1. Introduction
Supply chain collaboration has been the noteworthy topic of supply chain management, once the potential to improve supply chain performance is recognized by both academic researchers and industry practitioners. Accordingly, there have been various supply chain programs designed to build collaboration among supply chain members, and for example, Quick Response, Continuous Replenishment, and Vendor-Managed Inventory con-sequently lead to effective supply chain management. (Danese, 2006; Derrouiche et al., 2008; Sadeghi et al., 2014a; Xu & Leung, 2009). Meanwhile, many past studies about supply chain collaboration chose only particular programs and examined their performances in only special situations (Chakraborty et al., 2015; Egri & Vancza, 2013; Rad et al., 2014). Cross examination among different programs under various conditions is essential to identify the true nature of supply chain collaboration.
This study investigates how different supply chain collabo-ration programs perform. Among diverse collaboration programs, this study focuses on the consignment and Vendor-Managed Inventory (VMI) that have been broadly applied to industries. This study defines the consignment and VMI programs with three key elements of collaboration including information sharing, cost payment, and decision authority. Based on the claim that more portion of decision making or cost payment is done by one member under more advanced collaboration program, VMI is considered to be more advanced collaboration program than the consignment. This study intends to compare overall perform-ances of these two collaboration programs. The value of in-formation sharing is also evaluated under different collaboration programs. By directly comparing two collaboration programs and testing the impact of information sharing under different collabo-ration programs, this study can provide good ideas to develop a new collaboration program that brings better supply chain per-formance than any existing ones.
For the sophisticated analysis, this study employs simulation modeling based on time-phased operations. The simulation mod-el represents a two-stage supply chain system where a supplier sells a single item to a buyer, and it incorporates various oper-ations including demand forecasting, inventory controlling, manu-facturing, and delivering. The simulation models consider six op-erational and environmental factors, and they are the collabo-ration program, information sharing level, demand pattern, lead-time, production capacity, and costs. The supply chain per-formance is accessed in terms of profits, service levels, and bullwhip effect. Two collaboration programs are compared with the traditional non-collaborated system as a basis, and their per-formances are further analyzed to examine the impacts of op-erational and environmental factors on the supply chain performance. In general, the simulation outcomes support that more advanced collaboration programs outperform less advanced ones. The results also show that information sharing is not just a component of supply chain collaboration but it is essential for successful collaboration.
2. Theoretical Background
VMI is originally designed as the buyer-supplier partnership to improve the supply chain performance (Simchi-Levi et al., 2000). Under the VMI program, a supplier is fully responsible for man-aging buyer's inventory. Based on the detailed VMI contract, the supplier controls inventories stored at buyer's warehouse by placing orders and setting the inventory level based on the de-mand data that he receives directly from the buyer. After all, the VMI program is (Kim et al., 2014)beneficial to the buyer and he is free from the burden of replenishment and inventory control. The VMI also enables the supplier to synchronize re-plenishment, shipment, and production activities based on re-al-time information of market demand, and he can systematically save the relevant costs.
Past studies have considered diverse issues regarding VMI such as the power game among supply chain members (Yu et al., 2009b; Zheng et al., 2009), information sharing (Vigtil, 2007), contract terms (Guan & Zhao, 2010; Wong et al., 2009), and algorithm development for inventory and distribution prob-lems (Sadeghi et al., 2014b; Seifbarghy & Gilkalayeh, 2012). In the past studies, VMI is compared with other systems such as the non-collaborated system (Kiesmuller & Broekmeulen, 2010; Wang, 2009), Point-Of-Sales (Kim et al., 2011), and Collaborative Planning, Forecasting, and Replenishment(Sari, 2008a, b), and their performances are evaluated in terms of the inventory level (Moon & Kim, 2007; Sari, 2007) and economic values(Song & Dinwoodie, 2008; Yu et al., 2009a),.
The consignment is a conventional business practice with long history (Fenton & Sanborn, 1987), and it has been still broadly used in current businesses (Gerber, 1987). According to the consignment contract, the supplier holds his products on buyer’s premises before the buyer takes them for sales(Harding, 1999). Since the supplier still owns the inventories, the buyer needs to pay for the inventories once they are sold at the mar-ket(Gerber, 1991). Many past studies consider the consignment as one of supply chain collaboration programs and show that it improves supply chain performance and solves inherent prob-lems of the traditional system (Battini et al., 2010; Valentini & Zavanella, 2003).
