Volatility Of Financial Assets Research Articles

Dynamic Prediction Model of Financial Asset Volatility Based on Bidirectional Recurrent Neural Networks

Predicting financial market volatility is essential for investors and risk management. This study proposes a dynamic prediction model for financial asset volatility, with a Bi-directional Recurrent Neural Network (Bi-RNN) utilized to cleverly address market complexity. Our framework integrates Bi-RNN and gated recurrent units (GRU) to perform global optimization via particle swarm optimization algorithm (PSO). Bi-RNN combines historical data and future expectations, while GRU effectively solves long-term dependency issues through a gating mechanism, which enhances model generalization. Experimental results show that the model exhibits significant performance advantages on different financial datasets, along with strong learning and generalization capabilities superior to traditional methods. This research provides advanced and practical solutions for financial asset fluctuation prediction and is of positive significance for the greater accuracy of investment decisions and risk mitigation.

Rough Volatility: Fact or Artefact?

We investigate the statistical evidence for the use of ‘rough’ fractional processes with Hurst exponent H<0.5\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$H< 0.5$$\\end{document} for modeling the volatility of financial assets, using a model-free approach. We introduce a non-parametric method for estimating the roughness of a function based on discrete sample, using the concept of normalized p-th variation along a sequence of partitions. Detailed numerical experiments based on sample paths of fractional Brownian motion and other fractional processes reveal good finite sample performance of our estimator for measuring the roughness of sample paths of stochastic processes. We then apply this method to estimate the roughness of realized volatility signals based on high-frequency observations. Detailed numerical experiments based on stochastic volatility models show that, even when the instantaneous volatility has diffusive dynamics with the same roughness as Brownian motion, the realized volatility exhibits rough behaviour corresponding to a Hurst exponent significantly smaller than 0.5. Comparison of roughness estimates for realized and instantaneous volatility in fractional volatility models with different values of Hurst exponent shows that, irrespective of the roughness of the spot volatility process, realized volatility always exhibits ‘rough’ behaviour with an apparent Hurst index H^<0.5\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\widehat{H}<0.5$$\\end{document}. These results suggest that the origin of the roughness observed in realized volatility time series lies in the estimation error rather than the volatility process itself.

Volatility of financial assets in industrial sector companies listed on the Quito Stock Exchange

The study aimed to analyze the volatility of financial assets of companies in the industrial sector listed on the Quito Stock Exchange. An econometric model was used to determine the causes of price variation that affect the returns on variable income financial assets according to a quantitative-correlational strategy. Of the five companies taken as a sample, it was evident that most have a positive beta coefficient of less than 1. This result conveyed that asset fluctuations are below market demand, and they conceive behavior dependent on it. The study concluded that volatility has not shown notable fluctuations in the Ecuadorian stock market due to the minimal movements generated in the negotiations on the Quito Stock Exchange.

Features of realized volatility analysis and return predicting based on LGBM and RNN model

This paper proposes a method of predicting the realized volatility of financial assets using LGBM and RNN models. The study utilizes Convolutional Neural Networks to construct sub-indicators capturing the liquidity and volatility of financial assets. These sub-indicators are used to develop comprehensive measures of liquidity and volatility. Lognormal random walk theory is applied to each asset dimension to price volatility for multiple assets, and the value of European options independent of path is obtained via multiple integration. Monte Carlo method is applied to solve the integral, which becomes inefficient in the case of high dimension and orthogonality. This study also involves leveraging LGB and other models to efficiently exploit data to create high returns and achieve the highest sharp ratio. The current dataset, which comes from a recognized international market maker, includes stock market data that is important for trade execution in the financial markets, particularly snapshots of the order book and executed trades. The study shows that the proposed method can accurately predict the realized volatility of financial assets.

The impact of Covid-19 and Russia–Ukraine war on the financial asset volatility: Evidence from equity, cryptocurrency and alternative assets

This study investigates the volatility and external shock persistence within the financial and alternative assets markets during times of crises triggered by Covid-19 and the war in Ukraine. Univariate GARCH family models are used to capture the effect of financial turmoil caused by recent crises. Five different class of assets (which includes Islamic, ESG, Conventional, Crypto, FinTech, and commodities) have been chosen to represent a sample of the worldwide traditional financial market and alternative assets. The findings of this study revealed that almost all financial and alternative assets experienced an increase in volatility, except Bitcoin, across all observation periods. Islamic stock and ESG indexes exhibited high volatility before the Covid-19 outbreak. During the pandemic, all assets became more volatile. In addition, Islamic equities and ESG indexes showed relatively lower risk compared to conventional stocks and other alternative assets during the war. Multiple financial assets tend to be highly volatile during crises; however, global investors need to consider the advantages of incorporating Islamic stocks and ESG indexes as part of their investment portfolio innovation strategy, particularly in the presence of geopolitical risk.

