ARTICLE IN PRESS

Int. J. Production Economics 113 (2008) 587–597

www.elsevier.com/locate/ijpe

Mitigating the bullwhip effect by ordering policies and

forecasting methods
David Wright!, Xin Yuan
Telfer School of Management, University of Ottawa, 55 Laurier E, Ottawa, ON, Canada K1N 6N5
Received 1 December 2006; accepted 18 October 2007
Available online 10 March 2008

Abstract

The ‘‘bullwhip’’ effect, in which order variability increases as one moves up the supply chain, has been observed in a
range of industries, modeled by several authors and various remedies suggested. This paper provides a simulation of the
effect of improved forecasting methods, and finds that Holt’s and Brown’s methods substantially mitigate the bullwhip
effect across a range of performance metrics. The end result is to identify ordering policies that perform particularly well in
combination with these forecasting methods and indicate how they can be implemented in practice.
r 2008 Elsevier B.V. All rights reserved.

Keywords: Supply chain management; Bullwhip effect; Forecasting; Ordering policy; Simulation model.

1. Introduction The causes of the bullwhip effect can be divided

The ‘‘bullwhip effect’’, is an important phenom-
enon in supply chain management, in which the
order variability increases as one moves up the
supply chain. It has been observed in the commer-
cial operations of Campbell’s Soup (Fisher et al.,
1997), HP and Proctor & Gamble (Lee et al., 1997),
a clothing supply chain (Disney and Towill, 2003a),
and Glosuch (McCullen and Towill, 2000). The
bullwhip effect causes instability in the supply
chain, since a small change in orders received by a
retailer can result in larger changes in the resulting
orders received by a factory. It costs money, wastes
resources and results in a loss of market share.
!Corresponding author. Tel.: +1 613 562 5800x4784;
fax: +1 613562 5327.
E-mail addresses: dwright@uottawa.ca (D. Wright),
david.x.yuan@gmail.com (X. Yuan).
into two groups: time lags and planning and
behavioral aspects (Nienhaus et al., 2003). In the
first group, the remedy options include compressed
time delays (Blackburn, 1991), reduced lead time
(Metters (1997); Lee et al., 1997), echelon elimina-
tion and cycle time compression (McCullen and
Towill, 2000). In the second group, five major
causes of the bullwhip effect are identified by Lee
et al. (1997): lead time, demand signal processing,
order batching, price fluctuations, and rationing
and shortage gaming (flywheel effect). Of these
Disney and Towill (2003b) consider lead time and
demand signal processing to be of particular
importance. Remedies include synchronizing capa-
cities and lead times (Lee et al., 1997; Towill, 1997),
increased coordination among companies (Metters,
1997), vendor-managed inventory (Disney and
Towill, 2003b) and including demand variability in
pricing decisions (Naish, 1994).

