Search

This study focused on analyzing the complexities and risk spillovers that arise among financial institutions due to the development of financial markets. The research employed the conditional value at risk (CoVaR) methodology to quantify the extent of tail risk spillover and constructed a risk spillover network encompassing Chinese financial institutions. The study further investigated the characteristics, transmission paths, and dynamic evolution of this network under different risk conditions. The empirical findings of this research highlighted several important insights. First, financial institutions play distinct roles in the risk spillover process, with the securities and banking sectors as risk exporters and the insurance and diversified financial sectors as risk takers. The closest risk spillover relationships were observed between banking and insurance and between securities and diversified financial sectors. Second, in high-risk scenarios, there is significant intrasectoral risk transmission between banks and the diversified financial sector, as well as dual-sectoral risk contagion between banks and securities, with the most-common transmission occurring between diversified financial and securities sectors. Finally, the securities sector acts as the pivotal node for risk spillovers, being the main transmitter of intersectoral risks. The formation and evolution of risk spillover networks are influenced by endogenous mechanisms, in particular the convergence effect.

Purpose The purpose of this paper is to whether competition between platforms can be effective, thus leading to efficient allocations. Design/methodology/approach Based on the classic linear Hoteling model, this paper builds a two-period competition model for two competing platforms using two variants, namely, a discrimination pricing model and a unified pricing model. Findings In the case of the discrimination pricing model, the competition is moderate, and the two platforms split the market evenly in the first stage, while both platforms tended to offer preferential treatment to new users and set higher prices for regular customers in the second stage. Compared to the unified pricing model, in the first stage, the platform can provide a higher price that depends on the cross-network effect when it implements discrimination, and thus, obtains higher profits. However, in the second stage, fierce competition leads to the release of benefits, new and regular customers obtain lower prices and the platforms lose higher profits. In the long-run, discriminatory pricing is not the best option due to lower total profits. The two platforms will implement cooperative pricing or one platform becomes dominant. Originality/value Instead of focusing on the cross-network effects, this paper emphasizes the role of the same-side network effect on price discrimination regarding the platforms’ competition. The same-side network effects are investigated in relation to a discrimination pricing strategy and compared to a unified pricing strategy. Another innovative aspect is the study of these network effects in a dynamic setting based on a two-period competition model for two platforms.

Although transfer entropy can test the nonlinear causal relationship between sequences, it is mainly used for stationary data. For nonstationary sequences with large fluctuations, the traditional transfer entropy method has obvious defects. Based on traditional transfer entropy, this paper proposes a transfer entropy method with rolling windows. We verify that this new method can capture the dynamic order between sequences and better reveal the nonlinear causality between nonstationary time series. Furthermore, we construct an investor sentiment index based on principal component analysis, and based on the proposed dynamic transfer entropy model, we analyze the information transfer relationship among economic policy uncertainty (EPU), investor sentiment and stock markets. The results of the information flow analysis of EPU and investor sentiment show that EPU influenced investor sentiment mainly from August 2015 to June 2016. Among different policies, China’s exchange rate reform policy and ‘circuit-breaker’ policy in the stock market have played an important role. The analysis of the information flow between sentiment and stock price returns shows that investor sentiment is more a reflection of changes in stock price returns with a 1-month lag order and that the stock market has a significant bargainer effect and a weaker bandwagon effect. There is no significant information flow transmission relationship between EPU and stock market volatility, which indicates that stock market fluctuations are basically not affected by national policy fluctuations. Although investor sentiment is affected by changes such as exchange rate reform and stock market policies, many investors do not form consensus expectations.

The paper firstly studies the static tail dependence structure between the economic policy uncertainty (EPU) index and several financial markets (Brent Oil, CDS, VIX, SP500 and UK EPU) with Copula models. The results show significant negative upper tail dependence between the EPU index and the SP500 index in most years, but also show positive upper tail dependence during the 2016 period. Further, the negative tail dependence area between EPU and CDS also demonstrates that higher economic policy uncertainty may reduce real risk premium. Finally, based on Markov time-varying Copula functions, the dynamic time-varying structures between the EPU index and indexes including SP500, CDS, VIX and UK EPU indexes are studied. The significant regime switching characteristic is shown between the EPU index and the UK EPU index.