Forecasting stock market volatility: The sum of the parts is more than the whole

The volatility of financial assets can be decomposed into upside volatility and downside volatility. However, these two components have unique properties, so their predictability is completely different. In this paper, we explore a new forecasting method to predict the S&P 500 volatility by separately modeling upside volatility and downside volatility and summing the forecasts up. Our new method is proved to have better performance compared with directly modeling aggregate volatility. Moreover, the gains in forecast accuracy are robust concerning the individual and combined models.

Impact of geopolitical risk on the volatility spillovers among G7 and BRICS stock markets

Geopolitical risk (GPR) plays an important role in the international financial interactions. Previous studies have mostly analyzed the impact of the GPR on the return or volatility of financial assets, and less attention has been paid on the connection among international stock markets. Therefore, this study attempts to explore the impact of GPR on the volatility spillovers among G7 and BRICS stock markets using the impulse response function. To be specific, TVP-VAR model is utilized to capture the dynamic volatility spillovers among 12 stock markets, and we obtain the total and directional spillover indexes. Then, impulse response function based on the VAR model is applied to capture the dynamic effect of GPR on the spillover series. Based on the weekly price data ranging from August 16, 2002 to April 10, 2023, we find that the volatility spillover index increases when there are large shocks to the market such as war, financial crisis and COVID-19. In addition, the impact of GPR on the total spillover index is negative in the early and late periods, and it shows a more significant positive effect in the middle period. These results reflect the dynamic impact of GPR on the connection among international stock markets, which is essential for the international financial risk management.

Modeling asymmetric volatility of financial assets using univariate GARCH models: An Indian perspective

In recent years, numerous models with various amounts of variance have been developed to estimate and forecast important characteristics of time series data. While there are many studies on asymmetric volatility and accuracy testing of univariate Generalized Autoregressive Conditional Heteroscedasticity models, there are no parallel studies involving multiple financial assets and different heteroscedastic models and density functions. The objective of this study is to contrast the forecasting accuracy of univariate volatility models with Normal and Student-t distributions in forecasting the volatility of stock, gold futures, crude futures, exchange rate, and bond yield over a 10-year time span from January 2010 through December 2021 in Indian market. The results of exponential, threshold and asymmetric power models show that the volatility stock (–0.12047, 0.17433, 0.74020 for Nifty, and –0.1153, 0.1676, 0.7372 for Sensex), exchange rate (–0.0567, 0.0961,0.9004), crude oil futures (-0.0411, 0.0658, 0.2130), and bond yield (–0.0193, 0.0514 and –0.0663) react asymmetrically to good and bad news. In case of gold futures, an inverse asymmetric effect (0.0537, –0.01217, –0.1898) is discovered; positive news creates higher variance in gold futures than bad news. The Exponential model captures the asymmetric volatility effect in all asset classes better than any other asymmetric models. This opens the door for many studies in Indian financial market.

РЕАЛИЗАЦИЯ ТОРГОВОГО СОВЕТНИКА ДЛЯ МУЛЬТИРЫНОЧНОЙ ПЛАТФОРМЫ METATRADER 5

The article describes the process of creating a flexible trading strategy for algorithmic tradingin a specialized development environment MQL5 IDE in the MetaTrader 5 multi-asset platform.The advantages and expediency of using the MetaTrader 5, MetaTrader 4 platforms andtheir respective trading applications Trade Assistant, Forex Trade Manager, Trade Time Manager,CAP Gold Albatross EA and Fast Copy are shown. A comparative analysis of the existing implementationsof trading advisors based on various indicators, as well as those created using intelligenttechnologies, has been carried out. In the previously implemented trading advisors, for predictingthe prices of the volatility of financial assets, flexible learning algorithms, compensatoryfuzzy logic models, and technical analysis tools are mainly used, which entails high time costs, inconditions of high financial market volatility. To solve this problem, the authors propose an integratedapproach based on the use of technical analysis tools built into the MetaTrader 5 multimediaplatform and the trading strategy automation algorithm, which makes it possible to obtain aforecast of a given accuracy for the selected instrument in real time. The paper substantiates theneed to introduce elements of automatic trading when analyzing the quotes of financial instrumentsand managing a trading account in order to avoid mechanical, analytical, organizationaland psychological mistakes made by traders. The study shows step by step the process of creating,debugging, testing, optimizing and executing the implemented trading advisor. An algorithm forautomating a trading strategy has been developed and its block diagram has been presented.The initial data for the trading strategy automation algorithm are determined, and the mathematicalapparatus for calculating indicators of limit orders of the TakeProfit and StopLoss types isdescribed. Since exchange trading is associated with many risks, we analyzed the impact of differentvalues of lots of TakeProfit and StopLoss limit orders on possible profit and drawdown limit(loss). As a result, the EA worked correctly in real time without human intervention for eightweeks using two trading strategies. The results of testing the developed software allow us to drawthe following conclusions: when the EA shows a high degree of recommendation, the actual financialassets show high efficiency.