0925-5273/$ - see front matter r 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2007.10.022
 ARTICLE IN PRESS
588 D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597
In terms of management science techniques, Yao
and Dong-Qing (2001) indicates that demand
forecasting and ordering policies are two key
methods of controlling the bullwhip effect and Paik
and Seung-Kuk (2003), in a statistical study,
identified demand forecasting as one of the sig-
nificant variables for bullwhip control. Miyaoka
and Hausman (2004) also found that improved
forecasting could reduce fluctuations in manufac-
turing production levels.
Based on this research, we focus on two factors:
demand forecasting and ordering policies. Moving
average and single exponential smoothing methods
have been used by Graves (1999) and Chen et al.
(2000a, b). The first aim of the current paper is to
use more sophisticated forecasting methods, includ-
ing Holt’s method and Brown’s double-exponential
smoothing, DES, method.
Most authors use ‘‘order up-to’’ policy, including
Chen et al. (2000a, b) and Lee et al. (2004). Order
rate is decided by requisitions received from the
upstream trading partner, the gap between desired
and actual inventory, and delay in inventory. By
changing the definition of desired and actual
inventory, Kohli (2005) gave a new equation for
ordering policy including lead time, safety stock,
and order frequency, and Sterman (1989) applied
generic stock acquisition and an ordering heuristic
in his model.
The present paper is based on modifications to
Sterman’s model to study how different ordering
policies and forecasting techniques, either separately
or in combination, can control the bullwhip effect.
We identify a range of ordering policies for which
the bullwhip effect can be alleviated by using either
Holt’s or Brown’s forecasting method.
The remainder of the paper is organized as
follows. Section 2 describes the basic model and
how we have extended it to incorporate Holt’s and
Brown’s forecasting methods. Section 3 describes
the simulation methodology and the metrics used to
measure the extent of the bullwhip effect. It also
presents the results, identifying which ordering
policies can benefit from Holt’s and Brown’s
Order Order
Whole
Customer Retailer
Saler
Goods Goods
Fig. 1. Production–distribution system.
forecasting procedures. Section 4 describes how
these results can be used in practice to alleviate the
harmful effects of the bullwhip effect in supply
chain management.
2. Methods and models
Our model is based on the four-stage supply
chain, shown in Fig. 1. Orders for goods propagate
upstream from left to right, and goods are shipped,
downstream, in an opposite direction. We use a
Basic Model (BM) based on Sterman’s model, with
a customer and four trading partners: retailer,
wholesaler, distributor, and factory. Each trading
partner has its own stock management system and
ordering decision system. Extending the BM, we
build three other models: BM using moving average
forecasting (BM+MA), BM using Holt’s forecast-
ing (BM+Holt’s), and BM using DES or Brown’s
forecasting (BM+Brown’s).
Fig. 2 shows the flow relationships between two
trading partners. The order decision system comes
from expected demand (EDt), adjustment of
stock (ASt), and adjustment of supply line (ASLt).
Orders (Ot) increase the supply line while shipment
received (SRt) decreases the supply line. Goods flow
out of stock to decrease inventory while shipment
received (SRt) flows into the stock to increase
inventory.
2.1. Bullwhip model for retailer, wholesaler, and
distributor
Since the business process model for the factory is
different from other trading partners, the model for
retailer, wholesaler, and distributor is described
first. In Section 2.1.1, a basic conceptual model is
defined and in Section 2.1.2 we modify it to give a
computationally efficient formulation.
2.1.1. Basic conceptual model definition
The equations in Tables 1 and 2 formalize
heuristics based on the Sterman’s model (Sterman,
1989).
Order Order
Distributor Factory
Goods Goods
 ARTICLE IN PRESS
D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597 589

Thursday, July 20, 2006

Wholesaler

SL t St
SRt
Supply Goods
Stock
Line

Ot
ASLt

Order Ot ASt Retailer

SL t St
SR t
Supply Goods
Stock
Line
Ot
ASLt AS t
ED t Order Ot

OR t EDt

Fig. 2. Flow diagram for the heuristic stock management.

Table 1 Table 2
Original conceptual model Variables definition
Ot ¼ MAXð0; IOt Þ (1)
Variables Specification Initial
condition
IOt ¼ EDt þ ASt þ ASLt (2)
Ot Order quantity that trading O0 ¼ 0
ASt ¼ aS ðS ! & St&1 Þ, (3) partners place, should be non-
negative, which means trading
ASLt ¼ aSL ðSL! & SLt&1 Þ, (4) partners cannot cancel or reverse
order.
St ¼ St&1 þ SRt&1 & Lt&1 (5) IOt The indicated order rate is the
order the trading partners will
SLt ¼ SLt&1 þ IOt & SRt (6) place, which is based on the
anchoring and adjustment
EDt ¼ y ORt&1 þ ð1 & yÞEDt&1 ; 0pyp1 (7) heuristic (Tversky and Kahneman,
1974, pp. 1124–1131). In the
model, the anchor is a forecasted
demand at each time step, and we
2.1.2. Computationally efficient formulation adjust it based on the status of
stock and supply line.
The desired stock S ! and supply line SL! are
ORt Order received at time t. The
assumed to be constant. We define b ¼ aSL/aS and
quantity the downstream trading
S 0 ¼ S ! þ b $ SL! , yielding partner requests.
EDt EDt is the expected demand rate at ED0 ¼ 0
IOt ¼ EDt þ aS ðS 0 & St & b $ SLt Þ (8)
time t. We forecast the expected
Since S ! X0, SL! X0, aSLX0 and aSX0, therefore demand rate from the historical
0 demand from the downstream
S X0. Trading partners generally pay more atten-
trading partner. In (7), the expected
tion to the inventory than the supply line, so demand comes from a single-
that,aSLpaS, implying 0pbp1. exponential smoothing method.
The simulation is initialized in equilibrium. Each ASt Adjustment to correct
inventory contains 12 units S0 ¼ 12. discrepancies between the desired
 ARTICLE IN PRESS
590 D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597