Purpose The purpose of this study is to reveal the lead–lag structure between international crude oil price and stock markets. Design/methodology/approach The methods used for this study are as follows: empirical mode decomposition; shift-window-based Pearson coefficient and thermal causal path method. Findings The fluctuation characteristic of Chinese stock market before 2010 is very similar to international crude oil prices. After 2010, their fluctuation patterns are significantly different from each other. The two stock markets significantly led international crude oil prices, revealing varying lead–lag orders among stock markets. During 2000 and 2004, the stock markets significantly led international crude oil prices but they are less distinct from the lead–lag orders. After 2004, the effects changed so that the leading effect of Shanghai composite index remains no longer significant, and after 2012, S&P index just significantly lagged behind the international crude oil prices. Originality/value China and the US stock markets develop different pattens to handle the crude oil prices fluctuation after finance crisis in 1998.

We examine the multifractal scaling behavior and market efficiency of China’s clean energy stock indexes using an asymmetric multifractal detrended fluctuation analysis (A-MFDFA) and then investigate the tail correlation between this index and the crude oil market via an asymmetric multifractal detrended cross-correlation analysis (A-MFDCCA). First, we reveal that the overall, upward and downward trends of the clean energy stock indexes all have significant multifractal characteristics. The clean energy stock market is far from efficient regardless of whether the fluctuations are small or large. In addition, both upward and downward fluctuations exhibit considerable asymmetry. The significant gap between the downward and overall trends indicates that the downward trend following small-scale fluctuations implies weaker efficiency for investors. Furthermore,based on the sliding market deficiency measure (MDM),we find that the change in efficiency in the three trends significantly depends on the length of the window. In the short term, there is no significant efficiency difference among these three trends; however, in the long term, the asymmetry in the upward and downward trends has gradually increased,especially after December 2018. The results demonstrate that bear markets can offer considerably more opportunities for obtaining excess profits. Finally, we reveal that the cross-correlation between the trends of crude oil prices and low-carbon indexes exhibits significant multifractal characteristics. When the crude oil market is in a bull market or the low-carbon energy market is in a bear market, especially in a larger-scale fluctuation, investors should pay attention to the long-term influence of the counterparty market and carry out a hedging operation to avoid risks.

The employment of autocorrelation-based transformation to study the dynamics of the exchange rate system is meaningful because it benefits for chaotic prediction on the basis that the transformation from an exchange rate sequence to its associated autocorrelation sequence is reversible. This paper examines the influence of autocorrelation-based transformation on the systemic dynamics using exchange rates of CNY against different currencies among USD, EUR, JPY, GBP, MYR and RUB. First, we construct recurrence plots of exchange rate return sequences and autocorrelation sequences with a fixed sliding window length of 20. The recurrence quantification analysis (RQA) shows that the exchange rate return sequences exhibit lower degrees of determinism than the autocorrelation sequences. Further, by analyzing the RQA measures with bootstrap techniques and box plots, we reveal that the RQA measures of the exchange rate return systems and the autocorrelation sequence systems are mostly significant, and the vertical structures of recurrence plots of autocorrelation sequences are more sensitive to the shuffles of bootstrap techniques. Finally, we investigate the evolution of RQA measures with the changes of sliding window lengths. The analysis shows that appropriately adjusting the sliding window length can increase the systemic determinism.

