Multifractal Cross-correlation Research Articles

Asymmetric multifractality in China’s energy market based on improved asymmetric multifractal cross-correlation analysis

Based on the asymmetric multifractal detrending cross-correlation analysis (MF-ADCCA) method and the multifractal cross-correlation analysis (MFCCA) method, we propose a new method—the asymmetric multifractal cross-correlation analysis (MF-ACCA) method. The simulation results show that the algorithm can describe the asymmetric multifractal characteristics of two time series from qualitative and quantitative perspectives and we compare the MF-ACCA method and the MF-ADCCA method by using Binary ARFIMA model. In addition, the MF-ACCA method is used to analyze the asymmetric cross-correlation relationships among energy markets with different trends. We confirm that when a market experiences large fluctuations, the overall energy market in China exhibits multifractal characteristics, and the multifractal cross-correlations among energy markets are asymmetric.

Nexus of COVID-19 and carbon prices in the EU emission trading system: evidence from multifractal and the wavelet coherence approaches.

In this paper, we analyzed the role of carbon pricing as a vital tool for achieving the sustainable energy transition and discussed policy implications for reaching this goal in the era of the COVID-19 pandemic. To investigate the connection and cross-correlation between COVID-19 cases and carbon price returns in EU ETS, we applied two novel approaches: the multifractal detrending moving average cross-correlation (MF-X-DMA) and the wavelet coherence techniques. The coverage of data is from 2 March 2020 to 19 March 2021. The results of the MF-X-DMA method show that the cross-correlation is substantial and negatively correlated. Moreover, the results imply the presence of multifractal cross-correlations and that a significant change in the number of COVID-19 cases further impacts carbon price fluctuations. Concerning the wavelet coherence method results, we can also find a strong (negative) relationship between selected variables, and this strong correlation is observed throughout the whole study period. In addition, we observe that in the first wave of the COVID-19 outbreak, the changes in the COVID-19 cases are leading the CO2 price returns (particularly in the 16–32 time scale). The results of this study can have significant implications for policymakers and contributions for environmental experts and investors.

Multifractal Cross-Correlations Risk Among WTI and Financial Assets

Multifractal Cross-Correlations Risk Among WTI and Financial Assets

Asymmetry and conduction direction of the interdependent structure between cryptocurrency and US dollar, renminbi, and gold markets

The risk conduction mechanism between a cryptocurrency and the US dollar (USD), the renminbi (RMB), and gold markets is helpful for investors' risk management. Based on the non-linearity and asymmetry of the correlation between the fractal market hypothesis and the financial market, this study uses asymmetric multifractal cross-correlation analysis to study the correlation between a cryptocurrency and USD, RMB, and gold markets. First, using the MF-ADCCA method, an obvious asymmetric cross-correlation is found between a cryptocurrency and USD, RMB, and gold markets. USD, RMB, and gold markets are more effective than cryptocurrency markets. In particular, the impact of the RMB market on the cryptocurrency market is the largest among USD, RMB, and gold. When USD and RMB are rising, the risk conduction to cryptocurrencies is more significant. When gold is falling, the risk conduction to cryptocurrencies is more significant. When cryptocurrencies are falling, the risk conduction to USD is more significant. Among cryptocurrencies, Bitcoin, Ethereum, Litecoin, and New Economic Coin have more significant risk conduction to RMB when it falls. Through the DMF-ADCCA method, a two-way asymmetric conduction effect is found between the lagging cryptocurrency and the USD, RMB, and gold markets. Compared with USD and gold, the risk correlation and conduction between RMB and cryptocurrency are the highest. Preventing and controlling the risk impact of cryptocurrency on the RMB and further strengthening the research between cryptocurrency and the RMB are crucial to the internationalization of the RMB and the improvement of international competitiveness.

Cross-correlations between economic policy uncertainty and precious and industrial metals: A multifractal cross-correlation analysis

This study combines Seasonal-Trend decomposition using LOESS (STL) and multifractal cross-correlation analysis (MFCCA) to estimate the cross-correlations between the evolution of the economic policy uncertainty (EPU) of the USA and return of precious and industrial metals. Based on daily closing values of EPU and eight metals ranging from 01-Jan-2010 to 02-May-2021, the empirical findings confirm the presence of nonlinear dependencies, based on all EPU-Metal relationships, suggesting that EPU and metal markets are interlinked. Furthermore, there is evidence of multifractality, although with a varying degree, between EPU and metal markets. The strongest multifractality is found in EPU-Tin and EPU-Silver, while EPU-Platinum and EPU-Gold exhibit the weakest multifractality. Precious metals show persistent cross-correlations while industrial metals have anti-persistent cross-correlations with EPU, with that cross-correlation behavior being more persistent in the case of small fluctuations, which have different implications for the different market agents.

