ARCH) models, recurrent neural network (RNN), convolutional neural networks (CNN), backpropagation, Naïve Bayes, among others. Among these, Long Short-Term Memory (LSTM) stands out as particularly adept at handling time series data. The primary aim is to discern prevailing market trends and achieve accurate stock price forecasts. Leveraging LSTM and RNN , we strive for error free stock price predictions, with promising results.

South African Real Estate Investment Trusts Prefer Tuesdays

This study examines the day-of-the-week effect on the returns of different classifications of South African REITs. Ordinary least squares regression (OLS), generalized autoregressive conditional heteroskedasticity (GARCH) (1,1) (2,1), and Kruskal–Wallis (KW) tests were performed on data obtained from the IRESS Expert database from 2013 to 2021. We found statistical differences in the day-of-the-week effects for SAREITs; the best day to invest in office REITs is Friday, for diversified REITs Thursday, and for industrial REITs Friday. Generally, Wednesday was found to be the least profitable day to invest in all REIT classifications because it had the least average daily return. Tuesdays were the most profitable days for all REIT classifications, with the highest average daily return. REITs traded the most on Fridays, while REITs traded the least on Mondays. Returns were the most volatile on Monday, while volume was the least volatile on Thursday. The KW test revealed a statistically significant difference between the median returns across days of the week. Based on the above, profitability is expressed on Tuesdays in South African REITs. By recognizing the day-of-the-week effect, investors can buy and sell South African REITs more effectively. This study, apart from being the first in the context of South African REITs, provides updated evidence of the contested calendar anomaly issues.

Volatility forecasting incorporating intraday positive and negative jumps based on deep learning model

AbstractMost existing studies on volatility forecasting have focused on interday characteristics and ignored intraday characteristics of high‐frequency data, especially the asymmetric impact of positive and negative jumps on volatility. In this paper, 5‐min high‐frequency data are used to construct realized volatility which is decomposed into continuous components and jump components with positive and negative directions. Then, this information is combined with the long short‐term memory model for the realized volatility prediction. The empirical analysis demonstrates that negative jumps resulting from negative news have a more significant impact on market volatility than positive jumps. Additionally, the long short‐term memory model, which incorporates positive and negative jump volatility, outperforms traditional econometric and machine learning models in predicting out‐of‐sample volatility. Furthermore, applying the prediction results to value at risk yields a better measurement effect than the generalized autoregressive conditional heteroskedasticity model.

Time‐varying risk preference and equity risk premium forecasting: The role of the disposition effect

AbstractThis study examines whether the disposition effect can explain time‐varying risk preference and predict the equity risk premium. To do so, we propose an augmented general autoregressive conditional heteroskedasticity (GARCH)‐in‐Mean model unraveling the complex relationship between unrealized gains/losses, realized returns, and the equity risk premium. In our model, the risk aversion coefficient varies with the market state of unrealized gains/losses. Using data from the US stock markets, we show strong evidence that the disposition effect drives time‐varying risk preference: The risk aversion coefficient is significantly positive during periods of unrealized gains, but insignificant during periods of unrealized losses. These findings reconcile the conflicting results of the risk‐return trade‐off in existing literature. Moreover, our model shows significant predictability of the equity risk premium, both in‐sample and out‐of‐sample. Incorporating our model's predictions can yield substantial utility gains for a mean‐variance investor. Our results indicate that the disposition effect leads to time‐varying risk preference and thus induces equity risk premium predictability.

Investigating the Effects of the COVID-19 Pandemic on Stock Volatility in Sub-Saharan Africa: Analysis Using Explainable Artificial Intelligence

