Sage Journals: Discover world-class research

Abstract

This study aims to assess the ability of agricultural markets to operate stably in the face of uncertain shocks. This study conducts an empirical analysis using the spillover effect index, based on daily price data of 9 agricultural products from 259 wholesale markets across 12 provinces in China, covering the period from January 1, 2018, to August 31, 2022. The results reveal: First, the agricultural product market system demonstrates remarkable resilience. Although the overall price spillover effects surged sharply during the peak of the shock, they quickly declined as the pandemic normalized, indicating that the shock did not cause permanent structural damage. This finding provides empirical support for the “agricultural market resilience” theory from the perspective of a major public health crisis. Second, the impact of the shock exhibits significant heterogeneity across categories. Perishable agricultural products (e.g., cucumbers, cabbage) and basic staple foods (e.g., wheat, eggs) became key hubs for risk transmission. In contrast, the spillover effects of livestock and poultry products showed high instability. The government should establish a risk-based categorized reserve system, with policy resources tilted towards perishable agricultural products and staple foods. Third, risk transmission follows spatiotemporal evolution patterns. The spillover effects adhere to a dynamic path of “stable-rising-normal,” and interregional linkages are highly pronounced, with over half of the price fluctuations originating from cross-provincial spillovers. Given the high degree of regional interconnectedness, it is essential to break down the “market fragmentation” caused by administrative divisions, and strengthen cross-regional information sharing and logistics coordination.

Keywords

uncertainty shocks agricultural products market price spillover effects daily price data

Introduction

The market is a dynamic process in which numerous economic agents continuously collect, process, and disseminate dispersed information in an uncertain environment and make purposeful decisions. The stable operation of the market requires market participants to respond promptly to market information and adjust their behavior flexibly. Unforeseen external environmental changes that interfere with transactions are directly related to the market’s benign operation and security. In the face of uncertainty shocks, whether the market can quickly adjust and recover to a reasonable state while performing its basic functions is an important economic proposition for measuring market stability. However, empirical research on this proposition has mostly focused on virtual asset markets such as exchange rates, stocks, and finance (Kakran et al., 2025; Kang et al., 2017; Zhou et al., 2021; Zhu et al., 2025), and there are relatively few studies that quantitatively characterize the ability of physical commodity markets to withstand external shocks. Therefore, it is necessary to find suitable real-life scenarios to scientifically evaluate the market's ability to operate stably when facing uncertainty shocks and to provide theoretical references for accurately grasping market price fluctuations and enhancing market resilience.

The outbreak of COVID-19 in 2020 was one of the most severe and widespread infections in human history, posing not only a threat to people's health but also an unprecedented impact on the global economy and trade. According to the 2021 World Trade Report, the global economy experienced the worst recession since World War II in 2020 due to the pandemic, with a 9.6% decline in the value of goods and services trade and a 3.3% decrease in global GDP. Although China effectively contained the virus through strict isolation and control measures, the pandemic has become long-term and normalized due to its widespread, fast transmission and difficulty in prevention and control (Cheng et al., 2023). The occasional outbreaks in local areas of the pandemic have brought great unpredictability and uncertainty to market activities, which has an undeniable impact on the stable operation of the market. This provides a natural experiment for examining the volatility of physical markets in the face of uncertainty shocks. As a pure external shock, the impact of the COVID-19 epidemic on the market is short-term and does not change the medium to long-term operating trend of the market. However, under the normalization of the epidemic, the local repeated outbreaks have led to increased market uncertainty, such as price fluctuations, variable demands, and prediction deviations, which have changed the market participants' expectations and their consumption and investment behaviors, thereby constraining the market's medium- to long-term stable operation. From the perspective of consumer demand, the uncertainty of the epidemic has reduced consumers' demand for products and services, which has had a negative impact on the manufacturing and service industries (Al Mamun et al, 2025; Jung, 2022). From the perspective of consumer behavior, the epidemic has increased people's demand for delivery services and online shopping, thereby increasing the use of mobile payments (Liu et al., 2020). From the perspective of investment conditions, during the peak of the COVID-19 pandemic, gold played a leading role in hedging against epidemic uncertainty, and many investors sought to hold exchange-traded funds for precious metals (Mokni et al., 2021). The uncertainty associated with COVID-19 has had a negative impact on the returns of almost all industries, resulting in higher volatility (Ramsey et al., 2021; Szczygielski et al., 2022).

In the field of agriculture, given the characteristics of perishability, regional specificity, and vulnerability of agricultural products themselves, the agricultural product market is more susceptible to uncertainties such as natural disasters, policy adjustments, and macroeconomic fluctuations. From the perspective of agricultural production, epidemic prevention and control measures have made it difficult to supply labor, agricultural materials, machinery, and other production materials across regions. The mismatch between supply and demand has led to varying degrees of increase in production costs (Zhang et al., 2020). From the perspective of agricultural product circulation, during the pandemic, supply chains for agricultural products in many countries faced disruptions (T.-T. Nguyen et al., 2025). Under the normalized epidemic prevention and control mechanism, the cold storage and emergency response capabilities of agricultural product logistics are clearly insufficient (Li et al., 2023; Pu et al., 2021). From the perspective of agricultural product trade, the short-term impact of the epidemic has increased the volatility of agricultural product prices, which gradually return to normal levels thereafter (Balcilar et al., 2022; Yan et al., 2021). Undoubtedly, under the impact of uncertainty, the behaviors of micro market entities have undergone systematic adjustments.

The behavioral shifts of countless market participants ultimately manifested as significant changes in the macro-market conditions. First, the existing supply-demand balance in agricultural markets was disrupted. Precautionary hoarding and regional control measures leading to severe mismatches in the timing, geographic distribution, and structure of agricultural products (Cardoso & Malloy, 2021; Mishra et al., 2023). Second, while the value of reliable information surged, the spread of noise and rumors accelerated even faster, damaging the trust between buyers and sellers and drastically exacerbating information asymmetry (Jin et al., 2025; Xu et al., 2025). Such market chaos inevitably eroded the health of the price mechanism as a core signaling system. The price formation process no longer merely reflected immediate costs and utility but also incorporated a substantial risk premium for uncertainty (D. T. Nguyen, 2025; Su et al., 2025). As a result, price signals experienced severe volatility and frequent distortions.

What makes the situation more complex is that such uncertainty shocks do not act in isolation on a single market. Due to the price transmission effect, the agricultural product market is not an independent entity, and there is a certain degree of correlation between various product markets. The price fluctuations of a single agricultural product often lead to large fluctuations in related product prices. At the same time, due to the significant differences in the direct impact of uncertainty shocks on different agricultural products (Yu et al., 2020), it may cause different market performance of different agricultural products, and further lead to changes in the interaction between products. For example, data from the Development and Reform Commission of Guanyun County, Jiangsu Province, showed that on January 28, 2020, prices of vegetables such as celery, Chinese cabbage, white radish, eggplant, potato, and cabbage in the region generally increased, exhibiting clear co-movement characteristics. However, although the prices of all listed vegetables rose, the specific increases varied significantly, ranging from 12.5% to 50%. That is, there is an imbalance in the response degree among different agricultural products. Therefore, focusing solely on price fluctuations in individual markets is insufficient to fully understand the impact of uncertainty shocks on agricultural markets. Studying the price spillover effects across different agricultural markets under uncertainty shocks is crucial for preventing systemic market risks.

Currently, academia has conducted extensive research on market price volatility and spillover effects. Some scholars focus on the vertical transmission of market prices among different stages of the industry chain. Apergis and Rezitis (2003) used the generalized autoregressive conditional heteroscedasticity model (GARCH) to verify the positive volatility spillover effect among input price, output price, and retail price of bulk agricultural products. Buguk et al.(2003) take the catfish supply chain in the United States as the research object and find that there is a strong spillover effect between the price of feed raw materials, the price of feed, and the wholesale price of catfish. Wan and Li (2022) used the asymmetrical MGARCH-BEKK model to examine the structural breaks among piglet, hog, and pork prices in China, and their conclusions support the existence of asymmetric volatility transmission in the supply chain. Some scholars focus on the horizontal transmission between different agricultural markets. Szenderak et al. (2019) examine the spillover level of milk producer prices between EU member States and its development trend. Balli et al. (2019) analyze the overall association of 22 commodities during the global financial crisis and the oil price collapse of 2014 to 2016, based on the temporal and frequency connectivity of the uncertainty index.

