Int. J. Business Performance and Supply Chain Modelling, Vol. 1, No. 1, 2009

1

Effects of cascade information sharing in inventory

and service level in multi-echelon supply chains

F.T.S. Chan

Department of Industrial & Manufacturing Systems Engineering,

University of Hong Kong,

Pokfulam Road, Hong Kong, P.R. China

E-mail: ftschan@hkucc.hku.hk

H.K. Chan*

Norwich Business School,

University of East Anglia,

Norwich, Norfolk, NR4 7TJ, UK

E-mail: h.chan@uea.ac.uk

*Corresponding author

Abstract:

Information sharing has been regarded as an effective remedy to

counteract supply chain dynamics, especially in the current era with advanced

information technology. While sharing inventory information among supply

chain members would certainly help to improve a supply chain’s performance,

the scale of full information sharing is rather complex because the scope of

such integration is quite wide. This is because most of the supply chain

members are probably located in different countries as outsourcing is now not

uncommon. In this paper, a cascade information sharing approach is proposed

in multi-echelon supply chains in contrast to full information sharing. The

principle behind the proposed approach is to allow information sharing only to

a member’s immediate upstream member on its inventory information.

Simulation is employed to test the effectiveness of the proposed approach and

results indicate that the proposed approach is still able to minimise the problem

of supply chain dynamics. The achievement is even close to the full

information sharing approach subject to various service levels.

Keywords:

supply chain; information sharing; coordination; information

technology.

Reference

to this paper should be made as follows: Chan, F.T.S. and

Chan, H.K. (2009) ‘Effects of cascade information sharing in inventory and

service level in multi-echelon supply chains’,

Int. J. Business Performance and

Supply Chain Modelling,

Vol. 1, No. 1, pp.1–7.

Biographical notes:

Felix Chan received his BSc in Mechanical Engineering

from Brighton Polytechnic (now University), UK and obtained his PhD in

Manufacturing Engineering from the Imperial College of Science and

Technology, University of London, UK. Prior to joining The University of

Hong Kong, he was a Senior Lecturer at the School of Manufacturing and

Mechanical Engineering, University of South Australia. He is now an Associate

Professor at the Department of Industrial and Manufacturing Systems

Engineering, The University of Hong Kong. His current research interests are

logistics and supply chain management, distribution coordination, systems

modelling and simulation and supplier selection.

Copyright © 2009 Inderscience Enterprises Ltd.



2

F.T.S. Chan and H.K. Chan

H.K. Chan is a Lecturer in Operations, Logistics and Supply Chain

Management at Norwich Business School, University of East Anglia, UK. He

received his Bachelors in Electrical and Electronic Engineering, his MSc(Eng)

in Industrial Engineering and Industrial Management and his PhD from the

University of Hong Kong. He also received his BSc in Economics and

Management from London School of Economics and Political Science,

University of London. His research focuses on supply chain modelling and

management in uncertain environment, and applications of artificial

intelligence and soft computing in supply chains and advanced industrial

systems.

1

Introduction

A supply chain usually consists of members in multi-echelon configuration and they are

independent companies in most cases. Information of order often distorts along the chain.

This distortion amplifies towards the upstream members in the chain and then can make

the demand looks more fluctuated and hence unpredictable, even if the actual

consumption may be very stable. This is the famous bullwhip effect (Lee et al., 1997a).

This phenomenon results in excessive inventories which increases the operating cost of

the supply chain. Information sharing has been regarded as one of the approaches to

tackle this problem (e.g., Chen, 1998; Cachon and Fisher, 2000). However, achieving full

information in multi-echelon supply chains is not easy because the scope of integration is

quite wide and sometimes it may be even infeasible because of various barriers (like

trust, confidentiality of information, etc.) of setting up inter-organisation information

systems. In this connection, supply chain management ‘emphasises close coordination

among the various companies which are involved in the chain’ (Paik and Bagchi, 2007).