This study examines the impact of the COVID-19 pandemic on sector volatility in sub-Saharan Africa by drawing evidence from two large and two small stock exchanges in the region. The analysis included stock-specific data, COVID-19 metrics, and macroeconomic indicators from January 2019 to July 2022. This study employs generalized autoregressive conditional heteroskedasticity (GARCH) models to estimate volatility and Explainable Artificial Intelligence (XAI) in the form of SHapley Additive exPlanations (SHAP) to identify significant factors driving stock volatility during the pandemic. The findings reveal significant volatility increases at the onset of the pandemic, with government stringency measures leading to increased volatility in larger exchanges, while the introduction of vaccination programs helped to reduce volatility. Weaker macroeconomic fundamentals impact volatility in smaller exchanges. The healthcare sector has emerged as the most resilient, while non-essential sectors, such as consumer discretionary, materials, and real estate, face greater vulnerability, especially in smaller exchanges. The research findings reveal that the heightened stock market volatility observed was mainly a result of the government’s actions to combat the spread of the pandemic, rather than its outbreak. We recommend that governments introduce sound policies to balance public health measures and economic stability, and that investors diversify their investments to reduce the impact of pandemics.

The Volatility Spillover of Global Oil Price Uncertainty

This manuscript, for the first time, analyses the volatility spillover of oil price uncertainty in the world using data from oil price uncertainty recently developed by Abdul and Qureshi (2023), spanning the time 1996-2019 on a monthly frequency. ARCH/GARCH (Autoregressive Conditional Heteroskedasticity and Generalized Autoregressive Conditional Heteroskedasticity) models are employed as an econometric tool. The findings suggest that ARCH model is more consistent than GARCH model in assessing the volatility of oil price uncertainty in the world. The results show that the volatility of oil price uncertainty is high in the world. The transition to renewable energy sources is proposed as a way to resist unexpected oil shocks since the production of renewables does not depend on the fluctuations of oil prices. Consequently, uncertainties in the oil price do not hinder economic activities.

Time series clustering based on latent volatility mixture modeling with applications in finance

Modeling financial time series data poses a significant challenge in the realm of time series analysis. The Autoregressive Conditional Heteroskedasticity (ARCH) model stands out as a potent tool for capturing time-varying volatility and heteroskedasticity in financial data. However, conventional ARCH models display sensitivity to departures from normality, leading to the development of extensions employing more flexible distributions. In this context, we propose a robust enhancement to the mixture of ARCH (MoARCH) model by integrating normal mean–variance mixture (NMVM) distributions to model component errors. The stochastic representation of the proposed model allows for a straightforward implementation of an Expectation Conditional Maximization Either (ECME) algorithm for obtaining Maximum Penalized Likelihood estimates (MPL). To thoroughly evaluate the model, we conduct four simulation studies to explore finite-sample properties, assess MPL estimators, scrutinize model robustness, and evaluate the accuracy of our proposal in fitting, clustering, and forecasting. Practical applications further highlight the effectiveness of our methodology, showcasing successful implementations across diverse real datasets.

Online Muscle Activation Onset Detection Using Likelihood of Conditional Heteroskedasticity of Electromyography Signals.

The real-time, accurate detection of muscle activation onset (MAO) is significant for EMG-triggered control strategies in embedded applications like prostheses and exoskeletons. This paper investigates sEMG signals using the generalized autoregressive conditional heteroskedasticity (GARCH) model, focusing on variance. A novel feature, the likelihood of conditional heteroskedasticity (LCH) extracted from the maximum likelihood estimation of GARCH parameters, is proposed. This feature effectively distinguishes signal from noise based on heteroskedasticity, allowing for the detection of MAO through the LCH feature and a basic threshold classifier. For online calculation, the model parameter estimation is simplified, enabling direct calculation of the LCH value using fixed parameters. The proposed method was validated on two open-source datasets and demonstrated superior performance over existing methods. The mean absolute error of onset detection, compared with visual detection results, is approximately 65 ms under online conditions, showcasing high accuracy, universality, and noise insensitivity. The results indicate that the proposed method using the LCH feature from the GARCH model is highly effective for real-time detection of muscle activation onset in sEMG signals. This novel approach shows great potential and possibility for real-world applications, reflecting its superior performance in accuracy, universality, and insensitivity to noise.

Do oil price forecast disagreement of survey of professional forecasters predict crude oil return volatility?

