Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
"MANAGERIAL DECISION BEHAVIDURAND THE ESTIMATION OF
DYNAMIC SALES RESPONSE MODEL'
by
Wilfried VANHONACKF-R.
N° 90/76/MKT
* Associate Professor of Marketing, INSEAD, Boulevard de Constance,Fontainebleau, 77305 Cedex, France.
Printed at INSEAD,Fontainebleau, France.
Managerial Decision Behavior and the Estimation
of Dynamic Sales Response Models
Wilfried R. Vanhonacker"
Revised October 1990
* Professor of Marketing, INSEAD, Boulevard de Constance, 77305Fontainebleau, France.
2
Managerial Decision Behavior and the Estimation
of Dynamic Sales Response Models
ABSTRACT
It is well established that aggregated time-series data provide biased
estimates of carryover effects in dynamic sales response models. The origin of the bias is
a misspecification due to missing micro series. Supplementing the aggregate information
with the firm's decision behavior could alleviate this problem. Recent work suggests that
not considering this behavior would influence inference about carryover. This paper
elaborates on the necessity of knowing this behavior in assessing dynamic response
given temporally aggregated data. Considering a number of widely-used dynamic sales
response models, it is shown that when the carryover parameter measures only
advertising carryover, consistent estimation is entirely conditional on knowing the firm's
advertising decision behavior.
Key Words: SALES RESPONSE MODELS; CARRYOVER EFFECTS;
METHOD-OF-MOMENT ESTIMATION; DATA INTERVAL BIAS.
3
1. Introduction
Accurate measurement of the magnitude and duration of advertising effects on
sales underlies efficient and effective advertising budgeting and timing decisions. Hence,
it is not surprising that considerable efforts have been devoted to measuring these effects.
Although progress is being made, the complexity of advertising response continues to
press our modeling and measurement technology to the limits. The current attempts to
uncover response parameters given temporally aggregated data are cases in point (see,
e.g., Bass and Leone 1983, Blattberg and Jeuland 1981, Kanetkar, Weinberg and Weiss
1986, Rao 1986, Srinivasan and Weir 1988, Vanhonacker 1983, 1984, 1987, Weiss,
Weinberg and Windal 1983).
Temporal aggregation occurs frequently in practice. As researchers often have
little choice over the length of the discrete time periods over which variables are being
measured, the time-series observations available seldom coincide with the discrete time
periods characterising the response phenomenon one is modeling. This problem of fit
gives rise to the data interval bias which continues to challenge applied econometricians in
the sales response literature.
The data interval bias recognised and illustrated by Clarke (1976) refers to
biased estimation of the advertising carryover effect in temporally aggregated data l . As
has been demonstrated through explicit aggregation (Sasieni 1982, Vanhonacker 1983),
the origin of the bias lies in the fact that the aggregate analog of the disaggregate response
model is a misspecification because micro advertising terms are ignored. A variety of
modeling and/or estimation approaches have been suggested to recover the micro
response parameters (i.e., direct advertising response and carryover), often relying on
either explicit or implicit interpolation of the unobservable micro advertising series (see
Hanssens, Parsons and Schultz, 1990). Such interpolation requires knowledge on the
micro series which cannot be recovered from the aggregate discrete observations
themselves. For advertising, the firm's allocation rule at the micro (disaggregate) level
needs to be known to be able to accurately decompose the observed moving sums.
Russell (1988)'s contribution lies in showing that "... this prior information [the firm's
advertising decision behavior] is important, and ultimately dictates the quality of inference
about advertising effectiveness" (p. 268).
Once the firm's advertising decision behavior is known, explicit aggregation of
the micro response model results in a macro model which is straightforward to estimate.
What is striking, however, is that the derived specification is different depending on the
4
type of decision rule, and seldom resembles (either in variables or error structure) one of
the common specifications used in the sales response literature. As the researcher has
often incomplete or no knowledge of the firm's advertising budgeting and allocation
decisions, second guessing the rules can have serious consequences as erroneous
specifications could be relied upon. Recognizing this problem, efforts have been devoted
to developing parsimonious macro models whose parameter estimates robustly assess
advertising carryover irrespective of the underlying (unobserved) decision behavior.
Russell (1988) suggests an expanded Koyck-type model which provides reasonable
estimates of advertising carryover in a simulation framework.
This paper is not another attempt to develop a robust procedure, but elaborates on
the premise that decision behavior is needed to correctly assess carryover effects.
Looking at a number of widely-used sales response models, instances are identified
where a method-of-moment estimator can provide consistent estimates of carryover
without any knowledge of the firm's advertising decision behavior. Alternatively, the
rigorous analytic treatment suggests instances where the newly-developed robust
procedures should be most useful and why attempts should be made to enhance the
statistical properties of their estimates.
5
2. Modeling Framework
The basic sales response model used in this research is a stationary first-orderautoregressive model specified as2
s t = X. s t_ i + Pat + ut (1)
where st denotes sales at time t, and a t denotes advertising at time t. X is the carryoverparameter, and p is the direct advertising response parameter. u t denotes a random
disturbance whose properties will be discussed shortly. Throughout the paper, lower
case letters will refer to disaggregate or micro observations where upper case letters will
refer to the corresponding aggregate or macro observations.
Sales response model (1) is related to three different models all of which have
been used in applied econometric work in marketing: the Koyck model, the Partial
Adjustment model, and the Adaptive Expectations model. As these models span the area
of investigation in this research, they are discussed in some detail. Consistent with (1)
and without loss of generality, we'll suppress all intercept terms.
The Koyck model finds its origin in a distributed lag specification of sales
response to advertising, or
st = Nat+ P1at-1 + p2at-2 + + vt(2)
with E (vt) = 0, E (vtv t_ i) = cq, for i = 0 and E(viv t. i) = 0 for i 0. Assuming that the
effect of the lagged terms is a declining geometric sequence as
13i = (1-X)X, with 0<X<1, (3)
model (1) becomes a parsimonious representation of (2) with u t = vt - Xvt_ t . Hence, in a
Koyck scheme, the disturbance term u t in the basic response model has the followingproperties
E(ut) = 0,
and
6
E(utut_i) =0
= -
= 6i (1 + X.2) (4)
Modifications of (3) have been considered where the geometric process kicks in
after a number of lags. The net result is that additional lagged advertising terms show up
in (1) which is inconsequential for the results obtained here. Substantively, it is worth
noting that because of the underlying process in (2), X in (1) can be truly called the
advertising carryover parameter.
The Partial Adjustment model has a slightly different origin from the Koyck
model. It assumes that some optimal sales level is associated with current advertising, or
si = 004
(5)
However, because of ineritia, ignorance, etc., the optimal level is not achieved
immediately and sales will only move part of the way from their previous level (in the
direction of the optimum). This is captured in a reaction or adjustment function
st - s t_i T(s t* - 44) +
(6)
with 0 < 7 1 and the disturbance term v t having the properties as in (2). The closer y is
to one, the greater is the adjustment in the current period. Combining (5) and (6) results
in a model similar to (1) with X = (1-y) and, in contrast to the Koyck model, a white
noise disturbance term (i.e., u t = vt). Note that from (6), current sales can be expressed
as a function of lagged optimal sales with geometric declining weights identical to (3).
Hence, substantively the parameter X. captures dynamic adjustment in sales and measures
more than just advertising carryover.
The Adaptive Expectations model is based on the assumption that current sales
are driven not by current advertising but expected advertising at t, at Expected
advertising then captures period-to-period changes assumed to follow a process identical
to (6),
at* - at_ t = 8 (a t - at* i ) + vt (7)
7
with a < S 5 1 and v t as in (6). This expression together with s t = Poet implies a modelsimilar to (1) with X = (1-8), and u t = v t - Xv t with properties as defined in (4). It can be
shown that (7) implies
= S a t_ i + (1-8)844 + (1-8)28;_2 + + vt
or the expectations arise from a distributed lag model in actual advertising with the effects
of the lagged terms forming a geometric declining series as defined in (3). This illustrates
the close relationship between the Adaptive Expectations model and the Koyck model.
Hence, substantively the X parameter can again be accurately called the advertising
carryover parameter.
Although the three models arise from different schemes, their final specifications
show remarkable resemblance merely because each involve at some point declining.
geometric weights. Of importance for subsequent discussion are the ideosyncratic error
structures and the substantive interpretation of the lagged sales parameter X either as
advertising carryover or general carryover parameter encompassing more than advertising
response dynamics.
Sales response model (1) is the true micro response model whose parameters we
will attempt to estimate. The difficulty here is that only temporally aggregated
observations are available and none of the variables specified in (1) are observed at the
micro level t. Sales and advertising are available as moving sums. For simplicity of the
illustration and derivation (and without loss of generality), we limit ourselves to an
aggregation level of three (e.g., monthly into quarterly observations). Hence,S t= st + s t _ 1 + st _ 2 and A t =(at + at_i + a t_2). Note that because of the moving sums,S t = (St_t+st-s t-3). We observe the moving sums S3, S6, S9,... and A3, A6, A9 .... The
assumption carried through subsequent analytic work is that the exact level of aggregation
is known. In other words, given the observed moving sums, we know exactly the micro
level at which model (1) operates. This is necessary to recover the value of ?. at the micro
level. If the interest is simply in duration (or how long effects will last), the exact level
need not be known as duration formula express duration time in aggregate time units
(see, e.g., Clarke 1976).
