Solving dynamic discrete choice models using smoothing and sieve methods

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Standard

Solving dynamic discrete choice models using smoothing and sieve methods. / Kristensen, Dennis; Mogensen, Patrick K.; Moon, Jong Myun; Schjerning, Bertel.

I: Journal of Econometrics, Bind 223, Nr. 2, 2021, s. 328-360.

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Harvard

Kristensen, D, Mogensen, PK, Moon, JM & Schjerning, B 2021, 'Solving dynamic discrete choice models using smoothing and sieve methods', Journal of Econometrics, bind 223, nr. 2, s. 328-360. https://doi.org/10.1016/j.jeconom.2020.02.007

APA

Kristensen, D., Mogensen, P. K., Moon, J. M., & Schjerning, B. (2021). Solving dynamic discrete choice models using smoothing and sieve methods. Journal of Econometrics, 223(2), 328-360. https://doi.org/10.1016/j.jeconom.2020.02.007

Vancouver

Kristensen D, Mogensen PK, Moon JM, Schjerning B. Solving dynamic discrete choice models using smoothing and sieve methods. Journal of Econometrics. 2021;223(2):328-360. https://doi.org/10.1016/j.jeconom.2020.02.007

Author

Kristensen, Dennis ; Mogensen, Patrick K. ; Moon, Jong Myun ; Schjerning, Bertel. / Solving dynamic discrete choice models using smoothing and sieve methods. I: Journal of Econometrics. 2021 ; Bind 223, Nr. 2. s. 328-360.

Bibtex

@article{c4bd766dbdcc4247b19b9feb05c3c5f7,

title = "Solving dynamic discrete choice models using smoothing and sieve methods",

abstract = "We propose to combine smoothing, simulations and sieve approximations to solve for either the integrated or expected value function in a general class of dynamic discrete choice (DDC) models. We use importance sampling to approximate the Bellman operators defining the two functions. The random Bellman operators, and therefore also the corresponding solutions, are generally non-smooth which is undesirable. To circumvent this issue, we introduce smoothed versions of the random Bellman operators and solve for the corresponding smoothed value functions using sieve methods. We also show that one can avoid using sieves by generalizing and adapting the “self-approximating” method of Rust (1997b) to our setting. We provide an asymptotic theory for both approximate solution methods and show that they converge with N-rate, where N is number of Monte Carlo draws, towards Gaussian processes. We examine their performance in practice through a set of numerical experiments and find that both methods perform well with the sieve method being particularly attractive in terms of computational speed and accuracy.",

keywords = "Dynamic discrete choice, Monte Carlo, Numerical solution, Sieves",

author = "Dennis Kristensen and Mogensen, {Patrick K.} and Moon, {Jong Myun} and Bertel Schjerning",

note = "Publisher Copyright: {\textcopyright} 2020 Elsevier B.V.",

year = "2021",

doi = "10.1016/j.jeconom.2020.02.007",

language = "English",

volume = "223",

pages = "328--360",

journal = "Journal of Econometrics",

issn = "0304-4076",

publisher = "Elsevier",

number = "2",

}

RIS

TY - JOUR

T1 - Solving dynamic discrete choice models using smoothing and sieve methods

AU - Kristensen, Dennis

AU - Mogensen, Patrick K.

AU - Moon, Jong Myun

AU - Schjerning, Bertel

PY - 2021

Y1 - 2021

N2 - We propose to combine smoothing, simulations and sieve approximations to solve for either the integrated or expected value function in a general class of dynamic discrete choice (DDC) models. We use importance sampling to approximate the Bellman operators defining the two functions. The random Bellman operators, and therefore also the corresponding solutions, are generally non-smooth which is undesirable. To circumvent this issue, we introduce smoothed versions of the random Bellman operators and solve for the corresponding smoothed value functions using sieve methods. We also show that one can avoid using sieves by generalizing and adapting the “self-approximating” method of Rust (1997b) to our setting. We provide an asymptotic theory for both approximate solution methods and show that they converge with N-rate, where N is number of Monte Carlo draws, towards Gaussian processes. We examine their performance in practice through a set of numerical experiments and find that both methods perform well with the sieve method being particularly attractive in terms of computational speed and accuracy.

AB - We propose to combine smoothing, simulations and sieve approximations to solve for either the integrated or expected value function in a general class of dynamic discrete choice (DDC) models. We use importance sampling to approximate the Bellman operators defining the two functions. The random Bellman operators, and therefore also the corresponding solutions, are generally non-smooth which is undesirable. To circumvent this issue, we introduce smoothed versions of the random Bellman operators and solve for the corresponding smoothed value functions using sieve methods. We also show that one can avoid using sieves by generalizing and adapting the “self-approximating” method of Rust (1997b) to our setting. We provide an asymptotic theory for both approximate solution methods and show that they converge with N-rate, where N is number of Monte Carlo draws, towards Gaussian processes. We examine their performance in practice through a set of numerical experiments and find that both methods perform well with the sieve method being particularly attractive in terms of computational speed and accuracy.

KW - Dynamic discrete choice

KW - Monte Carlo

KW - Numerical solution

KW - Sieves

U2 - 10.1016/j.jeconom.2020.02.007

DO - 10.1016/j.jeconom.2020.02.007

M3 - Journal article

AN - SCOPUS:85094628639

VL - 223

SP - 328

EP - 360

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

IS - 2

ER -

ID: 270622185

Økonomisk Institut