Based on the current situation of the COVID-19 pandemic, this paper uses daily price data from domestic wholesale markets from January 1, 2018, to August 31, 2022, to examine both the overall and dynamic spillover effects between agricultural markets. Unlike studies that focus solely on single categories of agricultural markets (Lindsey et al., 2025; T.-T. Nguyen et al., 2025), this research aims to reveal price fluctuations in agricultural markets and the dynamic changes in information spillovers among different types of agricultural markets under major uncertainty shocks, thereby providing more systematic and comprehensive empirical evidence for understanding the operational mechanisms of agricultural markets.

This study employs the spillover index method by Diebold and Yilmaz (2012) to measure the overall spillover level (i.e., overall connectivity) of the entire agricultural market and identify whether individual markets act as “transmitters” or “receivers” of risks. Through rolling window analysis, this approach overcomes the limitations of static correlation coefficients and traditional VAR models in capturing the time-varying transmission characteristics of shocks across different phases, thereby enabling “phase-specific investigation.” Additionally, the dynamic smoothing nature of this method avoids the potential market “noise” that might be captured by high-frequency parameter fluctuations in time-varying models like TVP-VAR, making it more aligned with the relatively slower price fluctuations characteristic of agricultural markets compared to financial markets.

This study finds that agricultural markets exhibit strong resistance and self-adjusting capacity when facing external shocks, without suffering irreversible impacts, supporting the prevailing academic view of “agricultural market resilience” (Adenäuer et al., 2025; Hadachek et al., 2024). Such resilience does not imply market stagnation but manifests as a dynamic self-adjusting process: initially absorbing and dispersing risks through significantly enhanced price linkage effects during the shock phase, followed by a gradual return to normalcy.

We also found that the impact of uncertainty shocks exhibits significant category heterogeneity. The effects were primarily concentrated on perishable agricultural products with short storage periods, such as cucumbers and cabbage, as well as staple crops and protein sources like wheat and eggs. This indicates that the transmission mechanism of shocks is closely related to product characteristics and inelastic demand, providing micro-level evidence and theoretical explanations from the perspective of price linkages for the macro-level phenomenon of “shortages of staple products in global retail stores under uncertainty shocks” documented in existing literature (Aday & Aday, 2020).

Furthermore, we extended our investigation to examine the spatial correlations and interdependence of agricultural markets from a regional perspective. In terms of overall spillover effects, the price spillover index among regional agricultural markets demonstrated an initial increase followed by a subsequent decline. Market fluctuations across regions exhibited strong interconnectedness, with over half of the price volatility originating from interregional spillover effects. This indicates that during the initial phase of the shock, risks rapidly propagated through existing channels, leading to heightened connectivity. However, in the medium to late stages of the shock, as regions implemented adaptive policies or adjusted localized supply chains, interregional interdependence moderated. This “hump-shaped” evolution pattern provides new theoretical insights for understanding the spatiotemporal dynamics of agricultural supply chains during crises.

The findings of this study provide new empirical evidence for the applicability of price transmission theory and market integration theory under uncertain conditions, deepening the understanding of market operational mechanisms during unexpected events. For government departments, it helps identify key risk sources and vulnerable markets, providing a decision-making basis for formulating precise agricultural price regulation policies and supply chain stabilization measures during crises. For market participants (such as farmers, distributors, and investors), it enhances their understanding of risk transmission patterns, thereby facilitating better risk management, production planning, and investment decisions.

The rest of the paper is organized as follows. Section “Model Specification and Data Description” briefly introduces the research data and methods. Section “Empirical Result Analysis” analyzes and discusses the empirical results. The section “Conclusion and Discussion” presents conclusions, relevant policy recommendations, and prospects for future research.

Model Specification and Data Description

Model Specification

This paper employs the spillover index proposed by Diebold and Yilmaz (2012) for empirical analysis to examine the transmission of price volatility across different agricultural markets. The core logic of this method is as follows: when an unexpected price fluctuation (i.e., a “shock”) occurs in one market, the model quantifies the extent to which this fluctuation is “caused” by fluctuations in other markets. Using the generalized impulse response method to perform variance decomposition on the forecast errors of a vector autoregression (VAR) model, this approach allows for the simultaneous examination of interdependencies among all markets, revealing trends, cycles, and outbreaks of price spillover effects across multiple markets. This method enables the time-varying estimation of the volatility spillover index, providing insights into both the time-varying total spillover effects of the market and the time-varying net spillover and net receipt effects between different markets in terms of magnitude and direction. This helps visualize how the transmission mechanisms of various crises operate through different economic channels. By incorporating a rolling window analysis, it becomes possible to intuitively observe: during different phases, such as the outbreak period and the normalized prevention and control period, whether the overall market connectivity strengthens or weakens, which markets consistently act as sources of risk during crises, and which markets undergo the most drastic changes in their roles. Additionally, compared to the Cholesky orthogonal variance decomposition method, this approach eliminates the dependence of estimation results on the ordering of variables. The specific model construction steps are as follows.

Firstly, construct a covariance stationary N-dimensional VAR(P) model:

\begin{matrix} x_{t} = \sum_{i = 1}^{p} Φ_{i} x_{t - i} + ε_{t} \end{matrix}

(1)

Where $x_{t}$ is the N-dimensional column vector composed of prices of various agricultural products in the market. $ε_{t}$ is a disturbance vector that follows an independent and identically distributed distribution. The moving average form of VAR is:

\begin{matrix} x_{i} = \sum_{i = 0}^{\infty} A_{i} ε_{t - i} \end{matrix}

(2)

Where the N × N coefficient matrix $A_{i}$ follows the recursive formula:

\begin{matrix} A_{i} = Φ_{1} A_{i - 1} + Φ_{2} A_{i - 2} + \dots + Φ_{p} A_{i - p} \end{matrix}

(3)

Where $A_{0}$ is the identity matrix, and $A_{i} = 0$ when $i < 0$ . Based on the generalized variance decomposition, the self-variance share is defined as the proportion of the variance of the H-period forecast error of the product price that is explained by its own price shock, while the cross-variance share (i.e. spillover) is defined as the proportion of the variance of the H-period forecast error of the product price $x_{i}$ that is explained by the price shock of product j. For H = 1, 2 ⋯, the forecast variance decomposition can be expressed as:

\begin{matrix} θ_{ij}^{g} = \frac{σ_{ii}^{- 1} \sum_{h = 0}^{H - 1} {(e_{i}^{'} A_{h} \sum e_{j})}^{2}}{\sum_{h = 0}^{H - 1} (e_{i}^{'} A_{h} \sum A_{h}^{'} e_{j})} \end{matrix}

(4)

Where ∑ represents the covariance matrix of the forecast error vector ε. $σ_{ii}$ is the standard deviation of the i-th equation's forecast error variance. And $e_{i}$ is a column vector with the i-th element equal to 1 and all other elements equal to 0. Due to the fact that the impact on each variable in the generalized variance decomposition is not orthogonal, the sum of the contributions of the prediction error variance may not necessarily be equal to 1. Therefore, we standardize the variance decomposition matrix, and the row sum is represented as:

\begin{matrix} {\tilde{θ}}_{ij}^{g} (H) = θ_{ij}^{g} (H) / \sum_{j = 1}^{N} θ_{ij}^{g} (H) \end{matrix}

(5)

Where ${\tilde{θ}}_{ij} (H)$ represents the price spillover effect of agricultural product on agricultural product j, which can be decomposed into $\sum_{j = 1}^{N} {\tilde{θ}}_{ij}^{g} (H) = 1$ and $\sum_{j = 1}^{N} {\tilde{θ}}_{ij}^{g} (H) = 1$ . To measure the overall spillover effect of agricultural product prices, we construct the total spillover index using the volatility contribution of the variance decomposition as follows:

S^{g} (H) = \frac{\sum_{i, j = 1, i \neq j}^{N} {\tilde{θ}}_{ij}^{g} (H)}{\sum_{i, j = 1}^{N} {\tilde{θ}}_{ij}^{g} (H)} \times 100 = \frac{\sum_{i, j = 1, i \neq j}^{N} {\tilde{θ}}_{ij}^{g} (H)}{N} \times 100

(6)