In this paper we employ the conditional Granger causality model to estimate the information flow, and find that the improved model outperforms the Granger causality model in revealing the asymmetric correlation between stocks and futures in the Chinese market. First, we find that information flows estimated by Granger causality tests from futures to stocks are greater than those from stocks to futures. Additionally, average correlation coefficients capture some important characteristics between stock prices and information flows over time. Further, we find that direct information flows estimated by conditional Granger causality tests from stocks to futures are greater than those from futures to stocks. Besides, the substantial increases of information flows and direct information flows exhibit a certain degree of synchronism with the occurrences of important events. Finally, the comparative analysis with the asymmetric ratio and the bootstrap technique demonstrates the slight asymmetry of information flows and the significant asymmetry of direct information flows. It reveals that the information flows from futures to stocks are slightly greater than those in the reverse direction, while the direct information flows from stocks to futures are significantly greater than those in the reverse direction.

Purpose This paper aims to capture tail dependence between sentiment index and Shanghai composite index (SCI) by proposing a sentiment index based on text mining. Design/methodology/approach Online text mining and the Copula model were used in this study. Findings First, the paper finds herding effect in the expression of investors’ sentiment from online text data, and the usage occurrence frequency of most vocabulary is less correlative with SCI. Second, given these two features, the paper uses weighted divide-and-conquer algorithm to construct a sentiment index. Finally, because of multivariate non-Gaussian joint distribution between them, the paper uses the Copula model to detect their tail dependences, and finds that both upper and lower tail dependences could have a significant influence between positive sentiment and SCI, with a higher probability on the upper one. Additionally, only the upper tail dependence exhibits the significant influence between negative sentiment and SCI. Originality/value This paper proposes a framework of constructing investment sentiment index with the weighted conquer-and-divide algorithm, and characterizes tail dependence between sentiment index and SCI. The implication can measure the environment of investment market of China and provide an empirical ground for bandwagon effect and bargain shopper effect.

Interaction patterns among different warehouses could make the warehouse-out behavioral sequences less predictable. We firstly take a coupling detrended fluctuation analysis on the warehouse-out quantity, and find that the multivariate sequences exhibit significant coupling multifractal characteristics regardless of the types of steel products. Secondly, we track the sources of multifractal warehouse-out sequences by shuffling and surrogating original ones, and we find that fat-tail distribution contributes more to multifractal features than the long-term memory, regardless of types of steel products. From perspective of warehouse contribution, some warehouses steadily contribute more to multifractal than other warehouses. Finally, based on multiscale multifractal analysis, we propose Hurst surface structure to investigate coupling multifractal, and show that multiple behavioral sequences exhibit significant coupling multifractal features that emerge and usually be restricted within relatively greater time scale interval.

Based on industrial electricity consumption, we model industrial networks by Granger causality method and MST (minimum spanning tree), and then further stick onto an industrial coupling mechanism from energy-consumption perspective. First, we construct Granger causality networks of five provinces in South China of GD, GX, GZ, HN and YN based on their industrial electricity consumption data, and we demonstrate from a network-topology perspective: the distribution of weight of links of all industrial electricity-consumption Granger causality networks approximately follows power-law distribution, revealing a phenomenon that few industries may bring a tremendous influence on the rest. Moreover, correlation analysis between weight and degree of a node shows that in most Granger causality networks, both span and strength of influence of a given industry will significantly increase. Further, we analyze the relationship between the thresholds of Granger causality significance and density of corresponding networks. Results show GD and HN could be classified into a group with relatively greater global differentiation in industries and unbalanced industrial development, however, GX, GZ and YN are grouped as second cluster with relatively balanced industrial development. Furthermore, using Chu–Liu–Edmonds MST algorithm, we extract graphs of MSTs or maximal cliques from industrial electricity-consumption Granger causality networks, and research on energy transmission structure, feedback loop, and bootstrap reliability. By analyzing MSTs, we find that only GD, GX and YN can be extracted with MST graphs, and capture the probable transmission routes of key nodes. Besides we illustrate all three MST graphs are involved with feedback loops structures with various characteristics: GX has complete feed-forward section, feed-back section and feedback loop section; YN has only feed-forward section and feedback loop section; GD has multiple feedback loops section. Finally, we conduct bootstrap reliability tests on MSTs. Results show branches of reliability of GD, GX and YN are decentralized while those feedback loops sections are only concentrated on highly-reliable branches. As neither HN nor GZ can be extracted with MST graphs, we characterize them by maximal cliques. It is demonstrated that both industrial systems of HN and GZ are significantly grouped into two independent branches and those internal industries inside independent branches are highly correlated. However from perspectives of whole South China, the causal relationship is only converged on a few of industries. For regions without MST graphs like HN and GZ, the reliability of links is also greater than that of GD, GX and YN.