MULTIFRACTAL CROSS-CORRELATION ANALYSIS BETWEEN CARBON SPOT AND FUTURES MARKETS CONSIDERING ASYMMETRIC CONDUCTION EFFECT

The cross-correlation between carbon spot and futures markets reflects the risk conduction mechanism between the two markets. Deeply depicting and analyzing this risk conduction mechanism is of great significance for investors to carry out risk management strategies. Considering the nonlinear and asymmetric characteristics of cross-correlation between carbon spot and futures markets, this paper applies multifractal cross-correlation analysis method to investigate the cross-correlation between carbon spot and futures markets. Firstly, through Empirical Mode Decomposition (EMD)-Multifractal Detrended Cross-Correlation Analysis (MF-DCCA) detection, it is found that there is an obvious cross-correlation between carbon spot and futures markets, and the cross-correlation has multifractal characteristic. Secondly, by using EMD-time delay-DCCA method, we find the conduction direction between carbon spot and futures markets is bidirectional, and the futures market has a greater impact on the spot market in the short term. Thirdly, through using EMD-MF-ADCCA method, we find the cross-correlation between the two markets is asymmetric, and the cross-correlation between the two markets is more significant when carbon market is in a downward trend than in an upward trend. Fourthly, through constructing EMD-time delay-ADCCA model, we find that there is a two-way asymmetric conduction effect between carbon spot and futures markets when the lag period is short, and when the carbon market is in a downward trend, the conduction effect between the two markets is stronger. However, with the extension of time lag, the conduction effect of the two markets no longer presents obvious asymmetric characteristics.

MULTIFRACTAL TEMPORALLY WEIGHTED DETRENDED CROSS-CORRELATION ANALYSIS OF PM10, NOX AND METEOROLOGICAL FACTORS IN URBAN AND RURAL AREAS OF HONG KONG

Understanding of its correlation to some relevant factors is of paramount importance for modeling and predication of the air pollution process. Compared with the traditional cross-correlation analysis, multifractal detrended cross-correlation analysis (MFDCCA) was argued to be a more suitable method to analyze air pollutant time series due to their non-stationarity nature. Multifractal temporally weighted detrended cross-correlation analysis (MF-TWXDFA) was proposed to improve the shortcomings of MFDCCA. In this study, we apply MF-TWXDFA to investigate the cross-correlation between pollutants ([Formula: see text] and [Formula: see text]) and meteorological factors (temperature, pressure, wind speed (WS) and relative humidity (RH)). The results on the dataset from 1 January 2005 to 31 December 2014 in urban and rural areas of Hong Kong show the existence of multifractal cross-correlation between all pairs of pollutants and meteorological factors in both urban and rural areas. Different from the previous MFDCCA results, we found that the multifractal degree of cross-correlation between [Formula: see text] and (temperature, pressure) is more obvious in urban area. The multifractal strength of cross-correlation between [Formula: see text] and WS is very weak in either urban or rural area. Furthermore, the MF-TWXDFA cross-correlation coefficient [Formula: see text] can capture negative correlation between pollutants and meteorological factors. For [Formula: see text], [Formula: see text] in urban area is less than or close to that in rural area with respect to these four meteorological factors. The [Formula: see text] of [Formula: see text] in urban and rural areas shows more complex patterns for varied meteorological factors. Compared with MFDCCA, MF-TWXDFA can provide much richer information about the relationships between pollutants and meteorological factors, which is beneficial to further modeling and prediction of the air pollution process.

