Working Papers

Estimating dynamic diversion ratios in storable good industries

Online Appendix

This paper estimates diversion ratios capturing the influence of dynamic substitution patterns on forward-looking storable goods firms’ profits. These dynamic diverted value ratios are key inputs in a new dynamic upwards price pressure test, dGUPPI. This new method obviates the need to estimate consumers’ dynamic demand functions and is computable within a policy-making timeframe. To illustrate its practical use to policymakers, it is applied to the UK laundry detergent industry from 2002 to 2012. Estimated bounds on the dynamic diverted value ratios calculate set-valued dGUPPI that overturn standard static empirical policy tools that would incorrectly permit an anticompetitive hypothetical merger.

High dimensional high frequency retail price dynamics: accounting for missing prices and quantities joint with Lars Nesheim

We use scanner data to study the dynamics of prices and quantities in a high dimensional setting. In this setting, there are large missing data problems due to sample selection. We develop a method to solve these missing data problems. Our solution is to develop a “low rank” factor model to capture the dynamics. We assume that the dynamics of high dimensional prices and consumer demand are jointly driven by a common set of low-dimensional factors. These factors evolve according to a simple autoregressive model. In addition, we model the price process using a switching model with switching between a “regular price” process and a “sale price” process. We then analyse the implications for price index calculation and for demand estimation.

High-dimensional inventories and consumer dynamics: demand estimation for fast moving consumer goods

New version coming soon

This paper develops a high-dimensional dynamic discrete-continuous demand model for storable fast moving consumer goods. Assumptions of existing 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. In this paper I apply the model to the UK laundry detergent sector using household level purchase data.

Work in Progress

The effect of the financial crisis on bank lending to SMEs joint with Ben Hemingway