As the generalized variance decomposition is not dependent on the order of variables, we can use the standardized elements of the generalized variance decomposition matrix to calculate directional spillovers. The directional spillover of directed volatility that crop i receives from all other crop markets is given by:

\begin{matrix} S_{i \cdot}^{g} = \frac{\sum_{j = 1, i \neq j}^{N} {\tilde{θ}}_{ij}^{g} (H)}{\sum_{j = 1}^{N} {\tilde{θ}}_{ij}^{g} (H)} \times 100 \end{matrix}

(7)

Similarly, the directional spillover of oriented volatility transmitted by agricultural product i to all other agricultural markets is:

\begin{matrix} S_{\cdot i}^{g} = \frac{\sum_{j = 1, i \neq j}^{N} {\tilde{θ}}_{ji}^{g} (H)}{\sum_{j = 1}^{N} {\tilde{θ}}_{ji}^{g} (H)} \times 100 \end{matrix}

(8)

The net spillover index is the difference between the spillover effects transmitted to other agricultural commodity markets and the spillover effects received from other agricultural commodity markets, and is expressed as:

\begin{matrix} S_{i}^{g} (H) = S_{\cdot i}^{g} (H) - S_{i \cdot}^{g} (H) \end{matrix}

(9)

Similarly, we can calculate pairwise net spillover effects between two markets, defined as the difference between the spillover effect of agricultural product i on agricultural product j and the spillover effect of agricultural product j on agricultural product i. Specifically, it can be expressed as:

\begin{matrix} S_{ij}^{g} (H) = [\frac{{\tilde{θ}}_{ij}^{g} (H)}{\sum_{k = 1}^{N} {\tilde{θ}}_{ik}^{g} (H)} - \frac{{\tilde{θ}}_{ji}^{g} (H)}{\sum_{k = 1}^{N} {\tilde{θ}}_{jk}^{g} (H)}] \times 100 \end{matrix}

(10)

Data Description

For the price spillover effects of agricultural products, this article selected the wholesale price data of agricultural products in China for analysis. The data were obtained from various websites such as the China Price Information Network, the National Agricultural Products Wholesale Market Price Information System, and the official website of the Ministry of Commerce. After removing data with significant missing values, a relatively complete daily price data set of nine types of agricultural products in 259 wholesale markets across 12 provinces (autonomous regions/municipalities) was obtained. Individual missing values shall be filled with the linear interpolation method. These 12 provinces (autonomous regions/municipalities) are Inner Mongolia Autonomous Region, Beijing municipality, Sichuan province, Tianjin municipality, Anhui province, Shandong province, Shanxi province, Xinjiang Uygur Autonomous Region, Jiangsu province, Hebei province, Liaoning province, and Qinghai province, while the 9 types of agricultural products are rice, wheat, potatoes, Chinese cabbage, cucumber, eggs, pork, mutton, and beef.

For data from different wholesale markets in the same region, this study conducted weighted averaging by agricultural product category. When calculating the overall prices of agricultural products in different regions, this study used the production volume of different agricultural products in each region as weights to calculate the weighted average of agricultural product prices in each region. It should be noted that this study places greater emphasis on staple foods and vegetables, which are essential agricultural products in the daily diet of Chinese residents. In contrast, while fruits are also important, their consumption elasticity and substitutability are relatively higher, and their short-term absence has a comparatively limited impact on residents’ livelihoods. By focusing on staple food and vegetable markets, we aim to more clearly capture the direct effects of external shocks on fundamental agricultural markets, thereby more accurately assessing the potential risks they may pose to normal living conditions. Taking into account both data availability and the representativeness of the research subjects, we ultimately selected the aforementioned nine agricultural products.

The sample period for this study is from January 1, 2018, to August 31, 2022. Based on the development of the epidemic situation, we selected two key time points: January 20, 2020, when President of China Xi Jinping issued important instructions on epidemic prevention and control, and April 8, 2020, when Wuhan was lifted from lockdown. The study period was divided into three stages: a stable period before the outbreak of the epidemic, a strict control period after the outbreak of the epidemic, and a period of normalized epidemic control. This period fully encompasses the “normalcy-shock-response” research cycle, effectively capturing the entire process of the emergence, evolution, and gradual dissipation of uncertainty shocks. It provides an ideal real-world context for identifying structural changes and time-varying characteristics in price spillover effects within agricultural markets. Moreover, as the selected timeframe excludes the later phase of policy adjustments, the study can focus squarely on the impact of the pandemic itself without being confounded by other policy changes.

Empirical Result Analysis

To avoid the phenomenon of spurious regression and ensure the accuracy of the estimation model, this paper first conducted the ADF unit root test on each variable, and the results showed that all variables significantly rejected the null hypothesis of unit root and were stationary sequences. Secondly, this paper determined the optimal lag order of the VAR model as 4 based on information criteria such as AIC, HQIC, and SBIC. Finally, this paper constructed the spillover effect index based on the generalized variance decomposition.

Analysis of Overall Spillover Effects

The specific calculation results of the spillover effects in each stage are shown in Table 1. From the table, “From” represents the impact of spillover effects from other agricultural products on the price of the product; “To” represents the spillover effect of the product price on the prices of other products; “Net” represents the net spillover effect of the product; “Total” represents the total spillover effect that includes the impact of the product price on itself. Looking at the overall spillover effect of agricultural product market prices, the total spillover index shows a development trend of first rising and then falling. In the stable stage before the outbreak of the epidemic, the total spillover index was 15.25%, indicating that 15.25% of the fluctuations in agricultural product market prices could be explained by the price fluctuation correlations among related products. This result indicates that although there are differences in market demand for various agricultural products, due to reasons such as product substitutability, a relatively close correlation structure has been formed among different types of agricultural product markets. The price fluctuations of different products have obvious synergy. In the strict control stage after the outbreak of the epidemic, the total spillover index soared to 54.89%, indicating that more than half of the fluctuations in agricultural product market prices were caused by the spillover effects among products. This change reflects that external uncertainties have significantly strengthened the volatile correlation in the agricultural product market. One possible reason is that in the face of uncertain shocks, the substitution and complementary effects among agricultural products have strengthened, and the market price transmission mechanism has become more active. This is the market's stress response when facing shocks. After the epidemic entered the stage of normalized prevention and control, the total spillover index of agricultural product market prices fell back to 15.12%, which was similar to the stable stage before the outbreak of the epidemic. This indicates that although the uncertainty shock has significantly enhanced the volatility spillover effect of the agricultural product market in the short term, the agricultural product market has not been irreversibly affected. After entering the adaptation period, the agricultural product market can return to a normal level. This result confirms that the agricultural product market has demonstrated strong resistance and self-adaptability in the face of external shocks. This is consistent with the research results of Elleby et al (2020) and Kalli et al (2025).

Table 1.

Spillover Effect of Market Price of Agricultural Products in Different Stages.