Table 2 (continued ) Table 3

Basic model for retailer, wholesaler, and distributor
Variables Specification Initial
condition Ot ¼ MAXð0; IOt Þ (1)

and actual stock. The adjustment IOt ¼ EDt þ aS ðS0 & St & b $ SLt Þ (8)
for the stock creates a negative
feedback loop, which regulates the b ¼ aSL =aS (9)
stock. For simplicity the
adjustment is linear. S0 ¼ S! þ b $ SL! (10)
!
S Desired stock.
St Actual stock or inventory at time S0 ¼ 12 St ¼ St&1 þ SRt&1 & Lt&1 (5)
t. Inventory at time t is the
inventory at time t&1 plus goods SLt ¼ SLt&1 þ IOt & SRt (6)
received at time t&1 from
upstream trading partner minus EDt ¼ y $ ORt&1 þ ð1 & yÞEDt&1 ; 0pyp1 (7)
goods sent to downstream trading
partner at time t&1.
aS The fraction of the discrepancy 0paSp1
ordered in each period.
line is the volume of products ordered but not yet
ASLt Adjustment to correct
discrepancies between the desired produced. We assume it takes 3 time steps to
and actual supply line. As with produce goods, so that Eq. (5) in Table 3 becomes
ASt, the adjustment is linear and
creates a negative feedback loop. St ¼ S t&1 þ IOt&3 & Lt&1 (11)
SL! The desired supply line.
SLt The actual supply line records the SL0 ¼ 8
volume of goods that has been 2.3. BM+MA: basic model using moving average
ordered but not received from its forecasting
supplier. The supply line at time t
is the supply line at time t&1 plus
new orders placed by the trading Our first modification to Sterman’s model is to
partner itself minus the orders use the moving average forecasting method so that
fulfilled by supplier at time t&1. (7) becomes Eq. (12). We define n ¼ 6 in our
The initial condition for SL is set simulation model.
equal to twice the standard
deviation in customer demand. EDt ¼ ½ORt&1 þ ) ) ) þ ORt&n *=n (12)
aSL The fractional adjustment rate for 0paSLp1
the supply line.
SRt Shipment received from upstream 2.4. BM+Holt’s: BM using Holt’s forecasting
trading partners to increase stock method
or inventory at time t.
Lt Loss of the inventory, i.e. the When Holt’s forecasting method (Hanke and
amount of goods shipped to the
downstream trading partner. Lt of Reitsch, 1991, p. 128) is used, (7) is replaced by
a given trading partner is equal to EDt ¼ At&1 þ Bt&1 (13)
the SRt+1 of its downstream
trading partner. where
At ¼ l $ ORt þ ð1 & lÞðAt&1 þ Bt&1 Þ (14)
Using b the model can be efficiently simulated Bt ¼ gðAt & At&1 Þ þ ð1 & gÞBt&1 (15)
using the equations in Table 3.
At: smoothed value,
2.2. Bullwhip model for the factory l: smoothing constant for the data (0plp1),
g: smoothing constant for trend estimate
As the last node of the supply chain, the factory (0pgp1), and
does not place orders, but produces goods itself, and Bt: trend estimate.
delivers them directly to its own warehouse. In this
case the orders represent the volume of goods While forecasting the future demand using Holt’s
ordered from the production facility. The supply method, we minimize the mean-square error (MSE)
 ARTICLE IN PRESS
D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597 591

in order to determine the parameters l and g, based Customer Order Pattern One
50
on historical data. 45 Customer Order
40
2.5. BM+Brown’s: BM using Brown’s DES 35
forecasting method 30
25
When Brown’s DES method (Hanke and Reitsch, 20
15
1991, p. 128), is used, (7) is replaced by
10
EDt ¼ at&1 þ bt&1 (16) 5
0
where 1 10 19 28 37 46 55 64 73 82 91 100
At ¼ ab $ ORt þ ð1 & ab ÞðAt&1 Þ (17) Fig. 3. Retail customer demand pattern.