The excessive gaming and gambling during COVID-19

PurposeBy combining econometrics and multifractal methods, utilizing a financial framework, this paper will examine with objectivity the economic, financial and social impact of coronavirus disease 2019 (COVID-19) on society.Design/methodology/approachThrough Granger causality, the authors test the effect of the COVID-19 pandemic on excessive gaming and gambling activities, and through econometrics and multifractal methods, they combine the results to analyze a possible long-run relationship.FindingsThe COVID-19 confirmed cases Granger cause all examined stocks. Based on the co-integration technique, and the multifractal cross-correlation analysis, a long-run relationship exists between all examined stocks and COVID-19.Originality/valueThis is an empirical examination of a very important subject in the field of economics, namely, the consequences of the COVID-19-related events on the behavior of global citizens. It proposes a different and more objective approach (than the interviews and questionnaires) in the examination of this specific subject, through a financial framework, depicting the stock performance of the gaming and online gambling-related companies, and reflecting on the activity of these companies. It combines two different approaches from two different disciplines, namely econometrics and multifractal analysis, to test and describe the causal and the long-run relationship between the phenomena examined, combining the results to an overall and multidimensional view of this occurrence.

A multifractal cross-correlation investigation into sensitivity and dependence of meteorological and hydrological droughts on precipitation and temperature

A multifractal cross-correlation investigation into sensitivity and dependence of meteorological and hydrological droughts on precipitation and temperature

Exploring asymmetric multifractal cross-correlations of price–volatility and asymmetric volatility dynamics in cryptocurrency markets

Asymmetric relationship between price and volatility is a prominent feature of the financial market time series. This paper explores the price–volatility nexus in cryptocurrency markets and investigates the presence of asymmetric volatility effect between uptrend (bull) and downtrend (bear) regimes. The conventional GARCH-class models have shown that in cryptocurrency markets, asymmetric reactions of volatility to returns differ from those of other traditional financial assets. We address this issue from a viewpoint of fractal analysis, which can cover the nonlinear interactions and the self-similarity properties widely acknowledged in the field of econophysics. The asymmetric cross-correlations between price and volatility for Bitcoin (BTC), Ethereum (ETH), Ripple (XRP), and Litecoin (LTC) during the period from June 1, 2016 to December 28, 2020 are investigated using the MF-ADCCA method and quantified via the asymmetric DCCA coefficient. The approaches take into account the nonlinearity and asymmetric multifractal scaling properties, providing new insights in investigating the relationships in a dynamical way. We find that cross-correlations are stronger in downtrend markets than in uptrend markets for maturing BTC and ETH. In contrast, for XRP and LTC, inverted reactions are present where cross-correlations are stronger in uptrend markets.

Fluctuation Characteristics of Water Level and Water Temperature of Huize Well Based on MF-DCCA

In recent decades, the network of seismic subsurface fluid observatories is developing constantly, the observation data of subsurface fluids are enriched accordingly, which provides a favorable condition for the research on the formation, occurrence, and development of earthquakes. In the observation data of subsurface fluids, water level and water temperature changes are very important observation indicators, and their fluctuation sequences are quite complicated. Therefore, this paper employed a non-linear cross-correlation method to study the relationship between the water level and water temperature of Huize Well from 2004 to 2006, and found that there’s a significant cross-correlation between the time series of water level and water temperature; then, this study adopted DCCA (detrended cross-correlation analysis) to calculate the cross-correlation coefficient under different scales and explore the continuous changes of water level and water temperature; at last, this paper used the MF-DCCA (Multifractal-DCCA) method to prove that there’s multifractal cross-correlation between the time series of water level and water temperature. Before the M5.1 earthquake in Huize area, there’s an abnormal increase in the width of the multifractal spectrum of the water level and water temperature drawn with a sliding window of 500-hour, and this is a possible earthquake precursor.

Multifractal Cross-correlations between foreign exchange rates and interest rate spreads

We apply the Multifractal detrended moving average analysis (MF-DMA) and the Multifractal Cross-Correlation Analysis(MFCCA) to study the cross-correlation behaviors between foreign exchange markets and interest rate differentials. Our empirical results obtained from the cross-correlation test, qDCCA coefficient and MFCCA confirm the existence of strong multifractality cross-correlations between foreign exchange rates and interest rate differentials. Foreign exchange rate has a stronger cross-correlation with short-term interest rate differentials in Australia, Canada, Japan, the United Kingdom (UK), and European Union (EU), and with long-term interest rate differentials in China, respectively. The results of rolling window analysis suggest varying cross-correlations between the exchange rates and interest rate differentials during the sample period. In addition, the cross-correlations between exchange rates and interest spreads vary with period, country and maturity of the spreads.