As mentioned above, one of the most common forms of supply chain dynamics is the

bullwhip effect (Lee et al., 1997a). Small changes in demand at the customer end of a

supply chain will amplify towards the upstream members of the supply chain. Because of

that, companies at different echelons in the supply chain may wrongly interpret the

market demand. As a consequence, excessive inventory will build up and the situation is

even worse to upstream members. Lee et al. (1997b), after coining the term bullwhip

effect, claim that information sharing is an effective means to reduce the bullwhip effect

in supply chains: ‘one remedy (to the bullwhip effect problem) is to make demand data at

a downstream site available to the upstream site’. In addition, several researchers

supported his finding (e.g., Chen et al., 2000; Disney and Towill, 2003). By sharing

information, visibility among the echelons in the supply chain can be improved so that

coordination among echelons can be achieved (Chan and Chan, 2009). However, Chen et

al. (2000) found that ‘the bullwhip effect only can be reduced but not completely

eliminated’.

It is easy to conclude from the literature that the information sharing in the supply

chain is valuable. However, full information sharing may not always be possible and

partial information sharing may do well as compared to full information sharing

(Lau et al., 2002; Lau et al., 2004). In addition, upstream members usually gain more

benefit than downstream members if the downstream demand data is being shared (e.g.,

Chiang and Feng, 2007). In this connection, this paper proposes a partial information

sharing approach, namely cascade information sharing approach and compares the



Effects of cascade information sharing in inventory and service level

3

effectiveness of which to a full information sharing counterpart through a simulation

study. Unlike full information sharing, the proposed approach allows sharing inventory

information of a member only to its immediate upstream member. In other words, a

multi-echelon supply chain is broken into a number of overlapping two-stage supply

chain. The effect of this approach, as compared to the case with full information sharing,

is tested by a simulation study. The rest of this paper is organised as follows: Section 2

presents the research methodology, the simulation models and discusses the simulation

results. Section 3 concludes this paper.

2

Research methodology and results

2.1 Simulation models

Simulation is employed to compare the performance of the proposed (partial) information

sharing approach to a full information sharing counterpart on a multi-echelon supply

chain. In addition, the basic model without information sharing is also built as a form of

validating the results. Software package employed to build the models in this study is

Arena (Kelton et al., 2007). Demand input to the supply chain is normally distributed. For

each set of simulation experiment, ten random seeds were employed and the average

readings are recorded in order to reduce the error due to random error.

Figure 1

Simulation models, (a) a multi-echelons supply chain – the basic model

(b) multi-echelons supply chain with full information sharing (c) multi-echelons supply

chain with cascade information sharing

(Q

M

, L

M

)

(Q

W

, L

W

)

(Q

DC

, L

DC

)

(Q

R

, L

R

)

Supplier

Manufacturer

Wholesaler

Distribution

centre

Retailer

Demand

(a)

(Q

M

, L

M

)

(Q

W

, L

W

)

(Q

DC

, L

DC

)

(Q

R

, L

R

)

Supplier

Manufacturer

Wholesaler

Distribution

centre

Retailer

Demand

(b)

Notes:

Flow of order

Flow of material

Information sharing



4

Figure 1

F.T.S. Chan and H.K. Chan

Simulation models, (a) a multi-echelons supply chain – the basic model

(b) multi-echelons supply chain with full information sharing (c) multi-echelons supply

chain with cascade information sharing (continued)

(Q

M

, L

M

)

(Q

W

, L

W

)

(Q

DC

, L

DC

)

(Q

R

, L

R

)

Supplier

Manufacturer

Wholesaler

Distribution

centre

Retailer

Demand

(c)

Notes:

Flow of order

Flow of material

Information sharing

Sample ARENA models for the retailer, distribution centre and manufacturer

(see online version for colours)

Figure 2



Effects of cascade information sharing in inventory and service level

5

As mentioned above, three simulation models were developed to represent different

information sharing models of a multi-echelon supply chain, which consists of five

echelons: retail, distribution centre, wholesaler, manufacturer and supplier. The three

models under study are depicted in Figure 1. Figure 2 shows some sample models that

were extracted from the ARENA environment.