A0t ¼ ab $ At þ ð1 & ab ÞA0t&1 (18)

Order
140
at ¼ 2At & A0t (19) Retailer
120
ab Whole Saler
100
bt ¼ ðAt & A0t Þ (20) Distributor
1 & ab 80
Factory
60
A t: exponentially smoothed value of ORt at Customer
40
time t,
A 0 t: double exponentially smoothed value of 20
ORt at time t, 0
ab: smoothing constant, 1 9 17 25 33 41 49 57 65 73 81 89 97
at: difference between the exponentially Fig. 4. Example of orders placed by each trading partner
smoothed values, and illustrating the bullwhip effect.
bt: trend estimate.

Table 4
While forecasting the future demand using Brown’s Cost definition
method, we minimize the MSE in order to determine
Cost name Variable Definition
the parameter, ab, based on historical data.
Inventory CostI Holding cost in units of dollars
3. Simulation experiments and results cost per unit item per unit time
period.
Shortage cost CostS Penalty cost for lacking goods
3.1. Simulation experiments procedure in units of dollars per unit item
per unit time period.
The purpose of our research is to investigate the Fix cost for CostF Fixed cost for each order in
impact of ordering policies and forecasting methods order units of dollars per order.
Purchase cost CostP Transaction based cost for
on the bullwhip effect. Therefore, we conduct
goods in units of dollars per
simulations in ordering policy space, i.e. (a,b) space, unit.
from (8), combined with the four forecasting
techniques, described above. (For simplicity we
use a ¼ aS.) Fig. 3 shows the retail customer
demand pattern generated using local trends mod- computed from the current stock minus the order
ified by random disturbances taken from a normal backlog; and the cost. The cost for each trading
distribution with mean 0 and variance 4. Fig. 4 partner is defined by its inventory level and
shows an example of the bullwhip effect that this transaction volume according to (21) and (22).
data creates with a ¼ 0.4, b ¼ 0.3. Table 4 lists related cost definitions. In (21), IsOrder
We collect three data series from each simulation is 1 if there is a current order from its customer;
run: the order volume; the inventory level, which is otherwise, IsOrder equals 0.
 ARTICLE IN PRESS
592 D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597

costt ¼ costI $ S t þ costS $ backLogOrdert + The variance of order rate, which has also been
þ costF $ IsOrder þ costP $ IOt (21) used by Chen et al. (2000, pp. 436–443) and Lee
et al. (2004, pp. 1875–1886).
where + The root mean square of the discrepancies
backLogOrdert ¼ backLogOrdert&1 þ ORt & St between the trading partners order and the retail
customers original demand.
backLogOrder0 ¼ 0 + The total cost according to (22).

X
100
Fig. 5 illustrates the procedure in the simulation
Total Cost ¼ cos tt (22) experiments, where ordering policy represented by a
t¼1
and b uses increments of 0.1 between 0.1 and 1.0.
For each pair of (a,b), we have four forecasting
3.1.1. Simulation in ordering policy space options: Sterman’s adaptive forecasting method, the
In order to evaluate simulation results system- moving average forecasting method, Holt’s method,
atically, we identify three performance measures of and Brown’s DES method. Fig. 5 summarizes
the model, to measure the extent of the bullwhip whether the bullwhip effect can be alleviated by
effect as follows. using Holt’s and DES compared with moving

Simulation Results in ordering policy space

Y-Axis --Beta--

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

X-Axis --Alpha--

Bullwhip Effect can be alleviated by

Holt's and Brown's forecasting
methods

Fig. 5. Effect of ordering policy on bullwhip effect.