Cross-correlations between the P2P interest rate, Shibor and treasury yields

Using interest rate market data, this paper employs a multifractal cross-correlation analysis (MFCCA) to study the cross-correlations between market interest rate, treasury yields and policy interest rate. Cross-correlation statistics and coefficients verify the existence of cross-correlations, and the MFCCA method quantitatively confirms the presence of multifractality between the three time series for both the long and short term. Small fluctuations are persistent, while large fluctuations are generally anti-persistent in the long term. The results of the rolling window analysis reveal that cross-correlation scaling exponents are sensitive to external shocks.

Co-movements between Shanghai Composite Index and some fund sectors in China

In this article, we analyzed the cross-correlations between Shanghai Composite Index (SSEC) and some fund sectors in China. Four high-volume fund sectors such as finance, medicine, new energy, and consumption sectors were investigated. Multifractal Cross-Correlation Analysis (MFCCA) approach was conducted for the empirical researches of the long-range correlations for time series pairs. The obtained multifractal characteristics showed that the finance sector achieved the highest persistence of cross-correlations, then the new energy, consumption, and medicine sector. Furthermore, the Δλ of finance sector is the greatest among other sectors, which indicated that the multifractality of cross-correlations between SSEC and finance sector was the strongest, and then the medicine sector has the weakest multifractality of cross-correlations. In addition, we utilized one-tailed Student’s t-test to further evaluate the multifractality of cross-correlations, the results verified our conclusion.

Multifractal temporally weighted detrended partial cross-correlation analysis of two non-stationary time series affected by common external factors

When common factors strongly influence two cross-correlated time series recorded in complex natural and social systems, the results will be biased if we use multifractal detrended cross-correlation analysis (MF-DXA) without considering these common factors. In order to better study the time series of such cases, we extend the multifractal temporally weighted detrended cross-correlation analysis (MF-TWXDFA) proposed by our group (Wei et al., 2017) and propose multifractal temporally weighted detrended partial cross-correlation analysis (MF-TWDPCCA) to quantify intrinsic power-law cross-correlation of two non-stationary time series affected by common external factors in this paper. To test the performance of MF-TWDPCCA, we apply it and multifractal partial cross-correlation analysis (MF-DPXA) proposed by Qian et al. (2015) on simulated series. Numerical tests on artificially simulated series demonstrate that MF-TWDPCCA can more accurately detect the intrinsic cross-correlations for two simultaneously recorded series than MF-DPXA and MF-TWXDFA. To further show the utility of MF-TWDPCCA, we apply it on time series from stock markets and find that there exists significantly multifractal power-law cross-correlation between stock returns. In addition, a new partial cross-correlation coefficient is defined to quantify the level of intrinsic cross-correlation between two time series.

Multifractal analysis of air and soil temperatures.

Air and soil temperatures are important agrometeorological variables with several applications. Understanding the complex behavior of air and soil temperatures, as well as their interaction, will help in agricultural planning. Multifractal detrended fluctuation and multifractal cross-correlation analysis of air and soil temperatures were carried out in three locations (Akure, Abuja, and Bauchi) within a tropical country, Nigeria. Monthly and annual air and soil temperatures measured at 5 min intervals for a period of 1 year were obtained and analyzed for multifractality. There is evidence of seasonal dependence in the multifractal behavior of monthly soil temperature. Monthly temperatures (air and soil) were found to have higher degrees of multifractality than annual temperatures. Furthermore, latitudinal dependence was observed in the multifractal behavior of air and soil temperatures. The cross-correlation between air and soil temperatures also shows multifractality with persistence at the monthly scale and anti-persistence at the annual scale. This work has shed light on the complex relationship between air and soil temperatures, and the results will be useful in modeling the two variables.

Asymmetric multi-fractal cross-correlations of the price of electricity in the US with crude oil and the natural gas

The price of commodities are often difficult to forecast due to the underlying characteristics of the series such as autocorrelation, heteroskedasticity, and non-linearity. For many commodities, it is both socially and financially detrimental to have too much error in the forecasts. In the case of electricity production, both stock-outs and over production result in failure in the form of blackouts on one side and waste on the other. Therefore, it is desirable to improve the models that are used to forecast both the production and demand of electricity in order to optimize the match between supply and demand. To accomplish the aforementioned, understanding of the behavior of the commodity is necessary. In this paper, we apply a multi-fractal asymmetric detrended cross-correlation analysis to analyze the presence and asymmetry of the cross-correlations between the price of electricity in U.S. with respect to crude oil and natural gas markets. Our data draws from the major producers of electricity in the U.S. and evaluates the multi-fractal asymmetric detrended cross-correlation with respect to WTI and Natural Gas; both important inputs into the production of electricity. Our findings illustrate the fractal and cross-correlation relationship between electricity production and commodity prices.

