COVID-19, supply chain disruption and China’s hog market: a dynamic analysis

Yubin Wang (College of Economics and Management, China Agricultural University, Beijing, China)
Jingjing Wang (Department of Economics, University of New Mexico, Albuquerque, New Mexico, USA)
Xiaoyang Wang (Department of Economics, University of New Mexico, Albuquerque, New Mexico, USA)

China Agricultural Economic Review

ISSN: 1756-137X

Article publication date: 26 June 2020

Issue publication date: 20 August 2020

4634

Abstract

Purpose

The authors explicitly evaluate the dynamic impact of five most concerned supply chain disruption scenarios, including: (1) a short-term shortage and price jump of corn supply in hog farms; (2) a shortage of market hogs to packing facilities; (3) disruption in breeding stock adjustments; (4) disruption in pork import; and (5) a combination of scenario (1)–(4).

Design/methodology/approach

The agricultural supply chain experienced tremendous disruptions from the COVID-19 pandemic. To evaluate the impact of disruptions, the authors employ a system dynamics model of hog market to simulate and project the impact of COVID-19 on China hog production and pork consumption. In the model the authors explicitly characterize the cyclical pattern of hog market. The hog cycle model is calibrated using market data from 2018–2019 to represent the market situation during an ongoing African swine fever.

Findings

The authors find that the impacts of supply chain disruption are generally short-lived. Market hog transportation disruption has immediate impact on price and consumption. But the impact is smoothed out in six months. Delay in import shipment temporarily reduces consumption and raises hog price. A temporary increase of corn price or delay in breeding stock acquisition does not produce significant impact on national hog market as a whole, despite mass media coverage on certain severely affected regions.

Originality/value

This is the first evaluation of short-term supply chain disruption on China hog market from COVID-19. The authors employ a system dynamics model of hog markets with an international trade component. The model allows for monthly time step analysis and projection of the COVID-19 impact over a five-year period. The results and discussion have far-reaching implications for agricultural markets around the world.

Keywords

Citation

Wang, Y., Wang, J. and Wang, X. (2020), "COVID-19, supply chain disruption and China’s hog market: a dynamic analysis", China Agricultural Economic Review, Vol. 12 No. 3, pp. 427-443. https://doi.org/10.1108/CAER-04-2020-0053

Publisher

:

Emerald Publishing Limited

Copyright © 2020, Emerald Publishing Limited


1. Introduction

The pandemic coronavirus disease (COVID-19) has severely impacted Chinese economy. Started in December 2019, COVID-19 quickly became a national epidemic and peaked in February 2020. To curb its spread, the Chinese government has implemented strict regulations on social and economic activities, such as lockdown, regional traffic and travel restrictions. These policies have effectively curbed the spread of virus and brought the epidemic under control. However, these measures also impose costs on the economy with the idling production and transportation activity. With the impediments in transportation, a major concern arises in agricultural livestock industry. Due to the continuous nature of livestock markets, when production and consumption suffer temporary disruption in an early stage, the resulting impacts on commodity supply can aggravate over time and cause greater impact in the future.

Among agricultural products, the hog market can suffer from short-term transportation disruptions from COVID-19 because of two primary reasons. First, disruptions in hog production supply chain can lead to long-lasting market impacts. To curb the spread of virus, local authorities have set road obstacles to block hog transportation. The sudden shortage of hogs for packing facilities can result in surging hog prices and affect hog farms' ability to acquire piglets for future supply. The supply-side impact may not manifest until a few months later when piglets reach market for slaughter. Second, China hog production is still recovering from the African swine fever (ASF) that broke out in 2018. Production in 2020 is struggling at the lowest historical level. The recent outbreak of COVID-19 poses additional uncertainty to hog production recovery. The concern is particularly imminent under the background of an ongoing trade war that threatens the sourcing of international pork supply.

Because pork is the staple animal protein for Chinese households and amid the emerging concerns on food security, there is an urgent need to understand and evaluate the potential impact of COVID-19 on future hog market dynamics. However, aside from experiential judgments, there are limited tools available to model the dynamics of the complete hog market cycle and evaluate impacts from shocks at the different stages of production supply chain. Previous research has focused on estimating the isolated relationships at certain specific stages of the hog cycle, for example, farmers' production response to profitability or the price elasticity of pork consumption (Tan, 2010; Wu, 2011; Sun et al., 2014; Song, 2016). For a comprehensive assessment of the impact of COVID-19 on China hog market, an integrated model is needed to synthesize the isolated pieces together and incorporate the dynamic feedback loops and time lags featured in the hog market cycle.

In this article we numerically evaluate the impact of short-term supply chain disruptions from COVID-19 on China hog market. We use a system dynamics commodity cycle model developed by Wang and Wang (2020) and calibrate using 2018–2019 data. The system dynamics commodity market model was first introduced by Meadows (1970) to explain the cyclical patterns in commodities. The system dynamics framework has been employed to analyze the dynamics in various agricultural commodity markets, for example, Nicholson and Stephenson (2015) on US dairy market, Nicholson and Parsons (2012) on Mexico sheep market and Shamsuddoha et al. (2013) on poultry market. Wang and Wang (2020) developed a seasonal commodity cycle model to evaluate the long-term structural impact on hog market from COVID-19 shocks. The method we use explicitly characterizes the two prominent features of hog market, which are (1) the time lags that arise from biological growth cycle, time required for capacity expansion and price expectation formation process and (2) the feedback loops among price, consumption and production decisions. Other closely related models for commodity markets study only the supply responses of hog production (Chavas et al., 1985; Holt and Johnson, 1988; Chavas and Holt, 1991; Sun et al., 2014; Song, 2016) and the partial equilibrium system of structural equations used for long-term market projection and baseline evaluations (Hu et al., 2015). Compared to these models, the system dynamic model incorporates both supply and demand sides of the market and is flexible to capture the complex interactions across different stages of hog market. On the other hand, the system of structural equations model considers both supply and demand, but is mainly used for long-term annual projections, for example, the FAPRI baseline analysis (FAPRI, 2019). Our system dynamic model is capable of simulating short- to intermediate-term market dynamics, for example, monthly outcomes, which is important for a rapid and timely assessment of the COVID-19 impact through supply chain disruptions.

Specifically, we evaluate the dynamic impact of five supply chain disruption scenarios that are concerned by stakeholders. All shock scenarios are from transportation disruptions. First, disruption in transportation can lead hog farms to face a short-term shortage and price jump of corn, which is the main ingredient in hog feed. The shortage of feedstuff for some farms has appeared on news coverage and caused wide sensational concerns. Second, disruption in transportation can lead to a shortage of market hogs to packing facilities. The temporary unavailability of hog supply can result in pork supply shortage and a series of chain reactions in future supply and demand. Third, breeding stock replacement can suffer from disruption in transportation and results in difficulties at breeding stock adjustments. Fourth, there can be a delay in pork import arrival. As WHO declares global health emergency over the COVID-19, international freight to China has been disrupted with delays and canceling port calls [1]. Subsequent global spread of COVID-19 further stranded shipment in major European and North America economies, from where China sources the majority of pork import. Finally, we further create a fifth scenario of combining all four individual shocks because they are likely to happen together.

We report the following main findings. First, the supply chain disruption impacts are mainly transitory. As transportation restores after the COVID-19 is under control, hog market rebalances to the normal cycle pattern. Second, among the four individual scenarios, the disruption in market hog transportation has the biggest impact on hog price and pork consumption. Hog price immediately jumps in response to the shortage of market hog supply. But as packing facilities reopen and ramp up production, the impact is smoothed out in approximately six months, leaving little long-term aftermath. Third, delays in pork import temporarily reduce pork consumption and raise hog price. In the long-term, market reacts by increasing breeding capacity and the impact gradually dissipates. The increased breeding capacity brings down hog price over time and leads to slightly higher consumption two years after the epidemic. Finally, a short-term jump in the price of corn or delays in breeding stock acquisition do not produce significant marketwise impact despite the sensations in media coverage.

The rest of this article is organized as follows. In section 2 we present the model setup. Section 3 describes the data used for model calibration and parameter values. Section 4 shows the simulation results for different scenarios. Section 5 concludes the findings.

2. Model structure

The dynamic commodity cycle model consists of four modules that represent hog production, pork inventory, international trade and pork consumption [2]. We describe each module as follows.

2.1 Hog production

The module is described as in Eqs (1)-(10).

(1)EP=PtdtEPtdtADP

Eqn (1) specifies expected hog price EPt following an adaptive expectation process in differential equation form (Nerlove, 1958). The change in EPt is the difference between previous actual price Ptdt and expected price EPtdt, adjusted by the weighting factor ADP as the time delay of expectation adjustment.

(2)EHCRt=EPt/ECPt

In Eqn (2), expected hog–corn ratio EHCRt, a widely used indicator of hog production profitability, is defined as the ratio of EPt over expected corn price ECPt.

(3)dDBStdECHRt=ϕ,ϕ>0

In Eqn (3), the desired level of breeding herd DBSt is defined as an increasing function of EHCRt. To achieve the desired level of breeding stock, farmers continuously make the decision of how many new gilts should be kept for breeding versus for fattening. The breeding herd adjustment process is defined as the breeding stock acquisition BSARdt during dt.

(4)BSARdt=DBStBStADBS

Eqn (4) describes the process using similar adaptive process weighted by adjustment delay ADBS, recognizing that breeding herd adjustment is a dynamic process rather than a one-step shift.

(5)BS=BSAROSSR
(6)OSSRdt=BSt/spl

BSt is the actual breeding stock laid out in Eqs (5) and (6) with OSSRdt  being the old sow slaughter volume during dt, and spl the average sow productive life.

(7)BRdt=BStlpsmpsl

The supply of piglets (BRdt) is determined in Eqn (7) as the product of available breeding stock, litters per sow per month (lpsm) and pigs saved per litter (psl). Newly bred piglets are not immediately available for slaughter but delayed by the biological gestation–maturation process (gm).

(8)MRdt=BRdtsr(3gm)3t2e3tgm2

In Eqn (8) the matured hog MRdt is the product of piglet supply and survival rate (sr) weighted by the third-order delay function.

(9)MS=MRMSSRBSAR

Matured hogs build into the marketable stock at each period MSt, which is specified in Eqn (9) as maturing hogs subtract the monthly slaughter volume (MSSRdt) and breeding herd acquisition.

(10)MSSRdt=MSt1msfp

Finally MSSRdt is given in Eqn (10) and (msfp) is the matured hog feeding period.

2.2 Pork demand

The pork demand module simulates the consumer pork demand response to pork price, as in Eqs (11)-(13).

(11)RPt=MM+Pty

Eqn (11) specifies the retail pork price RPt as the live hog price divided by dress yield (y) plus a meat packing industry market margin (MM).

(12)d(ln(PCCPt))d(ln(RPt))=ε,ε<0

Eqn (12) defines per capita pork demand PCCPt as a function of retail pork price governed by its own price elasticity ε.

(13)Ct=popt(1+dtrpop)PCCPt

The aggregate pork demand Ct is then defined in Eqn (13) as depending on population popt, population natural growth rate rpop and per capita pork demand.

2.3 International trade

Net import of pork is another source of market supply in addition to domestic production. Eqn (14) defines net import NIt as an increasing function of domestic live hog price, with an exponentially weighted function for a delay in import arrival (im). This delay captures the time needed for booking purchase orders, scheduling production in packing plant and shipping freights.

(14)dNItdPt=ρ1imexp(tim),ρ>0

2.4 Pork inventory

The pork inventory module connects the production, demand and trade modules and closes the feedback loop. When pork inventory is above its desired level, hog price adjusts lower, and vice versa.

(15)I˙=(MSSR+OSSR)wy+NIC

We denote pork inventory by I, for which pork volume from slaughter and net import are inflows and pork demand is an outflow. In Eqn (15), I˙ is the change of pork inventory and w denotes average live hog weight. A negative value of pork inventory indicates a demand gap.

(16)COVt=ItEC

Eqn (16) defines pork inventory coverage COVt, which measures the relative abundance of pork supply, as the ratio of pork inventory over expected pork demand rate EC.

(17)RCOVt=COVtDCOV

Eqn (17) defines relative pork inventory coverage RCOVt, where DCOV denotes the desired level of pork inventory coverage.

(18)dPtdRCOVt=γ,γ<0

Thus Eqn (18) specifies live hog price to be inversely related to relative pork inventory coverage.

3. Model parameters

To reflect the current situation of an ongoing ASF disease impact, the model is calibrated using China hog market data from 2018–2019. All parameters values are presented in Table 1 and consistent with Wang and Wang (2020). We explain the source for each parameter value as follows.

The following parameters are directly obtained from literature or industry reports. For the litter per sow per month (lpsm), it is set at 0.18 because a sow can farrow on average 2.2 times per year, with each farrowing takes 3.8 months and two months of nursing and nonproductive period. Number of pigs saved per litter (psl) is set at 9 following Huatai Securities (2018). The average productive life of sow (spl) is 36 months or seven breeds, after which sows' productivity falls below new gilts so are liquidated for slaughter. [3] The total gestation–maturation delay (gm) is eight months for piglets to grow to market stage. The finishing survival rate (sr) is set to be 0.9 following Huatai Securities (2018). The matured hog feeding period (msfp) is set as two months because further feeding reduces weight gain efficiency. Pork import delay (im) is set to be two months to be consistent with reported industry average. [4] The adjustment delays of breeding stock (ADBS) and hog price expectation (ADP) are both set to be four months following Lan et al. (2019), which suggests that it takes farmers four months to adjust their expectations of hog price and sow inventory. Empirical estimates of the price elasticity of pork demand (ε) in China range from about −1.3 to 1.2 (Chen et al., 2015). We adopt the elasticity value of −0.34 from Zhuang and Abbott (2007), which is in the middle of the range. Population and growth rate are obtained from the 2019 National Statistics Bulletin.