For the basic model (Figure 1a), the retailer faces a customer demand stream and uses

a continuous review (Q

R

, L

R

) inventory control policy. Under this policy, a replenishment

of quantity Q

R

is ordered from the distribution centre whenever the inventory position

down-crosses level L

R

. Since the policy is standard bookwork knowledge, description of

the operations is omitted in this paper. If the distribution centre has sufficient inventory

on hand, then the order lead time consists only of transportation delay. But, if the

distribution centre has stock-out, there is an additional delay due to transportation delays

problem and possible stocks-out which happen in the upstream of the supply chain and

they have to order from the next level. Any excess demand at the retailer that cannot be

immediately satisfied from on-hand inventory is lost.

The demand stream at the distribution centre consists of orders from the retailer.

However, unlike the retailer, the unsatisfied demand is backordered. Similar to the

retailer, the distribution centre replenishes the stocks from the manufacturing plant, again

based on a (Q

DC

, L

DC

) continuous-review policy but based only on the ordering

information from retailer. The unsatisfied portions of orders place as the backorder for

the plant. The same building block repeats for the other stages (upstream) of the supply

chain as shown in Figure 1.

The full information sharing model [Figure 1(b)] is a variant of the basic form.

Operations are more or less similar and the only difference is that: end customer

inventory information is shared to every echelon from the downstream to the entire

supply chain as shown in Figure 1(b). By using the method as discussed in Lee et al.

(2000), the reorder policy can be improved. Unlike full information sharing, the cascade

information sharing model [Figure 1(c)] means that supply chain members share their

information only to their immediate upstream member. Therefore, each echelon can still

refine the reorder policy, only subject to the inventory information of the immediate

downstream member. In other words, a multi-echelon supply chain is broken into a

number of overlapping two-stage supply chain, which is exactly the configuration of the

supply chain studied by Lee et al. (2000).

2.2 Results

Different set of service levels, i.e., different reorder levels for the basic model are

simulated. Results of some of them are listed in Table 1. Average inventory positions of

each echelon are listed as well. For example, average inventory of the retailer,

distribution centre, wholesaler and manufacturer in basic model in Table 1(a) are 9.449,

18.9833, 26.2423 and 16.8994 respectively. It can be seen that the average number of

inventory is dramatically increasing along the supply chain from downstream to

upstream, which is in line with the bullwhip effect. Same pattern can be observed from

different settings and different information sharing modes. Another observation is that the

extent of the bullwhip effect is less prominent in the two information sharing models,

although the effect could not be eliminated [same conclusion from Chen (1998)].



6

Table 1

F.T.S. Chan and H.K. Chan

Simulation results

(a) High service level

Variables

Retailer average inventory

DC average inventory

Wholesaler average inventory

Manufacturer average inventory

Variables

Retailer average inventory

DC average inventory

Wholesaler average inventory

Manufacturer average inventory

Variables

Retailer average inventory

DC average inventory

Wholesaler average inventory

Manufacturer average inventory

Model A

9.4449

18.9833

26.2423

16.8894

(b) Medium service level

Model A

8.5448

18.7535

23.9496

16.9286

(c) Low service level

Model A

7.5576

18.8542

27.0824

16.8344

Model B

7.9538

11.3787

16.1957

17.9568

Model B

7.0110

10.4005

17.7019

18.3432

Model B

6.9716

10.8037

16.1558

16.6231

Model C

8.0019

10.7242

16.9478

19.2073

Model C

7.0479

10.9860

16.1472

20.2515

Model C

6.9850

11.9865

15.2302

19.3025

Notes: Model A = Basic model without information sharing

Model B = Full information sharing

Model C = Cascade information sharing.

The last two columns of each table summarise the main contribution of this paper: a

comparison between the proposed information sharing approach and the one with full

information sharing. It can be concluded that although the reduction in average inventory

levels of various supply chain members of the cascade information sharing approach is

less than the full information counterpart, the difference is relatively less significant. This

also means that if the scope of integration is taken into consideration, full information

sharing may not be the best option as compared to the cascade information sharing one.

