Use this URL to cite or link to this record in EThOS:
Title: Dynamic demand estimation for storable goods
Author: Crawford, Alan
ISNI:       0000 0004 7429 067X
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Restricted access.
Access from Institution:
The market for storable goods is worth more than £1 trillion and accounts for a large portion of household grocery expenditures. Their defining characteristics are that they can be stored for future consumption and are often sold using promotions. Both create inter-temporal links in consumer demand. Measuring both intra and inter-temporal substitution patterns is therefore central to understanding consumer behaviour for this large class of industries. Quantification of substitution patterns is made more challenging because choice sets facing consumers tend to be large (i.e. 100 products). Therefore, demand is inherently dynamic and high-dimensional. As a result, dynamic demand models suffer from the curse of dimensionality and are computationally intensive to estimate and solve. In this thesis I develop two complementary approaches to the estimation of demand models for storable goods that incorporate demand dynamics and apply them to the UK laundry detergent industry. The first is a computationally light approach that can be easily implemented within a policy making timeframe and only requires data routinely collected in antitrust investigations. It shows how accounting margins can be combined with a static demand model to estimate a set of price elasticities that are consistent with dynamic demand responses to price changes. I show how these can be input into empirical tools used by firms and policy makers. The second paper develops a high-dimensional dynamic discrete-continuous demand model for storable fast moving consumer goods. Assumptions of existing dynamic demand models are relaxed while retaining computational tractability. As a result, the model captures rich inter- and intra-temporal substitution patterns, allows for a detailed understanding of dynamic consumer behaviour, and provides a framework with wide applicability. To estimate and solve the dynamic demand model, I use techniques from approximate dynamic programming, large-scale dynamic programming in economics, machine learning, and statistical computing.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available