A Multifractal Cross-Correlation Analysis of Economic Policy Uncertainty: Evidence from China and US

In this study, we utilized the prevailing economic policy uncertainty index (EPU) as the proxy of state economic fluctuation and investigated Sino–US economic fluctuation long horizon cross-correlation with a multifractal detrended cross-correlation analysis (MF-DCCA). With the MF-DCCA approach, we found a reliable long-range cross-correlation between China and US EPU changes. In addition, we discovered that a power law cross-correlation existed for the variation of most scaling orders. However, no persistence of cross-correlations was detected within the Sino–US EPU change series. Additionally, we implemented Rényi exponent and spectrum singularity checks. Both the examination results proved series multifractality with the presented arch-shaped curves. We further calculated the Hölder exponent bounds within each series and found that the China EPU changes had maximal multifractality with the largest exponent difference.

Economic Policy Uncertainty and Stock Markets: A Multifractal Cross-Correlations Analysis

We utilize the multifractal detrended cross-correlation analysis (MF-DCCA) to investigate the cross-correlations between the US economic policy uncertainty (EPU) and US stock markets in the framework of Fractal Market Hypothesis (FMH). The data contain daily closing values of EPU, and the returns of Dow Jones Industrial Average Index (DJI), S&P 500 index (GSPC) and NASDAQ Composite Index (IXIC). Our empirical results show that changes in EPU and fluctuations in the US stock markets interact in a nonlinear way. Furthermore, there exists significant multifractality in the cross-correlations between EPU and stock markets. The cross-correlations exhibit dynamics and are affected by major international events. We capture the underlying mechanisms such as multifractality and nonlinear relation that dominate EPU-US stock markets nexus by means of FMH. The findings add a new dimension to the existing literature, and are important for market participants to adjust investment decisions.

Cross-correlations between price and volume in China's crude oil futures market: A study based on multifractal approaches

Abstract In this paper, we investigate the multifractal characteristics of the cross-correlations between price and volume in China's first crude oil futures market (INE), and compared the INE market with Brent and WTI market. Multifractal detrended cross-correlation analysis (MF-DCCA) was applied to comprehensively research the multifractal characteristics of the three different crude oil futures markets based on the data of the past five years. The results show that the cross-correlations between price change and volume change have significant multifractal characteristics in three markets, and the degree of multifractality is stronger in the WTI market for the full sample. Due to the INE market appeared late, the sample size of INE is not consistent with that of the other two markets. In order to solve this influence, we further compare the multifractal characteristics of the three crude oil futures markets in the same period (March 26, 2018 to November 20, 2020) and compares the results with those of the full sample period. The empirical results show that the degree of multifractality in both Brent and WTI markets is increased. To further verify whether this result is due to the emergence of the INE crude oil market, the entire sample of the Brent and WTI markets are divided into two periods according to the launch time of INE market. The result shows that the degree of multifractality and the risk of these two markets are all stronger in the last two years. Thus, there are reasons to believe that the emergence of INE market has a certain impact on the traditional Brent and WTI market, and a greater impact on Brent market. However, MFDCCA only consider the overall situation and cannot describe the multifractal cross-correlation between the upward and downward trends of the market. Thus, it probably won't accurately reflect the nonlinear cross-correlation. Furthermore, we study the characteristics of the asymmetric multifractal cross-correlations in different trends (upwards or downwards) by using the multifractal asymmetric detrended cross-correlation analysis (MF-ADCCA) method. Empirical results indicate that the cross-correlation between price and volume in three markets shows obvious asymmetry. Specifically, when the price changes series is on an upward trend, or when the volume changes series is on a downward trend, the degree of asymmetry in three markets is stronger, and the risk situation of the market is more complex. Therefore, for a portfolio of assets with different markets, investors should fully observe and understand different market trends and make reasonable investment based on their own risk preference.