Modeling Marketing Dynamics by Time Series Econometrics
- PDF / 173,960 Bytes
- 17 Pages / 595 x 842 pts (A4) Page_size
- 0 Downloads / 174 Views
Modeling Marketing Dynamics by Time Series Econometrics∗ KOEN PAUWELS † Assistant Professor, Tuck School of Business at Dartmouth
[email protected]
IMRAN CURRIM Chancellors Professor, Graduate School of Management, University of California, Irvine MARNIK G. DEKIMPE Professor, Catholic University Leuven and Erasmus University Rotterdam ERIC GHYSELS Bernstein Distinguished Professor of Finance, Kenan-Flagler Business School, UNC Chapel Hill DOMINIQUE M. HANSSENS Bud Knapp Professor of Marketing, UCLA Anderson School of Management NATALIE MIZIK Assistant Professor, Graduate School of Business, Columbia University PRASAD NAIK Associate Professor, Graduate School of Management, University of California, Davis
Abstract This paper argues that time-series econometrics provides valuable tools and opens exciting research opportunities to marketing researchers. It allows marketing researchers to advance traditional modeling and estimation approaches by incorporating dynamic processes to answer new important research questions. The authors discuss the challenges facing time-series modelers in marketing, provide an overview of recent methodological developments and several applications, and highlight fruitful areas for future research. This discussion is based on the First Annual Conference on ‘Modeling Marketing Dynamics by Time Series Econometrics’ at the Tuck School of Business at Dartmouth, Hanover, New Hampshire, USA on September 16–17, 2004. Keywords: time series models, marketing dynamics, data richness, Lucas critique, impulse response functions
Time-series econometrics has made several important contributions in fundamental areas of marketing. As reviewed in Dekimpe and Hanssens (2000, Table 1), time-series (TS) techniques were initially used in marketing (1) for forecasting purposes, (2) to determine the temporal ordering among variables through Granger-causality tests, or (3) to determine the over-time impact of marketing variables (e.g. through transfer-function analysis). Recently, ∗ Insights
from the First Annual Conference, Tuck School of Business at Dartmouth. author.
† Corresponding
168 Table 1.
PAUWELS ET AL.
Overview of Challenges and Proposed Approaches
Challenges 1. Data Richness: 1.1 Aggregation Over Consumers Over Time Periods
Approaches
1.2 Parameterization (Stores, SKUs) 1.3 Pruning
Segment-level response Optimal data interval Mixed data sampling Pooling parameters Dimension reduction Bias-reducing techniques
2. Lucas Critique
Super-exogeneity tests Varying-parameter models Spectral analysis
3. Broadening Techniques, and Marketing problems
4. Asymmetric Response
Kalman Filter Spectral band-pass analysis Bayesian error-correction Strategic Foresight Marketing-Finance interface Internet Bid analysis Add error correction terms
5. Definition Consistency
Define long-run elasticity
6. Changing Dynamics
Structural breaks Dynamic IRFs Moving Windows
Some relevant papers
Lim et al. (2004) Tellis and Franses (2004) Ghysels et al. (2003) Horv´ath et al. (2004) Pauwels et al. (2004
Data Loading...