 ARTICLE IN PRESS
D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597
average and Stermans’s adaptive forecasting meth-
od. If the variance of order rate, RMSE and cost,
are reduced, as described in detail in Section 3.1.2,
we use a dark shading to indicate that the bullwhip
effect has been alleviated. In general, these are
concentrated in the lower left of the grid, where a
and b are small, which is consistent with the
observation of Mosekilde et al. (1991) that
the supply chain is more stable in this region. The
bullwhip effect was not alleviated in the light-
shaded region of Fig. 5.
3.1.2. Criteria for optimal ordering policy
identification
Our three metrics, variance, RMSE and cost, are
illustrated in Figs. 7–9 for one ordering policy. It
can be seen that each metric is improved by both
Holt’s method and Brown’s method. This was
observed in general for other ordering policies.
Since Holt’s and Brown’s methods behave similarly
in this respect, we summarize the extent of the
improvement by
varHolts=varBM þ varHolts=varBMMA þ varBrowns=varBM þ varBrowns=varBMMA
varP ¼
4
rmseHolts=rmseBM þ rmseHolts=rmseBMMA þ rmseBrowns=rmseBM þ rmseBrowns=rmseBMMA
rmseP ¼
4
costHolts=costBM þ costHolts=costBMMA þ costBrowns=costBM þ costBrowns=costBMMA
costP ¼
4
using variables defined in Tables A1, A2, and A3 in
the Appendix A.
It is also clear from Figs. 7 to 9, that when one of
our metrics is improved, the other two are also
improved. This was observed in general for other
ordering policies. We are therefore able to define an
overall performance metric
varP þ rmseP þ costP
totalP ¼ (23)
3 forecasting is used, for one specific ordering policy:
From the functions above, the smaller the costP,
varP, rmseP, and totalP are, the better is the
performance of BM+Holt’s and BM+Brown’s
models over the other two models. The result of
optimal ordering policy identification is given in
Fig. 6. The darker the grid, the better is the
performance; for instance, the value in the grid
(0.2,0.4) is 0.456 that is better than that of grid
(0.4,0.5), which is 0.815.
593
3.1.3. Experimental results
This section illustrates sample simulation results
for one case a ¼ 0.2 and b ¼ 0.3, in order to
give an example of the behavior summarized in
Figs. 5 and 6.
Fig. 7 shows the variance of order rate; Fig. 8
shows the RMSE, and Fig. 9 shows the cost.
Fig. 10 shows the order rate of BM; Fig. 11
shows the order rate of BM+MA; Fig. 12
shows the order rate of BM+Holt’s, and Fig. 13
shows the order rate of BM+Brown’s. A visual
inspection of Figs. 10–13 indicates a clear
reduction of the bullwhip effect if Holt’s or Brown’s
forecasting method is used, which is confirmed
by each of the 3 quantitative measures shown in
Figs. 7–9.
3.2. Results analysis
3.2.1. Forecasting method analysis
We now focus on the shaded areas of Fig. 6
(corresponding to low values of a and b) in
which the ordering policy can be improved
by using Holt’s or DES forecasting compared
with Sterman’s adaptive forecasting. High
values of a and b give higher values of variance,
RMSE and cost, and are therefore inappropriate
in practice, independent of the forecasting method
used.
Fig. 7 shows that the variance of order rate at the
factory is alleviated by about 65% if Holt’s or DES
a ¼ 0.2 and b ¼ 0.3. Table 5 shows the variance
performance measure ‘‘varP’’; for all ordering
policies where Holt’s and DES are beneficial.
All the values are substantially o1, so the result
in Fig. 7 applies to a range of ordering policies.
Tables 6 and 7 show similar results for RMSE and
cost. The total performance measure, totalP, (23)
can be found in Fig. 6, again indicating a similar
result. We can therefore conclude that Holt’s and
 ARTICLE IN PRESS
594 D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597

Y-Axis --Beta--

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0
1.0

0.9

0.8

0.7

0.6 0.592 0.781

0.5 0.474 0.655 0.815

0.4 0.456 0.607

0.3 0.451 0.831

0.2 0.482 0.744 0.553 0.609 0.716 0.609

0.1

X-Axis --Alpha--

Fig. 6. Optimal ordering policy identification, with TotalP from Eq. (23).