Explicit aggregation of model (1) results in
s t = xs t _1+ p At + Ut (8)
8
or, alternatively,
S t = St .3 + (At + At + A t - 2) + ( Ut + Ut - 1 + Ut - 2 ) (9)
with U t = u t + ut + 2. With the available macro series, neither (8) nor (9) can be
estimated as both representations contain variables which are not directly observed.
Given the moving sum definition adopted above, if aggregates S t and At are observed,
then S t_ t and At_ t are not observed for i < nr where n = 1, 2, 3,... and r is the level of
aggregation. Knowing the firm's advertising decision behavior (e.g., the budgeting rule
over the macro periods and the allocation rule over the micro periods), one can derive the
observations on all variables specified in (9). Depending on the specific allocation rule,
the derived series will be different and, hence, the parameter estimates from the macro
model become ideosyncratic to the allocation rule. This illustrates the dependence of the
aggregated model upon the firm's advertising decision behavior.
Even when we know the firm's behavior, the estimation task remains
cumbersome because of a lagged dependent variable and complex error structures
involving unknown parameters. For example, OLS estimates will be inconsistent. A
slightly different estimation approach introduced next will be taken in this work. One has
to keep in mind, however, that because of the lagged dependent variable (a stochastic
regressor), at best good asymptotic properties will be achieved.
9
3. Estimation
The method-of-moment estimation approach adopted here was studied by Telser
(1967) for stationary autoregressive processes. It is extended here for stationary
autoregressive equations containing predictor variables. The estimation approach relies
on the autocovariances of the aggregate time-series observations. It is evident that given
the available data, consistent sample estimates of the autocovariances can be derived.
Since rational functions of consistent estimators are themselves consistent, it seems
reasonable to attempt to express the response parameters in the macro model as rational
functions of autocovariances.
Advertising decision behavior of two types is considered: deterministic versus
stochastic (or random). Although this basis of distinction is in line with the analytic
nature of this research, it might substantively (behavioraly) not be that appealing.
However, besides being in line with previous research in this area (Russell 1988), it will
become clear momentarily that the exact nature of the behavior (which substantively is
more interesting) beyond the analytic distinction is of no consequence to the results.
Deterministic Advertising Decision Behavior
Advertising decision behavior contains two components: the budget rule (how
much will the firm devote to advertising in a specific period of time) and the allocation
rule (how will that budget be spent over that time period). Deterministic advertising
decision behavior implies that these decision rules do not contain any random component.
Specifically, the micro series a t arise from a purely deterministic process.
We focus first on the autocovariances underlying the micro response model (1)
given the Partial Adjustment model error structure discussed in relation to (5) and (6).
Appendix 1 shows that the autocovariance between s t and s t-k equals
COV (St , St -k) = Xk 62 [1 /(1-X2 )1. (10)
The corresponding aggregated time series provide autocovariances which equal (see
Appendix 2)
St-k) = a.k a2 [11(1 -^ 2 )1 (X-2+ X-1 + 1) + X +X2),
10
which is consistent with the autocovariance generating functions discussed in Telser
(1967) and Tiao and Wei (1976). From (10) and (11), it is evident that
Coy (S t, S t_k) = [Coy (s t, st-k )](X-2+ 71-1 + 1) (1 + X + a.2)
which is an expression which can be generalised to any k and/or any level of temporal
aggregation.
Given the available observations S3, S6 , ..., etc, a consistent estimate of the
sample covariances Coy (St St-3) and Coy (S t St-6) can be obtained as
COv (S t St - = E (Si (t + 1) - (Sit -t =1
with S being the sample mean of the series, or
= (1/N) E Si t.t=
As is evident from (11), these autocovariances satisfy the following difference equation
Coy (S t S t_ 6) + q Coy (S t S t_3) = 0
(12)
Substituting the corresponding expressions from (11), we obtain that
q = - 3
It is interesting to note that since the power of X will always be equal to the level
of aggregation, duration effects can be directly derived from the q estimate without
knowing the actual level of temporal aggregation involved. This results from the fact that
duration formula measure duration in time units corresponding to the macro observations
(see, e.g., Russell 1988, p. 256).
If the carryover parameter X itself is needed, the level of aggregation is evidently
needed. Substituting the sample autocovariance estimates into expression (11), a
consistent estimate of X equals
= (-0/3 =[CoV (S t St-6 )/ CON, (S 1 So) 1/3. (13)
11
This estimate was derived without knowledge of the exact nature of the firm's advertising
decision behavior. The only assumption was that the behavior was deterministic. Hence,
the method-of-moment estimator can provide a consistent estimate of the carryover
parameter in a Partial Adjustment model without knowledge of the firm's actual decision
behavior.
Knowledge of the X parameter is not enough to interpolate the moving sums At.
Hence, although X can be estimated, the direct advertising effect (parameter fl in (1)) can
only be estimated with the explicit knowledge of the firm's decision behavior. Only with
that knowledge can we derive the observations A t_ t and A1.2 specified in (9).
When the Koyck or Adaptive Expectations scheme underly micro model (1), the
error structure is more complex than in the Partial Adjustment model. Specifically, the
properties of u t in (1) equal (4). Interestingly enough, Appendix 1 shows that the
autocovariances of the micro sales series are all equal to zero. This result carries through
in the Cov(S tS t_k). Hence, none of the parameters in (1) can be derived using the
method-of-moment estimation. The firm's decision behavior would have to be known in
order to derive micro advertising series, which subsequently could be incorporated in a
generalized least squares estimator on (9). The latter would provide consistent parameter
estimates.
Stochastic Decision Behavior
If decision behavior is random, the autocovariances are different. For illustrative
purposes, we consider the case where the advertising decision rule is first-order
autoregressive. This is the case studied by Russell (1988); furthermore, univariate time-
series analysis on disaggregated advertising series often identifies an AR (1) process.
Specifically, we assume
at = 0 a t_i + et
with E (Et) = 0, E (E t Et_ i) = 0 for i x 0, and E (E t Et_ i) o for i 0.
Given that 101 > X, Cov (at, at_k) =0k Gi {I / (1 - 02)1
and ut in (1) has the properties corresponding to the Partial Adjustment model, Appendix
1 shows that the autocovariance of the micro sales series equals
12
Cov (st, s t _k) = (f32 0k ai)/ (1 - 02) (1 - a 0) (1 - af0) + 02) / ( 1 - a.2). (14)
The first term on the right hand side of expression (14) refers to the implied
autocovariances of the advertising series and the second term (which equals (10)) refers
to the autoregressive process of the sales series themselves.
Using expression (14), Appendix 2 shows that the autocovariance of the macro
sales series equals
ok _2 _
Coy (S t , St-k) = uc(02+0 +1)(1 +0+02){1 /(1 -X.0)(1 -X/0)](1-02)
,k 2 n
+ (5 + X t + 1) (1 + X +A.2)(1-X2)
If the disturbance u t in (1) has properties equal to (4), Appendix 1 shows that the
autocovariance of the micro sales series equals the first term in expression (14). The
corresponding expression for Cov(S tS t_ k) equals the first term of (15). Hence, in this
case (corresponding to the Koyck and Adaptive Expectation schemes), the
autocovariances of both macro and micro sales are driven entirely by the autocovariance
of the advertising series given the random nature of the latter.
Given that the autocovariances in macro sales are complex non-linear functions of
parameters (despite the simplicity of the random advertising rule), parameter estimation
becomes more cumbersome. A number of different approaches can be pursued to derive
consistent estimates of the parameters in model (1). Irrespective of the avenue taken, the
firm's advertising decision behavior will have to be known to be able to interpolate the
micro advertising series. A straightforward method might be to derive a consistent
estimate of the decision rule's autoregressive parameter 0 using the method-of-moment
estimator. An expression similar to (12) in the sample autocovariances of the macro
advertising series could be relied upon. The resulting consistent estimate could then be
used to interpolate the micro advertising series which would enable the derivation of At_1
and At_2. With these observations, generalized least squares on (9) would result in
consistent estimates of the carryover and response parameter. Moreover, with random
advertising decision behavior, consistent estimation is conditional on the complete
knowledge of that decision behavior. This is the case for either the Partial Adjustment
model, the Koyck model, or the Adaptive Expectations model.
(15)
13
4. Summary and Discussion
The analytic results derived above are summarized in Table 1. As is shown there,
stochastic decision behavior requires knowledge of decision rules. Although the derived
results strictly apply to an assumed first-order autoregressive advertising decision rule,
the complexity involved illustrates that more elaborate random decision behavior is not
likely to change the conclusion. Moreover, if a random component is embedded in the
firm's advertising decision behavior, consistent estimation of carryover and direct
response effects requires knowledge of that behavior irrespective of the modeling
scheme.
With deterministic decision behavior, one instance has been identified where the
carryover parameter can be estimated consistently without knowledge of the firm's
decision behavior. This is the case when the sales response model is a Partial Adjustment
model. Interestingly enough (and in contrast to the Koyck model and the Adaptive
Expectations model), the carryover parameter in this model captures more than just
advertising carryover. In all other instances considered, consistent estimation is
conditional on the knowledge of the firm's actual advertising decision behavior. One
general result which arises from these findings is that when the carryover parameter
captures only advertising carryover, the firm's advertising decision behavior needs to be
known to ensure consistent parameter estimates.