Stage		Rice	Wheat	Potatoes	Cabbage	Cucumber	Pork	Beef	Mutton	Eggs	From
I	Rice	90.55	1.54	1.05	2.39	0.54	0.67	0.63	1.74	0.89	9.45
	Wheat	1.57	88.49	1.37	0.99	0.89	0.15	1.84	4.19	0.51	11.51
	Potatoes	2.36	1.91	71.48	3.24	1.99	12.53	4.63	1.32	0.53	28.52
	Cabbage	1.13	1.79	2.04	91.01	0.71	1.64	0.71	0.31	0.66	8.99
	Cucumber	0.41	0.68	2.27	0.77	93.88	0.54	0.44	0.79	0.22	6.12
	Pork	0.93	0.30	13.67	1.66	0.75	69.19	9.37	2.94	1.20	30.81
	Beef	0.50	1.49	4.35	0.66	0.73	7.86	81.62	1.72	1.06	18.38
	Mutton	0.52	5.02	0.72	0.80	0.32	3.55	3.21	85.34	0.53	14.66
	Eggs	0.62	0.51	1.10	1.90	0.95	1.89	1.17	0.72	91.14	8.86
	To	8.04	13.24	26.57	12.40	6.88	28.83	22.00	13.73	5.61	137.29
	Net	−1.41	1.73	−1.95	3.41	0.76	−1.98	3.62	−0.92	−3.25	−
	Total	98.59	101.73	98.05	103.41	100.76	98.02	103.62	99.08	96.75	15.25
II	Rice	41.33	2.32	10.77	13.02	11.66	4.44	9.38	2.58	4.51	58.67
	Wheat	2.60	38.09	6.74	19.34	7.64	8.51	8.67	1.91	6.50	61.91
	Potatoes	2.65	4.21	52.49	7.59	7.10	9.41	11.85	3.97	0.71	47.51
	Cabbage	1.69	12.65	9.17	36.24	5.24	6.93	13.59	5.74	8.74	63.76
	Cucumber	3.32	8.33	4.87	8.21	47.47	6.13	8.30	10.59	2.77	52.53
	Pork	5.24	4.53	12.93	16.42	3.94	37.62	2.75	5.53	11.05	62.38
	Beef	3.18	3.26	4.57	3.52	17.99	5.96	48.74	2.87	9.89	51.26
	Mutton	9.57	3.22	5.90	2.22	3.32	9.02	4.07	57.68	5.01	42.32
	Eggs	6.65	10.61	3.80	4.74	6.72	11.63	8.33	1.22	46.29	53.71
	To	34.91	49.12	58.75	75.06	63.62	62.03	66.94	34.42	49.19	494.05
	Net	−23.76	−12.79	11.25	11.31	11.10	−0.35	15.68	−7.91	−4.52	−
	Total	76.24	87.21	111.25	111.31	111.10	99.65	115.68	92.09	95.48	54.89
III	Rice	83.71	6.94	2.42	0.92	0.79	2.25	0.48	1.22	1.26	16.29
	Wheat	6.06	89.97	1.09	0.82	0.48	0.13	0.56	0.39	0.50	10.03
	Potatoes	1.11	0.45	75.62	2.09	8.98	5.48	0.72	3.98	1.57	24.38
	Cabbage	0.23	0.25	2.10	86.60	3.56	1.44	1.21	1.37	3.25	13.40
	Cucumber	0.35	1.03	8.18	3.45	82.93	0.61	0.91	1.07	1.47	17.07
	Pork	2.35	0.49	7.28	1.38	1.53	85.74	0.19	0.47	0.58	14.26
	Beef	0.42	0.59	0.78	0.73	2.03	0.13	89.38	4.12	1.82	10.62
	Mutton	1.16	1.28	5.25	2.03	2.28	1.92	4.01	80.97	1.09	19.03
	Eggs	1.36	0.89	1.44	2.94	1.00	0.91	2.29	0.15	89.01	10.99
	To	13.04	11.91	28.55	14.37	20.65	12.87	10.37	12.78	11.53	136.06
	Net	−3.25	1.88	4.17	0.97	3.58	−1.39	−0.25	−6.25	0.54	−
	Total	96.75	101.88	104.17	100.97	103.58	98.61	99.75	93.75	100.54	15.12

From the perspective of the spillover effects of other agricultural products' prices (“From”), in the stable period before the outbreak of the epidemic, pork and potatoes were more affected by the spillover effects of other agricultural product markets, accounting for 30.81% and 28.52% respectively, while cucumbers, Chinese cabbage, and eggs were less affected, accounting for 6.12%, 8.99%, and 8.86% respectively. During the strict control stage after the outbreak of the epidemic, the prices of various products were greatly affected by market spillover effects, all exceeding 40%, with some agricultural products even exceeding 60%. In terms of the magnitude of increase, the spillover effects of cucumbers and cabbage showed the most dramatic rises, increasing by factors of 7.58 and 6.09, respectively; followed by rice, eggs, and wheat, with increases of 5.21, 5.06, and 4.38 times, respectively; while the increases for pork, beef, mutton, and potatoes were relatively smaller. These results further verify that uncertainty shocks not only intensified price volatility across the agricultural market but also significantly strengthened the price linkage effects among products. Simultaneously, the observed differences reflect the heterogeneous impact of uncertainty shocks on different types of agricultural markets. The effects were more pronounced for perishable products (such as cucumbers, cabbage, and other fresh vegetables with short storage periods) as well as staple grains and protein products (such as wheat and eggs). This may be attributed to their lower supply and demand elasticity, higher storage costs, and inelastic consumption demand (Ghoshray, 2019). During the adaptation phase, the influence of market spillover effects on product prices declined rapidly, and the market gradually stabilized. However, the spillover levels for most products remained higher than during the pre-shock baseline period, particularly for cucumbers, cabbage, and eggs, indicating that the uncertainty shocks had a lingering short-term impact on agricultural markets.

From the perspective of the spillover effects of various agricultural products on other product prices (“To”), during the stable phase before the outbreak of the pandemic, pork, potatoes, and beef had greater price spillover effects, all exceeding 20%; followed by lamb, wheat, and cabbage with spillover effects on other product prices of 13.73%, 13.24%, and 12.40%, respectively; rice, cucumbers, and eggs had relatively small price spillover effects, all below 10%. Pork and beef are the most important meat consumption products for Chinese residents, and they are substitutes for each other with the most volatile prices, thereby creating a large spillover effect on the entire agricultural product market. As a dual-purpose crop functioning as both a grain and a vegetable, potato price fluctuations propagate to both the grain and vegetable markets, consequently exhibiting relatively high spillover effects. During the strict control phase after the outbreak of the pandemic, all types of products showed a significant increase in spillover effects on other product prices, especially cabbage, cucumbers, pork, beef, and potatoes, with spillover effects all exceeding 50%. From the perspective of the increase rate, cucumbers, eggs, and cabbage had higher spillover effects on the agricultural product market, increasing by 8.25, 7.77, and 5.05 times, respectively. This trend aligns closely with the price spillover effects they receive from other products, further indicating that uncertainty shocks have a particularly pronounced impact on price transmission for these categories of agricultural products. During the subsequent shock adaptation phase, the spillover effects of various agricultural products on other product prices gradually returned to normal levels, demonstrating that the direct impact of the uncertainty shocks gradually dissipated. This finding is consistent with certain theoretical studies (Deaton & Laroque, 1992; Williams & Wright, 1991), which posit that shocks to agricultural prices should be temporary in nature.

From the perspective of net spillover effects of various products (“Net”), in the stable period before the outbreak of the epidemic, rice, potatoes, pork, lamb, and eggs had negative net spillover effects, indicating that these agricultural products were more affected by other products than their impact on the prices of other products. Thus, they acted more as recipients of price fluctuations. Wheat, cabbage, cucumbers, and beef had positive net spillover effects, indicating that these agricultural products had a greater impact on the prices of other products. After the outbreak of the epidemic, the absolute values of the net spillover effects of most agricultural products increased but did not change direction, and gradually declined to normal levels after the epidemic prevention and control entered the normalization stage. It is worth noting that the net spillover values of potatoes and eggs shifted from negative to positive, indicating that under the shock, both transformed from being net recipients to net transmitters. This shift may stem from the heightened prominence of price formation mechanisms for potatoes as an important dual-purpose crop (food and vegetable) and eggs as a primary protein source in an environment of uncertainty.

Analysis of Dynamic Spillover Effects

The above analyzed in detail the spillover effects of agricultural market prices, but it did not reflect the temporal fluctuations of these effects. Therefore, this paper further examines the dynamic changes of various spillover indices in different periods from a dynamic perspective.

Dynamic Change Analysis of Total Spillover Effect

Figure 1 displays the dynamic changes of the total spill-over effects of agricultural product market prices during three different stages. One of the most core phenomena that we can observe from the figure is that the total spillover effect varies significantly at different stages, and the total spillover effects have obvious time-varying characteristics. During the stable period before the outbreak of the epidemic, the total spill-over effects mainly fluctuated slightly between 25% and 35%, reflecting that the overall operation of the agricultural product market was relatively stable during this period. During the strict control stage after the outbreak of the epidemic, the total spill-over effects consistently remained above 50% with narrowed fluctuations. This indicates that external shocks significantly strengthened the interconnectedness among different agricultural markets, leading to more synchronized price transmission and an increase in systemic risk. This observed change aligns with the typical pattern of stress responses in market mechanisms under uncertainty shocks. After entering the normalization stage of epidemic prevention and control, the total spill-over effects gradually declined to the conventional level of 30%, yet with significantly increased volatility. On the one hand, this suggests that as the uncertainty shock stabilized, the market stress response triggered by the sudden events gradually subsided, allowing agricultural product markets to return to a stable operation. On the other hand, the heightened market volatility may stem from sporadic rebounds of the shock and new uncertainties arising from regional adaptation policies. Notably, the total spillover effect in agricultural markets surged sharply in early 2022 (Figure 1c). This anomalous surge was primarily driven by the combined effects of localized resurgences of the shock and the outbreak of the international Russia-Ukraine conflict. The latter, in particular, intensified tensions in the global agricultural trade landscape, thereby amplifying domestic market fluctuations through channels such as import dependency and expectation transmission (Guo et al., 2023).