Root mean square of the errors’ of Order

Trading Partner View
Variance of Order-Trading Partner View 35
1200
30 BM
BM BM+MA
1000 BM+MA 25 BM+Holts
BM+Holts BM+Browns
800 20
BM+Browns
600 15

400 10

5
200
0
0 Retailer Whole Distributor Factory
Customer Retailer Whole Distributor Factory Saler
Saler
BM 7.038542 13.423755 21.148637 29.970608
BM 104.64434 149.99374 322.40219 651.23542 1106.8314

BM+MA 104.64434 139.56251 304.88606 638.91504 1105.6556 BM+MA 6.6784208 13.227441 21.818073 31.367546

BM+Holts 104.64434 107.49647 126.47865 171.807 284.21646 BM+Holts 5.2482415 6.7339632 9.4617702 14.124073

BM+Browns 104.64434 108.71322 123.46605 191.10868 361.51595 BM+Browns 5.3108915 6.9778109 10.903824 16.997989

Fig. 7. Variance of order rate for a ¼ 0.2 and b ¼ 0.3. Fig. 8. RMSE for a ¼ 0.2 and b ¼ 0.3.
 ARTICLE IN PRESS
D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597 595

Cost Order
60000 70
BM 60 Retailer
50000 BM+MA Whole Saler
50 Distributor
BM+Holts
40000 Factory
BM+Browns 40
Customer
30
30000
20
20000
10
10000 0
1 9 17 25 33 41 49 57 65 73 81 89 97
0
Retailer Whole Distributor Factory Total Cost Fig. 12. Order rate using BM+Holt’s for a ¼ 0.2 and b ¼ 0.3.
Seller

BM 8641.5439 10565.977 14450.293 18648.783 52306.596

Order
BM+MA 8716.8078 11068.65 15319.222 19764.638 54869.317
80
BM+Holts 6209.5031 6840.3063 7819. 0611 9100.1155 29968.986 Retailer
70
BM+Browns 6198.6164 6639.8185 7610.8977 9152.6507 29601.983
Whole Saler
60 Distributor
50 Factory
Fig. 9. Cost for a ¼ 0.2 and b ¼ 0.3. Customer
40
30
Order 20
140
Retailer 10
120 0
Whole Saler
100 1 9 17 25 33 41 49 57 65 73 81 89 97
Distributor
80 Factory Fig. 13. Order rate using BM+Brown’s for a ¼ 0.2 and b ¼ 0.3.
60 Customer

40
Table 5
20 Variance performance measure report
0
VarP at Value varP at Value
1 9 17 25 33 41 49 57 65 73 81 89 97
(0.2,0.2) 0.329 (0.3,0.5) 0.519
Fig. 10. Order rate using BM for a ¼ 0.2 and b ¼ 0.3.
(0.2,0.3) 0.292 (0.3,0.6) 0.675
(0.2,0.4) 0.293 (0.4,0.5) 0.745
(0.2,0.5) 0.318 (0.5,0.2) 0.454
Order
(0.2,0.6) 0.466 (0.6,0.2) 0.499
120
(0.3,0.2) 0.699 (0.7,0.2) 0.659
Retailer
100 (0.3,0.3) 0.810 (0.8,0.2) 0.494
Whole Saler (0.3,0.4) 0.475
80 Distributor
Factory
60
Customer
40 3.2.2. Ordering policy analysis
Tables 5–7 give a consistent result that small
20
values of a such as a ¼ 0.2 show good performance
0 with Holt’s and Brown’s forecasting, compared with
1 9 17 25 33 41 49 57 65 73 81 89 97 higher values such as a ¼ 0.8. For the same value of
Fig. 11. Order rate using BM+MA for a ¼ 0.2 and b ¼ 0.3.
a, e.g. a ¼ 0.2 or 0.3, a medium value of b such as
b ¼ 0.4 is preferable. Therefore, slow adjustment on
the discrepancy of stock with a slightly larger
DES have a substantial impact on reducing the adjustment on the discrepancy of the supply line is
bullwhip effect independent of the performance a best practice for alleviating the bullwhip effect.
measure. For example, ordering policies such as (a ¼ 0.2,
 ARTICLE IN PRESS
596 D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597

