Competitive reaction- and feedback effects based on VARX models of pooled store data

Horváth, Csilla; Leeflang, Peter S. H.; Wieringa, Jaap E.; Wittink,

doi:10.1016/j.ijresmar.2005.09.007

Cited by 39 publications

(33 citation statements)

References 33 publications

(32 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To estimate immediate, gross, and net effects, I use impulse response analyses (IRA). I illustrate the use of a VARX model by discussing a model specified and calibrated by Horvàth et al (2005). This model simultaneously considers advanced market response and advanced competitive reaction functions, and relies for calibration on pooled store data for each of three brands of tuna fish 11 .…”

Section: Varx Modelsmentioning

confidence: 99%

“…The SCAN*PRO model includes several own-and cross-brand promotional variables: price index, feature only, display only, and feature and display. Horvàth et al (2005) extend this model by including dynamic price promotion effects (delayed responses) and purchase reinforcement effects (through lagged sales). However, I do not include separate lagged non-price instruments to reduce concerns about the degrees of freedom.…”

Section: Response Functionsmentioning

confidence: 99%

“…Horvàth et al (2005) specify the market response function as: where lnS qi, is the natural logarithm of sales of brand i in store q in week t; lnPI qik,t is log price index (actual to regular price) of brand i in store q in week t; (k = 1 denotes the feature/display-supported price cuts and k = 2 denotes price cuts that are not supported); F qj,t is a feature-only dummy variable for non-price promotions of brand j in store q at time t; D qj,t is a display-only dummy variable for non-price promotions of brand j in store q at time t; FD qj, t is the combined use of feature and display supports of non-price promotions of brand j in store q at time t 12 ; α qi is a store-specific intercept for brand i and store q; β PIijk,t* = (pooled) elasticity of brand i's sales with respect to brand j's price index; φ ij,t* is the (pooled) substitution elasticity of brand i's sales with respect to competitive (j) sales in week t (i ≠ j); β Fij , β Dij , β FDij are the effects of feature-only (F ), display-only (D), and feature and display (FD); P ijk SP represents the number of lags for price index variable k of brand i included in the equation for brand j; P ij SP is the number of lags of the sales variable of brand i included in the equation for brand j; n is the number of brands in the product category; Q is the number of stores; and ε qi,t are disturbances. Horvàth et al (2005) test for the equality of slopes across stores and fail to reject this null hypothesis. Therefore, the specification of the demand model does not allow for slope heterogeneity.…”

Section: Response Functionsmentioning

confidence: 99%

“…However, I may doubt whether competitors react to price indexes. Horvàth et al (2005) use these indexes instead of regular or promotional prices because VARX models require the same set of model variables. The reaction functions also account for inertia in decision-making and coordination between own-brand instruments (internal decisions): where the variables are defined as in equation (21) (ℓ = 1, 2).…”

Section: Reaction Functionsmentioning

confidence: 99%

“…They may also track and react to other brands' performance (cross-feedback effects). Horvàth et al (2005) estimate a VARX model based on Equations (21) and (22) using two data sets covering two years of observations of weekly data. The data sets represent 24 (tuna fish) and 27 (shampoo) stores.…”

Section: Reaction Functionsmentioning

confidence: 99%

See 4 more Smart Citations

Modeling Competitive Reaction Effects

Leeflang

2008

Schmalenbach Bus Rev

View full text Add to dashboard Cite

abstRaCtin this study i critically review models that specify competitive reaction effects. i discuss different model structures and summarize my findings on competitive reaction effects and factors that explain competitive reactions. i discuss the many models of competitive market response that have been developed and classify them into twelve sets of models that are related to each other in a logical manner through the evolutionary model-building concept.JeL-Classification: C2, L13, m31.Keywords: Competition; marketing; models. evolutionaRy Model buildingIn this study I review models that specify competitive reaction effects. Many of the models that have been developed in the past 40 years are connected to models that were developed earlier. Here, I make the connections between different groups of models by using the evolutionary model-building concept. The basic idea of evolutionary model building is that the model builder begins with a relative simple model. Simplicity is paramount so that managers understand these models. Early applications may reveal shortcomings, and diagnostics can be used to guide further model development. The builder then expands the model, which will lead increasing the complexity of the model (Little (1970); Urban and Karash (1971)).The evolutionary model-building concept primarily has been applied in the context of marketing decision models. By gradually adding more complexities to simpler models, both model builders and model users jointly develop a more complete model by incorporating additional elements. The manager (model user) still understands this more complex

show abstract

Section: Varx Modelsmentioning

confidence: 99%

Section: Response Functionsmentioning

confidence: 99%

Section: Response Functionsmentioning

confidence: 99%

Section: Reaction Functionsmentioning

confidence: 99%

Section: Reaction Functionsmentioning

confidence: 99%

See 3 more Smart Citations

Modeling Competitive Reaction Effects

Leeflang

2008

Schmalenbach Bus Rev

View full text Add to dashboard Cite

show abstract

Time-Series Models in Marketing

Dekimpe

Franses²,

Hanssens³

et al. 2008

International Series in Operations Research &Amp; Management Science

View full text Add to dashboard Cite

identify three past stages in marketing model building and implementation, review the current status, and provide some intriguing thoughts on how the model-building process may evolve in response to ongoing and anticipated developments in the marketing environment. It is interesting to note that time-series techniques are not mentioned in their review of the past, receive considerable attention in their assessment of the current situation (mainly in the context of the insights these techniques can provide on marketing's long-run effectiveness), and are again absent in their speculations about the future. In this paper, we complement and extend on Leeflang and Wittink's discussion in three ways. First, we discuss why, in the past, time-series techniques received little attention from marketing model builders and users. Second, we show how each of the identified obstacles has recently been attenuated, and review more extensively the recent contributions of time-series applications to marketing science. Finally, we elaborate on some expected future developments in marketing research, and discuss opportunities and challenges to applied timeseries modelers.

show abstract

Modeling Competitive Responsiveness

Leeflang

2008

International Series in Operations Research &Amp; Management Science

View full text Add to dashboard Cite

Competitive reaction- and feedback effects based on VARX models of pooled store data

Cited by 39 publications

References 33 publications

Modeling Competitive Reaction Effects

Modeling Competitive Reaction Effects

Time-Series Models in Marketing

Modeling Competitive Responsiveness

Contact Info

Product

Resources

About