The paper studies the industrial energy transferring paths among the industries of China by distinguishing direct causality from the indirect. With complementary graphs, we propose that industrial causal relationship can be heterogeneous, and provide insights for refining robust industrial causality framework. First, by analyzing the in-weight and out-weight of the industries in Granger causality networks we find that public utilities have significant causality with other industries, and the industries with higher degree value tend to have stronger causality with others. Further, we eliminate the exogenous links by partial Granger causality model and find both Granger and partial Granger networks have consistent hub industries while some outliers emerge in partial Granger causality networks. Besides, compared with GX, GZ, HN and YN, the correlation between the volume of electricity consumption and the weight of each industry is more significant in the networks of GD and NF. By studying the characteristics of complementary graphs, we show that the industrial energy transferring paths in GD are more multidimensional, and the corresponding interdependent relationship among industries is more robust. Finally, using bootstrap method we verify the reliability of each industrial relationship network. Results exhibit that GD, GX and NF have more reliable causal relationship networks than other provinces, revealing their industrial structure to be more stable.

We aim to reveal the characteristics and mechanism of the Bitcoin bubble in 2019. First, we identify the period during which two important Bitcoin bubbles occurred based on the generalized supremum augmented Dickey-Fuller (GSADF) method. There are two significant bubble cycles. The first bubble lasted approximately 26 days from November 25, 2017, to December 21, 2017, while the second bubble lasted approximately one week from June 22 to June 29, 2019. The occurrence of the first bubble was related to the considerable expansion of initial coin offerings (ICOs) in 2017, while the formation of the second bubble was affected by the release of Libra. Second, as the GSADF method cannot be used to accurately infer the time at which a bubble bursts, we employ the log-periodic power law singularity (LPPLS) model for this purpose. We verify that the LPPLS method can not only infer the timing of a bubble burst but also shows stable results. Finally, we demonstrate the implications of the 2019 bubble. During the 2019 bubble, due to the increased supervision of European and American governments and the impact of hedging assets, the bubble’s duration was shorter, and the positive feedback mechanism was not as strong as that of the 2017 bubble. In addition, the oscillating frequency of the bubble in 2019 was low and unstable, which means that it would be more beneficial for investors to hold the currency for a long time.

In this paper, we first research on the distance distribution of human mobility with single vehicle based on the driving data from a taxi company in South China. Different from conventional exponential distribution, we discover the mobility distance with taxi follows power-law distribution. Further, we proposed a model which may explain the mechanism for the power-law distribution: mobility distance is constrained by time and fare. Specifically, the relationship between fare and mobility distance follows piecewise function, and responds to individual sensitivity; the relationship between time and mobility distance follows significant logarithmic relationship. These two factors, especially the logarithmic relationship between time and mobility distance, may contribute to a power-law distribution instead of an exponential one. Finally, with a simulation model, we verify the significant power-law distribution of human mobility behavioral distance with a single vehicle, by supplementing factors of waiting time and fare.