Figure 1.

(a) The dynamic changes of the total spillover effect of agricultural product market prices during the stable period before the shock. (b) The dynamic changes of the total spillover effect of agricultural product market prices during the period of uncertainty shock. (c) The dynamic changes of the total spillover effect of agricultural product market prices during the adaptation period after the shock.

Dynamic Analysis of Directional Spillover Effects

Figure 2 illustrates the dynamic trends of directional spillover effects among various agricultural product prices. Overall, although the price spillover effects of different agricultural products exhibit certain differences in magnitude and timing, they demonstrate similar fluctuation patterns throughout the entire sample period. Before the shock, the spillover effects of all products remained relatively stable. Following the uncertainty shock, a significant increase was observed, accompanied by more pronounced volatility. As the adaptation phase began, the spillover effects gradually returned to conventional levels. It is noteworthy that rice and wheat experienced a pronounced peak in spillover effects in early 2022 (Figure 2c), primarily due to rising global grain prices and disruptions in international import-export trade triggered by the Russia-Ukraine conflict, which subsequently drove up domestic grain prices and enhanced price spillover effects of staple food varieties.

Figure 2.

(a) The dynamic changes of the spillover effects of different agricultural product prices during the stable period before the shock. (b) The dynamic changes of the spillover effects of different agricultural product prices during the period of uncertainty shock. (c) The dynamic changes of the spillover effects of different agricultural product prices during the adaptation period after the shock.

In terms of volatility, livestock products such as pork, beef, mutton, and eggs displayed relatively high instability in price spillover effects during the sample period, with substantial variations across different stages. This phenomenon can be explained from two perspectives: On one hand, livestock products are more frequently influenced by factors such as feed costs and animal disease outbreaks, resulting in inherently volatile price formation processes. On the other hand, meat and egg products exhibit high consumption substitutability and interrelation, where price fluctuations in one product can rapidly propagate across the livestock product category. Additionally, differences in market structures and regulatory policies among various livestock products lead to varying price adjustment capacities and transmission rates (Bittmann et al., 2020; Loy et al., 2015), further exacerbating the instability of spillover effects.

Looking at the trend of how different agricultural products are affected by spillover effects from other product prices (as shown in Figure 3), it shows that different agricultural products exhibit the same trend in terms of spillover effects from other product prices during the sample period. That is, before being impacted by the pandemic, the spillover effects were relatively stable, but there was a clear increasing trend after entering the strict control phase, and then returned to a normal level as pandemic prevention and control entered the normalization phase. In terms of fluctuation amplitude, the spillover effects of various agricultural product prices from other products were generally stable during each phase, with relatively small fluctuations. Specifically, except for rice and wheat, which had a slight increase in spillover effects at the beginning of 2022, there was also a slight trend of increased fluctuation in the prices of eggs, cabbage, and cucumbers. This phenomenon indicates that although the overall level of spillover effects affecting some agricultural products remains relatively low, their dynamic stability still varies. These products may exhibit sensitivity changes during different periods due to external events or internal transmission mechanisms, thus warranting continuous attention in both policy formulation and market risk monitoring.

Figure 3.

(a) The dynamic changes of different agricultural products affected by the price spillover effect of other products during the stable period before the shock. (b) The dynamic changes of different agricultural products affected by the price spillover effect of other products during the period of uncertainty shock. (c) The dynamic changes of different agricultural products affected by the price spillover effect of other products during the adaptation period after the shock.

Dynamic Analysis of Net Spillover Effects

Figure 4 illustrates the dynamic changes of net spillover effects for various agricultural products in different stages, which is conducive to grasping the interdependence and operating characteristics between markets. From the results shown in the figure, the following observations can be made. Firstly, except for the time point when the Russia-Ukraine conflict erupted in early 2022, rice and wheat maintained negative net spillover effects during the stable period before the outbreak of the pandemic and during the normalized prevention and control period. It indicates that during these periods, they were mainly the recipients of market price fluctuations. However, during the uncertainty shock period, both temporarily shifted to positive net spillovers with significantly increased intensity. This reflects that their strategic importance as staple foods becomes prominent when the market is subjected to major shocks, making them drivers of market price changes. Second, cucumbers and cabbage consistently maintained positive net spillovers across various stages, demonstrating their dominant influence within the agricultural market price system. Particularly during periods of uncertainty, the spillover intensity of cucumbers increased notably. This result indicates that fresh vegetable products hold a relatively strong position in price formation. Therefore, when facing external shocks, efforts should be made to ensure the stability of fresh vegetable agricultural products in circulation, sales, and other segments, thereby enhancing the overall risk resilience and robustness of the agricultural market. Third, the net spillover effects of livestock and poultry products such as pork, beef, mutton, and eggs, while fluctuating across different stages, showed relatively limited overall variation. The possible reasons are as follows: On one hand, livestock and poultry products are more elastic in demand compared to necessities, giving the market relatively stronger self-adjustment capacity. On the other hand, storage serves as a primary tool for smoothing price fluctuations (Vercammen, 2012). During the pandemic, the government effectively stabilized market supply by releasing reserved meat and other regulatory measures, which helped mitigate price volatility and enhanced the ability of these product markets to withstand external shocks from the supply side.

Figure 4.

(a) The dynamic changes of the net spillover effects of different agricultural products during the stable period before the shock. (b) The dynamic changes of the net spillover effects of different agricultural products during the period of uncertainty shock. (c) The dynamic changes of the net spillover effects of different agricultural products during the adaptation period after the shock.

Analysis of Interregional Spillover Effects

The previous section analyzed the spillover effects among agricultural product markets in China. In order to further observe the spillover effects and market interdependence among different regions, this paper regards regions as relatively independent market units, and weighted average agricultural product prices by region based on the weight of various agricultural product yields. Based on this, a spillover effect index is constructed. The specific results are shown in Table 2.

Table 2.

Spillover Effect of Market Price of Agricultural Products in Different Stages and Regions.

Stage		Inner Mongolia	Beijing	Sichuan	Tianjin	Anhui	Shandong	Shanxi
I	Inner Mongolia	28.15	4.45	5.56	4.62	2.91	13.90	9.36
	Beijing	4.06	28.67	4.42	8.86	3.68	14.42	6.05
	Sichuan	7.03	6.20	42.70	5.39	5.65	5.59	4.79
	Tianjin	2.29	9.23	2.13	29.71	2.67	12.74	7.47
	Anhui	3.27	4.77	2.38	6.96	30.86	15.16	6.96
	Shandong	3.56	7.51	1.60	9.35	3.59	32.33	7.51
	Shanxi	4.45	4.28	3.09	7.67	3.13	17.11	27.35
	Xinjiang	6.37	6.45	4.05	6.01	3.83	10.40	3.90
	Jiangsu	1.33	7.41	2.17	10.02	4.57	14.22	4.24
	Hebei	3.04	8.31	2.18	11.89	3.20	15.46	5.14
	Liaoning	2.62	7.12	3.38	11.93	4.72	10.55	3.96
	Qinghai	4.14	0.88	2.43	5.02	0.57	7.28	6.22
	To	42.16	66.59	33.38	87.72	38.53	136.83	65.59
	Net	−29.70	−4.73	−23.92	17.43	−30.61	69.17	−7.06
	Total	70.30	95.27	76.08	117.43	69.39	169.17	92.94
		Xinjiang	Jiangsu	Hebei	Liaoning	Qinghai	From
	Inner Mongolia	4.17	1.45	5.91	12.39	7.16	71.85
	Beijing	2.62	3.67	7.68	14.31	1.55	71.33
	Sichuan	1.85	3.85	4.74	6.40	5.80	57.30
	Tianjin	2.58	4.20	9.15	16.03	1.81	70.29
	Anhui	1.38	5.95	4.26	15.82	2.24	69.14
	Shandong	1.98	5.05	8.19	16.13	3.19	67.67
	Shanxi	1.01	2.59	6.49	14.02	8.80	72.65
	Xinjiang	40.52	2.77	4.93	7.18	3.60	59.48
	Jiangsu	0.92	33.00	6.63	11.37	4.10	67.00
	Hebei	2.31	4.74	23.90	16.41	3.42	76.10
	Liaoning	2.15	2.72	6.13	42.01	2.72	57.99
	Qinghai	0.18	5.81	5.97	6.42	55.10	44.90
	To	21.15	42.79	70.09	136.48	44.38	785.69
	Net	−38.33	−24.20	−6.01	78.50	−0.53	–
	Total	61.67	75.80	93.99	178.50	99.47	65.47
II	Inner Mongolia	16.23	15.78	1.81	18.96	2.10	12.11	5.12
	Beijing	14.71	16.87	2.25	21.55	2.77	12.49	4.59
	Sichuan	10.12	9.60	15.63	11.57	6.68	11.15	3.52
	Tianjin	11.94	15.28	2.21	24.38	6.96	12.36	3.98
	Anhui	11.03	14.81	1.54	14.18	14.76	13.87	7.24
	Shandong	15.53	18.82	1.79	17.67	3.11	17.09	8.17
	Shanxi	13.61	16.58	1.86	22.67	2.66	13.78	6.81
	Xinjiang	14.27	10.83	3.09	14.47	2.50	8.17	3.33
	Jiangsu	10.87	15.67	3.12	15.80	7.19	15.34	6.50
	Hebei	14.67	16.96	0.94	13.61	5.44	11.85	5.70
	Liaoning	10.09	10.66	0.27	15.67	2.20	3.92	2.32
	Qinghai	3.90	3.69	3.07	13.59	7.47	4.16	5.55
	To	130.74	148.69	21.96	179.74	49.09	119.21	56.00
	Net	46.98	65.56	−62.41	104.13	−36.16	36.30	−37.19
	Total	146.98	165.56	37.59	204.13	63.84	136.30	62.81
		Xinjiang	Jiangsu	Hebei	Liaoning	Qinghai	From
	Inner Mongolia	6.69	1.61	9.57	9.55	0.46	83.77
	Beijing	6.85	1.24	7.87	7.76	1.04	83.13
	Sichuan	6.50	6.05	12.39	3.76	3.04	84.37
	Tianjin	5.23	1.75	9.13	6.18	0.59	75.62
	Anhui	4.37	4.76	9.85	2.87	0.71	85.24
	Shandong	6.08	1.03	6.57	3.63	0.50	82.91
	Shanxi	6.01	0.90	7.36	7.06	0.69	93.19
	Xinjiang	11.71	4.74	12.76	13.28	0.86	88.29
	Jiangsu	5.97	12.13	5.04	1.06	1.32	87.87
Stage		Inner Mongolia	Beijing	Sichuan	Tianjin	Anhui	Shandong	Shanxi
	Hebei	5.60	2.16	15.22	6.88	0.96	84.78
	Liaoning	4.13	0.61	19.43	29.26	1.44	70.74
	Qinghai	3.16	2.87	6.30	13.04	33.23	66.77
	To	60.59	27.72	106.28	75.04	11.62	986.69
	Net	−27.27	−60.15	21.49	4.30	−55.15	−
	Total	72.30	39.85	121.49	104.30	44.85	82.22
III	Inner Mongolia	42.80	4.94	0.58	17.91	2.47	10.32	8.56
	Beijing	9.97	31.43	0.54	20.31	2.08	11.88	7.56
	Sichuan	1.75	0.75	58.75	10.15	2.25	3.60	8.17
	Tianjin	10.44	6.99	0.82	45.43	2.47	10.84	6.45
	Anhui	3.86	3.89	1.30	7.45	64.21	6.39	5.55
	Shandong	8.91	5.97	0.13	19.04	2.25	34.41	9.18
	Shanxi	10.41	4.81	0.60	17.71	3.37	12.49	31.62
	Xinjiang	4.47	1.42	0.98	12.64	0.65	2.37	0.80
	Jiangsu	5.31	6.33	0.84	21.36	1.59	16.30	9.66
	Hebei	8.49	6.43	1.46	22.61	1.32	13.10	5.94
	Liaoning	13.91	4.82	2.63	18.72	0.78	8.35	8.01
	Qinghai	3.95	0.87	0.42	10.12	0.89	5.58	4.86
	To	81.46	47.20	10.31	178.38	20.12	101.22	74.73
	Net	24.26	−21.37	−30.94	123.81	−15.67	35.63	6.35
	Total	124.26	78.63	69.06	223.81	84.33	135.63	106.35
		Xinjiang	Jiangsu	Hebei	Liaoning	Qinghai	From
	Inner Mongolia	1.64	1.09	3.36	5.97	0.37	57.20
	Beijing	3.60	3.62	4.13	4.24	0.63	68.57
	Sichuan	0.79	7.76	0.68	3.36	1.97	41.25
	Tianjin	2.37	3.88	4.87	4.99	0.44	54.57
	Anhui	0.49	3.51	2.02	1.22	0.11	35.79
	Shandong	2.19	6.30	8.25	2.71	0.30	65.59
	Shanxi	1.41	4.93	5.04	7.04	0.58	68.38
	Xinjiang	61.66	1.02	8.92	4.89	0.19	38.34
	Jiangsu	1.26	30.42	5.09	1.32	0.54	69.58
	Hebei	3.86	2.33	27.14	5.90	1.42	72.86
	Liaoning	1.66	0.52	6.41	33.95	0.24	66.05
	Qinghai	1.57	3.52	4.21	2.28	61.75	38.25
	To	20.83	38.47	52.97	43.92	6.80	676.43
	Net	−17.51	−31.11	−19.89	−22.13	−31.44	−
	Total	82.49	68.89	80.11	77.87	68.56	56.37

In terms of the overall spillover effect, the spillover index of agricultural product prices in the entire market shows a trend of first increasing and then decreasing, and all exceed 50%, indicating a high degree of correlation between agricultural product markets among regions, and that more than half of the fluctuations are caused by price fluctuations in other regions. This reflects, to a certain extent, the level of regional integration of China's agricultural product market. Specifically, during the stable phase before the shock, the total spillover index stood at 65.47%. Following the uncertainty shock, it rose to 82.22%, indicating a further intensification of interregional price transmission under extreme external conditions. During the subsequent adaptation phase, the index declined to 56.37% but remained higher than the pre-shock level. Rather than leading to fragmentation between regions, the agricultural market demonstrated enhanced integration in response to the uncertainty shock. This phenomenon may be closely linked to the macro-control policies implemented by the Chinese government during the pandemic. To ensure the livelihood needs of residents, the government enacted extensive regulatory measures covering wholesale procurement, logistics, transportation, and distribution of agricultural products. These initiatives strengthened interregional market connections and, to some extent, counteracted the potential market segmentation effects that the pandemic itself might have caused. Precisely because of this, during the adaptation phase, as the government phased out emergency policy interventions in agricultural markets, the actual impact of uncertainty factors on the market became more apparent. The spillover level between regions consequently declined, even falling below the level observed during the stable period. Therefore, the potential disruption of external shocks to regional market coordination still warrants attention. In advancing the development of a nationally unified market, it is crucial to monitor market stability after the phasing out of short-term emergency policies and guard against systemic risks that may arise from weakened regional linkages. This aligns with the emphasis in the study by Liu and Tang (2023).

From the perspective of inter-regional agricultural market price spillover effects, the market price spillover effects between different regions are closely related to geographical location, and the spillover effects on adjacent regions are higher than on non-adjacent regions. In particular, the market connections between Beijing municipality, Tianjin municipality, Hebei province, Liaoning province, Shandong province, and Jiangsu province are particularly close, and the degree of market integration is relatively high. From the perspective of net spillover effects in each region, there has been a role change in Inner Mongolia Autonomous Region and Liaoning Province before and after the outbreak. Inner Mongolia has changed from a receiver of information to a driver, while Liaoning has changed from an information provider to a receiver. This change might be because Inner Mongolia is an important production base for livestock and poultry products. Under the impact of uncertainties, its position in the supply chain and market influence have been strengthened, thereby enhancing its price influence over other regions. The net spillover index of Beijing and Hebei was negative in the pre-outbreak stabilization period and the normal prevention and control period, and positive in the period of strict epidemic control. This might be due to the fact that the regulatory policies implemented by the government during the period of uncertainty shock (such as breaking through logistics bottlenecks and strengthening regional collaborative supply guarantee, etc.) significantly enhanced the interaction between the above-mentioned regions and the surrounding markets, making them temporarily become the regional diffusion centers of price fluctuations during the shock period.