This paper explores whether the dispersion in forecasted crude oil prices from the European Central Bank Survey of Professional Forecasters can provide insights for predicting crude oil return volatility. It is well-documented that higher disagreement among forecasters of asset price implies greater uncertainty and higher return volatility. Using several Generalized Autoregressive Conditional Heteroskedasticity with Mixed Data Sampling (GARCH-MIDAS) models, we find, based on the in-sample estimation results, the oil market experiences greater volatility when the forecasters’ disagreements increase. The model that integrates both historical realized variance and forward-looking forecaster disagreement into the conditional variance, along with the model focusing solely on pure forward-looking forecaster disagreement, exhibits a much superior fit to the data compared to the model relying solely on realized variance and the models considering forward-looking forecasted mean return. The out-of-sample forecasting results unequivocally illustrate that incorporating forecaster disagreement offers valuable insights, markedly enhancing the predictive accuracy of crude oil return volatility within the GARCH-MIDAS model. Moreover, we illustrate the economic benefit of considering forecasters’ disagreement when forecasting volatility, demonstrating its significance for VaR risk management.

Asset prices in a production network

The relative importance of sectoral and aggregate productivity shocks in asset pricing is examined using a nonlinear dynamic equilibrium model where heterogenous sectors interact in a production network. The model accounts for the heterogeneity in sectoral stock returns and endogenously generates conditional heteroskedasticity and fat tails. The equity risk premium is shown to be driven by sectoral shocks—specially to investment good producers and mining—with a limited contribution from the aggregate shock. SMM estimates of the elasticities of substitution between material inputs and between investment goods support the assumption of gross complementarity employed by previous network literature.

Is the exchange rate exposure puzzle really a puzzle? International evidence

PurposeGiven the difficulties in finding significant exchange rate exposure in the extant literature, this paper attempts to resolve the so-called “exposure puzzle” by investigating whether currency movements have any significant impact on international industry returns.Design/methodology/approachThis paper utilizes the multivariate Generalized AutoRegressive Conditional Heteroskedasticity (MGARCH) methodology to estimate both symmetric and asymmetric exchange rate exposures for each industry common across 12 countries simultaneously.FindingsThe empirical results show that exchange rate exposure is not only statistically significant but also economically important based on the estimation of an asymmetric three-factor exposure model using MGARCH methodology. This is an extremely important finding as it suggests that the “exposure puzzle” may not be a puzzle at all once a better methodology is utilized in the estimation.Research limitations/implicationsBecause this study tries to resolve the exchange rate exposure puzzle by focusing on whether exchange rate movements affect ex-post returns as opposed to ex ante expected returns and given the significant exposures with respect to different risk factors found in the study, it is interesting to see if any of these risk factors commands a risk premium. In other words, a natural extension of this study is to test whether any of these risk factors is priced in international industry returns.Practical implicationsThe findings of the study have interesting implications for international investors who would like to diversify their portfolios across different industries and are concerned about whether the unexpected movements in the bilateral exchange rates will affect their portfolio returns in addition to its interest rate and world market risk exposures.Originality/valueThe study utilizes the MGARCH methodology, which has not been fully exploited in the exchange rate exposure literature.

Calendar anomalies and asymmetric volatility of returns in the Indonesian stock market: conventional vs Islamic indices