Based on industrial electricity-consumption data of five southern provinces of China from 2005 to 2013, we study the industrial correlation mechanism with MST (minimal spanning tree) and HT (hierarchical tree) models. First, we comparatively analyze the industrial electricity-consumption correlation structure in pre-crisis and after-crisis period using MST model and Bootstrap technique of statistical reliability test of links. Results exhibit that all industrial electricity-consumption trees of five southern provinces of China in pre-crisis and after-crisis time are in formation of chain, and the “center–periphery structure” of those chain-like trees is consistent with industrial specialization in classical industrial chain theory. Additionally, the industrial structure of some provinces is reorganized and transferred in pre-crisis and after-crisis time. Further, the comparative analysis with hierarchical tree and Bootstrap technique demonstrates that as for both observations of GD and overall NF, the industrial electricity-consumption correlation is non-significant clustered in pre-crisis period, whereas it turns significant clustered in after-crisis time. Therefore we propose that in perspective of electricity-consumption, their industrial structures are directed to optimized organization and global correlation. Finally, the analysis of distance of HTs verifies that industrial reorganization and development may strengthen market integration, coordination and correlation of industrial production. Except GZ, other four provinces have a shorter distance of industrial electricity-consumption correlation in after-crisis period, revealing a better performance of regional specialization and integration.

In the paper, we research on the characteristics of RMB exchange rate time series fluctuation with methods of symbolization and coarse gaining. First, based on fluctuation features of RMB exchange rate, we define the first type of fluctuation mode as one specific foreign currency against RMB in four days’ fluctuating situations, and the second type as four different foreign currencies against RMB in one day’s fluctuating situation. With the transforming method, we construct the unique-currency and multi-currency complex networks. Further, through analyzing the topological features including out-degree, betweenness centrality and clustering coefficient of fluctuation-mode complex networks, we find that the out-degree distribution of both types of fluctuation mode basically follows power-law distributions with exponents between 1 and 2. The further analysis reveals that the out-degree and the clustering coefficient generally obey the approximated negative correlation. With this result, we confirm previous observations showing that the RMB exchange rate exhibits a characteristic of long-range memory. Finally, we analyze the most probable transmission route of fluctuation modes, and provide probability prediction matrix. The transmission route for RMB exchange rate fluctuation modes exhibits the characteristics of partially closed loop, repeat and reversibility, which lays a solid foundation for predicting RMB exchange rate fluctuation patterns with large volume of data.

This paper adopts multifractal methods to analyze the nonlinear correlations among economic policy uncertainty (EPU), the crude oil market and the stock market. First, using multifractal detrended fluctuation analysis (MF-DFA), we find that each of the three series shows multifractality, and the sources of multifractality are all from long-range correlations and fat-tailed distributions. However, for the stock and oil price series, the long-range correlation has a greater impact, while the fat-tailed distribution contributes more to EPU multifractality. Furthermore, we apply multifractal detrended cross-correlation analysis (MF-DCCA) to analyze the cross-correlation multifractal features among the three series. The generalized Hurst exponents of any two series are significantly greater than 0.5, and the stock and oil price series have the strongest cross-correlation values. Coupling detrended fluctuation (CDFA) analysis of the three series shows that the coupling correlation among the three sequences is also multifractal. The chi-squared statistic reveals that the contribution of the stock market to the multifractality of the coupling correlation is greater than that of other series, and EPU has the smallest influence on the coupling correlation. Finally, we employ multiscale multifractal analysis (MMA) to visualize the dynamic behaviors of the correlations among the series. The results show that the cross-correlation of stock with the oil price has some symmetry at small- and large-scale fluctuations. Additionally, the influence of EPU on the overall coupling features is mainly reflected under small-scale fluctuations, and at large scales, the coupling correlation of the three series remains similar to the Hurst surface of stock with the oil price. These results are useful for building predictive models.

This paper uses the thermal path method to study the lead–lag structure of sentiment and the stock market. First, based on principal component analysis, four indicators are selected to construct the sentiment index. Furthermore, the effectiveness of the thermal path method is verified by numerical simulation. Finally, the lead-lag characteristics of the Shanghai Stock Index and the sentiment index are studied via the symmetric thermal optimal path method. The analysis results show that in the short term, investor sentiment has a leading position in the stock market, which may be related to the herd effect and buying the winners behavior. However, over a longer period of time, investor sentiment is affected mainly by fluctuation in the market, which may be related to the existence of cyclical fluctuations in the market and futures arbitrage. In addition, the stock market's leading effect appears mainly from January 2006 to January 2012, with an average lead time of one month.