Modelling returns volatility: mixed-frequency model based on momentum of predictability

The estimation and prediction of financial asset volatility are important in terms of theoretical and practical applications. Considering that low-frequency and high-frequency information plays an important role in volatility prediction, this article proposes a mixed-frequency model based on the momentum of predictability (MF-MoP). To illustrate the advantages of the proposed model, comparative research is conducted on the prediction accuracy of volatility among the GARCH model, the Realized GARCH model and the MF-MoP model, by the loss function and MCS test. The empirical results show that the MF-MoP model has higher prediction accuracy than the other two models; especially based on skewed-t distribution, the MF-MoP significantly outperforms the competing models. Moreover, the MF-MoP model can improve the forecasting of volatility, regardless of different look-back periods (including 1, 3, 6 and 9 days), different data (including the CSI 300 index, the N225 index and the KS11 index), and realized measures (including RV, RRV and MedRV), indicating that the model is robust.

Modeling the stochastic volatility of MAD/EURO and MAD/USD the exchange rates by the Bayesian approach and the MCMC (Monte Carlo Markov Chain) algorithm

PurposeOne of the crucial issues in the contemporary finance is the prediction of the volatility of financial assets. In this paper, the authors are interested in modelling the stochastic volatility of the MAD/EURO and MAD/USD exchange rates.Design/methodology/approachFor this purpose, the authors have adopted Bayesian approach based on the MCMC (Monte Carlo Markov Chain) algorithm which permits to reproduce the main stylized empirical facts of the assets studied. The data used in this study are the daily historical series of MAD/EURO and MAD/USD exchange rates covering the period from February 2, 2000, to March 3, 2017, which represent 4,456 observations.FindingsBy the aid of this approach, the authors were able to estimate all the random parameters of the stochastic volatility model which permit the prediction of the future exchange rates. The authors also have simulated the histograms, the posterior densities as well as the cumulative averages of the model parameters. The predictive efficiency of the stochastic volatility model for Morocco is capable to facilitate the management of the exchange rate in more flexible exchange regime to ensure better targeting of monetary and exchange policies.Originality/valueTo the best of the authors’ knowledge, the novelty of the paper lies in the production of a tool for predicting the evolution of the Moroccan exchange rate and also the design of a tool for the monetary authorities who are today in a proactive conception of management of the rate of exchange. Cyclical policies such as monetary policy and exchange rate policy will introduce this type of modelling into the decision-making process to achieve a better stabilization of the macroeconomic and financial framework.

Forecasting stock volatility and value-at-risk based on temporal convolutional networks

In recent years, deep learning has attracted increasing popularity in modern financial fields. The volatility of financial asset returns as well as the Value-at-Risk (VaR) play a significant role in many applications such as risk management, investment portfolios and etc. Thus, it is extremely essential to accurately estimate volatility and VaR. Temporal convolutional networks (TCNs), a relatively new deep learning architecture for solving sequential modeling tasks, have demonstrated convincingly good performance in many applications. In this paper, we utilize TCNs to forecast stock volatility and VaR. To the best of our knowledge, this is the first attempt to address this task with TCNs. In the experiments conducted with both synthetic data and some real stock data, TCNs are compared with other twelve popular models which include nine conventional approaches (i.e., three GARCH-type models with each being considered three tail distributions) and three deep learning methods (i.e., LSTM, LSTM with attention mechanism and GRU). The Friedman test followed by the Nemenyi post-hoc test is also employed to analyze whether TCNs perform significantly better than the other methods across the real stock datasets. As for volatility modeling, experimental results show that TCNs outperforms all the other methods in terms of RMSE (root mean squared error) and MAE (mean absolute error). In the meantime, TCNs behave best in calculating VaR when evaluating their performance with several metrics. More importantly, the superiority of TCNs over GARCH-type methods are statistically significant. As a result, TCNs can be regarded as an important technique to forecast return volatility and the associated VaR.