3

Conclusions

With the help of a simulation study, this research confirmed that information sharing

could successfully alleviate the bullwhip effect. In addition to this, a partial information

sharing approach, namely the cascade information sharing is proposed. Its effectiveness,

as compared to the full information sharing counterpart, is simulated. Based on the result,

the bullwhip effect can also be significantly minimised and performance is close to the

full information sharing counterpart. However, the cost associate with information

sharing or investment in information technology has yet to be considered.

In the long run, it is worth investing for information sharing since the investment can

be covered by various cost reduction and can probably improve customer service level. In

reality, however, it is difficult, if not impossible, to implement such a complex



Effects of cascade information sharing in inventory and service level

7

inter-organisational information system for full information sharing in a dynamic

environment. Another reason that contributes to the difficulty is that traditional research

mainly focuses only on the benefits of information sharing and ignores the cost (both

investment costs and operation costs) to achieve information sharing. Therefore, partial

information sharing is recommended since the result of reducing the bullwhip effect is

also significant and the cost of implementing it is certainly less expensive than the full

information sharing counterpart.

References

Cachon, G. and Fisher, M. (2000) ‘Supply chain inventory management and the value of shared

information’,

Management Science,

Vol. 46, No. 8, pp.1032–1048.

Chan, H.K. and Chan, F.T.S. (2009) ‘Effect of information sharing in supply chains with

flexibility’,

International Journal of Production Research,

Vol. 47, No. 1, pp.213–232.

Chen, F. (1998) ‘Echelon reorder points, installation reorder points, and the value of centralized

demand information’,

Management Science,

Vol. 44, No. 12, pp.221–234.

Chen, F., Drezner, Z., Ryan, J.K. and Simchi-Levi, D. (2000) ‘Quantifying the bullwhip effect in a

simple supply chain: the impact of forecasting, lead-times and information’,

Management

Science,

Vol. 46, No. 3, pp.436–443.

Chiang, W.K. and Feng, Y. (2007) ‘The value of information sharing in the presence of supply

uncertainty and demand volatility’,

International Journal of Production Research,

Vol. 45,

No. 6, pp.1429–1447.

Disney, S.M. and Towill, D.R. (2003) ‘On the bullwhip and inventory variance produced by an

ordering policy’,

OMEGA,

Vol. 31, pp.157–167.

Kelton, W.D., Sadowski, R.P. and Sturrock, D.T. (2007)

Simulation with Arena,

McGraw-Hill,

New York.

Lau, J.S.K., Huang, G.Q. and Mak, K.L. (2002) ‘Web-based simulation portal for investigating

impacts of sharing production information on supply chain dynamics from the perspective of

inventory allocation’,

Integrated Manufacturing Systems,

Vol. 13, No. 5, pp.345–358.

Lau, J.S.K., Huang, G.Q. and Mak, K.L. (2004) ‘Impact of information sharing on inventory

replenishment in divergent supply chains’,

International Journal of Production Research,

Vol. 42, No. 5, pp.919–941.

Lee, H.L., Padmanabhan, V. and Whang, S. (1997a) ‘The bullwhip effect in supply chains’,

Sloan

Management Review,

Vol. 38, No. 3, pp.93–102.

Lee, H.L., Padmanabhan, V. and Whang, S. (1997b) ‘The bullwhip effect in supply chains’,

Sloan

Management Review,

Vol. 38, No. 3, pp.93–102.

Lee, H.L., So, K.C. and Tang, C.S. (2000) ‘The value of information sharing in a two-level supply

chain’,

Management Science,

Vol. 46, No. 5, pp.626–643.

Paik, S.K. and Bagchi, P.K. (2007) ‘Understanding the causes of the bullwhip effect in a supply

chain’,

The International Journal of Logistics Management,

Vol. 35, No. 4, pp.308–324.