By taking China’s beer industry as an example, this paper establishes a series of industrial competition–pressure networks and examines the correlation between competition structure and merger actions. We present a cascade dynamic-merger agent-based computational model driven by competition pressure diffusion to describe the forming process of industry merger wave. The empirical analyses and agent-based computational simulation results show that the competition structure among rivals has a strong effect on the scale, the duration time, and the stability of industry merger wave. We also give explanations on why there are different simulation results between in single market competition environment and in multi-market competition environment, as well as discuss the management implications for the industry-merger policy makers and the merger-tactics decision makers that are involved in merger wave.

In the paper, we research on the behavior dynamics for the large-scale logistics system warehouse-out operation systematically. First, we discover that steel products warehouse-out of different warehouses in a large-scale logistics system can be characterized by burst, and the warehouse-out inter-event time follows the power-law distribution with exponents close to α = 2.5 , which differs from the two classical models proposed by Barabasi (2005) and Vazquez (2005) respectively. By analyzing the warehouse-out inter-event time distribution of the products in one certain large-scale logistics system, we further discuss burst features and mechanisms of logistics system. Additionally, we find that in population behaviors, burst features can be explained by the priority that rooted in holidays and interior task scheduling. However, warehouse-out behaviors of active individuals do not show any features of burst. Further, we find that warehouse-out quantity of steel products follows Fractal Brownian motion with the HURST exponent higher than 0.5 by means of R / S , which infers that the quantity of products in a logistics system is not only guided by prices in the present market, but also related closely to the previous quantity of warehouse-out. Based on V statistic, we compare memory length of different products in warehouses. Finally, we apply complex networks visibility graphs for further validation of fractal features in a logistics system and find that almost every visibility graph exhibits small-world and scale-free features. Both R / S and complex networks visibility graphs reinforce that the warehouse-out quantity of products in a logistics system is not a random walk process, but contains intrinsic regularities and long-term correlation between present and previous warehouse-out quantity.

Buzzword-based viral marketing, known also as digital word-of-mouth marketing, is a marketing mode attached to some carriers on the Internet, which can rapidly copy marketing information at a low cost. Viral marketing actually uses a pre-existing social network where, however, the scale of the pre-existing network is believed to be so large and so random, so that its theoretical analysis is intractable and unmanageable. There are very few reports in the literature on how to design a spreading scheme for viral marketing on real social networks according to the traditional marketing theory or the relatively new network marketing theory. Complex network theory provides a new model for the study of large-scale complex systems, using the latest developments of graph theory and computing techniques. From this perspective, the present paper extends the complex network theory and modeling into the research of general viral marketing and develops a specific spreading scheme for viral marking and an approach to design the scheme based on a real complex network on the QQ instant messaging system. This approach is shown to be rather universal and can be further extended to the design of various spreading schemes for viral marketing based on different instant messaging systems.

we classify a bank's 15595 clients into 9 groups by their ages and bank balances, and then each group is classified further into two parts, namely active and inactive clients. The 1939 service channels, which are used in the transaction are divided into three transaction worlds for each group of clients: local world, the weak nonlocal world, and the strong nonlocal world. The service channels in regular use are defined as its local world, the weak nonlocal world, and the strong nonlocal world are the channels, which are used occasionally, but the use probability of the former is higher than the latter. For the bipartite weighted network, which consists of 15595 clients and 1939 service channels, we propose an evolving model, which has the mechanism of different active degrees of clients, preferential attachment in the local world, random attachment in the weak nonlocal world by higher probability and in the strong nonlocal world by lesser probability, to discuss the network generating mechanism. Finally, by the method of numerical simulation, we validate the generating mechanism. The work is significant for the generating mechanism study of the bipartite weighted network related to the geographical location and the management of bank service channels.