Information sharing has been the famous issue that many re-searchers address in their studies on the supply chain manage-ment(Kim & Song, 2013). While it is widely accepted that in-formation sharing provides the supply chain system with sig-nificant benefits, the results of past studies are somewhat mixed. For example, Cachon & Fisher (2000) and Raghunathan (2001) failed to show the significant benefit purely due to in-formation sharing. Chen (1998) and Lee et al. (2000)found out that information sharing gives significant benefits to the supply chain system only on the certain conditions. In Zhao & Xie (2002) study, information sharing gives benefits to only the sup-plier and hurts retailers. By implication, further investigation is still required to identify the true value of information sharing.
This study directly compares various performances of the consignment and VMI under diverse circumstances by using the simulation model, while most past studies primarily depended on analytical modeling that leads to limited findings about how these collaboration programs work(Chen, 2013; Mateen & Chatterjee, 2015). In addition, this study examines the value of information sharing under two different collaboration programs.
This study follows the series of previous research works on VMI and consignment. Some of them examine how two collabo-ration programs improve the supply chain performance and dem-onstrate the benefits of supply chain collaboration (Ryu, 2006b, 2007a, 2013b). The performance of VMI is examined under the different conditions (Ryu, 2013a) and even the combination of two programs is tested to look for more advanced collaboration program (Ryu, 2015). Meanwhile, those studies hold the limited concept of supply chain collaboration and ignore information sharing in the collaboration programs. The assumptions of their deterministic models oversimplify the complex supply chain operations. One study investigates the impact of different cost payment schemes on VMI's performance by using the simulation model with stochastic demand and lead-time (Ryu, 2012). Its simulation applies the time phased processes that are close to the operations occurring in the real supply chain system, and the supply chain model expands the scope of supply chain col-laboration by considering the impact of information sharing. This study shares the basic supply chain model and simulation de-sign with one of the previous studies (Ryu, 2012). By using the simulation, however, this study directly compares two different collaboration programs under the various conditions. In addition, this study examines the impacts of operational and environ-mental factors on the supply collaboration programs.
Since the supply chain collaboration became the main interest of researchers in the business area, there have been so many studies addressed this issue. Meanwhile, many researchers have chosen their own ways to define the supply chain collaboration, andevenit becomes extremely hardtofindthecommonfea-tures of specific collaboration programs such as VMI in their studies. Literally, there exist various ways to define the collabo-ration programs, and it may particularly cause the serious prob-lem when different collaboration programs are simultaneously evaluated.
Since this study directly compares distinct types of the supply chain collaboration, first of all, it needs to develop the funda-mental concept to define any supply chain collaboration programs. This study analyzes the supply chain collaboration programs based on three elements of collaboration, and they are information sharing, cost payment, and decision authority (Ryu, 2007b).
Information sharing is a typical feature that is found in most collaboration programs. Information sharing indicates that supply chain members actively share information with the purpose of improving the supply chain performance. Many past studies have focused on the value of information sharing (Cachon & Fisher, 2000; Raghunathan, 2001).
Supply chain collaboration is also realized with new cost pay-ment schemes. By changing the cost method such as price dis countor switching the member who pays particular cost items such as consignment, supply chain members can realize collab-oration and improve the overall supply chain performance. In general, the supply chain system performs better when a supply chain member who holds decision authority is responsible for paying more cost items (Gumus et al., 2008; Ryu, 2012).
Decision-making is also a key element of supply chain collaboration. In general, the supply chain collaboration happens when a single member holds the full authority to make deci-sions on the supply chain operations or the operational deci-sions are made by consensus among different members. Examples of the operational decisions that members make to-gether under the supply chain collaboration are sales price (Weng, 1995), production rate (Weng, 1997), lot size (Weng, 1995, 1997), discounting price (Chen et al., 2001), and inventory control policy (Ng et al., 2001).
Based on the proposed framework, a researcher can recog-nize the special features of a collaboration program that enables supply chain members to collaborate their operations in the sup-ply chain system. In particular, this framework allows anyone to figure out which is more advanced collaboration program than the others. According to the proposed frame work, the specific collaboration program is more advanced one as more in-formation is shared among supply chain members, as more cost items are paid by one member, as more decisions are made by one member.