Apart from the Partial Adjustment model, the distinction between deterministic
and stochastic decision behavior seems unnecessary. Although the analytic path leading
to the results was distinct, the error structures of the dynamic response models took over
and apparently pushed the inference issues into the stochastic domain. Moreover, the
philosophical question of whether or not advertising decision behavior is indeed
stochastic or whether we perceive it to be even though it might be fundamentally
deterministic seems less pressing.
More importantly, the results summarized in Table 1 seem to make a strong case
for supplementing the aggregated time series with knowledge about the firm's actual
decision behavior. This, however, depends on the statistical property of the estimates
pursued. The analytic work reported here was driven by consistency which is an
asymptotic property. Given that the aggregated series often available are finite, the
question remains whether the small-sample properties of these estimates are equally
desirable. Merging the analytic results obtained here with the numerical results (although
for a narrower modeling framework) on a parsimonious procedure which is robust to
different types of decision behavior reported in Russell (1988), this question becomes
Es Knowledge ofAdvertising Decision Behavior Needed?
True Micro SalesResponse Model
Partial Adjustment Model
Koyck Model
Adaptive Expections Model
Deterministic Stochastic
not needed needed
needed Q needed
needed needed
needed needed
needed needed
needed R needed
14
Table 1Consistent Estimation of Dynamic Sales Response
Given Temporally Aggregated Data
1 Carryover parameter.2 Direct advertising response parameter.
15
critical and fundamental to the future direction of this research. The procedure suggested
and analyzed empirically in Russell (1988) does not require the exact knowledge of the
firm's behavior but results in estimates with unknown statistical properties. However,
simulation results indicate that "reasonable" small-sample estimates can be obtained.
Hence, the estimates obtained here are asymptotically more desirable. Whether or not
they have more desirable small-sample properties remains to be seen. Empirical work
should be devoted to this question because if the reasonable small-sample performance of
the robust procedures can be improved upon, researchers should make every effort
possible to incorporate prior information on the firm's advertising decision behavior. At
the same time, as instances where such prior information is hard to come by will remain,
efforts should continue to be devoted to enhancing the robustness of these parsimonious
procedures and make them ammendable to a broader modeling framework.
1 6
Footnotes
1. Given the availability of single-source data (i.e., advertising exposure data and
purchase records for a panel of consumers on virtually a continuous basis), the
argument has been put forward that the data interval bias in dynamic sales response
is no longer an issue. Although this argument holds perhaps for a number of
frequently-purchased product categories, data for a larger number of other
categories (e.g. all durables) are not available in single-source format. Furthermore,
beyond the U.S., single-source data is currently non-existant.
2. Stability or stationarity is implied by Ikl<1.
17
Appendix 1: Autocovariance of the micro sales series st
Given response model in (1) and successive substitution, it is easily shown that
s t = 13 [a t + a t-1 +X2 a t _ 2 + ..] + [ Ut + X U t-1 + X2
Ut_2 +
and
St - k = o Fat-k + at - k -1 + X2 at -k -2 + •••]+ [U t_k + X U t - k - +X2 U t - k - 2 4- ...]• (1-2)
If at is deterministic, then
E (st) = p [a t + X at -1 + at- 2 + ••-]2
and
E (St ="_ 13 [a t -k X a t - k - 1 + X2
at - k - 2 +
Since Coy (st, s t _ = E [s t - E(st)] [st k - E (st - 0], the autocovariance equals
Cov (s t , s t - = E [(ut X ut - + Ut _ 2 + + _ k + Xk + 1 U t k + ...)
2(U t _k+XU t _k-1+ Ut-k-2 + )]
With the Partial Adjustment model error structure (i.e, E ( u t) = 0, E (u t 11 1_0 = 0 for i # 0 and
E (u, u t_ t ) = 62 for i = 0) and with infinite successive substitution, the autocovariance can be
expressed compactly as
Coy (St, st _ k) = Xk 02 [1/(1 - A.2)]
(1-3)
since I I < 1 or micro sales follow a stable autoregressive process.
With error structure (4) implied by the Koyck and Adaptive Expectation schemes, it can be
shown that Cov (s t , st - k) = 0 for all k # 0.
If at is stochastic, the autocovariance structure becomes more complex. For illustrative
purposes, we consider a stable first-order autoregressive process as the advertising decision
1 8
rule. Moreover, at = 0 at _ 1 + et with I 01 < 1, E (et) = 0, E (e t, e t i) = 0 for i 0, and
E (et e t i) = cq for i = OFor simplicity of derivation, we'll assume the advertising process to
be independent from the error process, or E ( u t Et- i) = 0 for all i.
With infinite successive substitution and steps similar to the ones taken above for the
micro sales series, we obtain
Coy (at , at - k) = E (at at - k) 0k [ i 02)]. (1-4)
With these expressions we can now derive the autocovariances of the micro sales series.
s t = [at + Xat _ + X2 at _ 2 +
at_k + xk + 1at - k - 1 +
+til t ut _ + A. U t _ 2 + + X U t _ k+ k+1 U t - k - 1 +
and
st _ k = 13 [a t _ k + X at _ k _ 4-X2 at _ k _ 2 + •••]
+ {ll t _k+ X U t _ k - 1 + X2 ut - k - 2 + ••• 1.
Considering the advertising terms in computing Cov (s t , s t _ k), we have
132[at at k + Xat at _ k - + X2 at at _ k - 2 + X3 a_t at _ k _ 3 +
4+ Xat _ at _ k + X
2at - 1 at _ k - 1 + 3 at _ 1 at _ k - 2 + A at _ 1 at _ k - 3+ •••
+ X2at _ 2 at _k+A.3 at _2 at -k_i+ X4 at _2 a t -k-2+X5at - 2 at -k-3+••-
k ,k+1 k+ 3A. at - k a t -k+ A. at -kat-k-1+A. at -k at -k-2 + A k at k ' • _ 3+ •••
+ at - k-1 at - k+ 2 at
k tat _ k _ 1 + Xk + 3 at _ k I at _ k _ 2+xk+4 at k at k_3+...1
19
which, taking expectations, given (1-4) and ignoring the term 132
for a moment, can be
expressed as
62 k k+1 2 k+2 3 k+3
c 0 +X0 +X0 +X +...( 1 - 02)
(1-
+
02)
02e r 2 k- 2 3 k1 4 k 5 k+1LA. 0 +X 0 +A. 0 +X 0 +...](1 - 02)
6Fr
k k +1 +3+ A.
k+2 02+ Xk+X 03
+ ...](1 - 02)
[ k+1 -1 k+2 k+3r 0 +X +X 0+Xk
+ 40
2+-1 ....( 1 - 02)
Dividing each term in brackets by its first component (i.e. taking the first component outside the
brackets), each term in brackets becomes a series in A . 0 which, given infinite successive
substitution, can be condensed to [1/ (1 - 0)].
Moreover, we obtain for the advertising terms
132 aE 1 [0k+ x 0k - 1 + x2 ok - 2 ... xk xk + 1 0- 1
(1 - 02) (1 - 0)
or
= 132 a? ØkkAl- 02)][1/(1- X0)1{1 + X 0- 1+ X.2 0- 2 +
which for IX/0 < 1 can be written compactly as
02 cq Økk/ - 02)] [1/ (1 - x 0)] [ii(1 - x/0)]. (1-5)
Cr( [x 0k- 1 + x2 0k+ x3 0k + 1 + x40k +2+
2 0
Note that the simplification requires the absolute value of the autoregressive parameter of themicro advertising series to be larger than the autoregressive parameter of the micro sales series
(i.e., > X). Moreover, with the Partial Adjustment model error structure the autocovariancesin the micro sales series equal the sum of (1-4) and (1-5), or
E (st st _ k) = 132 ozy? 01` [1/(1 - 02)] [1/(1 - X 0 [1/ (1 - X/0)1+ X k 62 {W(1 - X2)}-
With error structure (4), the autocovariance is simply equal to (1-4).
2 1
Appendix 2 : Autocovariance of the macro sales series St
From (2), we have
S t = X S t _t +13 At + Ut
which after successive substitution results in
AS t =13(A t +2A t _ 1 +X2 A t _ 2 +A.
3 .ft t _ 3+ ••.)
ut + x ut _ + X2 - 2 + X3 ti t - 3 + + A-k Ut - k + A-k 1 - k - 1+
and, likewise,
S t _ k = (A t - k + X A, k _ 1 + A t _ k - 2 + X3 At - k - 3 + ••-)
Ut _ k + Lit - k - + - k - 2 + A.3- k - 3+ •-•
Ha t (and, hence, A t) is deterministic
Coy (S t , S t _ E [S t - E (S t)] ES t - k E (St - equals
Cov (S t , S t _ k) = E [Ut + Ut - 1 "2 ut _2 xk ut k k + 1
Ut - k - +
[Ut _k +XUt-k_1+ a.2- k - 2 + .•1.