Table 6 which orders are designed to stabilize a combination

RMSE performance measure report of stock levels and supply line levels. It is shown that
rmseP at Value rmseP at Value
a relatively slow adjustment of stock levels, com-
bined with a slightly more rapid adjustment of
(0.2,0.2) 0.543 (0.3,0.5) 0.791 supply line levels provides the most stability when
(0.2,0.3) 0.508 (0.3,0.6) 0.899 combined with either Holt’s or Brown’s forecasting
(0.2,0.4) 0.499 (0.4,0.5) 0.957
method.
(0.2,0.5) 0.501 (0.5,0.2) 0.692
(0.2,0.6) 0.618 (0.6,0.2) 0.723 To implement these results in practice, either
(0.3,0.2) 0.849 (0.7,0.2) 0.793 Holt’s or Brown’s forecasting method should be
(0.3,0.3) 0.911 (0.8,0.2) 0.685 used to forecast demand at each level in the supply
(0.3,0.4) 0.737 chain, based on the historical demand pattern. The
parameters of the forecasting method should be
chosen so as to minimize the MSE of forecasts made
Table 7 during the historical period. Then orders should be
Cost performance measure report placed with suppliers based on the forecasted
CostP at Value Costp at Value
demand plus two adjustments for how far (a) the
stock levels and (b) the supply line are from desired
(0.2,0.2) 0.573 (0.3,0.5) 0.653 values. The difference from the desired values
(0.2,0.3) 0.556 (0.3,0.6) 0.769 should be multiplied by about 0.2 for the stock
(0.2,0.4) 0.575 (0.4,0.5) 0.741
levels and by a number in the range 0.2–0.4 for the
(0.2,0.5) 0.604 (0.5,0.2) 0.513
(0.2,0.6) 0.691 (0.6,0.2) 0.605 supply line.
(0.3,0.2) 0.683 (0.7,0.2) 0.695
(0.3,0.3) 0.773 (0.8,0.2) 0.648
(0.3,0.4) 0.607 Appendix A