This study focuses on decision making and cost payment in the comparison of the consignment and VMI programs. According to the proposed framework, this study claims that VMI that let a single member make more decisions and pay greater portion of the cost is more advanced collaboration program than the consignment.
3. Simulation Model and Design
The simulation model represents a two stage supply chain system where one supplier sells a single item to a buyer (Ryu, 2012). Both buyer and supplier make the decisions on their own operations in a way to maximize their profits. The buyer's profit includes sales revenue at the retail market, ordering cost, in-ventory carrying cost, and shortage cost. The supplier’s profit contains sales revenue from the buyer, set-up cost, inventory carrying cost, backorder cost, production cost, capacity ex-pansion cost, and transportation cost.
The simulation is designed to examine three different supply chain models including traditional system, consignment, and VMI. In each supply chain system, the buyer and supplier have different decision authority and information structure from the other system.
3.1. Three Supply Chain Models in Simulation
In the traditional system, the buyer and supplier manage their processes by making independent decisions based on a limited set of accessible information. With the record of past demands at the retail market, the buyer conducts demand forecasting to prepare ordering. In the replenishment plan, the order quantity and frequency are determined based on the transfer price in-formed by the supplier. The buyer determines the retail price before forecasting the demand, and the original price is adjusted depending on the newly updated forecasted demands. The sup-plier forecasts buyer’s orders based on the past orders, and he develops production plans to fulfill the orders placed by the buyer. The supplier set the transfer price before the buyer pla-ces orders, and then he modifies the original price to reflect the updated information of the replenishment schedule. Both supplier and buyer use the periodic review system (R,Q) as the in-ventory control policy (Angulo et al., 2004). The buyer forecasts market demands to decide the orders by using Holt’s linear ex-ponential smoothing method (Holt, 2004).
In the consignment, the buyer does not have to make pay-ment to the supplier for products until they are consumed. Accordingly, the buyer still has the full authority to make any decisions on replenishment under the consignment. In the simu-lation model, the buyer is responsible for paying only the stock-ing cost out of the entire cost to hold inventories at his warehouse. The inventory holding cost is separated into stocking and financing costs (Valentini & Zavanella, 2003). The stocking cost indicates the costs to store, handle, and insure inventories. The financing cost is an opportunity cost that the supplier pays while he invests finances in producing the inventories. Under the consignment, the buyer pays only stocking cost and the supplier is responsible for the financing cost of inventories stor-ed at buyer’s warehouse.
Under VMI, the supplier controls inventories stored at buyer’s warehouse. Buyer's demand forecast are directly transferred to the supplier who is responsible for the orders. The buyer still decides the retail price and informs this price to the supplier. Based on the retail sales price and forecasted market demand obtained from the buyer, the supplier decides the lot size and order frequency to control the inventories stored at buyer's warehouse. The supplier also develops the production plan ac-cording to the order schedules that he makes. The suppler de-cides the transfer price based on the replenishment plan and in-ventory levels at two warehouses.
3.2. Operational and Environmental Factors
In the simulation runs, this study considers the operational and environmental factors including the collaboration program, information sharing level, demand pattern, lead-time, production capacity, and costs. By considering different levels of these fac-tors in simulation runs, this study seeks to obtain the general-izable outcomes in diverse circumstances.
Six factors associated with operations and environments of the general supply chain system are chosen to significantly af-fect supply chain performance. One factor that is considered in the simulation is the collaboration program. The collaboration program represents a specific way that all members in the sup-ply chain system operate their functions and make decisions. In the simulation models, total three different supply chain systems are considered including the traditional system. The traditional system serves as the base compared with the consignment and VMI.
Many past studies support that the collaborative supply chain system out performs the traditional non-collaborative system (Aviv, 2001; Mc Carthy & Golicic, 2002). VMI is considered to be more advanced collaboration program than consignment, be-cause the decision authority (related to ordering) is more con-centrated on the one member (supplier).