Since
- k = Ut - k + Ut - k - 1 + Ut - k - 2,
Ut-k-1 = Ut-k-l+Ut-k-2+Ut-k-3, etc.
the covariance can be written as
Cov(St,St-k) = E[Ut+ XUt-1+...+a k- 2(Ut-k+2+Ut-k+1+Ut- k)
xk 1(U t _ k + 1 + Ut k + U t - k - 1) + A-k (Ut -k + Ut - k - 1+ Ut - k - 2)
22
+ xk + 1 (Ut - k - + Ut - k - 2 +Ut - k - 3) + •..]
[(li t _ k + Ut k - 1 + Ut - k - 2 ) + X (Ut -k -1 + Ut - k -2 + Ut - k - 3 )
+ X2 (U t - k - 2 + Ut - k -3 + Ut - k - 4) +---. 1. (2 -1)
With the Partial Adjustment model error structure, E (u t ut_;) = 02 for i = 0 and
E ( u t u t _ i) = 0 for i # 0, expression (2 - 1) can be written as
Coy (S t, St - k) = (Xk - 2 + Xk - 1 + Xk) 62
+ + (?k - 1 + xk + Xk + 1) 62
+ (x2 + + 1) (xk + xk + 1 + + 2) (32
+ (X/ - 2
+X / -3 + X1 - (X
/ - 2 + 2.
/- 3 +X
/ - 62 +
Alternatively,
COV (Sr, S t _ = Xk (X 2 + X- 1 + 1) cy2
+ A.k + 0: 1 + 1 + 62
+Xk(X2-0.+1)(1+X+X2)62
or
COV (S t, St - k) xk2(1+x+x2)(32
+ xk 1 a + 1) + + x2) 0.2
+ x.k (X2 + X + (1 + X + A.2) 02
+ + 1 (X2 + X + (1 + X + X2) 02 + .
23
More compactly, the autocovariance expression becomes
(S[(1 + (xk - 2 + xk + xlc)1
COV t , S t .1t)
+ (1 + + x2)2(xk + xk +2 + xrk +4 + ...)1
= 432 { (1 + x+ x2) (xk -2 + xk - 1 +x11
(i x + x2)2(xk (1 x2))
k + + x2) (x- 2 + x- 1 +
(1- x2)
which is consistent with the autocovariance generating functions discussed in Telser (1967) and
Tiao and Wei (1976). Given expressions (1-3) and (2-2), it is clear that
Coy (S t, S t _ k) =[ Coy (S t, St - (A- 2 + 1 + 1) ( X
2 + X + 1)
which establishes the relationship between the autocovariances of the macro and micro sales
series given a Partial Adjustment model error structure. With error structure (4) implied by the
Koyck and Adaptive Expectation schemes, it can be shown that cov (S t, S t _ k) = 0 for all
k O.
If a t (and, hence, A t) is stochastic, the results are more complex. Following up on the
discussion in Appendix 1, it is evident from the above results that
,2
k 2-
CON, (A t, A t _k) = (0 + - I + 1) (1 + 0 + 02).
(1 - 02)
With this autocovariance and an approach similar to the one followed in Appendix 1 for the
micro series, it can be shown that with the Partial Adjustment model error structure, the
covariance equals
Coy (S t, S t _ = [32 0k at
(0 -2 +0 + 1) (1 + + 02) [1/ (1 - 0)111/ (1 - ),./0)](1 - 02)
(2-2)
24
xka2 oc2+x_i+0(1+x+x2).
(1 - X2)(2 - 3)
Remember that this concise representation implies that <For error structure (4) implied by the Koyck and Adaptive Expectation schemes, the macro
autocovariance equals the first term in (2-3).
25
References
Bass, F.M., and R.P. Leone (1983), "The Data Interval Bias and Empirical Estimation of
Bi-Monthly Relations from Annual Data," Management Science, 29, 1-11.
Blattberg, R. and A. Jeuland (1981), " Micromodeling Approach to Investigate the
Advertising Sales Relationship, "Management Science, 27, 988-1005.
Clarke, D. (1976), "Econometric Measurement of the Duration of Advertising Effect on
Sales," journal of Marketing Research, 13, 345-357.
Hanssens, D.M., L.J. Parsons, and R.L. Schultz (1990), Market Response Models:
Econometric and Time Series Analysis. Boston, Mass: Kluwer.
Kanethkar, V., C. Weinberg, and D. Weiss (1986), "Recovering Micro Parameters from
Aggregate Data for the Koyck and Brand Loyal Models," Journal of Marketing
Research, 23, 298-304.
Rao, R. (1986), "Estimating Continuous Time Advertisng - Sales Models," Marketing
Science, 5, 125-142.
Russell, G.J. (1988), "Recovering Measures of Advertisng Carryover From Aggregate
Data: The Role of the Firm's Decision Behavior." Marketing Science. 7, 252-270.
Sasieni, M. (1982)," The Effects of Combining Observation Periods in Time Series,"
Journal of the Operational Research Society, 33, 647-653.
Srinivasan, V. and H. Weir (1988), "A Direct Aggregation Approach to Inferring
Microparameters of the Koyck Advertisng-Sales Relationship from Macro Data,
"Journal of Marketing Research, 25, 145-156.
Telser, L.G. (1967), "Discrete Samples and Moving Sums in Stationary Stochastic
Processes, "Journal of the American Statistical Association, 62, 484-499.
Tiao, G.C. and W.S. Wei (1976), "Effect of Temporal Aggregation on the Dynamic
Relationship of Two Time Series Variables," Biometrika, 63, 513-523.
2 6
Vanhonacker, W.R. (1983), "Carryover Effects and Temporal Aggregation in a Partial
Adjustment Model Framework," Marketing Science, 2, 297-317.
Vanhonacker W.R. (1984), 'Estimation and Testing of a Dynamic Sales Response Model
With Data Aggregated Over Time: Some Results for the Autoregressive Current
Effects Model," Journal of Marketing Research, 21,445-455.
Vanhonacker, W.R. (1987), "Estimating the Duration of Dynamic Effects with
Temporally Aggregated Observations," Journal of Statistical Computation and
Simulation, 27, 185-209.
Weiss, D., C. Weinberg, and P. Windal (1983), "The Effects of Serial Correlation and
Data Aggregation on Advertising Measurement, "Journal of Marketing Research,
20, 268-279.
1986
86/01 Arnoud DE MEYER
86/02 Philippe A. NAERT
Marcel WEVERBERGH
and Guido VERSWUVEL
86/03 Michael BRIMM
86/04 Spyros MAKRIDAKIS
and Mich2Ie HIBON
86/05 Charles A. WYPLOSZ
86/06 Francesco GIAVAllI,
Jeff R. SHEEN and
Charles A. WYPLOSZ
86/08 Jose de la TORRE and
David H.NECICAR
86/09 Philippe C. HASPESLAGH
86/10
R. MOENART,
Arnoud DE MEYER,
J. BARBE and
D. DESCHOOLMEESTER.
"The R & D/Production interface".
"Subjective estimation in integratingcommunication budget and allocationdecisions: a case study", January 1986.
"Sponsorship and the diffusion of organizationalinnovation: a preliminary view".
"Confidence intervals: an empiricalinvestigation for the series in the M-Competition".
"A note on the reduction of the workweek",July 1985.
"The real exchange rate and the fiscalaspects of a natural resource discovery",Revised version: February 1986.
"Forecasting political risks forinternational operations", Second Draft: March 3,1986.
"Conceptualizing the strategic process indiversified firms: the role and nature of thecorporate influence process", February 1986.
"Analysing the issues concerningtechnological de-maturity".
Philippe A. NAERT
and Alain BULTEZ
86/12 Roger BETANCOURT
and David GAUTSCHI
86/13 S.P. ANDERSON
and Damien J. NEVEN
86/14 Charles WALDMAN
86/15 Mihkel TOMBAK and
Amoud DE MEYER
86/16 B. Espen ECKBO and
Herwig M. LANGOHR
86/17 David B. JEMISON
86/18 James TEBOUL
and V. MALLERET
86/19 Rob R. WEITZ
86/20 Albert CORHAY,
Gabriel HAWAWINI
and Pierre A. MICHEL
86/21 Albert CORHAY,
Gabriel A. HAWAWINI
and Pierre A. MICHEL
86/22 Albert CORHAY,
Gabriel A. HAWAWINI
and Pierre A. MICHEL
"From "Lydiainetry" to "Pinkhamization":misspecifying advertising dynamics rarelyaffects profitability".
"The economics of retail firms", RevisedApril 1986.
"Spatial competition 91a Cournot".
"Comparaison internationale des marges brutes ducommerce", June 1985.
"How the managerial attitudes of firms withFMS differ from other manufacturing toms:survey results", June 1986.
"Les primes des offres publiques, la noted'information et le mamba des transferts decontrale des socidtds".
"Strategic capability trader in acquisitionintegration", May 1986.
"Towards an operational definition ofservices", 1986.
"Nostradamus: a knowledge-basedforecasting advisor".
"The pricing of equity on the London stockexchange: seasonality and size premium",June 1986.
"Risk-premia seasonality in U.S. andEuropean equity markets", February 1986.
"Seasonality in the risk-return relationshipssome international evidence", July 1986.
INSEAD WORKING PAPERS SERIES
86/11
86/07 Douglas L. MacLACHLAN "Judgmental biases in sales forecasting",and Spyros MAKRIDAKIS February 1986.