See Tables A1–A3 for definitions used in calcula-

tion of performance metrics.
b ¼ 0.3), or (a ¼ 0.2, b ¼ 0.4), or (a ¼ 0.3, b ¼ 0.4),
or (a ¼ 0.3, b ¼ 0.5) work well, when combined
with Holt’s or Brown’s forecasting methods. Table A1
Definition of factory variance performance function
4. Conclusions Variable Variable meaning
name
This paper has explored the bullwhip effect in the
supply chain using simulation analysis to investigate VarP Final performance for variance
VarHolt’s (Factory variance)/(Customer variance) in
the potential benefit of improved forecasting meth-
BM+Holt’s
ods, viz., Holt’s and Brown’s methods. In all our VarBM (Factory variance)/(Customer variance) in BM
simulations, the bullwhip effect is still present, i.e. VarBMMA (Factory variance)/(Customer variance) in
the variability of order volumes increases as one BM+MA
moves up the supply chain from retailer to factory. VarBrown’s (Factory variance)/(Customer variance) in
BM+Brown’s
However, it can be significantly alleviated, by up to
55% overall, by choosing an appropriate ordering
policy and forecasting method.
Previous authors have modeled the bullwhip Table A2
effect using single-exponential smoothing and mov- Definition of factory RMSE performance function
ing average methods to forecast demand. The
Variable name Variable meaning
present paper has shown that the bullwhip effect
can be substantially reduced, and hence the supply RmseP Final performance for RMSE
chain can be stabilized, by using Holt’s or Brown’s RmseHolt’s Factory RMSE in Model BM+Holt’s
forecasting techniques. In order to achieve these RmseBM Factory RMSE in Model BM
RmseBMMA Factory RMSE in Model BM+MA
improvements, forecasting must be done in con- RmseBrown’s Factory RMSE in Model BM+Brown’s (DES)
junction with an appropriate ordering policy, in
 ARTICLE IN PRESS
D. Wright, X. Yuan / Int. J. Production Economics 113 (2008) 587–597
Table A3
Cost performance function definition
Variable name Variable meaning
CostP Final performance for cost
CostHolt’s Total cost in Model BM+Holt’s
CostBM Total cost in Model BM
CostBMMA Total cost in Model BM+MA
CostBrown’s Total cost in Model BM+Brown’s (DES)
References
Blackburn, J.D., 1991. The Quick Response Movement in the
Apparel Industry: A Case Study in Time-Compressing Supply
Chain, in Time-Based Competition: The Next Battleground in
American Manufacturing. Irwin, Homewood, IL.
Chen, F., Drezner, Z., Ryan, J.K., Simchi-Levi, D., 2000a.
Quantifying the bullwhip effect in a simple supply chain: The
impact of forecasting, lead times, and information. Manage-
ment Science 46 (3), 436–443.
Chen, F., Ryan, J.K., Simchi-Levi, D., 2000b. The impact of
exponential smoothing forecasts on the bullwhip effect. Naval
Research Logistics 47 (0894-069x), 269–286.
Disney, S.M., Towill, D.R., 2003a. The effect of vendor managed
inventory (VMI) dynamics on the bullwhip effect in supply
chains. International Journal of Production Economics 85 (2),
199–215.
Disney, S.M., Towill, D.R., 2003b. Vendor-managed inventory
and bullwhip reduction in a two-level supply chain. Interna-
tional Journal of Operations and Production Management 23
(5–6), 625–651.
Fisher, M., Hammond, J., Obermeyer, W., Raman, A., 1997.
Configuring a supply chain to reduce the cost of demand
uncertainty. Production and Operations Management 6 (3),
211–225.
Graves, S.C., 1999. A single-item inventory model for a
nonstationary demand process. Manufacturing and Service
Operations Management 1 (1), 50–61.
Hanke, J.E., Reitsch, A.G., 1991. Moving averages and smooth-
ing methods. In: Business Forecasting. Prentice-Hall, Engle-
wood Cliffs, NJ (Chapter 5).
597
Kohli, A.S., 2005. A dynamic simulation study to assess the
impact of collaboration on the performance of a supply chain.
Ph.D. Dissertation, University of Louisville.
Lee, H.L., Padmanabhan, V., Seungjin, W., 1997. The bullwhip
effect in supply chains. Sloan Management Review 38 (3),
93–102.
Lee, H.L., Padmanabhan, V., Seungjin, W., 2004. Information
distortion in a supply chain: The bullwhip effect/comments on
‘information distortion in a supply chain: The bullwhip
effect’. Management Science 50 (12), 1875–1886.
McCullen, P., Towill, D.R., 2000. Practical ways of reducing
bullwhip: The case of Glosuch global supply chain. IOM
Control 27 (1), 24–30 (December–January).
Metters, R., 1997. Quantifying the bullwhip effect in supply
chains. Journal of Operations Management 15 (2), 89–100.
Miyaoka, J., Hausman, W., 2004. How a base stock policy using
‘stale’ forecasts provides supply chain benefits. Manufactur-
ing and Service Operations Management 6 (2), 149–162.
Mosekilde, E., Larsen, E., Sterman, J., 1991. Coping with
complexity: Deterministic chaos in human decisionmaking
behavior. In: John, L.C, Anders, K. (Eds.), Beyond Belief:
Randomness, Prediction and Explanation in Science. CRC
Press, Boca Raton, Ann Arbor, Boston, pp. 119–229.
Naish, H.F., 1994. Production smoothing in the linear quadratic
inventory model. The Economic Journal 104 (425), 864–875.
Nienhaus, J., Ziegenbein, A., Duijts, C., 2003. How Human
Behaviour Amplifies the Bullwhip Effect a Study Based on the
Beer Distribution Game Online. Centre for Enterprise
Sciences (BWI), Swiss Federal Institute of Technology
(ETH), Zurich.
Paik, S.-K., 2003. Analysis of the causes of ‘bullwhip’ effect in a
supply chain: A simulation approach. Ph.D. Dissertation, The
George Washington University.
Sterman, J.D., 1989. Modelling managerial behavior: Mispercep-
tions of feedback in a dynamic decision experiment. Manage-
ment Science 35 (3), 321–339.
Towill, D.R., 1997. Forridge—Principles of good material flow.
Production Planning and Control 8 (7), 622–632.
Tversky, A., Kahneman, D., 1974. Judgement under uncertainty:
Heuristics and biases. Science 185, 1124–1131.
Yao, D.-Q., 2001. Study of bullwhip effect and channel design in
supply chains. Ph.D. Dissertation, The University of Wiscon-
sin-Milwaukee.