The rest of the parameters are estimated using the following data. We collect China hog market data from January 2018 to December 2019. We mainly rely on official data because of its consistency and objectiveness. We collect monthly national average live hog price, pork price, corn price, slaughter volume of inspected packing plants from China Ministry of Agricultural and Rural Affairs (MARA). Because the slaughter volume of inspected packing plants typically accounts for only a third of national total, we also collect the annual total marketed hogs and pork production from China Bureau of Statistics Annual Bulletins. Monthly pork import is obtained from China Customs Monthly Bulletin. Figure 1 plots the monthly pork, hog, corn prices, inspected slaughter volume and pork import. Panel (a) shows that live hog price increased to above 20 RMB/kg in August 2019 and soared to beyond 35 RMB/kg in two months. Pork price moves in a consistent pattern with live hog price. Panel (b) shows that corn price was relatively stable around 2 RMB/kg throughout the period, benefiting from strong government intervention policies. The corn price variation is within 2% of sample average. The expected corn price (ECP) is set at 2 RMB/kg following sample average. Panel (c) shows the inspected slaughter volume with continued decline in 2019 due to the hog inventory loss from ASF. Panel (d) shows the pork imports surged to a historical high of 0.375m metric tons in December 2019, driven by high hog price and by an approximately two-month lag.

The live hog price response to relative pork inventory coverage (γ) is estimated in the following procedures. First, we estimate monthly pork production from reported annual production data, using monthly share of pork production derived from monthly inspected slaughter volume. The long-term expected pork consumption (EC) is calculated as the monthly average pork production of 2018–2019. Second, we calculate monthly pork demand using monthly pork price and price elasticity of pork demand. Third, we estimate monthly pork inventory (I) as cumulative difference between monthly production and demand, which can be a positive value representing actual inventory or a negative value indicating a market shortage. Fourth, we calculate relative pork inventory coverage (RCOVt) following Eqs (16)-(17). Lastly, we linearly regress live hog price with respect to relative pork inventory coverage to estimate the coefficient γ in Eqn (18).

The desired level of breeding inventory response to hog–corn ratio (ϕ) is estimated through the following procedures. First, we estimate breeding inventory because it is not directly reported in official portal. MARA conducts a monthly survey on 400 major hog producing counties and only publishes the percentage changes. However, Yu and Abler (2014) argue that there is overreporting in the survey data. We calculate the inferred sow inventory from monthly pork production using Eqs (6)-(8). Second, we estimate the long-term desired level of breeding stock as the sample average monthly breeding inventory. The difference between the short-term desired level and the long-term one depends linearly on hog–corn ratio. Given the data on hog–corn ratio and the calculated long-term desired level of breeding inventory, we estimate the short-term desired level of breeding inventory via ordinary least squares (OLS) regression. Lastly, we use OLS to estimate the coefficient ϕ in Eqn (3), given the estimated short-term desired level of breeding inventory and hog–corn ratio.

The dress yield (y) and market margin (MM) are estimated via OLS using the monthly data on live hog price and pork price according to Eqn (11). Average hog weight (w) is then calculated as annual pork production divided by annual hog slaughter volume and by dress yield. In the international trade module, we linearly regress net pork import with respect to live hog price with a two-month delay to estimate the coefficient ρ in Eqn (14).

4. Simulation and results

We simulate China hog market dynamics for five years. The start point of simulation is August 2019 because it is the time that hog market started to sharply increase to record level and the concern of food security emerged. The end of simulation projection is July 2024. We present the simulation results in three steps. First, we show the baseline results. Next, we introduce the four supply chain disruptions and assess their dynamic impact. Finally, we present the combined impact of multiple disruptions. Out of all defined variables, we select to show the five variables that can best exemplify the impact on hog market, which are the live hog price, pork inventory (or demand gap), breeding stock, consumption demand and net import.