Robustness Test

To verify the accuracy of the findings in this study, the forecast horizons were set to 30 days, 90 days, and 180 days (approximately 1 month, 3 months, and 6 months, respectively), and changes in the spillover indices under these settings were compared. Table 3 reports the robustness test results. Both the total spillover index and the directional spillover indices of various agricultural products exhibit a high degree of stability across different forecast horizons. Moreover, under all forecast horizons, the spillover indices demonstrate a “low-high-low” dynamic pattern across the three phases of pandemic development. Neither the numerical values nor the trends show substantial differences compared to the baseline regression results. These findings indicate that the core conclusion of this study—regarding the dynamic patterns of volatility spillovers across agricultural markets—remains unchanged even as the model’s forecast horizon is extended, thereby strongly confirming the reliability and robustness of the baseline regression results.

Table 3.

Results of the Robustness Test.

		The prediction period is 30 days			The prediction period is 90 days			The prediction period is 180 days
	Directional spillover	I	II	III	I	II	III	I	II	III
Rice	From	9.50	60.45	16.32	9.50	60.45	16.32	9.50	60.46	16.32
Wheat	To	8.10	36.47	13.06	8.10	36.49	13.06	8.10	36.49	13.06
Potatoes	From	11.54	63.34	10.10	11.54	63.35	10.10	11.54	63.35	10.10
Cabbage	To	13.27	53.07	11.94	13.27	53.12	11.94	13.27	53.12	11.94
Cucumber	From	28.55	49.60	24.38	28.55	49.62	24.38	28.55	49.62	24.38
Pork	To	26.59	59.08	28.53	26.59	59.11	28.53	26.59	59.11	28.53
Beef	From	9.00	65.08	13.43	9.00	65.09	13.43	9.00	65.09	13.43
Mutton	To	12.45	77.09	14.42	12.45	77.10	14.42	12.45	77.10	14.42
Eggs	From	6.13	54.90	17.05	6.13	54.92	17.05	6.13	54.92	17.05
Rice	To	6.90	63.37	20.74	6.90	63.36	20.74	6.90	63.36	20.74
Wheat	From	30.82	63.95	14.31	30.82	63.96	14.31	30.82	63.96	14.31
Potatoes	To	28.84	63.20	12.89	28.84	63.21	12.89	28.84	63.21	12.89
Cabbage	From	18.40	52.49	10.63	18.40	52.51	10.63	18.40	52.51	10.63
Cucumber	To	22.02	70.64	10.37	22.02	70.62	10.37	22.02	70.62	10.37
Pork	From	14.70	44.03	19.06	14.70	44.04	19.06	14.70	44.04	19.06
Beef	To	13.74	35.98	12.80	13.74	35.99	12.80	13.74	35.99	12.80
Mutton	From	8.89	54.62	11.00	8.89	54.62	11.00	8.89	54.62	11.00
	To	5.61	49.55	11.55	5.61	49.57	11.55	5.61	49.57	11.55
Total		15.28	56.49	15.14	15.28	56.51	15.14	15.28	56.51	15.14

Conclusion and Discussion

To explore the changes in spillover effects of agricultural market prices under uncertainty shocks, this paper takes the different stages of the COVID-19 epidemic as the real scenario. Using daily wholesale price data of 9 types of agricultural products from 12 provinces from January 1, 2018, to August 31, 2022, the paper employs the spillover effect index to analyze the spillover effects and time-varying characteristics of agricultural product market prices across products and regions. Research findings indicate that although uncertainty shocks have significantly enhanced the volatility spillover effect of the agricultural product market in the short term, the agricultural product market has strong resistance and self-adaptive capabilities in the face of external shocks and has not been irreversibly affected.

We also found that the impact of uncertainty not only intensified the price fluctuations in the entire agricultural product market but also significantly enhanced the price linkage effect among products. The impact of uncertain shocks on the agricultural product market mainly focuses on fresh vegetable products with short storage time, such as cucumber and Chinese cabbage, as well as major grain crops and protein source products such as wheat and eggs. Furthermore, from a dynamic perspective, the price spillover effects of livestock and poultry products such as pork, beef, mutton, and eggs are highly unstable, exhibiting significant variations across different periods and stages. Therefore, for market regulators, it is crucial to pay closer attention to agricultural products that dominate the price system when addressing uncertainty shocks, while fully considering the price transmission mechanisms and interconnections among different products and regions. When price fluctuations of a single product become particularly volatile, close monitoring of price changes in closely related products is essential. Simultaneously, the government should establish a risk-based, categorized reserve system, with policy resources tilted toward perishable agricultural products and staple foods, to stabilize supply expectations for key categories.

Furthermore, we have extended our analysis to examine the spatial linkages and interdependence of agricultural markets from a regional dimension. In terms of overall spillover effects, the price spillover index among regional agricultural markets shows a trend of initial increase followed by a decline. Market fluctuations across regions exhibit strong interconnectedness, with more than half of the price volatility originating from interregional spillover effects. In response to sudden uncertainty shocks, the government should implement well-designed macro-control strategies, including interventions in logistics, transportation, wholesale procurement, and market supply. These measures can help leverage the substitutability and complementarity among products and regional markets, thereby containing risk fluctuations within manageable levels. At the same time, the government should remain vigilant about market stability after the withdrawal of short-term emergency policies, guarding against potential systemic risks that may arise from weakened interregional linkages.

It should be noted that this study provides a preliminary exploration of the changing mechanisms of spillover effects in agricultural markets under uncertainty shocks. While the findings possess a certain degree of representativeness and general applicability, this research still has some limitations. First, in terms of the selection of agricultural product categories, this study focuses more on essential staple foods and vegetables for residents. Due to data availability constraints, it does not cover several important categories, such as fruits, nor does it conduct specialized selections within each category. Future research could expand the data scope by incorporating more diversified agricultural product categories to enhance the comprehensiveness of the system. Second, to concentrate on the core impact of the pandemic shock itself, the selected timeframe covers the stable period before the outbreak, the phase of intense volatility during the initial outbreak, and the normalized prevention and control phase, but does not extend to the later phase of pandemic policy adjustments. While this design helps isolate the confounding effects of other policies, it makes it difficult to fully capture the long-term impact of uncertainty shocks on agricultural markets. Subsequent studies could extend the observation window, particularly by including policy transition phases, to more systematically assess the persistent effects of the shocks. Finally, the spillover index adopted in this study provides an effective benchmark for understanding the average linkage structure of agricultural markets during the pandemic period. Future research could employ more advanced dynamic econometric methods, such as TVP-VAR, to capture the dynamic evolution of market linkage structures and deepen the understanding of the transmission mechanisms in agricultural markets under uncertainty shocks.

Footnotes

ORCID iD

Yangqi Fu

Ethical Considerations

This article does not contain any studies with human or animal participants.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We gratefully acknowledge the financial support from the Jiangsu Provincial Social Science Foundation Youth Project (25GLC029) and the Yangzhou “Lüyang Jinfeng” Program for Outstanding Doctors.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

I have research data associated with this article, but I have no right to disseminate it. Those in need can download it from the official website mentioned in the text by themselves.

References

Aday

M. S.

(2020). Impact of COVID-19 on the food supply chain. Food Quality and Safety, 4(4), 167–180.

Adenäuer

Frezal

Chatzopoulos

(2025). Weathering the storm: Trade openness and the resilience of agricultural markets under increasing weather extremes. Journal of Agricultural Economics, 76, 241–250.

Al Mamun

T. G. M.

Ullah

Hassan

M. S.

Yusof

M. F.

Islam

M. A.

Rab

N. B.

Rahman

M. A.

(2025). Fiscal policy and household savings in Central Europe: A Markov switching VAR with COVID shock. Cogent Economics & Finance, 13(1), 2537233.

Apergis

Rezitis

(2003). Agricultural price volatility spillover effects: The case of Greece. European Review of Agricultural Economics, 30(3), 389–406.