Purpose This study aims to identify calendar anomalies that can affect stock returns and asymmetric volatility. Thus, the objective of this study is twofold: on the one hand, it examines the impact of calendar anomalies on the returns of both conventional and Islamic indices in Indonesia, and on the other hand, it analyzes the impact of these anomalies on return volatility and whether this impact differs between the two indices. Design/methodology/approach The authors apply the GJR-generalized autoregressive conditional heteroskedasticity model to daily data of the Jakarta Composite Index (JCI) and the Jakarta Islamic Index for the period ranging from October 6, 2000 to March 4, 2022. Findings The authors provide evidence that the turn-of-the-month (TOM) effect is present in both conventional and Islamic indices, whereas the January effect is present only for the conventional index and the Monday effect is present only for the Islamic index. The month of Ramadan exhibits a positive effect for the Islamic index and a negative effect for the conventional index. Conversely, the crisis effect seems to be the same for the two indices. Overall, the results suggest that the impact of market anomalies on returns and volatility differs significantly between conventional and Islamic indices. Practical implications This study provides useful information for understanding the characteristics of the Indonesian stock market and can help investors to make their choice between Islamic and conventional equities. Given the presence of some calendar anomalies in the Indonesia stock market, investors could obtain abnormal returns by optimizing an investment strategy based on seasonal return patterns. Regarding the day-of-the-week effect, it is found that Friday’s mean returns are the highest among the weekdays for both indices which implies that investors in the Indonesian stock market should trade more on Fridays. Similarly, the TOM effect is significantly positive for both indices, suggesting that for investors are called to concentrate their transactions from the last day of the month to the fourth day of the following month. The January effect is positive and statistically significant only for the conventional index (JCI) which implies that it is more beneficial for investors to invest only in conventional assets. In contrast, it seems that it is more advantageous for investors to invest only in Islamic assets during Ramadan. In addition, the findings reveal that the two indices exhibit lower returns and higher volatility, which implies that it is recommended for investors to find other assets that can serve as a safe refuge during turbulent periods. Overall, the existence of these calendar anomalies implies that policymakers are called to implement the required measures to increase market efficiency. Originality/value The existing literature on calendar anomalies is abundant, but it is mostly focused on conventional stocks and has not been sufficiently extended to address the presence of these anomalies in Shariah-compliant stocks. To the best of the authors’ knowledge, no study to date has examined the presence of calendar anomalies and asymmetric volatility in both Islamic and conventional stock indices in Indonesia.

Quantitative Assessment and Impact Analysis of Land Surface Deformation in Wuxi Based on PS-InSAR and GARCH Model

Land surface deformation, including subsidence and uplift, has significant impacts on human life and the natural environment. In recent years, the city of Wuxi, China has experienced large-scale surface deformation following the implementation of a groundwater abstraction ban policy in 2005. To accurately measure the regional impacts and understand the underlying mechanisms, we investigated the spatiotemporal characteristics of surface deformation in Wuxi from 2015 to 2023 using 100 Sentinel-1A SAR images and the Persistent Scatterer InSAR (PS-InSAR) technique. The results revealed that surface deformation in Wuxi exhibited significant spatial and temporal variations, with some areas experiencing alternating trends of subsidence and uplift rather than consistent unidirectional change. To uncover the factors influencing this volatility, we conducted a comprehensive analysis focusing on groundwater, precipitation, and soil geology. This study found strong correlations between the groundwater level changes and surface deformation, with the soft soil geology of the area, characterized by alternating layers of sand and clay, further increasing the surface volatility. Moreover, we innovatively applied the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model, typically used in financial analyses, to analyze the subsidence displacement time series in Wuxi. Based on this model, we propose a new “Amplitude Factor” index to evaluate overall surface deformation volatility in the city. Our qualitative assessment of surface stability based on the Amplitude Factor was consistent with research findings, demonstrating the accuracy and effectiveness of the proposed model. These results provide valuable insights for urban planning, construction, and safety control, highlighting the importance of continuous monitoring and analysis of surface deformation volatility for the city’s future development and safety.

Exploring the relationship between Chinese crude oil futures market efficiency and market micro characteristics

This paper uses daily frequency data spanning from March 26, 2018 to February 3, 2023, and combines the exponential generalized autoregressive conditional heteroskedasticity in mean (EGARCH-M) model with the time-varying parameters state-space (TVPSS) model to dynamically test the efficiency of the Chinese crude oil futures market (CCOFM). Additionally, the paper examines the relationship between market efficiency and market micro characteristics such as market depth, market liquidity, noise trading, speculation rate, and information asymmetry. The results show that the CCOFM exhibited weak-form efficiency in June 2020, which lasted until the end of February 2022, and after that, the efficiency showed a poor development trend. Analyzed from the macro level, the Market Maker System and Trader Suitability System have a positive effect on market efficiency, while the Russian-Ukrainian conflict has a negative effect on market efficiency. Analyzed from the micro level, market depth and information asymmetry have a positive effect on market efficiency, while the speculation rate has a negative effect on market efficiency, and noise trading and market liquidity have no significant effect on market efficiency. Market efficiency has a positive effect on market depth and a negative effect on speculation rate. Market liquidity and noise trading have a positive effect on speculation rate, while both information asymmetry and market liquidity have a negative effect on noise trading. In addition, speculation rate has a weak negative effect on information asymmetry, and market depth and speculation rate have a weak positive effect on market liquidity. This study provides theoretical support and decision-making foundation for the development of CCOFM by dynamically exploring the relationship between CCOFM efficiency and micro characteristics.