Information sharing has been addressed as a main topic in supply chain management due to its potential value in supply chain operations. Past studies examined the impact of in-formation sharing when the supply chain members share various kinds of information such as customer demand (Lee et al., 2000; Raghunathan, 2001), forecasted demand (Aviv, 2001; Cachon & Lariviere, 2001), and inventory (Cachon & Fisher, 2000; Zhao & Xie, 2002).In general, extended information shar-ing enhances visibility of operations in the supply chain system, and consequently allows supply chain members to make better decisions on their operations. According to past studies, when the supply chain members share information more extensively, the supply chain system achieves better performance (Cachon & Fisher, 2000).
The simulation model considers three levels of information sharing (Gavirneni et al., 1999). No information sharing implies that the buyer informs only orders to the supplier, and no other information is transferred between the buyer and supplier. With partial information sharing, the only selected portion of in-formation is shared by the supplier and buyer, and the buyer provides either forecasting or market sensitivity parameters to the supplier. Full information sharing indicates that the buyer and supplier share full information including both market demand and forecast parameters.
In general, the demand pattern is considered as an important input of simulation runs to evaluate the supply chain system. Past studies use diverse demand patterns in their simulation models in-cluding the simple probability distribution (Waller et al., 1999), par-ticular demand patterns of trend or seasonality (Zhao & Xie, 2002; Zhao et al., 2002a), autocorrelation (Aviv, 2002; Raghunathan & Yeh, 2001), price-sensitive demand (Mishra & Raghunathan, 2004), and combinations of more than one demand models (Zhao et al.,2002c). This study considers three different demand patterns in-cluding the steady state uniform distribution, demand with a linear trend, and the autocorrelation (AR(1)).
The lead-time is the time length required for a certain routine operation such as moving, queuing, set-up, and run times (Hopp & Spearman, 2000). In general, the lead-time lessens the cer-tainty of information. Even though the lead-time is known to af-fect the supply chain performance, many past studies simplify their supply chain models by assuming no lead-time (Robinson et al., 2005). In the simulation model, this study considers two different lead-times. Buyer’s lead-time represents the time dura-tion between the order and arrival of products at buyer's place. Supplier’s lead-time refers the time length between when the production plan schedules a product for manufacturing and when manufacturing the product is complete. In the simulation runs, the lead-time is generated based on the specific proba-bility distribution to allow uncertainty of the lead-time.
The production capacity represents the maximum amount of production throughout, and normally it restricts the flexibility of production process to fulfill unstable demands (Hopp & Spearman, 2000). Since the collaboration programs like VMI are known to lessen the demand uncertainty, they can maintain the limited production capacity and do not require the high inventory level (Wijngaard, 2004).Due to its significant impact on the sup-ply chain performance, many past studies considered the pro-duction capacity to be the important condition of collaboration programs (Gavirneni et al., 1999; Waller et al., 1999).
Since the arbitrarily chosen parameters are used in the simu-lation runs, this study varies the cost parameters to secure gen-eralizability of the simulation outcomes. In particular, the in-ventory holding, shortage, and ordering costs are chosen among many different cost items, because they are significantly different among distinct industries (Zhao et al., 2002b).By varying the ra-tio of costs with supplier’sand buyer’s costs, the simulation mod-el maintains the same value of total supply chain cost(Simpson, 2007; Yao et al., 2007).
3.3. Performance Measurement
In the simulation runs, this study accesses the supply chain performance in terms of economic values, customer service lev-els, and bullwhip effect. Most past studies measured the eco-nomic values when they evaluate the supply chain performance due to their practical significance in most businesses (Aviv, 2002; Raghunathan, 1999). This study also uses the economic values including profit as well as cost. In addition, the detailed cost items are calculated to identify the exact causes of overall outcomes.
The customer service level refers the fraction of customer de-mands fulfilled on time (Hopp & Spearman, 2000).In the past studies, the customer service level has been measured in differ-ent scales such as the fill rate (Braglia & Zavanella, 2003; Disney & Towill, 2002) and backorder level (Waller et al., 1999). This study measures the fill rate, which is the fraction of orders fulfilled by available inventory, to indicate both buyer’s and sup-plier's customer service levels.