86/23 Arnoud DE MEYER "An exploratory study on the integration of
information systems in manufacturing",July 1986.
"A methodology for specification and
aggregation in product concept testing",July 1986.
86/34 Philippe HASPESLAGHand David JEMISON
86/35 Jean DERMINE
86/36 Albert CORHAY and
Gabriel HAWAWINI
'Acquisitions: myths and reality", July 1986.
"Measuring the market value of a bank, aprimer", November 1986.
"Seasonality in the risk-return relationship:
some international evidence", July 1986.
86/24
David GAUTSCHIand Vithala R. RAO
86/25
H. Peter GRAY
"Protection", August 1986.and Ingo WALTER 86/37 David GAUTSCHI and "The evolution of retailing: a suggested
Roger BETANCOURT economic interpretation"."The economic consequences of the FrancPoincare", September 1986.
"Negative risk-return relationships in
business strategy: paradox or truism?",October 1986.
"Interpreting organizational texts.
"Why follow the leader?".
"The succession game: the real story.
"Flexibility: the next competitive battle",October 1986.
"Flexibility: the next competitive battle",Revised Version: March 1987.
Performance differences among strategic
group members", October 1986.
"The role of public policy in insuringfinancial stability: a cross-country,
comparative perspective", August 1986,Revised November 1986.
86/38 Gabriel HAWAWINI
86/39 Gabriel HAWAWINIPierre MICHELand Albert CORHAY
86/40 Charles WYPLOSZ
86/41 Kasra FERDOWSand Wickham SKINNER
86/42 Kasra FERDOWSand Per LINDBERG
86/43 Damien NEVEN
86/44 Ingemar DIERICKX
Carmen MATUTESand Damien NEVEN
1987
87/01 Manfred KETS DE VRIES
"Financial innovation and recent
developments in the French capital markets",Updated: September 1986.
"The pricing of common stocks on the
Brussels stock exchange: a re-examination ofthe evidence", November 1986.
"Capital flows liberafiXition and the EMS, aFrench perspective", December 1986.
"Manufacturing in a Now perspective", July1986.
"FMS as indicator of manufacturing
strategy", December 1986.
"On the existence of equilibrium in
hotelling's model", November 1986.
"Value added tax and competition",December 1986.
"Prisoners of leadership".
86/26
Barry EICHENGREENand Charles WYPLOSZ
86/27
Karel COOLand Ingemar DIERICKX
86/28
Manfred KETS DE VRIESand Danny MILLER
86/29
Manfred KETS DE VRIES
86/30
Manfred KETS DE VRIES
86/31
Arnoud DE MEYER
86/31
Arnoud DE MEYER,Jinichiro NAKANE,Jeffrey G. MILLERand Kasra FERDOWS
86/32
Karel COOLand Dan SCHENDEL
86/33
Ernst BALTENSPERGERand Jean DERMINE
87/02 Claude VIALLET "An empirical investigation of internationalasset pricing", November 1986.
87/03 David GAUTSCHIand Vithala RAO
87/04 Sumantra GHOSHAL andChristopher BARTLETT
87/05 Arnoud DE MEYERand Kasra FERDOWS
87/06 Arun K. JAIN,Christian PINSON andNaresh K. MALHOTRA
87/07 Rolf BANZ andGabriel HAWAWINI
87/08 Manfred ICETS DE VRIES
87/09 Lister VICKERY,Mark PILKINGTONand Paul READ
87/10 Andre LAURENT
87/11 Robert FILDES andSpyros MAKRIDAKIS
"A methodology for specification andaggregation in product concept testing",Revised Version: January 1987.
"Organizing for innovations: case of themultinational corporation", February 1987.
"Managerial focal points in manufacturingstrategy", February 1987.
"Customer loyalty as a construct in themarketing of banking services", July 1986.
"Equity pricing and stock marketanomalies", February 1987.
"Leaders who can't manage", February 1987.
"Entrepreneurial activities of EuropeanMBAs", March 1987.
"A cultural view of organizational change",March 1987
"Forecasting and loss functions", March1987.
"The Janus Head: learning from the superiorand subordinate faces of the manager's job",April 1987.
"Multinational corporations as differentiatednetworks", April 1987.
"Product Standards and CompetitiveStrategy: An Analysis of the Principles", May1987.
"METAFORECASTING: Ways of improvingForecasting. Accuracy and Usefulness",May 1987.
"Takeover attempts: what does the languagetell us?, June 1987.
"Managers' cognitive maps for upward anddownward relationships", June 1987.
"Patents and the European biotechnologylag: a study of large European pharmaceuticalfirms", June 1987.
"Why the EMS? Dynamic games and theequilibrium policy regime", May 1987.
"A new approach to statistical forecasting",June 1987.
"Strategy formulation: lie impact of nationalculture", Revised: July 1987.
"Conflicting ideologies: structural andmotivational consequences", August 1987.
"The demand for retail products and thehousehold production model: new views oncomplementarily and substitutability".
"The internal and external careers: atheoretical and cross-cultural perspective",Spring 1987.
"The robustness of MDS configurations inthe face of incomplete data", March 1987,Revised: July 1987.
"Demand complementarities, householdproduction and retail assortments", July1987.
87/19 David BEGG andCharles WYPLOSZ
87/20 Spyros MAKRIDAKIS
87/21 Susan SCHNEIDER
87/22 Susan SCHNEIDER
87/23 Roger BETANCOURTDavid GAUTSCHI
87/16 Susan SCHNEIDERand Roger DUNBAR
87/17 Andre LAURENT andFernando BARTOLOME
87/18 Reinhard ANGELMAR andChristoph LIEBSCHER
87/15 Spyros MAKRIDAKIS
87/12 Fernando BARTOLOMEand Andre LAURENT
87/13 Sumantra GHOSHALand Nitin NOHRIA
87/14 Landis GABEL
87/25 A. K. JAIN,N. K. MALHOTRA andChristian PINSON
87/26 Roger BETANCOURTand David GAUTSCHI
87/24 C.B. DERR andAndre LAURENT
"The dark side of CEO succession",November 1987.
"Product compatibility and the scope ofentry", November 1987.
"Seasonality, size premium and therelationship between the rink and the returnof French common stocks", November 1987
87/27 Michael BURDA "Is there a capital shortage in Europe?",
August 1987.
87/28
Gabriel HAWAWINI
"Controlling the interest-rate risk of bonds:an introduction to duration analysis andimmunization strategies", September 1987.
87/29
Susan SCHNEIDER and
"Interpreting strategic behavior: basicPaul SHRIVASTAVA assumptions themes in organizations",
September 1987
87/39
87/40
87/41
Manfred KETS DE VRIES
Carmen MATUTES and
Pierre REGIBEAU
Gabriel HAWAWINI and
Claude VIALLET
87/42
Damien NEVEN and
"Combining horizontal and vertical87/30
Jonathan HAMILTON
"Spatial competition and the Core", August
Jacques-F. THISSE
differentiation: the principle of max-minW. Bentley MACLEOD
1987. differentiation", December 1987.and J. F. THISSE
87/43
Jean GABSZEWICZ and
"Location", December 1987.87/31
Martine QUINZII and
"On the optimality of central places", Jacques-F. THISSEJ. F. THISSE
September 1987.
87/44
Jonathan HAMILTON,
Jacques-F. THISSE
and Anita WESKAMP
87/45
Karel COOL,
David JEMISON and
Ingemar DIERICKX
87/46
Ingemar DIERICKX
and Karel COOL
1988
88/01
Michael LAWRENCE and
Spyros MAKRIDAKIS
/02 Spyros MAKRIDAKIS
88/03
James TEBOUL
"Spatial discrimination: Bertrand vs.Cournot in • model of location choice",December 1987.
"Business strategy, market structure andrisk-return relationshipm a causalinterpretation", December 1987.
"Asset stock accumulation and sustainabllityof competitive advantage", December 1987.
"Factors affecting judgemental forecasts andconfidence intervals", January 1988.
"Predicting recessions and other turningpoints", January 1988.
"De-industrialize service for quality", January1988.
87/32
Arnoud DE MEYER
"German, French and British manufacturingstrategies less different than one thinks",September 1987.
87/33
Yves DOZ and
"A process framework for analyzingAmy SHUEN cooperation between firms", September 1987.
87/34
Kaara FERDOWS and
"European manufacturers: the dangers ofArnoud DE MEYER
complacency. Insights from the 1987European manufacturing futures survey",October 1987.
87/35
P. J. LEDERER and
"Competitive location on networks underJ. F. THISSE
discriminatory pricing", September 1987.
87/36
Manfred KETS DE VRIES
"Prisoners of leadership", Revised version
October 1987.
87/37
Landis GABEL
"Privatization: its motives and nicelyconsequences", October 1987.
87/38
Susan SCHNEIDER
"Strategy formulation: the impact of nationalculture", October 1987.
88/04 Susan SCHNEIDER
88/05 Charles WYPLOSZ
88/06 Reinhard ANGELMAR
88/07 Ingemar DIERICKX
and Karel COOL
88/08 Reinhard ANGELMAR
and Susan SCHNEIDER
88/09 Bernard SINCLAIR-
DESGAGNE
88/10 Bernard SINCLAIR-
DESGAGNE
88/11 Bernard SINCLAIR-
DESGAGNE
88/12 Spyros MAKRIDAKIS
88/13 Manfred KEFS DE VRIES
88/14 Alain NOEL
88/15
Anil DEOLALIKAR and
La rs-H end rik ROLLER
"National vs. corporate culture: implications
for human resource management", January1988.