Balcilar

Sertoglu

Agan

(2022). The COVID-19 effects on agricultural commodity markets. Agrekon, 61(3), 239–265.

Balli

Naeem

M. A.

Shahzad

S. J. H.

de Bruin

(2019). Spillover network of commodity uncertainties. Energy Economics, 81, 914–927.

Bittmann

Loy

J. P.

Anders

(2020). Product differentiation and cost pass-through: Industry-wide versus firm-specific cost shocks. Australian Journal of Agricultural and Resource Economics, 64(4), 1184–1209.

Buguk

Hudson

Hanson

(2003). Price volatility spillover in agricultural markets: An examination of US catfish markets. Journal of Agricultural and Resource Economics, 28(1), 86–99.

Cardoso

Malloy

(2021). Impact of the first wave of the COVID-19 pandemic on trade between Canada and the United States. Canadian Public Policy / Analyse de Politiques, 47(4), 554–572.

10.

Cheng

Z. J.

Zhan

Xue

Zheng

Lyu

Zhang

X. D.

Luo

Huang

Zhang

Wang

Zhong

Sun

(2023). Public health measures and the control of COVID-19 in China. Clinical Reviews in Allergy & Immunology, 64(1, SI), 1–16.

11.

Deaton

Laroque

(1992). On the behavior of commodity prices. Review of Economic Studies, 59, 1–23.

12.

Diebold

F. X.

Yilmaz

(2012). Better to give than to receive: Predictive directional measurement of volatility spillovers. International Journal of Forecasting, 28(1), 57–66.

13.

Elleby

Domínguez

I. P.

Adenauer

Genovese

(2020). Impacts of the COVID-19 pandemic on the global agricultural markets. Environmental and Resource Economics, 76(4), 1067–1079.

14.

Ghoshray

(2019). Are shocks transitory or permanent? An inquiry into agricultural commodity prices. Journal of Agricultural Economics, 70(1), 26–43.

15.

Guo

Zhang

Zhao

(2023). How are climate risk shocks connected to agricultural markets? Mathematical Problems in Engineering, 2023, 100367.

16.

Hadachek

Sexton

R. J.

(2024). Market structure and resilience of food supply chains under extreme events. American Journal of Agricultural Economics, 106(1), 21–44.

17.

Jin

Kim

(2025). Economic freedom and market resilience: Safeguarding liquidity in times of crisis. Journal of Multinational Financial Management, 79, 100918.

18.

Jung

(2022). A study on changes in the food service industry about keyword before and after COVID-19 using big data. The International Journal of Internet, Broadcasting and Communication, 14(3), 85–90.

19.

Kakran

Bajaj

P. K.

Pandey

D. K.

Kumar

(2025). Interconnectedness and return spillover among APEC currency exchange rates: A time-frequency analysis. Research in International Business and Finance, 73, 102572.

20.

Kalli

Karali

Liu

Gopinath

(2025). Return and volatility spillovers among major cotton markets. Agribusiness. https://doi.org/10.1002/agr.70008

21.

Kang

S. H.

McIver

Yoon

(2017). Dynamic spillover effects among crude oil, precious metal, and agricultural commodity futures markets. Energy Economics, 62, 19–32.

22.

Chen

Huang

(2023). How do logistics disruptions affect rural households? Evidence from COVID-19 in China. Sustainability, 15(1), 465.

23.

Lindsey

O’Brien

Clark

(2025). Perceived effects of Brexit and the COVID-19 epidemic on the UK lamb value chain: A qualitative interview-based study. British Food Journal, 127(4), 1287–1304.

24.

Liu

Pan

Yin

(2020). Pandemic, mobile payment, and household consumption: Micro-evidence from China. Emerging Markets Finance and Trade, 56(10), 2378–2389.

25.

Liu

Tang

(2023). Research on the impact of emergencies on the poultry market integration in China. China Agricultural Economic Review, 15(1), 197–213.

26.

Loy

J. P.

Holm

Steinhagen

Glauben

(2015). Cost pass-through in differentiated product markets: A disaggregated study for milk and butter. European Review of Agricultural Economics, 42(3), 441–471.

27.

Mishra

Golaviya

Panchal

Hinsu

Yadav

Fournié

Barnett

Koringa

Paleja

H. U. I.

Dasgupta

(2023). Exploring the impact of COVID-19 and the associated lockdown on the production, distribution, and consumption of poultry products in Gujarat, India: A qualitative study. Poultry, 2, 395–410.

28.

Mokni

Al-Shboul

Assaf

(2021). Economic policy uncertainty and dynamic spillover among precious metals under market conditions: Does COVID-19 have any effects? Resources Policy, 74, 102238.

29.

Nguyen

D. T.

(2025). Geopolitical risks and crypto exchange rate premium: Implications for capital flight. Finance Research Letters, 85, 108287.

30.

Nguyen

T.-T.

Pham

C. M.

Thai

V. V.

Tan

J. Y.

Pham

H.-V.

Thi

T. H. T.

(2025). How has the aquaculture supply chain’s competitiveness changed after the COVID-19 pandemic in emerging countries? The case of Vietnam. Sustainability, 17(4), 1451.

31.

Chen

Zhong

(2021). Overstocked agricultural produce and emergency supply system in the COVID-19 pandemic: Responses from China. Foods, 10(12), 3027.

32.

Ramsey

A. F.

Goodwin

B. K.

Hahn

W. F.

Holt

M. T.

(2021). Impacts of COVID-19 and price transmission in U.S. meat markets. Agricultural Economics, 52(3), 441–458.

33.

C. W.

Luo

Qin

(2025). Does internet concern act as a catalyst for crude oil price shifts? Energy Sources, Part B: Economics, Planning and Policy, 20(1), 2540094.

34.

Szczygielski

J. J.

Brzeszczyński

Charteris

Bwanya

P. R.

(2022). The COVID-19 storm and the energy sector: The impact and role of uncertainty. Energy Economics, 109, 105258.

35.

Szenderak

Popp

Harangi-Rákos

(2019). Price and volatility spillovers of the producer price of milk between some EU member states. German Journal of Agricultural Economics, 68(2), 61–76.

36.

Vercammen

(2012). Agricultural marketing: Structural models for price analysis (pp. 35–59). Routledge.

37.

Wan

(2022). Asymmetric price volatility transmission in agricultural supply chains: Evidence from the Chinese pork market. Mathematical Problems in Engineering, 2022, 4801898.

38.

Williams

J. C.

Wright

B. D.

(1991). Storage and commodity markets. Cambridge University Press.

39.

Zhu

Gao

Xing

(2025). The price of realized extreme climate events in the implied cost of equity capital: International evidence. Journal of Banking & Finance, 180, 107525.

40.

Yan

Cai

Lin

Amhaw

D. T.

(2021). The impacts of trade restrictions on world agricultural price volatility during the COVID-19 pandemic. China & World Economy, 29(6), 139–158.

41.

Liu

Wang

Feil

(2020). The impact of COVID-19 on food prices in China: Evidence of four major food products from Beijing, Shandong and Hubei provinces. China Agricultural Economic Review, 12(3), 445–458.

42.

Zhang

Wang

Yuan

Liu

Gong

(2020). The impact of epidemics on agricultural production and forecast of COVID-19. China Agricultural Economic Review, 12(3), 409–425.

43.

Zhou

Zhang

(2021). How does news flow affect cross-market volatility spillovers? Evidence from China’s stock index futures and spot markets. International Review of Economics & Finance, 73, 196–213.

44.

Zhu

Y. L.

R. H.

Liu

Y. H.

(2025). How connected is the multilayer risk spillover network between worldwide climate policy uncertainty and the Chinese stock market? Journal of Environmental Management, 393, 127141.

The Impact of Uncertainty Shocks on Agricultural Price Spillover Effects

Abstract

Keywords

Introduction

Model Specification and Data Description

Model Specification

Data Description

Empirical Result Analysis

Analysis of Overall Spillover Effects

Analysis of Dynamic Spillover Effects

Dynamic Change Analysis of Total Spillover Effect

Dynamic Analysis of Directional Spillover Effects

Dynamic Analysis of Net Spillover Effects

Analysis of Interregional Spillover Effects

Robustness Test

Conclusion and Discussion

Footnotes

ORCID iD

Ethical Considerations

Funding

Declaration of Conflicting Interests

Data Availability Statement

References