The bullwhip effect is the phenomenon that orders to the supplier have higher variance than sales to the buyer (Lee et al., 1997). Many researchers have focused on the bullwhip ef-fect, because its well-known negative effect on the supply chain performance. While many past studies have addressed the is-sues regarding the bullwhip effect, relatively a small number of them have used it to evaluate the collaboration programs(Chen et al., 2000; Disney & Towill, 2003). This study measures the bullwhip effect in the simulation runs to examine how well the supply chain collaboration programs lessen it. According to the past studies(Chen et al., 2000; Disney & Towill, 2003), the bull-whip effect is measured as the ratio of demand variance at the upstream echelon over one at the downstream echelon.
3.4. Parameter Setting
In the simulation, three different supply chain collaboration programs and three levels of information sharing are considered (Gavirneni et al., 1999), and the experiment is designed to com-pare total six supply chain systems including the traditional sys-tem with no information sharing, traditional system with partial information sharing, consignment with no information sharing, consignment with partial information sharing, VMI with partial in-formation sharing, and VMI with full information sharing. The other three cases from the full factorial design of collaboration programs and information sharing levels are not considered in the experiment, because a certain level of information sharing is not applicable to a particular collaboration program.
The level of information sharing is determined depending on whether the buyer and supplier apply the same parameters to their demand forecasting and estimating market demand (Aviv, 2001; Cachon & Lariviere, 2001). The values of specific param-eters used for demand forecasting and market demand in the casesareshownin[Table1].
[Table 1] Parameter Setting in Different Levels of Information Sharing
Source: Ryu(2006a).
In the simulation runs, three types of demand patterns includ-ing steady state demand, demand with trend, and autocorrela-tion are considered. The steady state demand indicates the case that the maximum demand is a random number from the uniform probability distribution between 1 and 4 units. The de-mand with trend is generated by using Holt’s linear exponential smoothing method as follows.
\(\begin{array}{l} L_{t}=\alpha \cdot Z_{t}+(1-\alpha) \cdot\left(L_{t-1}+T_{t-1}\right) \\ T_{t}=\beta \cdot\left(L_{t}-L_{t-1}\right)+(1-\beta) \cdot T_{t-1} \\ D_{t+m}=L_{t}+m \cdot T_{t} \end{array}\)
where,α = 0.01, β= 0.01, m = 100, Zt ~ Uniform (1,3), and Dt is theactual demandat time t.
The demand with autocorrelation follows the standard AR(1) model, and it is generated by the following equation.
\(D{t+1} = 2.5 + \rho \times D_t -2.5 \times Z_{t+1} \) where, \(\rho = 0.5 \) and \(Z_t ~ Uniform (-1,1)\)
[Table 2] shows the values of the specific parameters used for different levels of lead-time, production capacity, and cost structure. The other parameters used in the basic simulation model are shown in [Table 3].
[Table 2] Parameters Used in Simulation
Source: Ryu (2006a).
[Table 3] Parameters Used in Base Case
Source: Ryu (2006a).
After all, the number of total cases considered in the simu- lation are counted as 6 (collaboration programs and information sharing) 3(demand patterns) 2(buyer’s lead-time) 2(supplier’s lead-time) 2(production capacity) 2(inventory holding cost ratio) 2(ordering cost ratio) 2(shortage cost ratio) = 1,152.The warm-up period is determined after checking how both buyer and suppli-er’s inventory levels change over the time. Since their inventory levels become stabilized after 35 periods, each simulation run has 35 warm-up period.
Total simulation running time is 135 periods including 35 warm-up period. Meanwhile, the simulation performance is re-corded during the half-period to keep the rule of i.i.d.(independently and identically distributed) on the simulation outcomes. Accordingly, 50 points of the data in each simulation run are observed for the analyses ((135-35)/2 = 50).
4. Simulation Results
The performances of the individual supply chain systems are obtained from the simulation runs as shown in [Table4].
[Table4] Average Performances of Supply Chain Systems
According to the simulation outcomes, the supply chain sys-tem achieves higher supply chain profit than the others in the order of VMI with full information sharing, consignment with par-tial information sharing, traditional system with partial information sharing, traditional system without information sharing, consign-ment without information sharing, and VMI with partial in-formation sharing. Under the assumption that the supply chain system applies more advanced collaboration programs than the others in the order of VMI, consignment, and traditional system, this result supports that more advanced collaboration programs result in greater supply chain profit than less advanced ones.