4.1 Baseline

The baseline scenario shows the current market situation of tight hog supply caused by ASF without COVID-19 outbreak. In Figure 2 the solid line denotes the baseline results. The first hog price peak is projected to arrive in the summer of 2020 around August to the high of 38.6 RMB/kg. The price increase arises from a decline in pork production. On the other hand, breeding stock adjustment lags behind. Compared to the projected peak price and supply gap that will arrive in Q3 2020, breeding stock continues to build up until April 2021, which lags behind by 6–7 months. Consumption demand tanked following the price surge but benefits from an increased pork import. Net pork import increases with hog price but lags by two months, which reaches a peak import level in November 2020 at 0.39m MT. Following the expansion of breeding capacity, the first hog price trough is projected to arrive in March 2022 at the low of 17.9 RMB/kg. Pork supply/gap will recover to almost breakeven level following the rebuilding of breeding stock. However, due to the productivity and capacity loss from ASF, pork supply will remain tight in the current hog cycle and limit the downside room for hog price. The trough in breeding stock will appear seven months later in October 2020. Following the decline of hog price, pork consumption sees a recovery to a high of 4.9m MT while net import drops its lowest volume of 0.17m MT. After market reaches the trough, market dynamics will move toward the next round of uptrend, with a peak hog price of 36.9 RMB/kg scheduled to arrive in early 2024.

The baseline results demonstrate the model can successfully capture the cyclical pattern of hog market. Simulation results exhibit a period length of 40 months or 3.3 years. This is in line with the empirically observed hog cycle of 3–4 years gap between hog price peaks or troughs (Mao and Zeng, 2008). In the current parameter setup, the cycle amplitude is slightly decreasing, suggesting it is a converging cycle. This is similar to the case of a converging cobweb model. In the intermediate five-year projection range, the model predicts that hog market will remain in a tight supply status. This is a reasonable outlook by assuming that an effective vaccine for ASF is not yet available. Thus, the loss in hog farm productivity cannot recover in the foreseeable future. In connection to the market behavior since August 2019, key variable outputs closely track the reality in Figure 1. Hog price increased from 22 RMB/kg to the level of 35 RMB/kg during the August–December period. Pork import also increases as projected to above 0.3m MT. Overall, the model is a good platform for analyzing the COVID-19 impact.

4.2 Supply chain disruption scenarios

We explore the following five scenarios of supply chain disruption to the hog market. Notably each scenario is not necessarily unique while other scenarios can be considered for evaluation as well.

4.2.1 Scenario 1: corn shortage and price jump

A major concern on the supply chain disruption emerges from feed transportation. Local authorities imposed traffic restrictions to combat the spread of COVID-19. The side effect is a sudden feed supply cutoff for some livestock farms [5]. There are news reporting starvation losses in local livestock finishing facilities and caused sensational impact [6]. To investigate the impact of a sudden supply scarcity, we simulate a 10% increase in national corn price for two months, which then falls back, as illustrated in Figure 2 panel (a). This is a dramatic change because in sample period corn price has been steady with the maximum price change of less than 5% from 2 RMB/kg. Despite certain local shortages as reported, nationwide supply issue is on average small after central government interventions [7]. The 10% increase in corn price is capable of representing the overall impact for a corn supply shock.

Figure 2 shows the impact. Interestingly, despite sensations in media coverage, a temporary corn price increase fails to generate significant differences from the baseline outcome. The amplitude of hog cycle is slightly widened. The hog price peak is projected to be 38.8 RMB/kg in August 2020, 0.2 RMB/kg higher than the baseline. Price trough is projected to be 17.7 RMB/kg in March 2022, 0.2 RMB/kg lower than the baseline. The wider range of hog price variation leads to a similarly widened range of breeding stocks variation by 0.1m heads to the baseline, yet no change in the timing of peak and trough. Pork consumption demand marginally changes from the baseline by a maximum of 4981 MT/month, or 0.1%, which is almost indistinguishable. Pork import similarly experiences slightly wider variation by a maximum of 702 MT/month, or 0.2%. In sum the sudden jump of corn price due to temporary scarcity will lead to an increase in hog price peak in the short run, but the magnitude is negligibly small.

4.2.2 Scenario 2: market hog transportation disruption

Similar to the feed supply, the traffic cutoff also restrained the market hog transportation to packing plants. Packing plants have to bid up purchase price for hogs or are forced to shut down due to the lack [8] of accessible hog supply. In February 2020, the slaughter volume of inspected packing plants dropped by 36% over a year ago according to MARA [9]. Notably, the 36% drop was based on a 17% lower hog inventory in February 2020 compared to a year ago due to ASF. To simulate the impact, we model a 30% drop in slaughter volume in February. Though this seems a smaller drop than actual inspected volume, in reality it is likely to be a more stringent scenario. Then in March after transportation impediments are lifted, slaughter volume recovered back with a 19% increase to absorb the accumulated hogs. The production overshoot will exponentially decay through April and May with 8 and 3% higher than normal slaughter volume. The overrecovery happens because plants can increase capacity utilization to catch up in production. The total percentage deviation of slaughter volume from normal level in the February–May period sums to zero to ensure that the accumulated hogs are eventually cleared from inventory. The shock path is provided in Figure 2 panel (a).