"The swinging dollar: is Europe out of
step?", January 1988.
"Les conflits dans les canaux de
distribution", January 1988.
"Competitive advantage: a resource basedperspective", January 1988.
"Issues in the study of organizational
cognition", February 1988.
"Price formation and product design throughbidding", February 1988.
"The robustness of some standard auction
game forms", February 1988.
"When stationary strategies are equilibriumbidding strategy: The single-crossing
property", February 1988.
"Business firms and managers in the 21st
century", February 1988
"Alexithymia in organizational life: the
organization man revisited", February 1988.
"The interpretation of strategies: a study ofthe impact of CEOs on the
corporation", March 1988.
"The production of and returns from
industrial innovation: an econometric
analysis for a developing country", December1987.
88/16 Gabriel HAWAWINI
/17 Michael BURDA
88/18 Michael BURDA
88/19 M.J. LAWRENCE and
Spyros MAKRIDAKIS
88/20
Jean DERMINE,Damien NEVEN and
J.F. THISSE
88/21
James TEBOUL
22 Lars-Hendrik ROLLER
88/23
Sjur Didrik FLAM
and Georges ZACCOUR
88/24
B. Espen ECKBO andHerwig LANGOHR
88/25
Everette S. GARDNER
and Spyros MAKRIDAKIS
"Market efficiency and equity Pricing:
international evidence and implications for
global investing", March 1988.
"Monopolistic competition, costs of
adjustment and the behavior of European
employment", September 1987.
"Reflections on "Wait Unemployment" inEurope", November 1987, revised February
1988.
"Individual bias in judgements of
confidence", March 1988.
"Portfolio selection by mutual funds, an
equilibrium model", March 1988.
"De-industrialize service tor quality", March
1988 (88/03 Revised).
"Proper Quadratic Functions with anApplication to AT&T", May 1987 (Revised
March 1988).
"Equffibres de Nash-C-onrnot dans le marchd
enropilen du gaz: an cas oh les solutions en
boucle ouverte et en feedback coincident",Mere 1988.
"Information disclosure, means of payment,and takeover premia. Public and Private
tender offers in France", July 1985, Sixthrevision, April 1988.
"The future of forecasting", April 1988.
88/26
Sjur Didrik FLAM
"Semi-competitive Cournot equilibrium inand Georges ZACCOUR multistage oligopolies", April 1988.
88/27 Murugappa KRISHNANLars-Hendrik ROLLER
88/28 Sumantra GHOSHAL andC. A. BARTLETT
88/29 Naresh K. MALHOTRA,Christian PINSON andArun K. JAIN
88/30 Catherine C. ECKELand Theo VERMAELEN
88/31 Sumantra GHOSHAL andChristopher BARTLETT
88/32 Kasra FERDOWS andDavid SACKRIDER
88/33 Mihkel M. TOMBAK
34 Mihkel M. TOMBAK
88/35 Mihkel M. TOMBAK
88/36 Vikas TIBREWALA andBruce BUCHANAN
88/37 Murugappa KRISHNANLars-Hendrik ROLLER
88/38 Manfred KETS DE VRIES
"Entry game with resalable capacity",April 1988.
"The multinational corporation as a network:perspectives front interorganizationaltheory", May 1988.
"Consumer cognitive complexity and thedimensionality of multidimensional scalingconfigurations", May 1988.
"The financial fallout from Chernobyl: riskperceptions and regulatory response", May1988.
"Creation, adoption, and diffusion ofinnovations by subsidiaries of multinationalcorporations", June 1988.
"International manufacturing: positioningplants for success", June 1988.
"The importance of flexibility inmanufacturing", June 1988.
"Flexibility: an important dimension inmanufacturing", June 1988.
"A strategic analysis of investment in flexiblemanufacturing systems", July 1988.
"A Predictive Test of the NBD Model thatControls for Non-stationarity", June 1988.
"Regulating Price-Liability Competition ToImprove Welfare", July 1988.
"The Motivating Role of Envy : A ForgottenFactor in Management", April 88.
88/39 Manfred KETS DE VRIES
/40 Josef LAKONISHOK andTheo VERMAELEN
88/41 Charles WYPLOSZ
88/42 Paul EVANS
88/43 B. SINCLAIR-DESGAGNE
88/44 Hamm MAHMOUD andSpyros MAICRIDAMS
88/45 Robert KORAJCZYKand Claude VIALLET
88/46
Yves DOZ andAmy SHUEN
88/47
Alain BULTEZ,Els GUSBRECHTS,Philippe NAERT andPiet VANDEN ABEELE
88/48Michael BURDA
88/49 Nathalie DIERKENS
/50 Rob WEITZ andArnoud DE MEYER
"The Leader as Mirror : ClinicalReflections", July 1988.
"Anomalous price behavior aroundrepurchase tender offers", August 1988.
"Assymetry in the EMS: intentional orsystemic?", August 1988.
"Organizational development in thetransnational enterprise", June 1988.
"Group decision support systems implementBayesian rationality", September 1988.
"The state of the art and future directionsin combining forecasts", September 1988.
"An empirical investigation of internationalasset pricing", November 1986, revisedAugust 1988.
"From intent to outcome: a processframework for parbmedrips", August 1988.
"Asymmetric camribarnm between substituteitems listed by redefine", September 1988.
"Reflections on 'Wait unemployment' inEurope, II", April 1988 revised September1988.
"Information asymmetry and equity issues",September 1988.
"Managing expert systems: from inceptionthrough updating", October 1987.
/51 Rob WEITZ "Technology, work, and the organization:the impact of expert systems", July 1988.
Susan SCHNEIDER andReinhard ANGELMAR
Manfred KETS DE VRIES
Lars-Hendrik ROLLERand Mihkel M. TOMBAK
Peter BOSSAERTSand Pierre MILLION
Pierre BILLION
Wilfried VANHONACKERand Lydia PRICE
B. SINCLAIR-DESGAGNEand Mihkel M. TOMBAK
Martin KILDUFF
Michael BURDA
Lars-Hendrik ROLLER
Cynthia VAN HULLE,Theo VERMAELEN andPaul DE WOUTERS
"Cognition and organizational analysis:
who's minding the store?", September 1988.
"Whatever happened to the phrlosopher-
king: the leader's addiction to power,September 1988.
"Strategic choice of flexible production
technologies and welfare implications",October 1988
"Method of moments tests of contingent
claims asset pricing models", October 1988.
"Size-sorted portfolios and the violation of
the random walk hypothesis: Additional
empirical evidence and implication for tests
of asset pricing models", lune 1988.
"Data transferability: estimating the response
effect of future events based on historical
analogy", October 1988.
"Assessing economic inequality", November1988.
"The interpersonal structure of decisionmaking: a social comparison approach to
organizational choice", November 1988.
"Is mismatch really the problem? Some
estimates of the Chelwood Gate II model
with US data", September 1988.
"Modelling cost structure: the Bell Systemrevisited", November 1988.
"Regulation, taxes and the market forcorporate control in Belgium", September1988.
88/63 Fernando NASCIMENTOand Wilfried R.VANHONACKER
88/64 Kasra FERDOWS
88/65 Arnoud DE MEYERand Kasra FERDOWS
66 Nathalie DIERKENS
88/67
Paul S. ADLER andKasra FERDOWS
1989
89/01
Joyce K. BYRER andTawfik JELASSI
89/02
Louis A. LE BLANCand Tawfik JELASSI
89/03
Beth H. JONES andTawfik JELASSI
89/04
Kasra FERDOWS andArnoud DE MEYER
89/05
Martin KILDUFF andReinhard ANGELMAR
89/06 Mihkel M. TOMBAK andB. SINCLAIR-DESGAGNE
"Strategic pricing of differentiated consumer
durables in n dynamic duopoly: a numerical
analysis", October 1988.
"Charting strategic roles for internationalfactories", December 1988.
"Quality up, technology down", October 1988
"A discussion of exact measures of
information assymetry: the example of Myers
and Majiuf model or the importance of the
asset structure of the firm", December 1988.
"The chief technology officer", December1988.
"The impact of languagertheories on DSS
dialog", January 1989.
"DSS software selection: a multiple criteriadecision methodology", January 1989.
"Negotiation support: the effects of computer
intervention and conflict level on bargainingoutcome", January 1989.
"Lasting improvement in manufactunng.
performance: la search of • new theory",January 1989.
"Shared history or shared culture? The
effects of time, culture, and performance on
institutionalization in simulated
organizations", January 1989.
"Coordinating manufacturing and businessstrategies: I", February 1989.
89/07 Damien J. NEVEN
89/08
Amoud DE MEYER and
Hellmut SCHUTTE
89/09
Damien NEVEN,Carmen MATUTES andMarcel CORSTJENS
89/10 Nathalie DIERKENS,Bruno GERARD andPierre HILLION
89/11
Manfred KETS DE VRIESand Main NOEL
89/12
Wilfried VANHONACKER
89/13
Manfred KETS DE VRIES
89/14
Reinhard ANGELMAR
89/15
Reinhard ANGELMAR
89/16
Wilfried VANHONACKER,Donald LEHMANN andFareena SULTAN
89/17
Gilles AMADO,Claude FAUCHEUX andAndri LAURENT
"Structural adjustment in European retailbanking. Some view from industrial
organisation", January 1989.
