Revisiting the price transmission from international commodity to Chinese agricultural products based on an industry chain perspective | Humanities and Social Sciences Communications

Humanities and Social Sciences Communications volume 12, Article number: 936 (2025) Cite this article

The previous researches proved that the price fluctuations of international commodity can significantly influence the price of agricultural products. In fact, the affected agricultural products may further transmit the impacts to other products along the agricultural industry chain, and cause direct and indirect effects from international commodity to agriculture-related products, which is defined as cascaded transmission in this paper. To reveal the cascaded transmission effects, this paper designs a novel model combining the econometric methods and the network analysis method. Taking the Commodity Research Bureau index (CRB) and the price indexes of 21 Chinese agriculture-related products as empirical data, the following results were obtained. (1) The transmission path of price fluctuations from CRB to Chinese agricultural products is complex, rather than along the agricultural chain from upstream to downstream. (2) The indirect effects of CRB on agricultural products cannot be ignored, and the transmission relationships are tighter than the nonlinear relationships. (3) The roles of agricultural products are different in the cascaded transmission process. Price regulatory authorities should categorize agricultural products to achieve decentralized management and mitigate the impacts of the international commodity market volatility.

The price fluctuations of international commodity have been regarded as an important external risk affecting the macroeconomy. As one of the world’s largest commodity consuming and trading countries, China’s demand for international commodities continues to increase, along with its dependence on imports. Besides, as China’s primary industry, agriculture has a fundamental position in the national economy, and is of great significance to food security and economic security. Existing researches show that price fluctuations of international commodity mainly affect China’s agricultural prices through the following three ways. The first is international trading. From 2006 to 2024, China’s importing volume of agricultural products increased from $32.07 billion to $215.66 billionFootnote 1, with an average annual growth rate of 11.17%. Therefore, price fluctuations of international commodities can intensify the impact on China’s agricultural prices through the international trading (Alexander and Wyeth 1994). The second is financial market. The acceleration of financial integration has resulted in a closer connection between the international commodity market and the financial market (Liao and Li 2023). Price fluctuations of international commodity could cause the volatility of exchange rates (Zheng et al. 2024), thereby exerting a significant impact on agricultural prices (Roache 2010). The third is the expectations of market participants. Changes in international commodity price can affect the expectations of domestic producers, sellers and other market participants, who will adjust their production and sales plan, altering the supply-demand relationships in the Chinese agricultural market and ultimately leading to fluctuations of agricultural prices.

In addition, due to the supply-demand relationships and substitution effects between products in the agricultural industrial chain, the prices of each product are interrelated. The agriculture-related products affected by international commodity may further transmit the price fluctuations to others, exerting indirect influence, similar to the “domino effect” (Sun et al. 2018). Studying the indirect impacts of international commodity on Chinese agriculture-related products from the perspective of industry chain can deepen our understanding of price fluctuations in the Chinese agricultural market (Xi et al. 2023), which is of great implications for food security and price risk management of the agricultural industry chain.

The related literature can be divided into two categories, namely the spatial transmission from international commodity to agricultural products in a country and the vertical transmission in the agricultural industrial chain. In the aspect of the spatial transmission, there are two dominant opinions. One is the one-price principle, assuming that there exists a long-term linear and stable equilibrium relationship among the variables (Sephton 2003). The error correction model (ECM) (Baquedano and Liefert 2014) and Granger causality test method (Ma and Hou 2019) are popularly used to reveal the linear influence of price fluctuations from international commodity to agricultural products. For example, Arnade et al. (2017) estimated an ECM model and found that Chinese soybeans were significantly affected by international commodity, while rice was barely influenced. The other is that there exists a nonlinear relationship between international commodity and agricultural market due to the government interventions, such as inventory holding behaviors of government (Deaton and Laroque 1995; Takahashi 2012), floor price policies (Rapsomanikis and Hallam 2006) and tariff policies (Baltzer 2013; Thow et al. 2010). Researchers usually adopt the nonlinear autoregressive distributed lag model (Mila et al. 2023) and nonlinear Granger causality test model (Zhang and Liu 2020) to study the nonlinear effects. For instance, Nazlioglu (2011) estimated a nonlinear Granger causality test model and proved the nonlinear impacts of oil prices on corn and wheat prices.

In the aspect of vertical transmission, the price transmission among the links of the agricultural industry chain is described. The earliest theoretical research of vertical transmission in agricultural industry chain was conducted by Gardner (1975), who analyzed the price transmission effects between the wholesale price and retail price of food. Subsequently, scholars began to research the vertical transmission of agricultural industry chains, such as wheat (Djuric and Goetz 2016; Haile et al. 2017; Kharisma and Indrawan 2023; Ricci et al. 2019), meat (Capitanio et al. 2019; Chaudhry and Miranda 2020), and aquatic products (Asche et al. 2014; Singh et al. 2015). For example, Xu et al. (2012) studied the price transmission effect within the Chinese swine industry chain, regarding the price of corn, compound feed for fattening pig, piglet as the upstream, the price of pig as the midstream and the pork price as the downstream. The results showed that the downstream was more significantly affected by the upstream.

The two types of researches mentioned above were conducted independently. However, certain agriculture-related products affected by international commodity may further transmit price fluctuations to other products along the agricultural industry chain. As far as we know, there is little research to integrate the international commodity and agricultural industry chain into a whole system and study the indirect influence of international commodity on Chinese agricultural products from a cascaded view. Since the complicated interactions among various variables, a single econometric approach cannot reach a comprehensive analysis of the price cascaded transmission effects.

Thus, this paper designs a cascaded transmission network model combining econometric methods and complex network methods to explore the direct and indirect price transmission effects from international commodity to Chinese agricultural industry chain. We select Commodity Research Bureau (CRB) index as the international commodity price index, and 21 price indexes are selected from the agricultural means of production market, wholesale market and retail market to represent the upstream, midstream and downstream of the Chinese agricultural industry chain. Firstly, linear and nonlinear Granger causality tests are employed to confirm the direct price transmission relationships among the CRB and the agriculture-related products. Secondly, we test the degree of influence among CRB and agricultural products using the generalized forecast error variance decomposition (GFEVD) model. According to the above results, we construct linear and nonlinear variance decomposition networks (LVDN and NLVDN), respectively, with CRB index and 21 agriculture-related product price indexes as nodes, the causal relationships among the nodes as edges and the degree of influence as the weights of edges. The networks offer a platform for the price cascaded transmission. Thus, we explore the cascaded transmission paths, the direct and indirect influence of price volatilities of international commodity on agriculture-related products, and the role of agricultural products in the industry chain.

The main contributions are as follows. Firstly, to the best of our knowledge, this paper is the first one to combine the spatial transmission from international commodity to Chinese agricultural products and the vertical transmission in the China’s agricultural industrial chain into a systematic research framework, and mainly focuses on the cascading transmission effects. Secondly, it is difficult to comprehensively analyze the cascading price transmission effects by using a single econometric method (such as error correction model and Granger causality test). In this paper, we design a cascading transmission network model that combines econometric methods and complex network methods. This model aims to quantitatively measure the direct and indirect effects of international commodity market on Chinese agricultural market. Thirdly, by revealing the cascading transmission path from international commodity market on Chinese agricultural market, this paper provides a new analytical tool for food security and price risk management. Besides, the model can also be applied to the study of other similar complex economic systems.

The rest of this paper is presented below. The section “Data and methodology” offers the data and methodology. The section “Results and discussion” introduces the empirical results and the discussion. The section “Robustness” presents the results of robustness test. The section “Conclusions and policy implications” gives the conclusions and policy implications.

In this paper, the agricultural means of production market, wholesale market and retail market are identified as the upstream, midstream and downstream of the Chinese agricultural industry chain. For the brevity of this paper, the specific products and their abbreviations are shown in Table 1. Besides, based on the considerations of historical data continuity, industry representativeness, liquidity requirements, and the capacity for economic forecasting, we select the CRB index to represent the price of international commodityFootnote 2. The index not only helps investors assess market trends and manage risks, but also indicates the trend of inflation or deflation, which is an important reference indicator for policy makers and economists and has been widely recognized by the academic community.

The price indices of agricultural products are released monthly by the China’s National Bureau of Statistics, reflecting the overall price level in China. Since the monthly data on the price index of agricultural means of production market is only updated up to April 2020, the selected data in this paper spans from January 2006 to March 2020. The CRB index is the daily data from January 1, 2006 to March 31, 2020. All the above data are acquired from the Wind database (http://www.wind.com). To make the frequency of the data consistent, we convert the daily data of the CRB index into monthly data. Therefore, each variable contains 171 observations. The results of descriptive statistics and stationarity tests of each variable are shown in Table 1. We can conclude that some original time series are not stationary at the 10% significance level, and all series are stationary at the 1% significance level after the first-order difference.

The paper constructs cascaded transmission networks to measure the impact of CRB on agriculture-related products. Specifically, firstly, the linear and nonlinear Granger causality tests are used to identify the causal relationships between each two pair of variables, respectively. We then quantitatively measure the impact of one variable on another by setting a GFEVD model based on LASSO (Least absolute shrinkage and selection operator) method. Finally, we construct cascaded transmission networks to deeply detect the transmission range and degree of the CRB on the Chinese agricultural industry chain.

The Granger causality test is a common method to test whether there exists a linear causal relationship between two time series, which is widely used to study the price transmission relationship between two economic variables. The detailed method is shown in the Appendix A. In this study, there are 22 variables under analysis, thus a total of 231 pairwise linear Granger causality tests (calculated as 22*(22-1)/2) were conducted, generating a linear Granger causality matrix LG, as formalized in Eq. (1):

where $L{G}_{{ij}}$ represents the linear Granger causality from i to j, and i ≠ j. If there is a significant linear causality from i to j, then $L{G}_{{ij}}=1$; otherwise, $L{G}_{{ij}}=0$. The N denotes the number of variables, namely 22. The first variable represents CRB and others represent agriculture-related products.

The time series of economic variables may produce structural changes because of the shocks from the economic policies and other factors, making the economic variables present significant nonlinear characteristics in the interaction process (Chen 2007)Footnote 3. Diks and Panchenko (2006) proposed a nonlinear Granger causality method. This method extracts the residuals of the VAR model to filter out the linear factors and then performs the nonlinear Granger test on residual series. The detailed method is provided in the Appendix A. Similar to the linear Granger causality tests, we conduct 231 nonlinear Granger causality tests, and then obtain a nonlinear Granger causality matrix ${NG}$, as shown in Eq. (2):

where $N{G}_{{ij}}$ represents the nonlinear Granger causality from i to j, and i ≠ j. If there exists a significant nonlinear causality from i to j, then $N{G}_{{ij}}=1$; otherwise, $N{G}_{{ij}}=0$.