The impact of hog transportation disruption is presented in Figure 2. There is an immediate impact on hog market. Packing plants bid live hog price up to as high as 36.5 RMB/kg in February 2020 to maintain production operation. This is in line with the actual observation that in February average hog price jumped to 37 RMB/kg [10]. When transportation is restored, hog price drops quickly to a low of 36 RMB/kg in March–April because packing plants ramp up production to clear accumulated hogs. After the accrued hog inventory is cleared, market quickly reverts to the baseline track by July 2020. But because of the earlier disruption, hog price peak is projected to reach up to 38.2 RMB/kg, 0.4 RMB/kg lower compared to the baseline. The decrease is primarily due to the lower than baseline hog price in March–April. When hog price falls, it leads to less pork import and contributes to higher domestic price. Pork inventory gap and consumption demand follow the same path of changes, which are declining at the beginning and restore afterward. Interestingly, breeding stock is little affected because of the short-lived nature of disruption that corrects back within six months.

4.2.3 Scenario 3: delay in breeding stock replacement

In the current hog cycle, farmers are still building breeding inventory driven by expected higher hog prices. The transportation disruption affects not only market hogs but also breeding stocks. For hog farms intending to add new breeding herd, transportation impediments can delay this process. We model this shock as an additional one-month delay (or 25% increase) in obtaining new sows. The additional delay lasts for two months and returns to normal length in April as transportation rehabilitates. The shock is illustrated in Figure 3 panel (a).

Results shown in Figure 3 suggest that this impact is not significant. Similar to the impact of corn price jump, the amplitude of hog price cycle is extended slightly wider by 0.1 RMB/kg. The breeding stock building process is slightly delayed during March–April but a maximum of 0.8m heads. But it catches up in the second half of 2020 when hog price reached higher. Breeding stock peak eventually reaches a slightly higher level of 42.8m heads, up by 0.1m heads than the baseline and arrives almost at the same time. Overall the impact of a two-month delay is minimal because farms can escalate the inventory building pace when hog price increases faster.

4.2.4 Scenario 4: delay in pork import

Due to the global outbreak of COVID-19, international shipping is disrupted. We simulate the scenario that from February to April 2020, shipping disruptions cause the import delay time to double to four months. Starting in May, the import delay returns to normal length. The impact path is illustrated in Figure 3 panel (a). Though we only consider a disruption of three months, as the pandemic develops in EU and North America, import delays can be extended for a longer period.

As shown in Figure 3 panel (f), net pork import sees an immediate drop in February and continued to stay below the baseline through April. After shipping capacity returns to normal, there will be an immediate catch-up starting from May and overshoot to nearly half million MT. After the peak level in May, import starts to decline exponentially because more import generates downside pressure on hog price. Import volume will eventually converge to baseline level in November 2020, after six months' adjustment. The sudden dwindling of import volume pushes up hog price higher to 36 RMB/kg in April. On average hog price is projected to be 1 RMB/kg higher than the baseline during February–April period. From May to November, hog prices will be lower than the baseline scenario by an average of 0.5 RMB/kg, benefiting from higher pork import during this period. The impact on breeding inventory is very small by a maximum of 0.2m more sows in May–June period, brought from the higher hog price during import disruption.

4.2.5 Scenario 5: combination of all supply chain disruptions

The four disruption scenarios are highly likely to happen together instead of independently. In this section we investigate the impact of combined shocks. The simulation results are illustrated in Figure 4.

The difference between Scenario 5 and baseline results mainly resides in the first half of simulation, which is the period before April 2022. For hog price, the biggest difference from baseline emerges between February and September 2020, when market is going through the supply chain disruptions. Due to the combined effect of disruptions in market hog transportation and import shipments, hog price increases to a short-term peak of 36 RMB/kg. In subsequent months of slaughter volume and import ramp-up, hog price adjusts to a temporary low of 35.5 RMB/kg before rebounding higher to 38.2 RMB/kg following the normal cycle. The hog price arrives two months later than baseline in October. After that hog price begins to converge toward the baseline outcome. When hog price reaches its trough in April 2022 at 17.9 RMB/kg, it almost merges with baseline results yet lags behind by about one month.