Modeling and Forecasting Volatilities of Financial Assets with an Asymmetric Zero-Drift GARCH Model

Abstract In this study, we extend the zero-drift generalized autoregressive conditional heteroskedasticity (GARCH) model to incorporate the well-known asymmetric effects of shocks on financial volatility and propose an asymmetric zero-drift GARCH (AZD-GARCH) model. Relevant asymptotics of the new model, including those for the quasi-maximum-likelihood estimator and the powers of the stability test and the model misspecification test, are comprehensively discussed with simulation evidence. Our empirical studies focus on the daily Brent oil price, the AUD/USD exchange rate, and the S&amp;P 500 returns covering the recent 2019–2020 period. The results demonstrate the usefulness of the AZD-GARCH model in understanding the volatility features of financial assets and the model’s superiority to a range of competitors in precisely forecasting volatilities. Robustness checks on data for extended sample periods (2017–2020 and 2009–2020) further provide highly consistent results. Therefore, the proposed AZD-GARCH model can help policymakers and market participants in various applications, such as monitoring asset volatility and hedging relevant risks.

How Does Uncertainty Affect Volatility Correlation between Financial Assets? Evidence from Bitcoin, Stock and Gold

ABSTRACT This paper deciphers the correlation of volatility between Bitcoin, stock and gold, in the context of uncertainty. The wavelet analysis results indicate that the selected assets are primarily positively correlated with each other, specifically in periods when the economic policy uncertainty (EPU) is high. Furthermore, the logit regression confirms that the EPU and categorial EPU indices have heterogeneous effects on the interdependence between Bitcoin, the S&P 500 and gold. Therefore, our findings provide insights for policy-makers to reduce the adverse impact of uncertainty on financial asset volatility.

Popular cryptoassets (Bitcoin, Ethereum, and Dogecoin), Gold, and their relationships: volatility and correlation modeling

Cryptoassets have experienced dramatic volatility in their prices, especially during the COVID-19 pandemic era. This pilot study explores the volatility asymmetry and correlations among three popular cryptoassets (Bitcoin, Ethereum, and Dogecoin) as well as Gold. Multiple Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models are analyzed. We find that positive shocks have a greater impact on the volatility of these financial assets than negative shocks of the same magnitude, perhaps a manifestation of the fear of missing out (FOMO) effect. Our research is one of the first to use COVID-19-period volatility of financial assets (in-sample data) to forecast their later COVID-19-period volatility (out-of-sample data). This forecast accuracy is compared to that produced by forecasts using the same out-of-sample data and a longer in-sample data. Our results indicate that generally, the larger in-sample dataset gives a higher forecast accuracy though the smaller in-sample dataset is from the same regime as the out-of-sample data. We also evaluate the correlations among the assets using the Dynamic Conditional Correlation (DCC) framework and find that there is an elevated positive correlation between Gold and Bitcoin during the past two years. The Gold-Bitcoin correlation hit its peak during the peak of the COVID-19 pandemic and then fell back to around zero in July 2021 when the pandemic crisis eased. Unsurprisingly, there is a strong positive correlation among the cryptocurrencies. Pairwise correlation among all four assets was stronger during the COVID-19 pandemic. Such continuing analysis can inform portfolio asset allocation as well as general financial policy decisions.

Specification tests for univariate diffusions

A new class of specification tests for stochastic differential equations (SDE) is proposed to determine whether the probability integral transform of the estimated model generates an independent and identically distributed uniform random variable. The tests are based on Neyman’s smooth test, appropriately adjusted to correct for both the size distortion arising from having to estimate the unknown parameters of the SDE and possible dependence in the uniform random variable. The suite of tests is compared against other commonly used specification tests for SDEs. The finite sample properties of the tests are investigated using a range of Monte Carlo experiments. The tests are then applied to testing the specification of SDEs used to model the spot interest rate and financial asset volatility.

Forecasting volatility by integrating financial risk with environmental, social, and governance risk

AbstractThe study aims to verify whether the consideration of a risk measure based on environmental, social, and governance (ESG) factors can reduce the difference between the ex‐ante financial risk and ex‐post volatility of financial assets. The statistical models are run on 17,996 firm‐year observations (3332 active firms from 55 countries and 10 industries, listed on the ECPI Global Ethical Equity index) in 2007–2015. According to our main results, the forecasting effectiveness of traditional financial risk measures can be improved by integrating financial risk with an ESG risk measure that considers the ESG entropy. We found that the dispersion of ESG scores within a country, sector and year is a risk factor that would be helpful in predicting the volatility of financial assets. Other similar long‐run risk measures, such as issuers' credit ratings, do not reveal the same forecasting power. By reducing unexpected volatility, especially in the medium term, the ESG risk measure provides investors and fund managers with a useful metric for decision making.