The variance decomposition method can examine the degree to which one variable impacts other variables. But the result of traditional variance decompositions based on the VAR model contains lagged terms for the variables, which can significantly reduce the freedom degrees and affect the accuracy of the model. Tibshirani (3.0.Co;2-3" href="/articles/s41599-025-05238-4#ref-CR33" id="ref-link-section-d57095236e2764">1997) proposed a LASSO method that can solve the above problem by constructing a penalty function. Combining the LASSO method and VAR model, Nicholson et al. (2017) proposed a BigVAR model to solve the parameter estimation problem of the high-dimensional VAR model. Based on the framework of Nicholson et al. (2017), we apply the GFEVD model to quantitatively estimate the contribution of price fluctuations of a variable to another variable. Compared with the traditional variance decomposition method, GFEVD has the following advantages. Firstly, the GFEVD method effectively addresses the ordering dependency issue inherent in conventional methodologies, thereby eliminating the potential bias associated with variable ordering. Secondly, the GFEVD method is robust to common economic shocks (e.g., supply-demand imbalances and policy changes) and provides accurate estimates of the contribution of each variable to forecast error variance. Thirdly, the GFEVD method has high flexibility, being applicable to both linear and nonlinear systems without requiring restrictive structural assumptions. The contribution of the variable i to the forecast error variance of the variable j at the forecast horizon H can be expressed as ${{GED}}_{ij}\left(H\right)$, as shown in Eq. (3). And its normalized result is shown in Eq. (4):

where ${e}_{j}$ is a vector, which the jth element is 1 and the others is 0; ${C}_{h}$ is the product of the $h$-lagged shock vector and the coefficient matrix; $\varphi$ denotes the covariance matrix of the shock vector in the non-orthogonal VAR model; and ${\sigma }_{{ii}}$ is the ith diagonal element of $\varphi$.

Therefore, we can obtain a GFEVD matrix ${GD}$, as shown in Eq. (5). Based on the extant literature (Shahzad et al. 2021; Sun et al. 2019), we take the value of H as 10. In the robustness tests, we compare the results for H from 3 to 12:

In step 1, the matrices ${LG}$ and ${NG}$ represent the price transmission relationships between variables. In step 2, the matrix ${GD}$ represents the degree of influence between variables. In this step, based on the above results, the price transmission networks from CRB to agriculture-related products is constructed, namely, LVDN and NLVDN. Specifically, we regard the variables as nodes and the price transmission relationships between variables as edges. We use the degree of influence between variables to represent the weight of edge. The networks are defined as matrix ${LGGD}$ and ${NGGD}$, as shown in Eqs. (6) and (7):

If $L{{GGD}}_{{ij}}=0$, it means there has no linear causal relationship from i to j. If ${L{GGD}}_{{ij}} > 0$, it indicates that i Granger causes j and the contribution of i on the fluctuations of j is $L{{GGD}}_{{ij}}$:

If ${N{GGD}}_{{ij}}=0$, it means that there is no nonlinear causal relationship from $i$ to $j$. If ${N{GGD}}_{{ij}} > 0$, it means that $i$ nonlinear Granger causes $j$ and the contribution of $i$ on the fluctuations of $j$ is $N{{GGD}}_{{ij}}$.

To improve the readability of the process of network construction, we add an illustrative graph, please see Fig. 1.

The process of network construction.

According to network analysis method, we describe the topological characteristics of LVDN and NLVDN, and analyze the role of each agriculture-related product in the industrial chain. In addition, the transmission range and impact degree of CRB on each agriculture-related product are measured. As the network indicators of LVDN and NLVDN are calculated in a same way, we take NLVDN as an example to introduce each network indicator.

Network density

The closeness of the relationships among CRB and agriculture-related products in the NLVDN is defined as network density. The greater the value of network density, the tighter the relationship among the variables. The density of NLVDN can be expressed as Eq. (8):

Average shortest path length

The average shortest path length indicates the average of the shortest paths between all pairs of variables in the network, which is expressed as Eq. (9):

where the shortest path length ${S}_{{ij}}$ is the distance between nodes $i$ and ${j}$. In this study, a small average path length of the networks means that it is easy for products to transmit the fluctuations to others.

CRB might directly affect agriculture-related products because CRB contains commodities that are raw materials for production, such as oil, non-ferrous metals and basic agricultural products. Besides, the directly affected agriculture-related products can further transmit the price volatilities to others due to the supply-demand relationships and substitution relationships in the agricultural industry chain (Sun et al. 2019). Thus, the fluctuations of CRB not only directly affect the price of an agriculture-related product but also indirectly affect it through others in the chain. The transmission range indicates the number of products that are impacted by the fluctuations of CRB. Since the existence of indirect impact, this paper puts forward the concepts of direct transmission range and cascaded transmission range. To make it easier, we draw a sample network model, as shown in Fig. 2, which contains six variables, namely CRB and five agricultural products. We can see that there are edges from CRB to A and C, meaning that CRB can directly impact these two nodes, and the direct transmission range of CRB is 2. Besides, CRB also indirectly impacts B, D and E in the subsequent transmission. In detail, although there has no direct path from CRB to E, there are two cascaded paths from CRB to E, namely “${CRB}\to C\to E$” and “${CRB}\to A\to B\to D\to E$”. Thus, the shortest path length from CRB to E is 2 (“${CRB}\to C\to E$”), which means CRB indirectly influences E after two transmission steps. Analogously, there are also two cascaded paths from CRB to D, namely “${CRB}\to C\to {\rm{E}}\to D$” and “${CRB}\to A\to B\to D$”, and CRB can indirectly impact D after three transmission steps. Synthetically, considering both direct and indirect influence, CRB can impact all agricultural products. Therefore, we define the cascaded transmission range of CRB as 5.

A simple model of NLVDN. The node CRB represents the international commodity, and the nodes A, B, C, D and E represent agricultural products. The direction of the arrow indicates the direction in which price fluctuations are transmitted.

In the first transmission step, CRB might directly affect some agriculture-related products. The direct impact of the CRB to agriculture-related product $j$ is denoted as ${NLDIC}1$, as shown in Eq. (10):

In the second transmission step, the fluctuations of CRB may have an indirect effect on agriculture-related product ${j}$ through an intermediate product. For example, in Fig. 2, CRB can indirectly influence B through A. To measure the indirect influence degree of CRB on B, we need to calculate both the impact of CRB on A and the impact of A on B. Therefore, the indirect influence of CRB to B is ${N{GGD}}_{1A}\times {N{GGD}}_{AB}$. We can generalize the second transmission step from ${CRB}$ to agriculture-related product $j$ to a more general situation, as shown in Eq. (11):

where ${N{GGD}}^{2}$ denotes the square of the matrix $N{GGD}$. Similarly, we can extend this effect to step $s$, as shown in Eq. (12):

Then the total impact from CRB to agriculture-related product $j$ is the sum of the degree of direct and indirect influence, defined as ${{NL}T{otal}}_{{CRB},j}$, as shown in Eq. (13):

We refer to the research of Feng et al. (2022) to measure the role of each agriculture-related product from aspects of influence, sensitivity and intermediary ability. In the NLVDN, the out-degree of a variable reveals how many variables are influenced by it, which can describe the influence of the variable, shown in Eq. (14). The in-degree of a variable indicates how many variables can influence it, which can explain the sensitivity of the variable, shown in Eq. (15):

The betweenness centrality can quantify the intermediary ability of each agriculture-related product in the process of cascaded price transmission, that is the capacity as a bridge in the network. The variable with high betweenness centrality plays an important role in cascaded transmission processes, which is the convergence point of the transmission path among other variables. If these variables are absent from the network, the impacts cannot spread rapidly and widely. The betweenness centrality of each agriculture-related product in the NLDVN can be expressed as Eq. (16):

where $i$, $j$ and $k$ represent variables; ${g}_{{jk}}$ is the number of the shortest paths between $j$ and $k$; ${g}_{{jk}}\left(i\right)$ is the number of the shortest paths between $j$ and $k$ that pass-through $i$.

In the context of economic globalization and expanding scale of China’s foreign trade, the influence of price volatilities of international commodities on Chinese agricultural markets has become increasingly prominent. To guarantee the national food security and counteract the volatility of international food price, China introduced several agricultural protection policies, such as temporary grain storage policies and agricultural support and protection subsidy policies. These polices may mitigate the impact of international commodity on Chinese agricultural markets to some extent. For example, the surge in international crude oil prices can elevate the production cost of agricultural products such as fertilizer and farm diesel, thereby pushing up consumer market prices for agricultural products. However, China’s agricultural support policies, such as grain subsidies and agricultural machinery subsidies policies, can reduce agricultural production costs and stabilize agricultural prices. These protective measures partially offset the effects of international commodity market volatility, resulting in a nonlinear effect of international commodity market on Chinese agricultural market. Therefore, it is urgent to research the linear and nonlinear influence of international commodity on the Chinese agricultural industrial chain. We conduct linear and nonlinear Granger causality tests between each two pair of variables, and the empirical results are shown in Fig. 3. For example, CRB is a linear Granger cause of ALC, but CRB cannot nonlinearly Granger cause ALC. According to Section “Network construction method”, the LVDN and NLVDN are constructed, shown in Fig. 4. To reveal the transmission characteristics from CRB to agriculture-related products in more details, we investigate the structure of the networks.

a Linear causal relationships between variables, and b Nonlinear causal relationships between variables. The blue rectangle represents that there exists a price transmission relationship from the abscissa variable to the ordinate variable. The white rectangles represent that there is no transmission relationship from the abscissa variable to the ordinate variable. Besides, there is no meaning in the gray rectangles.

a Linear cascaded transmission network, and b nonlinear cascaded transmission network. The blue, yellow, and green nodes represent the upstream, midstream, and downstream products in agricultural industry chain, respectively. The thicker the edge, the greater the degree of influence between the variables.

In the LVDN and NLVDN, both of the two networks involve 22 variables, with 166 and 121 price transmission relationships (or causal relationships), respectively. The density of LVDN and NLVDN is 0.359 and 0.262, respectively, meaning that the linear transmission relationships among CRB and agriculture-related products are tighter than the nonlinear causal relationships. Besides, the average shortest path length of the LVDN and NLVDN is 1.707 and 1.722, respectively, indicating that the average transmission steps from one variable to other variables is only about 1.7, and the price transmission in the LVDN is easier and faster than that in the NLVDN.

Compared to the LVDN, the NLVDN exhibits fewer price transmission relationships and slower transmission speeds. The major reason may be the agricultural support policies introduced by China. When international commodity price volatility exacerbates agricultural price volatility, China’s agricultural support policies, such as agricultural support protection subsidy policy and minimum purchase price policy for grain, can reduce agricultural production costs and stabilize agricultural prices. Thus, these protective policies partially offset the impact of the international commodity market volatility on Chinese agricultural market. Moreover, policy interventions may also alter the transmission path and direction of volatility, leading to an increase in the transmission steps and a decrease in the transmission speed. To verify the effectiveness of the policy intervention, we conduct additional experiments, and more details are presented in the Appendix C.

To shed light on the transmission process of CRB, we measure the shortest path from CRB to each agriculture-related product, shown in Table 2. It can be concluded that the transmission range of CRB in the LVDN is wider than that in the NLVDN. To be more specific, in the LVDN, the fluctuations of CRB can directly influence all agriculture-related products in the upstream agricultural market, and the direct transmission range of CRB is 15. Besides, CRB can indirectly impact 5 agricultural products at the second transmission step and 1 agricultural product (namely MEA) at the third transmission step. Thus, the fluctuations of CRB can influence all the prices of agriculture-related products by direct and indirect transmission paths, and the cascaded transmission range of CRB is 21. However, in the NLVDN, we can see that the CRB impact 10 agriculture-related products in the first transmission step, namely, MFT, FEZ, PPD, IND, JAP, LGR, NGR, FOD, OIL and VEG. Thus, the direct transmission range of CRB is 10, which is smaller than that in the LVDN. Moreover, CRB can indirectly impact 7 agriculture-related products in the second transmission step. It is also worth noticing that not all agriculture-related products can be directly or indirectly impacted by the fluctuations of CRB. There is no transmission path from CRB to 4 agriculture-related products, namely FER, AEO, MEA and EGG. Therefore, the cascaded transmission range of CRB is only 17 in the NLVDN.

Overall, it suggests that price fluctuations of CRB are not always transmitted along the chain from upstream to downstream, and its transmission characteristics are complex. As a comprehensive index with high industry coverage, the fluctuation of CRB index is closely linked to the upstream, midstream and downstream of the agricultural industry chain. Firstly, the influence of CRB fluctuation can be transmitted along the industrial chain from upstream to downstream, forming a cost-push transmission path. For example, higher prices of international commodities, such as crude oil, fertilizers, and metals, can raise the production cost of upstream products in the agricultural industry chain, thereby further elevating the product prices in the midstream and downstream markets. Secondly, due to the influence of various factors such as supply and demand relationships within the industrial chain, cost transmission mechanisms, and market expectations, price fluctuations in downstream commodities could have intricate impacts on the prices of midstream and upstream commodities. For instance, rising grain prices in the international commodity market can drive up feed costs, thereby elevating the price of livestock meat in the downstream. Part of the demand for grains may be transferred to the domestic market, which can increase grain prices in the midstream and further raise the price of fodder and fertilizer in the upstream, thus forming a demand-driven transmission path. Therefore, the fluctuations of CRB can directly influence the price of some agriculture-related products in the upstream, midstream and downstream. Then, the influenced products could further spread the impact of the CRB fluctuations along the industry chain. This result also confirms the cascaded price transmission mentioned in the introduction.

To enhance the comprehension of the influence of price volatility of CRB on the Chinese agricultural industry chain, this paper calculates the direct, indirect and total influence of CRB fluctuations on the industry chain, as shown in Table 3. At the same time, we also calculate the impact of CRB on each agriculture-related product, as shown in the Fig. 5.

Direct influence

The downstream is more affected by price fluctuations of CRB than the upstream. The reason might be the higher demand elasticity of agricultural products in downstream, where consumers are more likely to alter their purchasing behavior due to price fluctuations. Thus, agricultural products in downstream are more sensitive to price fluctuations in international commodity. The linear and nonlinear impacts of CRB on the downstream are 8.83% and 11.17%, respectively. In the downstream agriculture-related products, CRB has the greatest impact on OIL (Cooking oil) with both linear and nonlinear effects of 8.58%. The main reason may be that China is highly dependent on imports of palm oil and soybean oil, and price fluctuations in international commodity market can be transmitted directly to cooking oil market in China through international trade. In addition, CRB also has a strong linear influence on midstream, while the nonlinear influence is relatively weak. The products of midstream and downstream are closely related to the daily consumption of residents, indicating that price fluctuations of international commodity directly affect residents’ welfare.

Indirect influence

In terms of linear price transmission, the fluctuations of CRB prices have a similar impact on upstream, midstream and downstream, ranging from 3 to 4%. The indirect impact of CRB on the upstream accounts for the highest proportion, reaching 61.52%. In the upstream agriculture-related products, FEZ (Chemical fertilizers) and PPD (Pesticides and pesticide machinery) are not directly affected by CRB but experience an indirect impact of 0.70% and 0.54%, respectively. Regarding the nonlinear price transmission, CRB price fluctuations have the largest indirect impact on midstream at 3.99%, accounting for 81.93% of the total impact. This is probably because that most of the agricultural products in the midstream are import-dependent. Therefore, prices fluctuations of international commodities can affect the supply and demand of midstream agricultural products, thus influencing the agricultural prices in midstream. In the midstream agriculture-related products, IND (Indica rice) is not directly affected by CRB but has an indirect impact of 0.77%. It is worth noting that the influence of price fluctuations of CRB gradually decreases with the increase of transmission steps, reaching zero after the fourth stage. This is consistent with the three-step transmission theory of Fowler and Christakis (2008).

Total influence

The midstream and downstream are most affected by CRB price fluctuations. In the midstream of the industrial chain, SOY (Soybeans) is the most linearly impacted by CRB at 9.97%, mainly as a result of the direct impact. LGR (Long-grained rice) is the most nonlinearly influenced by CRB at 1.35%, and this influence is mainly indirect. In the downstream, OIL is the most affected, with linear and nonlinear total influence of 9.51% and 8.81%, which the indirect effects contributing 0.93% and 0.23% respectively. The total impact of CRB on agriculture-related products is the sum of direct and indirect impacts. Ignoring the indirect impact would underestimate the degree of the influence of international commodity on agricultural products.

a Linear influence degree, and b nonlinear influence degree. The X-axis represents the transmission step of fluctuations of CRB. LDIC1 and NLDIC1 represent the direct linear and nonlinear shock arising from price fluctuations of CRB, LIIC2- LIIC5 and NLIIC2- NLIIC5 represent the 2–5 steps of linear and nonlinear indirect shock arising from price fluctuations of CRB. LTotal and NLTotal respectively denote the total degree of linear and nonlinear impact of CRB price fluctuations. Y-axis is the impact of CRB on each agriculture-related product.

What role does each agriculture-related product play in the cascaded price transmission process? We explore the influence, sensitivity and intermediary ability of each agriculture-related product in the industry chain. As shown in Fig. 6, each agriculture-related product occupies different roles in the LVDN and NLVDN. We can conclude that the influence, sensitivity and intermediary ability of products in midstream and downstream are obviously positively correlated. The intermediary ability of products upstream is significantly lower than other products in both LVDN and NLVDN. Besides, in the LVDN, the influence, sensitivity and intermediary ability of the agricultural products downstream are generally stronger than that in upstream and midstream. However, in the NLVDN, the intermediary ability of products in midstream is obviously larger than others.

a Role of each product in the LVDN, and b role of each product in the NLVDN. Points in the graph represent agriculture-related products. The X-axis, Y-axis and Z-axis display the influence, sensitivity and intermediary ability of agriculture-related products, respectively. The colors of the products represent the agricultural industrial chain link attribute. Red means upstream, blue represents midstream, and green indicates downstream.

More specifically, we display the top 6 agricultural products ranked through the above three indicators, shown in Table 4. AQU (Aquatic products) is the most important agricultural product in the LVDN. The ranking of AQU in influence, sensitivity and intermediary ability is 1st, 3rd, and 1st respectively. However, in the NLVDN, AQU is not as important as it is in the LVDN, and NGR (Non glutinous rice) plays an important role with a high value of influence, sensitivity and intermediary ability. In addition, SOY (Soybeans) has an important position both in the LVDN and NLVDN. This is because soybean is not only an important food crop but also an important cash crop. Soybean can be used to extract oil, and the by-products can be used as feed. The price fluctuations of SOY would influence relevant agriculture-related products. We also find that although some agricultural products do not have a strong influence, they also play important roles as intermediaries. For example, VEG (Fresh vegetables) ranks 16th and 13th in influence and sensitivity, respectively, but it ranks 5th in intermediary ability and plays a crucial intermediary role in the agricultural industry chain. If there is dramatic price volatility in the chain, the prices of these agriculture-related products with high intermediary ability can be supervised to weaken the transmission of the price fluctuations.

The results in Section “Results and discussion” are based on the forecast horizon H of 10. Referring to (Diebold and Yilmaz 2012), this section checks for the sensitivity of the results to the choice of forecast horizon. We select the forecast horizon H ranging from 3 to 12 and construct the cascaded transmission networks from CRB to agriculture-related products under each H. Furthermore, the direct, indirect and total impacts of CRB on each agricultural product are calculated under different H, shown in Fig. 7. We can see that the gray area is very small, indicating that there is no significant difference in the impact degree of international commodity on agricultural products under different H. In other words, the differences in the degree of influence are relatively small when H takes different values. Therefore, it can be concluded that the results of this paper are robust.

a, c and e display the linear total impact, direct impact and indirect impact of CRB on each agricultural product, respectively. b, d and f display the nonlinear total impact, direct impact and indirect impact of CRB on each agricultural product, respectively. We extract the maximum and minimum values of the impacts of CRB on agricultural products when H takes different values, and the gray area represents the difference between the maximum and minimum values. The black solid line indicates the impacts for H = 10.

This paper studies the direct and indirect price transmission effects of international commodity on the Chinese agricultural industry chain by combining econometric methods and complex network methods. The following conclusions can be obtained.

Price fluctuations of international commodity are not always transmitted along the chain from upstream to downstream, and its transmission characteristics are complex. It may lead to the price volatilities of all agriculture-related products in the Chinese agricultural industry chain. Besides, the linear transmission relationships among international commodity and agriculture-related products are closer than the nonlinear transmission relationships, and the linear transmission range of international commodity is wider than the nonlinear transmission range. It may be because that China’s price intervention policies partially offset the effects of international commodity market volatility, and alter the transmission path and direction of volatility.

Some agriculture-related products are not directly affected by commodity price shocks but they could be indirectly affected. The upstream is most indirectly impacted by CRB, accounting for over 61.52%. In addition, international commodity has the highest indirect impact on agriculture-related products in the second step. After the third price transmission step, the indirect impact of international commodity on agriculture-related products is very small and can be ignored.

The roles of agriculture-related products are different in the agricultural industry chain. Some products have a strong influence in transmitting price fluctuations, such as aquatic product, while some products are highly susceptible to price fluctuations of others, such as cooking oil. There are also products with strong intermediary ability that are “bridges” for cascaded price transmission, such as soybeans.

According to the above conclusions, this paper provides targeted policy implications as below.

From the perspective of the agricultural industrial chain, the influence of price fluctuations of international commodity on an agricultural product is not independent. The price volatility of one product may be transmitted to all products in the agricultural industry chain. Therefore, when the price of a certain product fluctuates, the price regulators should also pay attention to the price volatilities of other products in the industrial chain. Besides, the price regulators should not only focus on the linear influence of price volatility of international commodity, but also focus on the nonlinear influence. It is suggested to build a price monitoring platform, which can integrate the price data of agriculture-related products across the whole industry chain and their influencing factors, such as international commodities and policy interventions. This platform not only visualizes the cascading transmission path of price fluctuations and provide real-time updates, but also sends warning signals to products with abnormal fluctuations.

When measuring the effect of international commodity on the Chinese agricultural industrial chain, the price regulators need to consider the indirect influence of price fluctuation of international commodity on agriculture-related products. If only the direct impacts are considered, the impact of international commodity would be underestimated, resulting in the ineffective functioning of the price regulation mechanisms. Thus, it is suggested that the agricultural sectors integrate resources in the upstream, midstream, and downstream of the agricultural industry chain. This measure can reduce transaction costs and enhance the efficiency of price information transmission, thereby mitigating the indirect impacts of international commodity price fluctuations on Chinese agricultural market.

Different products play different roles in the industry chain, so the government should take targeted measures on these products. Some agriculture-related products play an intermediary role in the agricultural industrial chain and greatly drive the transmission of price fluctuations of international commodity. When the price of international commodity changes, the regulators should concentrate on these products with high intermediary ability to impede the cascaded price transmission. The regulators should also conduct comprehensive supervision on the production, processing and sales of these agricultural products. This can effectively block the further price transmission in the industrial chain by stabilizing price expectations and strengthening market access management. In addition, some products may be easily influenced by other products, and the regulators can monitor them from a preventive perspective to reduce the influence of international commodity on such products. Regarding this kind of products, the government is suggested to increase the reserve of agricultural product by building warehouse and logistics facilities, provide price subsidies, and improve the agricultural insurance system, thus enhancing the ability to resist the risks of price fluctuations.