Other market indicators generally follow the same pattern as hog price. Pork consumption demand and inventory fluctuate in the disruption period and then stay slightly below baseline level by an average of 20,000 MT per month. By the time of April 2022, both indicators converge with baseline. Breeding stock falls below baseline during the disruption period by an average of 0.8m heads a month. But fueled by higher hog prices, hog farms build up inventory faster in the postdisruption period and catche up with baseline by February 2021. Pork import mainly suffers from international shipping disruptions and behaves almost the same as in Scenario 4.

5. Conclusions

The global outbreak of COVID-19 pandemic has significantly impacted Chinese and world economy. Disruptions in supply chain have been observed in almost all industries. Among the impacted industries, the loss in agriculture and particularly the livestock sector is seen to be especially concerning for two reasons. First, the damages in agricultural production can be detrimental to national food security, social stability and household life. Second, the livestock sector is biologically vulnerable to outside shocks because it relies on smooth circulation of feed input and market output to keep functioning. For these reasons, in this article we evaluate the impact of supply chain disruptions from COVID-19 outbreak on China hog market, which is the largest livestock sector and a staple meat protein for Chinese household.

To assess the dynamic impact of COVID-19, we employ a system dynamic model on China hog market to capture the full picture of market responses. The model is calibrated using 2018–2019 China hog market data to reflect the current tight supply situation because of the ongoing ASF disease. The model well incorporates the featured cyclicity and feedback relationships in hog market. A total of five supply chain disruption scenarios derived from realistic public, policy and industrial concerns are proposed and evaluated.

The following findings are concluded from simulation results. First, despite sizable impacts during the period of disruption, the overall effects of the considered scenarios are transitory. After disruptions dissipate, the long-term aftermath is projected to last another 18 months. However, the magnitude is very small and negligible in economic sense. The facts suggest hog market has strong self-correction capacity. Second, among the individual disruption scenarios, market hog transportation has the most significant immediate impact on price and consumption. Packing plants raise hog and pork price to maintain production, which in turn reduces consumption demand and raises pork import and breeding capacity. As traffic reopens and packing plants ramp up production, the impact dissipates in six months. All key market indicators return to the baseline cycle after the shock passes. Third, delay in import shipment temporarily reduces consumption and raises hog price. However, market reacts by slight increase of the breeding capacity, which generates more pork supply for two years later. After shipping schedule returns to normal situation, net import immediately overshoots to catch up with backlogged orders. The price effect gradually fades in about six months. Fourth, despite mass media coverage over certain severely affected regions, a sudden jump of corn price or delay in breeding stock acquisition does not produce significant impact on national hog market as a whole. The impact mostly resides in the nearest one-year horizon and the magnitude is negligibly small. This study complements the findings of Wang and Wang (2020) that long-term structural demand and supply shocks can lead to significant changes in hog cycle and food security. More emphasis should be focused on preventing short-term supply chain disruption from lasting for a prolonged time period.

The model is a powerful tool to conduct impact analysis of supply chain disruptions on hog market. The method can be applied to other livestock markets including cattle and poultry. Further analysis can be performed to understand alternative shock scenarios and interactions among different markets. Further efforts can focus on evaluating relevant scenarios for improved analysis. Another limitation is we only considered the impact of supply chain shocks. Consumption demand is also heavily hit by the COVID-19 outbreak and social distancing regulations. When demand dampens, the impact of a temporary shortage of hog supply is likely to be ameliorated as in Scenario 2. Further analysis can be performed to consider the demand shock scenarios.

Figures

Data summary

Figure 1

Data summary

Supply chain disruptions: corn and market hog transportation

Figure 2

Supply chain disruptions: corn and market hog transportation

Supply chain disruptions: breeding stock and pork import transportation

Figure 3

Supply chain disruptions: breeding stock and pork import transportation

Combined supply chain disruptions

Figure 4

Combined supply chain disruptions

Model parameters

ParameterNoteValueUnit
Lpsmlitters per sow per month0.18
Pslpigs saved per litter9head
Splproductive life of sow36month
Gmgestation and maturation delay8month
Srfinishing survival factor0.9
Msfpmature stock feeding period2month
Impork import delay2month
ADPhog price expectation adjustment delay4month
ADbsbreeding stock adjustment delay4month
εprice elasticity of pork demand−0.34
ECPexpected corn price2RMB/kg
Popconsumer population1.40billion
rpoppopulation growth rate0.334%annual
ECexpected monthly pork consumption4.1million MT
DCOVdesired coverage ratio0.72
γhog price response to RCOV−11.5
yhog dress yield0.68
MMmarket margin2.62RMB/kg
whog live weight115kg/head
ϕdesired breeding stock changes to EHCR1.78million heads
ρpork import demand response to hog price0.13million MT

Notes

1.