SKEW NORMAL AND SKEW STUDENT-T DISTRIBUTIONS ON GARCH(1,1) MODEL

The Generalized AutoRegressive Conditional Heteroskedasticity (GARCH) type models have become important tools in financial application since their ability to estimate the volatility of financial time series data. In the empirical financial literature, the presence of skewness and heavy-tails have impacts on how well the GARCH-type models able to capture the financial market volatility sufficiently. This study estimates the volatility of financial asset returns based on the GARCH(1,1) model assuming Skew Normal and Skew Student-t distributions for the returns errors. The models are applied to daily returns of FTSE100 and IBEX35 stock indices from January 2000 to December 2017. The model parameters are estimated by using the Generalized Reduced Gradient Non-Linear method in Excel’s Solver and also the Adaptive Random Walk Metropolis method implemented in Matlab. The estimation results from fitting the models to real data demonstrate that Excel’s Solver is a promising way for estimating the parameters of the GARCH(1,1) models with non-Normal distribution, indicated by the accuracy of the estimation of Excel’s Solver. The fitting performance of models is evaluated by using log-likelihood ratio test and it indicates that the GARCH(1,1) model with Skew Student-t distribution provides the best fitting, followed by Student-t, Skew-Normal, and Normal distributions.

Contagion of COVID-19 pandemic between oil and financial assets: the evidence of multivariate Markov switching GARCH models

PurposeIn this paper, we investigate empirically the time-frequency co-movement between the recent COVID-19 pandemic, G7stock markets, gold, crude oil price (WTI) and cryptocurrency markets (bitcoin) using both the multivariate MSGARCH models.Design/methodology/approachThis paper examines the relationship between the volatilities of oil, Chinese stock index and financial assets (cryptocurrency, gold, and G7 stock indexes), for the period January 17th 2020 to December 10th 2020. It tests the presence of regime changes in the GARCH volatility dynamics of bitcoin, gold, Chinese, and G7 stock indexes as well as oil prices by using Markov–Switching GARCH model. Also, the paper estimates the dynamic correlation and volatility spillover between oil, Chinese and financial assets by using the MSBEKK-GARCH and MSDCC-GARCH models.FindingsOverall, we find that all variables display a strong volatility concentrated in the first four months of Covid-19 outbreak. The paper conducts different backtesting procedures of the 1% and 5% Value-at-Risk forecasts of risk. The results find that gold has the lowest VaR. However, the Canadian and American indices have the highest VaR, for respectively 1% and 5% confidence level. The estimation results of MSBEKK-GARCH prove the volatility spillover between Chinese index, oil and financial assets. Although, the past news about shocks in the Chinese index significantly affects the current conditional volatility of financial assets. Moreover, for the high regime, the correlation increased between Chinese and G7 stock indexes which proving the contagion effect of the COVID-19 pandemic. On the contrary, the correlation decreased between Chinese-gold and Chinese-bitcoin, which confirming that gold and bitcoin can be considered as an alternative hedge for some investors during a crisis. During the COVID-19 pandemic, the correlations for the couples oil-gold and oil-bitcoin peaked. Contrary to gold, bitcoin cannot be considered as a safe haven during the global pandemic when investing in crude oil.Originality/valueIn contrast, comparative analysis in terms of responses to US COVID-19 pandemic, the US Covid-19 confirmed cases have relative higher impact on the co-movement in WTI and bitcoin. This paper confirms that gold is a safe haven during the COVID19 pandemic period.

Borsa Endeksi, Döviz Kuru, Faiz Oranları ve CDS Primleri Arasındaki Oynaklık Yayılımları: Türkiye Örneği

In recent years, the volatility of financial assets has significantly increased. High volatility causes instability in financial markets and increases portfolio risks. Under these conditions, modeling the volatility, determining the volatility relationships, and volatility spillovers are important for market actors. In this study, for the period January 2, 2010 - April 10, 2020, it was investigated the volatility spillover and correlation relationship between the Borsa Istanbul (BIST) 100 Index, the exchange rates ($/ ), interest rates, and credit default swap (CDS) premiums. It was also examined the effects of economic, political, and social factors on the correlation relationship. In the study, it was seen that the bidirectional volatility spillover between the BIST 100 index - exchange rates, interest rates - exchange rates, and CDS premiums – interest rates, on the other hand, unidirectional volatility spillover from CDS premiums to the exchange rates. A positive volatility relationship was determined between CDS premiums - interest rates and CDS premiums - exchange rates while the negative volatility relationship between interest rates - BIST 100 index and CDS premiums - BIST 100 index. It was also determined that political, economic, and social factors affect the relationships of correlation between variables.