The limitations of this paper are threefold. Firstly, we ignore the time lag of the price transmission in this paper. In the future, we will design a new transmission model to research the time lag of price transmission from international commodity to the Chinese agricultural industry chain. Secondly, the present model does not take into account exogenous variables. We will construct models that incorporate exogenous variables such as policy and trade in future studies to strengthen the reliability of the research. Finally, our analyses are mainly based on the theory of complex networks, and we will strengthen the analysis of economic theory in our future research.

Data are provided within the Supplementary Information files.

The data is obtained from the Ministry of Agriculture and Rural Affairs of the People’s Republic of China (http://www.moa.gov.cn).

The CRB Index, developed by Commodity Research Bureau, covers six categories of commodities, namely food, fat and oil, livestock, metal, industrial material, and textile.

A structural break test is performed for the variables involved in this study. Please see Appendix B for details.

Alexander C, Wyeth J (1994) Cointegration and market integration: an application to the Indonesian rice market. J Dev Stud 30(2):303–334. https://doi.org/10.1080/00220389408422317

Article Google Scholar

Arnade C, Cooke B, Gale F (2017) Agricultural price transmission: China relationships with world commodity markets. J Commod Mark 7:28–40. https://doi.org/10.1016/j.jcomm.2017.07.001

Article Google Scholar

Asche F, Dahl RE, Valderrama D, Zhang D (2014) Price transmission in new supply chains—the case of salmon in France. Aquac Econ Manag 18(2):205–219. https://doi.org/10.1080/13657305.2014.903309

Article Google Scholar

Baltzer K (2013) International to domestic price transmission in fourteen developing countries during the 2007-08 food crisis. https://doi.org/10.1093/acprof:oso/9780198718574.003.0002

Baquedano FG, Liefert WM (2014) Market integration and price transmission in consumer markets of developing countries. Food Policy 44:103–114. https://doi.org/10.1016/j.foodpol.2013.11.001

Article Google Scholar

Capitanio F, Adinolfi F, Goodwin BK, Rivieccio G (2019) A copula-based approach to investigate vertical shock price transmission in the Italian hog market. N Medit 18(1):3–14. https://doi.org/10.30682/nm1901a

Article Google Scholar

Chaudhry MI, Miranda MJ (2020) Price transmission in Pakistan’s poultry supply chain. J Agric Resour Econ 45(2):282. https://doi.org/10.22004/ag.econ.302455

Article Google Scholar

Chen SW (2007) Measuring business cycle turning points in Japan with the Markov Switching Panel model. Math Comput Simul 76(4):263–270. https://doi.org/10.1016/j.matcom.2006.11.003

Article MathSciNet Google Scholar

Deaton A, Laroque G (1995) Estimating a nonlinear rational expectations commodity price model with unobservable state variables. J Appl Econ 10(S1):S9–S40. https://doi.org/10.1002/jae.3950100503

Article Google Scholar

Diebold FX, Yilmaz K (2012) Better to give than to receive: predictive directional measurement of volatility spillovers. J Forecast 28(1):57–66. https://doi.org/10.1016/j.ijforecast.2011.02.006

Article Google Scholar

Diks C, Panchenko V (2006) A new statistic and practical guidelines for nonparametric Granger causality testing. J Econ Dyn Control 30(9-10):1647–1669. https://doi.org/10.1016/j.jedc.2005.08.008

Article MathSciNet Google Scholar

Djuric I, Goetz L (2016) Export restrictions—Do consumers really benefit? The wheat-to-bread supply chain in Serbia. Food Policy 63:112–123. https://doi.org/10.1016/j.foodpol.2016.07.002

Article Google Scholar

Feng S, Sun Q, Liu X, Xu T (2022) Spillover network features from the industry chain view in multi-time scales. Entropy 24(8):1108. https://doi.org/10.3390/e24081108

Article PubMed PubMed Central Google Scholar

Fowler JH, Christakis NA (2008) Dynamic spread of happiness in a large social network: longitudinal analysis over 20 years in the Framingham Heart Study. Br Med J 337:a2338. https://doi.org/10.1136/bmj.a2338

Article Google Scholar

Gardner BL (1975) The farm‐retail price spread in a competitive food industry. J Agr Econ 57(3):399–409

Article Google Scholar

Haile MG, Kalkuhl M, Algieri B, Gebreselassie S (2017) Price shock transmission: evidence from the wheat-bread market value chain in Ethiopia. Agric Econ 48(6):769–780. https://doi.org/10.1111/agec.12373

Article Google Scholar

Kharisma B, Indrawan ZMS (2023) Analysis of rice price transmission in West Java, Indonesia. Cogent Food Agr 9(2). https://doi.org/10.1080/23311932.2023.2266198

Liao X, Li W (2023) Research on the tail risk contagion in the international commodity market on the China’s financial market: based on a network perspective. Kybernetes. https://doi.org/10.1108/k-06-2023-1001

Ma ZW, Hou WJ (2019) The interactions between Chinese local corn and WTI crude oil prices: an empirical analysis. Pet Sci 16(4):929–938. https://doi.org/10.1007/s12182-019-0339-1

Article CAS Google Scholar

Mila FA, Nahar A, Hossain ME, Amin MR (2023) Spatial price transmission in the onion markets of Bangladesh: an application of NARDL approach. PLoS ONE 18(4):e0284555. https://doi.org/10.1371/journal.pone.0284555

Article CAS PubMed PubMed Central Google Scholar

Nazlioglu S (2011) World oil and agricultural commodity prices: evidence from nonlinear causality. Energy Policy 39(5):2935–2943. https://doi.org/10.1016/j.enpol.2011.03.001