DAO Dynamics: Treasury and Market Cap Interaction

This study examines the dynamics between treasury and market capitalization in two Decentralized Autonomous Organization (DAO) projects: OlympusDAO and KlimaDAO. This research examines the relationship between market capitalization and treasuries in these projects using vector autoregression (VAR), Granger causality, and Vector Error Correction models (VECM), incorporating an exogenous variable to account for the comovement of decentralized finance assets. Additionally, a Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model is employed to assess the impact of carbon offset tokens on KlimaDAO’s market capitalization returns’ conditional variance. The findings suggest a connection between market capitalization and treasuries in the analyzed projects, underscoring the importance of the treasury and carbon offset tokens in impacting a DAO’s market capitalization and variance. Additionally, the results suggest significant implications for predictive modeling, highlighting the distinct behaviors observed in OlympusDAO and KlimaDAO. Investors and policymakers can leverage these results to refine investment strategies and adjust treasury allocation strategies to align with market trends. Furthermore, this study addresses the importance of responsible investing, advocating for including sustainable investment assets alongside a foundational framework for informed investment decisions and future studies in the field, offering novel insights into decentralized finance dynamics and tokenized assets’ role within the crypto-asset ecosystem.

Crude Oil Prices Forecast Based on Mixed-Frequency Deep Learning Approach and Intelligent Optimization Algorithm.

Precisely forecasting the price of crude oil is challenging due to its fundamental properties of nonlinearity, volatility, and stochasticity. This paper introduces a novel hybrid model, namely, the KV-MFSCBA-G model, within the decomposition-integration paradigm. It combines the mixed-frequency convolutional neural network-bidirectional long short-term memory network-attention mechanism (MFCBA) and generalized autoregressive conditional heteroskedasticity (GARCH) models. The MFCBA and GARCH models are employed to respectively forecast the low-frequency and high-frequency components decomposed through variational mode decomposition optimized by Kullback-Leibler divergence (KL-VMD). The classification of these components is performed using the fuzzy entropy (FE) algorithm. Therefore, this model can fully exploit the advantages of deep learning networks in fitting nonlinearities and traditional econometric models in capturing volatilities. Furthermore, the intelligent optimization algorithm and the low-frequency economic variable are introduced to improve forecasting performance. Specifically, the sparrow search algorithm (SSA) is employed to determine the optimal parameter combination of the MFCBA model, which is incorporated with monthly global economic conditions (GECON) data. The empirical findings of West Texas Intermediate (WTI) and Brent crude oil indicate that the proposed approach outperforms other models in evaluation indicators and statistical tests and has good robustness. This model can assist investors and market regulators in making decisions.

Spillovers across global stock markets before and after the declaration of Russia’s invasion of Ukraine

Since the financial meltdown, studies on systemic risk and financial contagion have gained currency. Events like the COVID pandemic and the Russian invasion of Ukraine have fueled such an importance. This study examines the impact of the invasion on volatility transmissions across major stock markets worldwide. The stock indices considered in this study are ASX 200, ESTOXX 40, FTSE 100, HNGSNG, NIFTY 50, NIKKIE, and S&P 500. The work uses Vector Auto Regression (VAR) to study the transmission of returns. Later, the work performs Dynamic Conditional Covariance-Generalized Auto Regression Conditional Heteroskedasticity (DCC-GARCH) on the residuals where the transmission of returns was significant. The DCC-GARCH (E-GARCH) shows that all the asymmetric transmissions are negative. The study finds that co-movements of stock returns for the following pairs: ESTOXX 50-S&P 500, NIFTY 50-FTSE100, NIFTY 50-NIKKIE, NIKKIE-ESTOXX 50, S&P 500-NIFTY 50, and SP500-HNGSNG significantly intensified after the declaration of invasion. Such intensification of co-movements does establish the contagion effect triggered by invasion. The study shows that ESTOXX 50, which has the closest geographical proximity to the war zone, happens to be the highest generator of spillovers.