"Trends in the development of technology
and their effects on the production structure
in the European Community", January 1989.
"Brand proliferation and entry deterrence",February 1989.
"A market based approach to the valuation
of the assets in place and the growth
opportunities of the firm", December 1988.
"Understanding the leader-strategy interface:
application of the strategic relationship
interview method", February 1989.
"Estimating dynamic response models when
the data are subject to different temporal
aggregation", January 1989.
"The impostor syndrome: a disquieting
phenomenon in organizational life", February1989.
"Product innovation: a tool for competitive
advantage", March 1989.
"Evaluating a firm's product innovation
performance", March 1989.
"Combining related and sparse data in linearregression models", February 1989.
"Changement organisationnel et rialitOs
culturefies: contrastes franco-am4ricains",March 1989.
89/18 Srinivasan BALAK-RISHNAN andMitchell KOZA
89/19 Wilfried VANHONACKER,Donald LEHMANN andFareena SULTAN
89/20 Wilfried VANHONACKERand Russell WINER
89/21 Arnoud de MEYER andKasra FERDOWS
89/22 Manfred KETS DE VRIESand Sydney PERZOW
89/23 Robert KORAJCZYK andClaude VIALLET
89/24 Martin KILDUFF andMitchel ABOLAFIA
89/25
Roger BETANCOURT andDavid GAUTSCHI
89/26
Charles BEAN,Edmond MALINVAUD,Peter BERNHOL2,Francesco GIAVAllIand Charles WYPLOSZ
89/27
David KRACKHARDT andMartin KILDUFF
89/28
Martin KILDUFF
"Information asymmetry, market failure andjoint-ventures: theory and evidence",March 1989.
"Combining related and sparse data in linearregression models", Revised March 1989.
"A rational random behavior model ofchoice", Revised March 1989.
"Influence of manufacturing improvement
programmes on performance", April 1989.
"What is the role of character in
psychoanalysis?" April 1989.
"Equity risk premia and the pricing offoreign exchange risk" April 1989.
"The social destruction of reality:Organisational conflict as social drama"zApril 1989.
"Two essential characteristics of retail
markets and their economic consequences"March 1989.
"Macroeconomic policies for 1992: the
transition and after", April 1989.
"Friendship patterns and cultural
attributions: the control of organizationaldiversity", April 1989.
"The interpersonal structure of decisionmaking: a social comparison approach to
organizational choice", Revised April 1989.
89/29 Robert GOGEL and "The battlefield for 1992: product strength 89/42
Robert ANSON and
"A development framework for computer-Jean-Claude LARRECHE and geographic coverage", May 1989. Tawfik JELASSI supported conflict resolution", July 1989.
89/30
Lars-Hendrik ROLLER "Competition and Investment in Flexible 89/43
Michael BURDA
"A note on firing costs and severance benefitsand Mihkel M. TOMBAK Technologies", May 1989. in equilibrium unemployment", June 1989.
89/31
Michael C. BURDA and
"Intertemporal prices and the US trade 89/44
Balaji CHAKRAVARTHY
"Strategic adaptation in multi-businessStefan GERLACH
balance in durable goods", July 1989. and Peter LORANGE
firms", June 1989.
89/48
89/49
89/50
89/51
89/52
89/32
Peter HAUG andTawfik JELASSI
89/33
Bernard SINCLAIR-DESGAGNE
89/34
Sumantra GHOSHAL andNittin NOHRIA
89/35 Jean DERMINE andPierre HILLION
89/36 Martin KILDUFF
89/37 Manfred KETS DE VRIES
89/38 Manfred KETS DE VRIES
89/39 Robert KORAJCZYK andClaude VIALLET
89/40 Balaji CHAKRAVARTHY
"Application and evaluation of a multi-criteria decision support system for thedynamic selection of U.S. manufacturinglocations", May 1989.
"Design flexibility in monopsonisticindustries", May 1989.
"Requisite variety versus shared values:managing corporate-division relationships inthe M-Form organisation", May 1989.
"Deposit rate ceilings and the market valueof banks: The case of France 1971-1981",May 1989.
"A dispositional approach to social networks:the case of organizational choice", May 1989.
"The organisational fool: balancing aleader's hubris", May 1989.
"The CEO blues", June 1989.
"An empirical investigation of internationalasset pricing", (Revised June 1989).
"Management systems for innovation andproductivity", June 1989.
Rob WEITZ and
Arnoud DE MEYER
Marcel CORSTJENS,
Carmen MATUTES and
Damien NEVEN
Manfred KETS DE VRIES
and Christine MEAD
Damien NEVEN and
Lars-Hendrik ROLLER
Jean DERMINE
Jean DERMINE
Spyros MAKRIDAKIS
Antoud DE MEYER
Spyros MAKRIDAKIS
S. BALAKRISHNAN
and Mitchell KOZA
"Managing expert systems: a framework andcase study", June 1989.
"Entry Encouragement", July 1989.
"The global dimension in leadership and
organization: issues and controversies", April1989.
"European integration and trade flows",August 1989.
"Home country control and mutualrecognition", July 1989.
"The specialization of financial institution,the EEC model", August 1989.
"Sliding simulation: a new approach to timeseries forecasting", July 1989.
"Shortening development cycle times: amanufacturer's perspective", August 1989.
"Why combining works?", July 1989.
"Organisation costs and a theory of jointventures", September 1989.
89/45
89/46
89/47
89/53
89/54
89/41 B. SINCLAIR-DESGAGNEand Nathalie DIERKENS
"The strategic supply of precisions", June1989.
89/55 H. SCHUTTE "Euro-Japanese cooperation in information 89/67 Peter BOSSAERTS and "Market microstructure effects oftechnology", September 1989. (FIN) Pierre HILLION government intervention in the foreign
exchange market", December 1989.89/56 Wilfried VANHONACKER
and Lydia PRICE
"On the practical usefulness of meta-analysis
result?, September 1989.
199089/57 Taekwon KIM,
Lars-Hendrik ROLLER
and Mihkel TOMBAK
"Market growth and the diffusion of
multiproduct technologies", September 1989. 90/01
TM/EP/AC
B. SINCLAIR-DESGAGNE "Unavoidable Mechanisms", January 1990.
89/58 Lars-Hendrik ROLLER "Strategic aspects of flexible production 90/02 Michael BURDA "Monopolistic Competition, Costs of(EP,TM) and Mihkel TOMBAK technologies", October 1989. EP Adjustment, and the Behaviour of European
Manufacturing Employment", January 1990.89/59
(COB)
Manfred KETS DE VRIES,
Daphne ZEVADI,
Alain NOEL and
Mihkel TOMBAK
"Locus of control and entrepreneurship: a
three-country comparative study", October
1989.90/03TM
Arnoud DE MEYER "Management of Communication in
International Research and Development",January 1990.
89/60 Enver YUCESAN and "Simulation graphs for design and analysis of 90/04 Gabriel HAWAWINI and "The Transformation of the European
(TM) Lee SCHRUBEN discrete event simulation models", October FIN/EP Eric RAJENDRA Financial Services Industry: From1989. Fragmentation to Integration", January 1990.
89/61 Susan SCHNEIDER and "Interpreting and responding to strategic 90/05 Gabriel HAWAWINI and "European Equity Markets: Toward 1992(AB) Arnoud DE MEYER issues: The impact of national culture". FIN/EP Bertrand JACQUILLAT and Beyond", January 1990.
October 1989.
90/06 Gabriel HAWAWINI and "Integration of Enrollees' Equity Markets:89/62
(TM)
Arnoud DE MEYER "Technology strategy and international R&D
operations", October 1989.FIN/EP Eric RAJENDRA Implications of Structural Change for Key
Market Participants to and Beyond 1992",January 1990.
89/63 Enver YUCESAN and "Equivalence of simulations: A graph
(TM) Lee SCHRUBEN approach", November 1989. 90/07 Gabriel HAWAWINI "Stock Market Anomalies and the Pricing of
89164
(TM)
Enver YUCESAN and
Lee SCHRUBEN
"Complexity of simulation models: A graph
theoretic approach", November 1989.
FIN/EP Equity on the Tokyo Stock Exchange",
January 1990.
90/08 Tawfik JELASSI and "Modelling with MCDSS: What about89/65
(TM,
AC, FIN)
Soumitra DUITA and
Piero BONISSONE"MARS: A mergers and acquisitionsreasoning system", November 1989.
'FM/EP
90/09EP/FIN
B. SINCLAIR-DESGAGNE
Alberto GIOVANNINI
and Jae WON PARK
Ethics?", January 1990.
"Capital Controls and International TradeFinance", January 1990.
89/66(TM,EP)
B. SINCLAIR-DESGAGNÈ "On the regulation of procurement bids",November 1989. 90/10 Joyce BRYER and "The Impact of Language Theories on DSS
TM Tawfik JELASSI Dialog", January 1990.