We study information transfer routes among cross-industry and cross-region electricity consumption data based on transfer entropy and the MST (Minimum Spanning Tree) model. First, we characterize the information transfer routes with transfer entropy matrixes, and find that the total entropy transfer of the relatively developed Guangdong Province is lower than others, with significant industrial cluster within the province. Furthermore, using a reshuffling method, we find that driven industries contain much more information flows than driving industries, and are more influential on the degree of order of regional industries. Finally, based on the Chu-Liu-Edmonds MST algorithm, we extract the minimum spanning trees of provincial industries. Individual MSTs show that the MSTs follow a chain-like formation in developed provinces and star-like structures in developing provinces. Additionally, all MSTs with the root of minimal information outflow industrial sector are of chain-form.

By taking China’s beer industry as an example, this paper establishes a series of industrial competition pressure networks and examines the correlation between competition structure and merger actions. We present a cascade dynamic-merger agent-based computational model driven by competition pressure diffusion to describe the forming process of industry merger wave. The empirical analysis and simulation results show that the competition structure among rivals has a strong effect on the scale, the duration time, and the stability of industry merger wave. We also give explanations on why there are different simulation results between in single market competition environment and in multi-market competition environment, as well as discuss the management implications for the industry-merger policy makers and the merger-tactics decision makers that are involved in merger wave.

This paper presents a novel method to explore the relationship between dynamics mergers and the evolution of industrial competition by introducing complex network tool. By taking China's beer industry as an example, we established Markets-Firms bipartite time series networks, weighted Markets-Firms time series networks and industrial competition time series networks by using the data from 1992 to 2009. Through analyzing the changes of topology index of these networks, we find that dynamic mergers play a key important role in the evolution of industrial competition. The results show that dynamic mergers promote the fragmented local markets to be consolidated into a global market for China's beer industry, and it also shows the process that competitive relationship among the rivals turns from some single segmented markets to a global cross-market. This paper gives a new view to observe the changing of industrial competition driven by dynamic mergers.

Based on cross-correlation algorithm, we analyze the correlation property of warehouse-out quantity of different warehouses, respectively, and different products of each warehouse. Our study identifies that significant cross-correlation relationship for warehouse-out quantity exists among different warehouses and different products of a warehouse. Further, we take multifractal detrended cross-correlation analysis for warehouse-out quantity among different warehouses and different products of a warehouse. The results show that for the warehouse-out behaviors of total amount, different warehouses and different products of a warehouse significantly follow multifractal property. Specifically for each warehouse, the coupling relationships of rebar and wire rod reveal long-term memory characteristics, no matter for large fluctuation or small one. The cross-correlation effect on long-range memory property among warehouses probably has less to do with product types,and the long-term memory of YZ warehouse is greater than others especially in total amount and wire rod product. Finally, we shuffle and surrogate data to explore the source of multifractal cross-correlation property in logistics system. Taking the total amount of warehouse-out quantity as example, we confirm that the fat-tail distribution of warehouse-out quantity sequences is the main factor for multifractal cross-correlation. Through comparing the performance of the multifractal detrended cross-correlation analysis (MF-DCCA), centered multifractal detrending moving average cross-correlation analysis (MF-X-DMA) algorithms, the forward and backward MF-X-DMA algorithms, we find that the forward and backward MF-X-DMA algorithms exhibit a better performance than the other ones.

It has been shown that many complex networks share self-similar properties. To reveal the ubiquitous characteristics that many real networks may have in common, we first improve the box covering algorithm and use it to measure several real industrial competition networks. As a result, we find that some real industrial competition networks illustrate the fractal properties. As the improved box covering algorithm could not reflect the structure characteristic of the industrial competitive network efficiently, and the scale of the network namely the average length path is not enough for quantifying the fractal properties, we use the clustering coefficient as the scale measure index and present a systematic analysis of some real industrial competition networks from the perspective of multifractal features. We find that many real industrial competition networks can also be characterized by the multifractal features. Finally, with a motif-hierarchical model, we simulate how the fractal structure of the industrial networks may be formed, and to some extent verify that the disassortative process is an important self-organizing mechanism in industry system.