Article Google Scholar

Nicholson W, Matteson D, Bien J (2017) Bigvar: tools for modeling sparse high-dimensional multivariate time series. https://doi.org/10.48550/arXiv.1702.07094

Rapsomanikis G, Hallam D (2006) Market integration and price transmission in selected food and cash crop markets of developing countries: review and applications. FAO Commodity and Trade Policy Research Working Papers, pp 1–90

Ricci EC, Peri M, Baldi L (2019) The effects of agricultural price instability on vertical price transmission: a study of the wheat chain in Italy. Agriculture 9(2):36. https://doi.org/10.3390/agriculture9020036

Article Google Scholar

Roache SK (2010) What explains the rise in food price volatility? IMF Working Papers, pp 1–29. https://doi.org/10.5089/9781455201129.001

Sephton PS (2003) Spatial market arbitrage and threshold cointegration. Am J Agr Econ 85(4):1041–1046. https://doi.org/10.1111/1467-8276.00506

Article Google Scholar

Shahzad SJH, Bouri E, Kristoufek L, Saeed T (2021) Impact of the COVID-19 outbreak on the US equity sectors: evidence from quantile return spillovers. Financ Innov 7(1):14. https://doi.org/10.1186/s40854-021-00228-2

Article PubMed PubMed Central Google Scholar

Singh K, Dey MM, Laowapong A, Bastola U (2015) Price transmission in Thai aquaculture product markets: an analysis along value chain and across species. Aquac Econ Manag 19(1):51–81. https://doi.org/10.1080/13657305.2015.994236

Article Google Scholar

Sun QR, An HZ, Gao XY, Guo S, Wang Z, Liu SY, Wen SB (2019) Effects of crude oil shocks on the PPI system based on variance decomposition network analysis. Energy 189:116378. https://doi.org/10.1016/j.energy.2019.116378

Article Google Scholar

Sun QR, Gao XY, Wen SB, Chen ZH, Hao XQ (2018) The transmission of fluctuation among price indices based on Granger causality network. Phys A 506:36–49. https://doi.org/10.1016/j.physa.2018.04.055

Article Google Scholar

Takahashi D (2012) The distributional effect of the rice policy in Japan, 1986-2010. Food Policy 37(6):679–689. https://doi.org/10.1016/j.foodpol.2012.07.008

Article Google Scholar

Thow AM, Swinburn B, Colagiuri S, Diligolevu M, Quested C, Vivili P, Leeder S (2010) Trade and food policy: case studies from three Pacific Island countries. Food Policy 35(6):556–564. https://doi.org/10.1016/j.foodpol.2010.06.005

Article Google Scholar

Tibshirani R (1997) The lasso method for variable selection in the Cox model. Stat Med. 16(4):385–395. 10.1002/(sici)1097-0258(19970228)16:4<385::Aid-sim380>3.0.Co;2-3

3.0.CO;2-3" data-track-item_id="10.1002/(SICI)1097-0258(19970228)16:43.0.CO;2-3" data-track-value="article reference" data-track-action="article reference" href="https://doi.org/10.1002%2F%28SICI%291097-0258%2819970228%2916%3A4%3C385%3A%3AAID-SIM380%3E3.0.CO%3B2-3" aria-label="Article reference 33" data-doi="10.1002/(SICI)1097-0258(19970228)16:43.0.CO;2-3">Article CAS PubMed Google Scholar

Xi Z, Yu J, Sun Q, Zhao W, Wang H, Zhang S (2023) Measuring the multi-scale price transmission effects from crude oil to energy stocks: a cascaded view. Int Rev Financ Anal 90:102891. https://doi.org/10.1016/j.irfa.2023.102891

Article Google Scholar

Xu S-W, Li Z-M, Cui L-G, Dong X-X, Kong F-T, Li G-Q (2012) Price transmission in China’s swine industry with an application of MCM. J Integr Agric 11(12):2097–2106. https://doi.org/10.1016/s2095-3119(12)60468-7

Article Google Scholar

Zhang X, Liu Y (2020) The dynamic impact of international agricultural commodity price fluctuation on Chinese agricultural commodity prices. Int Food Agribus Manag Rev 23(3):391–409. https://doi.org/10.22434/ifamr2019.0172

Article Google Scholar

Zheng T, Zhao GQ, Chu SW (2024) A study on the impact of external shocks on the resilience of China’s grain supply chain. Sustainability 16(3):956. https://doi.org/10.3390/su16030956

Article Google Scholar

Download references

This work was supported by the Humanities and Social Sciences Youth fund project under the Ministry of Education of the PRC (No. 24YJCZH264) and the Social Science Cultivation Project of Hebei University (No. 2023HPY014).

Hebei University, Baoding, China

Qingru Sun, Yaojun Shan, Shukun Hu, Mingting Ding & Han Zhang

Hebei Vocational University of Technology and Engineering, Xingtai, China

Jinfeng Miao

Search author on:PubMed Google Scholar

QS wrote the original draft. JM and YS put forward the framework of the paper and revised the draft. SH calculated the network results and prepared all figures. MD prepared all tables and conducted content discussion. HZ collected the data and constructed the network model. All authors reviewed the manuscript.

Correspondence to Jinfeng Miao or Yaojun Shan.

The authors declare no competing interests.

This article does not contain any studies with human participants performed by any of the authors.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Sun, Q., Miao, J., Shan, Y. et al. Revisiting the price transmission from international commodity to Chinese agricultural products based on an industry chain perspective. Humanit Soc Sci Commun 12, 936 (2025). https://doi.org/10.1057/s41599-025-05238-4

Download citation

Received: 22 October 2024

Accepted: 05 June 2025

Published: 01 July 2025

DOI: https://doi.org/10.1057/s41599-025-05238-4

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

News

Revisiting the price transmission from international commodity to Chinese agricultural products based on an industry chain perspective | Humanities and Social Sciences Communications