In addition, the simulation results indicate that higher level of information sharing leads to greater supply chain profit regard-less of the collaboration program applied to the system. In par-ticular, without high level of information sharing, more advanced collaboration systems such as VMI and consignment achieve lower supply chain profit than less advanced ones. This result implies that the collaboration program can be successful only if the proper amount of information is shared among supply chain members.
The main cause of the simulation results can be figured out by observing the demand and order quantity. Overall, the collab-oration programs enable the members to increase their revenues by increasing the total through put in the entire supply chain system and they can reduce the cost by controlling the order quantity efficiently. In particular, under VMI, the supplier has the full authority over ordering products and is able to reduce the cost significantly by properly balancing his production rate and buyer's order.
Information sharing allows the supply chain members to figureout the exact market sensitivity and obtain correct forecasted demands. With accurate information, they tend to lower their pri-ces to increase the total demand, and it leads to the increased supply chain revenue.
In general, the buyer obtains greater profit when more ad-vanced collaboration program is applied to the system and the level of information sharing is higher. On the other hand, the supplier’s profit is lower under more advanced collaboration systems. This result implies that the buyer receives most of the benefits from the supply chain collaboration program. With the consideration of the key nature of the supply chain collabo-ration, which requires every member’s participation, the proper program to compensate supplier’s loss must be prepared to en-sure successful application of the collaboration program. In one exceptional case of VMI with full information sharing, both buyer and supplier improve their performances, and again, it implies the importance of information sharing in applying the collabo-ration program.
The buyer’s and supplier’s service levels appeared in [Table 4] show no significant difference under the distinct collaboration pro-grams or with different levels of information sharing. Meanwhile, supplier’s service level is less under VMI compared with the other systems.
The bullwhip effect, which is another indicator of the supply chain performance, is not significantly different between the tra-ditional system and consignment. VMI, however, results in less bullwhip effect than the other systems. VMI enables the supplier to determine ordering and prevents the excessive changes of order quantity over time.
[Table 5] shows buyer’s, supplier’s, and supply chain profit under different levels of the operational and environmental factors. The simulation outcomes indicate that the traditional system and consignment make greater profit when buyer’s lead-time is shorter, but it shows a marginal difference from when buyer’s lead-time is longer. Meanwhile, VMI’s perform-ances are not significantly affected by buyer’s lead-time.
[Table 5] Average Performances of Supply Chain Systems in Individual Cases
When supplier’s lead-time becomes longer, buyer’s profit does not change, but both supplier’s profit and supply chain profit increase. This outcome is commonly observed regardless of supply chain collaboration programs.
Supplier’s production capacity has no significant impact on buyer’s profit. However, greater production capacity significantly decreases supplier’s profit. This outcome is also shared by ev-ery collaboration system. This result happens due to the in-creased cost caused by unnecessarily high supply with ex-cessive production capacity and the consequent increase in the order quantity.
When the inventory holding cost ratio is high, buyer’s profit is lower and supplier’s profit is higher in the traditional system. With higher inventory holding cost ratio, buyer’s profit is greater and supplier’s profit is less under the consignment and VMI. When the ratio is high, the traditional system obtains less profit, and the consignment and VMI have greater profit.
With high ordering cost ratio, both buyer and supplier’s profits are lower in the traditional system. Under the consignment, buy-er’s profit is lower and supplier’s profit is higher when the ratio is high. Meanwhile, the ordering cost ratio has no significant im-pact on buyer’s and supplier’s profits under VMI. Except for VMI, the supply chain profit is greater with high ordering cost ratio.
The shortage cost ratio has the marginal impact on buyer’s profit in every collaboration system. With the high ratio, suppli-er’s profit is higher in the traditional system but it is less under the consignment and VMI. However, they show just marginal differences. Consequently, the shortage cost ratio has an insig-nificant impact on the supply chain profit regardless of collabo-ration programs.
In general, the traditional system generates higher supply chain profit with shorter buyer’s lead-time, longer supplier’s lead-time, lower production capacity, lower inventory holding cost ratio, lower ordering cost ratio, and higher shortage cost ratio. Compared with the traditional system, however, the consignment system’s profit is greater with higher inventory cost ratio and higher shortage cost ratio. Meanwhile, VMI achieves higher profit with longer supplier’s lead-time, higher inventory cost ratio, and lower shortage cost ratio.