90/21 Roy SMITH and "Reconfiguration of the Global Securities90/11 Enver YUCESAN "An Overview of Frequency Domain FIN Ingo WALTER Industry in the 1990's", February 1990.TM Methodology for Simulation Sensitivity
Analysis", January 1990. 90/22 Ingo WALTER "European Financial Integration and ItsFIN Implications for the United States", February
90/12EP
Michael BURDA "Structural Change, Unemployment Benefitsand High Unemployment: A U.S.-European
1990.
Comparison", January 1990. 90/23 Damien NEVEN "EEC Integration towards 1992: SomeEP/SM Distributional Aspects", Revised December
90/13 Soumitra DUTTA and "Approximate Reasoning about Temporal 1989TM Shashi SHEKHAR Constraints in Real Time Planning and
Search", January 1990. 90/24 Lars Tyge NIELSEN "Positive Prices in CAPM", January 1990.FIN/EP
90/14 Albert ANGEHRN and "Visual Interactive Modelling and IntelligentTM Hans-Jakob LUTHI DSS: Putting Theory Into Practice", January
1990.90/25FIN/EP
Lars Tyge NIELSEN "Existence of Equilibrium in CAPM",January 1990.
90/15 Amoud DE MEYER, "The Internal Technological Renewal of a 90/26 Charles KADUSHIN and "Why networking Fails: Double Binds andTM Dirk DESCHOOLMEESTER,
Rudy MOENAERT andBusiness Unit with a Mature Technology",January 1990.
OB/BP Michael BRIMM the Limitations of Slarmlfm Networks",February 1990.
Jan BARBE90/27 Abbas FOROUGHI and "NSS Solutions to Major Negotiation
90/16 Richard LEVICH and "Tax-Driven Regulatory Drag: European TM Tawfik JELASSI Stumbling Blocks", February 1990.FIN Ingo WALTER Financial Centers in the 1990's", January
1990. 90/28 Amoud DE MEYER "The Manufacturing Contribution toTM Innovation", February 1990.
90/17FIN
Nathalie D1ERKENS "Information Asymmetry and Equity Issues",Revised January 1990. 90/29 Nathalie DIERKENS "A Discussion of Correct Measures of
FIN/AC Information Asymmetry", January 1990.90/18 Wilfried VANHONACKER "Managerial Decision Rules and theMKT Estimation of Dynamic Sales Response 90/30 Lars Tyge NIELSEN "The Expected Utility of Portfolios of
Models", Revised January 1990. FIN/EP Assets", March 1990.
90/19 Beth JONES and "The Effect of Computer Intervention and 90/31 David GAUTSCHI and "What Determines U.S. Retail Margins?",TM Tawfik JELASSI Task Structure on Bargaining Outcome",
February 1990.MKT/EP Roger BETANCOURT February 1990.
90/32 Srinivasan BALAK- "Information Asymmetry, Adverse Selection90/20TM
Tawfik JELASSI,Gregory KERSTEN and
"An Introduction to Group Decision andNegotiation Support". February 1990.
SM RISHNAN andMitchell KOZA
and Joint-Ventures: Theory and Evidence",Revised, January 1990.
Stanley ZIONTS 90/33OB
Caren SIEHL,David BOWEN and
"The Role of Rites of Integration in ServiceDelivery", March 1990.
Christine PEARSON
90/45 Soumitra MITA and "Integrating Case Based and Rule Based90/34FIN/EP
Jean DERMINE "The Gains from European BankingIntegration, a Call for a Pro-Active
TM Piero BONISSONE Reasoning: The Possibilistic Connection",May 1990.
Competition Policy", April 1990.90/46 Spyros MAKRIDAIUS "Exponential Smoothing: The Effect of
90/35 Jae Won PARK "Changing Uncertainty and the Time- TM and Mich2Ie HIBON Initial Values and Loss Functions on Post-EP Varying Risk Premia in the Term Structure
of Nominal Interest Rates", December 1988,Revised March 1990. 90/47 Lydia PRICE and
Sample Forecasting Accuracy".
"Improper Sampling in NaturalMKT Wilfried VANHONACKER Experiments: Limitations on the Use of
90/36 Arnaud DE MEYER "An Empirical Investigation of Meta-Analysis Results in BayesianTM Manufacturing Strategies in European Updating", Revised May 1990.
Industry", April 1990.90/48 Jae WON PARK "The Information in the Term Structure of
90/37TM/OB/SM
William CATS-BARIL "Executive Information Systems: Developingan Approach to Open the Possibles", April
EP Interest Rates: Out-of-Sample ForecastingPerformance", June 1990.
1990.90/49 Soumitra DUITA "Approximate Reasoning by Analogy to
90/38 Wilfried VANHONACKER "Managerial Decision Behaviour and the TM Answer Null Queries", June 1990.MKT Estimation of Dynamic Sales Response
Models", (Revised February 1990). 90/50EP
Daniel COHEN andCharles WYPLOSZ
"Price and Trade Effects of Exchange Ratesfluctuations and the Design of Policy
90/39TM
Louis LE BLANC andTawfik JELASSI
"An Evaluation and Selection Methodologyfor Expert System Shells", May 1990.
Coordination", April 1990.
90/51 Michael BURDA and "Gross Labour Market flows in Europe:90/40 Manfred KETS DE VRIES "Leaden on the Couch: The case of Roberto EP Charles WYPLOSZ Some Stylized Facts", June 1990.OB Calvi", April 1990.
90/52 Lars Tyge NIELSEN "The Utility of Infinite Means", June 1990.90/41FIN/EP
Gabriel HAWAWINI,Itzhak SWARY and
"Capital Market Reaction to theAnnouncement of Interstate Banking
FIN
lk HWAN JANG Legislation", March 1990. 90/53EP
Michael Burda "The Consequences of German Economicand Monetary Union", June 1990.
90/42 Joel STECKEL and "Cross-Validating Regression Models inMKT Wilfried VANHONACKER Marketing Research", (Revised April 1990). 90/54 Damien NEVEN and "European Financial Regulation: A
El' Colin MEYER Framework for Policy Analysis", (Revised90/43 Robert KORAJCZYK and "Equity Risk Premia and the Pricing of May 1990).FIN Claude VIALLET Foreign Exchange Risk", May 1990.
90/55 Michael BURDA and "Intertemporal Prices and the US Trade90/44OB
Gilles AMADO,Claude FAUCHEUX and
"Organisational Change and CulturalRealities: Franco-American Contrasts", April
EP Stefan GERLACH Balance", (Revised July 1990).
Andre LAURENT 1990.
90/56 Damien NEVEN and "The Structure and Determinants of East-WestEP Lars-Hendrik ROLLER Trade: A Preliminary Analysis of the 90/68 Soumitra DUTTA "A Framework and Methodology for Enhancing the
Manufacturing Sector", July 1990 TM/SE Business Impact of Artificial IntelligenceApplications", September 1990
90/57 Lars Tyge NIELSEN Common Knowledge of a Multivariate AggregateFIN/EP/ Statistic", July 1990 90/69 Soumitra DUTTA "A Model for Temporal Reasoning in MedicalTM TM Expert Systems", September 1990
90/58FIN/EP/TIN
Lars Tyge NIELSEN "Common Knowledge of Price and Expected Costin an Oligopolistic Market", August 1990
90/70TM
Albert ANGEHRN "'Triple C': A Visual Interactive MCDSS",September 1990
90/59 Jean DERMINE and "Economies of Scale and 90/71 Philip PARKER and "Competitive Effects in Diffusion Models: AnFIN Lars-Hendrik ROLLER Scope in the French Mutual Funds (SICAV) MKT Hubert GATIGNON Empirical Analysis", September 1990
Industry", August 199090/72 Enver YUCESAN "Analysis of Marko, Chains Using Simulation
90/60 Peri a and "An Interactive Group Decision Aid for TM Graph Models", October 1990TM Tawfik JELASSI Multiobjective Problems: An Empirical
Assessment", September 1990 90/73TM
Arnoud DE MEYER andKasra FERDOWS
"Removing the Barriers in Manufacturing",October 1990
90/61 Pankaj CHANDRA and "Models for the Evlauation of ManufacturingTM Mihkel TOMBAK Flexibility", August 1990
90/74 Sumantra GHOSHAL and "Requisite Complexity: Organising Headquarters-90/62 Damien NEVEN and "Public Policy Towards TV Broadcasting in the SM Nitin NOHRIA Subsidiary Relations in MNCs"EP Ma= VAN DUK Netherlands", August 1990
90/63SM
Sumantra GHOSHAL andEleanor WESTNEY
"Organising Competitor Analysis Systems",August 1990
90/75MKT
Roger BETANCOURT andDavid GAUTSCHI
"The Outputs of Retail Activities: Concepts,Measurement and Evidence"
90/64 Sumantra GHOSHAL "Internal Differentiation and CorporateSM Performance: Case of the Multinational
Corporation", August 1990
90/65 Charles WYPLOSZ "A Note on the Real Exchange Rate Effect ofEP German Unification", August 1990
90/66 Soumitra DUTrA and "Computer Support for Strategic and TacticalTM/SE/FIN Piero BONISSONE Planning in Mergers and Acquisitions",
September 1990
90/67TM/SE/FIN
Soumitra DU'rrA andPiero BONISSONE
"Integrating Prior Cases and Expert Knowledge Ina Mergers and Acquisitions Reasoning System",September 1990