Heterogeneity in pricing behavior in hybrid DSGE-ABM macrodynamics - Supplementary Material

Published: 6 November 2025| Version 1 | DOI: 10.17632/tcxc62mc82.1
Contributors:
Nikolas Schiozer,
,

Description

Evidence demonstrates that pricing behavior is persistently heterogeneous across firms. We developed a hybrid dynamic stochastic general equilibrium (DSGE)-agent-based model (ABM) to explore how pricing heterogeneity evolves within which firms periodically revise (and possibly switch) pricing strategies. Firms choose a pricing heuristic from a set of alternatives encompassing a cost-based heuristic with a standard markup, a cost-based heuristic with a quality-adjusted markup, and a heuristic of following competitors' pricing behavior. For illustration, we employ a simulated method of moments to calibrate the model with quarterly data for the United States. The model qualitatively replicates several empirical facts about firms’ strategic pricing behavior, particularly with respect to the frequency and magnitude of price adjustments. Two key findings are price stickiness and persistent heterogeneity in pricing behavior, with no single pricing rule becoming dominant. A local sensitivity analysis was conducted on selected key micro-level and macro-level parameters, with results indicating that price stickiness and persistent heterogeneity in firms’ pricing behavior are driven by a combination of both micro- and macroeconomic factors. Embedding some tenets of the ABM approach such as introducing heterogeneous firms evolutionarily learning and adapting to the environment into the DSGE model enables us to qualitatively reproduce the persistent heterogeneity in pricing behavior across firms without compromising its ability to reproduce several stylized macroeconomic facts. A key broad policy implication is that the microdiversity and macrodynamics of the economy are inherently co-evolutionary phenomena.

Files

Steps to reproduce

The repository contains two R scripts and one RDS file: ABM-DSGE_EM.R Main script used to run the simulations and generate all figures and sensitivity tables. GA_ABM-DSGE_EM.R Script used to calibrate the model (genetic algorithm / grid search). It shares the same core model code, so you can work entirely from a single file if preferred. TABELAS.rds The model includes a local sensitivity analysis routine that systematically varies key structural parameters around their baseline values. The results are summarized in the file TABELAS.rds, which is automatically generated by the script “ABM-DSGE_EM.R”. To run the R scripts, the following are required: • R (≥ 4.0), along with a standard R IDE such as RStudio (recommended) or VS Code with R extensions. • Common R packages (e.g., ggplot2, dplyr, etc.). If any are missing, R will prompt you to install them automatically. The two scripts are designed to demonstrate different exercises carried out during the development of the model presented in “Heterogeneity in Pricing Behavior in Hybrid DSGE-ABM Macrodynamics.” In the first R file, “ABM-DSGE_EM.R”, you will find the initial conditions, parameters, functions, simulation loop, plotting routines, and the code used to generate the sensitivity-analysis tables described in the paper. The code is fully documented, with clear comments explaining the purpose of each section. References are included to indicate which commands generate the figures and tables reported in the article. To run “ABM-DSGE_EM.R”: 1. Open “ABM-DSGE_EM.R” in RStudio (or another R IDE). 2. Source the entire script (Ctrl/Cmd + Shift + Enter in RStudio). 3. The script will: o Run the baseline simulations, o Produce the graphs shown in the paper, and o Compute the sensitivity-analysis tables. By default, plots display in the Viewer/Plots pane. You can also save all outputs to disk (see “Saving outputs” below). The second R file, “GA_ABM-DSGE_EM.R”, provides the code used for model calibration. All relevant instructions and model details are fully documented within the script. To explore the calibration procedure, open this file. Since both scripts include the same core model code, users may work exclusively with one of them—depending on whether they are interested in running simulations or performing calibration. To run “GA_ABM-DSGE_EM.R”: 1. Open “GA_ABM-DSGE_EM.R” in RStudio (or another R IDE). 2. Source the script to run the calibration routine. 3. Use the calibrated parameters to re-run “ABM-DSGE_EM.R” if you wish to regenerate the figures using the calibrated values. Saving outputs • Figures and tables can be saved directly to your computer. • Look for the lines in the code that contain the command “ggsave” in ABM-DSGE_EM.R and enable them by removing the # at the beginning of each line. • Then, set the desired file paths and output options to specify where and how the figures will be saved.

Institutions

Categories

Macroeconomics, Agent-Based Modeling, Pricing, Dynamic Stochastic General Equilibrium, Population Heterogeneity

Funders

Related Links

Article
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5030953
is related to this dataset

Licence