5. Conclusion
Since the supply chain management became one of the key issues in the business, both researchers and practitioners have been searching for the specific programs that bring efficient col-laboration among supply chain members and consequently im-prove the supply chain performance. This study examines the consignment and VMI as well-known supply chain collaboration programs. By evaluating their performances under various cir-cumstances, this study intends to find out how these collabo-ration programs perform under different conditions.
For the analysis of the distinct supply chain systems, this study applies simulation modeling with time-phased operations. The simulation model represents the multiple steps of supply chain operations in a two-stage supply chain system where a supplier manufactures and sells a single product item to one buyer. In the simulation runs, six operational and environmental factors are considered and different collaboration programs are evaluated in terms of profits, service levels, and bullwhip effects. Based on the analysis on the simulation outcome, two supply chain collaboration programs are compared with the traditional non-collaborated system, and their performances are further ex-amined to figure out the impacts of operational and environ-mental factors.
The simulation outcomes lead to the following main findings and managerial implications. First, in general, the simulation re-sults support that more advanced supply chain collaboration pro-grams result in better performance than less advanced ones. More advanced collaboration program holds stronger collabo-ration among supply chain members and wider scope of their participation on the operations enables the supply chain system to achieve better performance. This result implies the potentials to develop new collaboration program that can lead to better performance than the existing ones such as VMI and consignment. The companies need to keep looking for additional collaborative features to design new collaboration programs.
Second, the increased total throughput in the entire system is the main reason that the supply chain collaboration system out-performs the traditional system. The enlarged throughput is mainly caused by joint decision making based on cooperation between supply chain members. This result indicates that the companies should count the revenue as well as cost to identify the true value of supply chain collaboration.
Third, the simulation results indicate that there are specific conditions that the particular collaboration programs perform better. By implication, full knowledge about the specific opera-tional and environmental conditions is quite required to obtain maximum outcomes from the collaboration program.
Fourth, the supply chain system achieves greater perform-ances as its members share more information. This result sup-ports that information sharing brings a significant value to the supply chain system (Cachon & Fisher, 2000; Lee et al., 2000). In particular, the simulation outcomes indicate that information sharing is essential for successful supply chain collaboration.
Finally, the buyer receives most benefits from the supply chain collaboration (Yao et al., 2007). This result can be the major obstacle for practical application of the supply chain col-laboration programs. The proper program that compensates sup-plier’s loss is required to apply the collaboration programs in practices.
The key contributions of this research are threefold. First, this study directly compares two distinct collaboration programs. Most past studies show the advantages of individual collaboration sys-tems over the traditional non-collaborated system (Mateen & Chatterjee, 2015; Rad et al., 2014). There have been only a few studies that directly compare different collaboration programs (Cachon, 2001; Gumus et al., 2008). By analyzing two different collaboration systems at the same time, this study attempts to validate that more advanced collaboration program outperforms less advanced one.
Second, this study figures out the circumstances that the spe-cific collaboration program performs well. By evaluating the per-formances of two collaboration programs under different levels of operational and environmental factors in the simulation, this study identifies the particular conditions that the collaboration programs can fully exploit their advantages.
Finally, the outcome of this study can answer the important question of whether the supply chain members should share in-formation with others at their expense. By examining more than one collaboration programs having different levels of information sharing, this study evaluates the value of information sharing under distinct collaboration forms.
This study has some limitations that may become potential research issues. Since the simulation runs in this study rely on the arbitrarily chosen parameters, the simulation outcomes may not be applicable to every case. The future study can acquire generalizable results by using the data obtained from the empiri-cal study on real cases.
This study shows how operational and environmental factors affect the performances of two collaboration programs, but it does not provide the comprehensible reasons of their impacts. The thorough experiment with extensive ranges of parameter setting is required to identify the exact causes of the simulation outcomes, and it can lead to the reasonable guideline on which collaboration program to select under the specific condition. This issue is rendered to the future study.
Cited by
- How Quick Response affects the Supply Chain Performance vol.17, pp.7, 2016, https://doi.org/10.15722/jds.17.7.201907.87