CNN, “Global shipping has been hit by the coronavirus. Now goods are getting stranded.” https://www.cnn.com/2020/02/05/business/shipping-coronavirus-impact/index.html

2.

Aside from the supply-response-related cyclical pattern, hog market also exhibits a relatively regular annual seasonality due to biological factor and consumption preference. In this analysis, we do not consider the seasonality pattern.

References

Chavas, J.P. and Holt, M.T. (1991), “On nonlinear dynamics: the case of the pork cycle”, American Journal of Agricultural Economics, Vol. 73 No. 3, pp. 819-828.

Chavas, J.P., Kliebenstein, J. and Crenshaw, T.D. (1985), “Modeling dynamic agricultural production response: the case of swine production”, American Journal of Agricultural Economics, Vol. 67 No. 3, pp. 636-646.

Chen, D., Abler, D., Zhou, D., Yu, X. and Thompson, W. (2015), “A meta-analysis of food demand elasticities for China”, Applied Economic Perspectives and Policy, Vol. 38 No. 1, pp. 50-72.

FAPRI (2019). “2019 baseline update for international dairy, livestock and biofuel markets”, Food and Agricultural Policy Research Institute FAPRI-MU Report #04-19. University of Missouri, available at: https://www.fapri.missouri.edu/wp-content/uploads/2019/09/2019-International-Baseline-Update.pdf.

Holt, M.T. and Johnson, S.R. (1988), “Supply dynamics in the US hog industry”, Canadian Journal of Agricultural Economics, Vol. 36 No. 2, pp. 313-335.

Hu, X., Wang, M. and Shi, Z. (2015), “Study on pork supply and demand in China based on market model”, Chinese Rural Economy No. 4, pp. 14-28.

Huatai Securities (2018), Research Report on Hog Farming Industry. Huatai Securities Equity Research Report, December, available at: http://pdf.dfcfw.com/pdf/H3_AP201812281280188869_1.pdf.

Lan, H., Zhao, J. and Xiao, Y. (2019), “An empirical study of the relationship between expected price, feeding cost and live pig price–based on statistics of sichuan province”, West Forum on Economy and Management, Vol. 30 No. 6, pp. 12-25.

Mao, X. and Zeng, Y. (2008), “Pork price period identification based on time series decomposition”, Chinese Rural Economy No. 2, pp. 4-13.

Meadows, D.L. (1970), Dynamics of Commodity Production Cycles, Wright Allen Press, Cambridge, Mass.

Nerlove, M. (1958), “Adaptive expectations and cobweb phenomena”, The Quarterly Journal of Economics, Vol. 72 No. 2, pp. 227-240.

Nicholson, C. and Parsons, D. (2012), “Dynamic analysis of policy options for Mexico's sheep sector”, 2012 International Conference of the System Dynamics Society, pp. 1-28.

Nicholson, C.F. and Stephenson, M.W. (2015), “Milk price cycles in the US dairy supply chain and their management implications”, Agribusiness, Vol. 31 No. 4, pp. 507-520.

Shamsuddoha, M., Quaddus, M. and Klass, D. (2013), “Poultry supply chain: a system approach”, International Conference of the System Dynamics Society.

Song, C. (2016), “Study on fluctuation of pork price under nonlinear and disquilibrium cobweb model–based on empirical analysis of time-varying parameter model”, Journal of Huazhong Agricultural University (Social Sciences Edition) No. 6, pp. 1-7.

Sun, X., Wu, X. and Qiao, J. (2014), “Study on supply response of pork in China based on Nerlove model”, On Economic Problems No. 8, pp. 109-112.

Tan, Y. (2010), “An empirical analysis on the volatility of aggregate pork supply and its influencing factor”, Journal of Huazhong Agricultural University (Social Sciences Edition) No. 3, pp. 24-29.

Wang, J. and Wang, X. (2020), “The dynamic impact of COVID-19 on commodity cycle: the case of China’s hog market”, Working paper, Department of Economics, University of New Mexico, Albuquerque, New Mexico.

Wu, X. (2011), “Economic reasons for periodic fluctuations of pork price in China from 2004 to 2009 -- Based on analysis of elasticity of pig supply and demand”, China Price No. 11, pp. 30-34.

Yu, X. and Abler, D. (2014), “Where have all the pigs gone? Inconsistencies in pork statistics in China”, China Economic Review, Vol. 30, pp. 469-484.

Zhuang, R. and Abbott, P. (2007), “Price elasticities of key agricultural commodities in China”, China Economic Review, Vol. 18 No. 2, pp. 155-169.

Corresponding author

Xiaoyang Wang can be contacted at: xiaoyangwang@unm.edu

Related articles