Skip to content

Latest commit

 

History

History
124 lines (97 loc) · 6.3 KB

credits.md

File metadata and controls

124 lines (97 loc) · 6.3 KB

Credits & Acknowledgments

pylcm stands on a large body of methodological and applied research, and on a vibrant open-source ecosystem for solving discrete-continuous dynamic models. This page credits the methods, models, and software pylcm draws on. We are grateful to all of these authors and maintainers.

Authors & maintainers

pylcm is developed under the OpenSourceEconomics organization by Tim Mensinger, Maximilian Jahn, Janos Gabler, and Hans-Martin von Gaudecker.

Numerical methods

The endogenous grid method (EGM)

The endogenous grid method — inverting the Euler equation on an exogenous end-of-period grid instead of root-finding on a dense grid — originates with Carroll (2006), "The Method of Endogenous Gridpoints for Solving Dynamic Stochastic Optimization Problems," Economics Letters 91(3), 312–320, doi:10.1016/j.econlet.2005.09.013.

Discrete-continuous EGM (DC-EGM)

pylcm's DCEGM solver implements the discrete-continuous endogenous grid method of Iskhakov, Jørgensen, Rust & Schjerning (2017), "The endogenous grid method for discrete-continuous dynamic choice models with (or without) taste shocks," Quantitative Economics 8(2), 317–365, doi:10.3982/QE643. The deterministic retirement model from that paper is shipped as a closed-form test oracle (see below).

Upper-envelope refinement

pylcm refines the (non-monotone) EGM candidate correspondence into the upper envelope with the Fast Upper-Envelope Scan (FUES) of Dobrescu & Shanker (2022), "Fast Upper-Envelope Scan for Discrete-Continuous Dynamic Programming," SSRN working paper 4181302. The treatment of non-smooth, non-concave value functions traces to Fella (2014), "A Generalized Endogenous Grid Method for Non-Smooth and Non-Concave Problems," Review of Economic Dynamics 17(2), 329–344, doi:10.1016/j.red.2013.07.001.

pylcm's FUES backend is adapted from OpenSourceEconomics/upper-envelope (Apache-2.0, © The Upper-Envelope Authors), then substantially modified for the JAX kernel. The reference implementation by the method's authors is akshayshanker/FUES.

Single-asset EGM

The EGM solver is the plain endogenous grid method above, applied to a regime with one continuous (Euler) state and no discrete choice. Inverting the Euler equation on the post-decision savings grid solves such a period exactly, so no upper envelope is involved — see the Carroll (2006) reference at the top of this file.

Multidimensional & nested EGM

pylcm solves models with more than one continuous (Euler) state by nesting, through its NEGM solver. Those methods are due to:

  • Druedahl (2021), "A Guide on Solving Non-convex Consumption-Saving Models," Computational Economics 58(3), 747–775, doi:10.1007/s10614-020-10045-x — nested EGM (NEGM). Code: NumEconCopenhagen/ConsumptionSaving.
  • Druedahl & Jørgensen (2017), "A General Endogenous Grid Method for Multi-Dimensional Models with Non-Convexities and Constraints," Journal of Economic Dynamics and Control 74, 87–107, doi:10.1016/j.jedc.2016.11.005 — G2EGM. Code: JeppeDruedahl/G2EGM.
  • Dobrescu & Shanker (2024), "Using Inverse Euler Equations to Solve Multidimensional Discrete-Continuous Dynamic Models: A General Method," SSRN working paper 4850746 — the inverse-Euler / Rooftop-Cut (RFC) multidimensional envelope. Code: akshayshanker/InverseDCDP.

Stochastic-process discretization

pylcm's AR(1) / Markov discretization follows Tauchen (1986), Rouwenhorst (1995), and Kopecky & Suen (2010), with the Tauchen implementation following QuantEcon. The choice of Rouwenhorst as the default for non-stationary lifecycle processes follows Fella, Gallipoli & Pan (2019), "Markov-Chain Approximations for Life-Cycle Models," Review of Economic Dynamics 34, 183–201, doi:10.1016/j.red.2019.03.013, against whose results pylcm's discretization accuracy is validated. Code & data: RePEc red/ccodes/17-149.

Replicated example models

Models shipped with pylcm (src/lcm_examples/) that replicate published work:

  • Iskhakov, Jørgensen, Rust & Schjerning (2017) — the deterministic retirement model (lcm_examples.iskhakov_et_al_2017), used as a closed-form test oracle.
  • Mahler & Yum (2024), "Lifestyle Behaviors and Wealth-Health Gaps in Germany," Econometrica 92(5), 1697–1733, doi:10.3982/ECTA20603 — a finite-horizon discrete-continuous lifecycle model (lcm_examples.mahler_yum_2024). The replication package is available via the Econometrica supplemental materials.

Pedagogy

pylcm's design and documentation are informed by Sargent & Stachurski's Dynamic Programming (QuantEcon, 2024).

The open-source ecosystem

pylcm interoperates with, learns from, and is grateful to the broader ecosystem for solving discrete-continuous dynamic models: