Maximum Simulated Likelihood Estimation Research Articles

Fitting mixed random regret minimization models using maximum simulated likelihood

In this article, we describe the mixrandregret command, which extends the randregret command introduced in Gutiérrez-Vargas, Meulders, and Vandebroek (2021, Stata Journal 21: 626–658) by allowing random coefficients in random regret minimization models. The newly developed mixrandregret command allows the user to specify a combination of fixed and random coefficients in the regret function of the classical random regret minimization model introduced in Chorus (2010, European Journal of Transport and Infrastructure Research 10: 181–196). In addition, the user can specify normal and lognormal distributions for the random coefficients using the appropriate command’s options. The models are fit by maximum simulated likelihood estimation using numerical integration to approximate the choice probabilities.

Simulated maximum likelihood estimation of the sequential search model

AbstractWe propose a new approach to simulate the likelihood of the sequential search model. By allowing search costs to be heterogeneous across consumers and products, we directly compute the joint probability of the search and purchase decisions when consumers are searching for the idiosyncratic preference shocks in their utility functions. Under the assumptions of Weitzman’s sequential search algorithm, the proposed procedure recursively makes random draws for each quantity that requires numerical integration while enforcing the conditions stipulated by the algorithm. In an extensive simulation study, we compare the proposed method with existing likelihood simulators that have recently been used to estimate the sequential search model. The proposed method attributes the uncertainty in the search order to the consumer-product-level distribution of search costs and the uncertainty in the purchase decision to the distribution of match values across consumers and products. This results in more precise estimation and an improvement in prediction accuracy. We also show that the proposed method allows for different assumptions on the search cost distribution and that it recovers consumers’ relative preferences even if the utility function and/or the search cost distribution is mis-specified. We then apply our approach to online search data from Expedia for field-data validation. From a substantive perspective, we find that search costs and “position” effects affect products in the lower part of the product listing page more than they do those in the upper part of the page.

Biases in the Maximum Simulated Likelihood Estimation of the Mixed Logit Model

In a recent study, it was demonstrated that the maximum simulated likelihood (MSL) estimator produces significant biases when applied to the bivariate normal and bivariate Poisson-lognormal models. The study’s conclusion suggests that similar biases could be present in other models generated by correlated bivariate normal structures, which include several commonly used specifications of the mixed logit (MIXL) models. This paper conducts a simulation study analyzing the MSL estimation of the error components (EC) MIXL. We find that the MSL estimator produces significant biases in the estimated parameters. The problem becomes worse when the true value of the variance parameter is small and the correlation parameter is large in magnitude. In some cases, the biases in the estimated marginal effects are as large as 12% of the true values. These biases are largely invariant to increases in the number of Halton draws.

The impacts of food certificate adoption on e-commerce income among small online agri-food sellers

PurposeThis paper aims to assess the determinants of food certificate adoption and analyse the impacts of food certificates on e-commerce income among small online agri-food sellers in Indonesia.Design/methodology/approachThe authors used data from an online survey of 228 small-online agri-food sellers in East Java, Indonesia. This study aims to focus on two food certificates: a mandatory Halal (Islamic dietary law) certificate and the P-IRT certificate, a food safety certificate for home-based businesses. A maximum simulated likelihood (MSL) estimator was employed to account for selection bias and endogeneity.FindingsThe study highlights the continued importance of certification in agri-food markets, including e-commerce and the need to consider the degree of substitutability and resource allocation between multiple food certificates. It finds that online agri-food sellers adopting the Halal certificate earn two to three times higher compared to non-adopters. Conversely, the gross income per month from e-commerce sales is 78% lower among those adopting the P-IRT certificate. Moreover, access to regulatory information sources motivates the likelihood of adopting food certificates. In contrast, the business size, marketing channels, contractual relationship and management capabilities are insignificant factors for the adoption of any of the Halal and P-IRT certificate combinations.Research limitations/implicationsResults from this research might be specific to the context of the focus study area, thereby reducing their generalisability. In addition to gathering representative samples, future research should also capture more complex dimensions of food certificates. These include the cost of acquiring food certificates, online sellers' perceptions of food certificate adoption, and emerging topics such as group certification and the use of technology.Originality/valueTo the authors' knowledge, this research is one of the first studies investigating the adoption of food certificates within the e-commerce setting. This study also contributes to the small number of studies looking at multiple certificate adoption and food certificate issues from the retailers' perspectives

Estimation of Realized Asymmetric Stochastic Volatility Models Using Kalman Filter

Despite the growing interest in realized stochastic volatility models, their estimation techniques, such as simulated maximum likelihood (SML), are computationally intensive. Based on the realized volatility equation, this study demonstrates that, in a finite sample, the quasi-maximum likelihood estimator based on the Kalman filter is competitive with the two-step SML estimator, which is less efficient than the SML estimator. Regarding empirical results for the S&P 500 index, the quasi-likelihood ratio tests favored the two-factor realized asymmetric stochastic volatility model with the standardized t distribution among alternative specifications, and an analysis on out-of-sample forecasts prefers the realized stochastic volatility models, rejecting the model without the realized volatility measure. Furthermore, the forecasts of alternative RSV models are statistically equivalent for the data covering the global financial crisis.

Climate-Smart Conservation Agriculture, Farm Values and Tenure Security: Implications for Climate Change Adaptation and Mitigation in the Congo Basin

Background and Research Aims: Agriculture through deforestation is an important threat to biodiversity conservation in the Congo Basin’s tropical forest. The policy challenge is not only to promote adaptation to perceived climate change but also to promote forest conservation. The aim of this study is to provide empirical evidence on the impact of farm-level investments in climate-smart agricultural practices related to conservation agriculture in some Congo Basin countries. The hypothesis is that property rights to land and trees play a fundamental role in governing the patterns of investment, forestland management for conservation, as well as in the profitability of agriculture. Methods: A Simulated Maximum Likelihood Estimation using a Mixed Logit model is used to test farmers’ choice of agricultural system and a farmland value model for each agricultural system which includes determinants of tenure or property rights, climate, soils, and socioeconomic variables such as education and gender. The data was collected from more than 600 farms covering 12 regions and 45 divisions in 3 countries, Cameroon, the Central African Republic and the Democratic Republic of Congo. Results: Farmers choose one of three agricultural systems to maximize farm profit mindful of the current tenure regime and environmental conditions. Conservation agriculture techniques within climate-smart practices show benefits for smallholder farmers through improvements in soil health, soil moisture retention and enhanced crop yields. The rights to access, withdraw, manage, as well as exclude others from land and trees affect both the farmers' choice of system and the profit earned from the chosen system. Conclusion: Farm-level investments improve farm incomes and enhance conservation effort for farmers perceiving climate change. Implications for Conservation: Climate change adaptation through planting of trees improves soil stability, restores ecosystems and creates a safe haven for biodiversity. Secure land tenure promotes better forestland management and reduces land degradation in vulnerable communities.

Is Contract Farming with Modern Distributors Partnership for Higher Returns? Analysis of Rice Farm Households in Taiwan

This study provides empirical evidence of the economic effect of contract farming for the agriculture sector dominated by smallholder farms. In light of the association between contract farming and modern food distribution channels, we categorize the adoption decisions of contract farming and modern marketing channels into four mutually exclusive choices and investigate their economic effects through the simulated maximum likelihood estimation of the multinomial treatment effects model. The results provide empirical evidence supporting higher returns from the dual partnerships as choosing modern distributors generates more revenues for the those participating in contract farming than for those with no contract farming, and contract farming is more likely to help generate more revenues for those who have taken modern distributors as their major marketing channel compared with those relying on traditional channels. Moreover, we examine whether any distributional pattern of marginal economic effects, of either contract farming or modern marketing channel, is present among farmers at various scales by using the conditional and unconditional quantile regression models. Our findings suggest that the marginal treatment effects are generally in an increasing trend as the quantile increases, implying that the economic effects of contract farming or partnership with modern distributors are more pronounced for higher returns among rice farmers in Taiwan. This finding has great policy implications for developing sustainable agriculture and food supply when facing greater uncertainties due to global warming in the future, especially in an agriculture sector with most smallholder farmers.

Risk preferences, gender effects and Bayesian econometrics

Gender differences in decision making is a topic that has attracted much attention in the literature and the debate seems to be inconclusive. A method that is often used in the economics literature to account for gender effects is by estimating econometric structural models and testing the significance of the estimated parameters. In this paper we focus on estimations of preference models and we show how omitting to account for behavioural heterogeneity can lead to failures in identifying potential differences. Using data from risky choice experiments, we compare the traditional representative agent Maximum Likelihood Estimation approach against two more flexible inference methods that allow for heterogeneity at the individual level, the Maximum Simulated Likelihood Estimation, and the Hierarchical Bayesian modelling. We show how ignoring heterogeneity may lead to failures capturing gender differences and we suggest the use of Bayesian modelling to effectively estimate the underlying parameters.

Jackknife bias reduction for simulated maximum likelihood estimator of discrete choice models

We propose to reduce asymptotic biases of simulated maximum likelihood estimators (SMLE) by using a jackknife method similar to Dhaene and Jochmans (2015). Because the jackknife method does not require an explicit characterization of the bias, it may be a practically attractive alternative to Lee’s (1995) estimator.

Scaling Bayesian inference of mixed multinomial logit models to large datasets

Variational inference methods have been shown to lead to significant improvements in the computational efficiency of approximate Bayesian inference in mixed multinomial logit models when compared to standard Markov-chain Monte Carlo (MCMC) methods without increasing estimation bias. However, despite their demonstrated efficiency gains, existing methods still suffer from important limitations that prevent them to scale to large datasets, while providing the flexibility to allow for rich prior distributions and to capture complex posterior distributions. To effectively scale Bayesian inference in Mixed Multinomial Logit models to large datasets, we propose an Amortized Variational Inference approach that leverages stochastic backpropagation, automatic differentiation and GPU-accelerated computation. Moreover, we show how normalizing flows can be used to increase the flexibility of the variational posterior approximations. Through an extensive simulation study and real data for transport mode choice from London, we empirically show that the proposed approach is able to achieve computational speedups of multiple orders of magnitude over traditional maximum simulated likelihood estimation (MSLE) and MCMC approaches for large datasets without compromising estimation accuracy.

Biases in Maximum Simulated Likelihood Estimation of Bivariate Models

Abstract This paper finds that the maximum simulated likelihood (MSL) estimator produces substantial biases when applied to the bivariate normal distribution. A specification of the random parameter bivariate normal model is considered, in which a direct comparison between the MSL and maximum likelihood (ML) estimators is feasible. The analysis shows that MSL produces biased results for the correlation parameter. This paper also finds that the MSL estimator is biased for the bivariate Poisson-lognormal model, developed by Munkin and Trivedi (1999. “Simulated Maximum Likelihood Estimation of Multivariate Mixed-Poisson Regression Models, with Application.” The Econometrics Journal 2: 29–48). A simulation study is conducted, which shows that MSL leads to serious inferential biases, especially large when variance parameters in the true data generating process are small. The MSL estimator produces biases in the estimated marginal effects, conditional means and probabilities of count outcomes.

Designed quadrature to approximate integrals in maximum simulated likelihood estimation

SummaryMaximum simulated likelihood estimation of mixed multinomial logit models requires evaluation of a multidimensional integral. Quasi-Monte Carlo (QMC) methods such as Halton sequences and modified Latin hypercube sampling are workhorse methods for integral approximation. Earlier studies explored the potential of sparse grid quadrature (SGQ), but SGQ suffers from negative weights. As an alternative to QMC and SGQ, we looked into the recently developed designed quadrature (DQ) method. DQ requires fewer nodes to get the same level of accuracy as QMC and SGQ, is as easy to implement, ensures positivity of weights, and can be created on any general polynomial space. We benchmarked DQ against QMC in a Monte Carlo and an empirical study. DQ outperformed QMC in all considered scenarios, is practice ready, and has potential to become the workhorse method for integral approximation.

WITHDRAWN: A Simulation on lattice reduction techniques for wireless communication MIMO QAM 64x64 and above

Abstract In wireless communication system, signal those pass through in transmission and receiving antenna in MIMO (Multi-input multi- output) channel to lattice reduce over a QAM (quadrature-amplitude-modulation) system, which is simulated maximum likelihood (ML) estimation and minimum mean square error (MMSE) . In this paper “A Simulation on Lattice Reduction Techniques for wireless Communication MIMO with QAM 64x64 and above”, here the minimum likelihood (ML) estimation and MMSE for multi-input multi-output (MIMO) channels. Generally, in OFDM (orthogonally frequency division multiplexing) system of a MMSE system completely eliminates interference and the results in suboptimal performance due to noise enhancement in symbol. The main aim simulated MMSE detector that the transmitter must spend within M QAM quadrature- amplitude-modulation () alphabet and the ML, MMSE allow to transmitted signal and remove the interference which comes from in receiver and pass through it. Which is QAM 64x64 above simulated BER (bit error rate) with SNR.

Maximum simulated likelihood estimation of the seemingly unrelated stochastic frontier regressions

In this paper, we use the maximum simulated likelihood (MSL) approach to estimate multiple stochastic frontier (SF) models with random effects and correlated composite errors. We show that the separate estimation of the single equation ignores the correlation between the composite errors and causes significant efficiency loss in estimation. In addition to using Monte Carlo simulation to investigate the finite sample performance of the simulated estimator, we demonstrate the usefulness of our approach in estimating the technical efficiency of Taiwan’s international hotels based on their accommodation and restaurant divisions.

Heterogeneous Demand and Supply for an Insurance‐linked Credit Product in Kenya: A Stated Choice Experiment Approach

AbstractWe employ a discrete choice experiment to elicit demand and supply side preferences for insurance‐linked credit, a promising market‐based tool for managing agricultural weather risks and providing access to credit for farmers. We estimate preference heterogeneity using primary data from smallholder farmers and managers of lenders/insurers combined with household socio‐economic survey data in Kenya. We analyse the choice data using maximum simulated likelihood and Hierarchical Bayes estimation of a mixed logit model. Although there are some similarities, we find that there is conflicting demand and supply side preferences for credit terms, collateral requirements, and loan use flexibility. We also analyse willingness to buy and willingness to offer for farmers and suppliers, respectively, for the risk premium for different attributes and their levels. Identifying the preferred attributes and levels for both farmers and financial institutions can guide optimal packaging of insurance and credit providing market participation and adoption motivation for insurance‐bundled credit product.

Estimation of dynamic models of recurrent events with censored data

Summary In this paper, we consider estimation of dynamic models of recurrent events (event histories) in continuous time using censored data. We develop maximum simulated likelihood estimators where missing data are integrated out using Monte Carlo and importance sampling methods. We allow for random effects and integrate out this unobserved heterogeneity using a quadrature rule. In Monte Carlo experiments, we find that maximum simulated likelihood estimation is practically feasible and performs better than both listwise deletion and auxiliary modelling of initial conditions. In an empirical application, we study ischaemic heart disease events for male Maoris in New Zealand.

Youth Awareness and Expectations about GMOs and Nuclear Power Technologies within the North American Free Trade Bloc: A Retrospective Cross-Country Comparative Analysis

This study reports on the cross-country heterogeneity in youth awareness and expectations about genetically modified organisms (GMOs) and nuclear power technology (NPT) within the North American free trade area (NAFTA). Models are estimated with data on youth respondents from the USA, Canada and Mexico, using seemingly unrelated bivariate weighted ordered probit regression, with maximum simulated likelihood estimation. Our findings show that the diffusion of technology and information within the trade bloc, for the 20 years prior to the 2015 data collection period, did not significantly contribute to cross-country convergence in youth awareness and expectations about GMOs and NPTs. Indeed, with regard to awareness, compared to youth from the USA, those from Canada show 15% (GMOs) and 7.1% (NPT) more awareness, respectively; while youth from Mexico show 34.4% and 19.5% less awareness about GMOs and NPT, respectively. With respect to expectations about future developments of the two technological artifacts, compared to youth from the USA, those from Canada and Mexico are 34.4% and 39.9% more optimistic about GMOs, respectively, while 15% and 49.7% are more optimistic about NPT. Overall, our findings show that the youth population within NAFTA is 2.5% and 6.7% more optimistic about GMOs and NPT for each level of increase in their awareness about the two technologies, respectively. Theoretically, our results seem to reject the hypothesis of NAFTA being a technology convergence country club in the Schumpeterian view, while seemingly supporting the existence of heterogeneous growth regimes within NAFTA.

Risk aversion, moral hazard, and gender differences in health care utilization

This paper uses truncated count model with endogeneity and simulated maximum likelihood estimation technique to estimate gender differences in moral hazard in health care insurance. We use the dataset that consists of invoices for all outpatient services from a regional hospital in Croatia. Our theoretical model predicts that higher risk aversion is associated with smaller ex-post moral hazard effect. If women are more risk averse than men, then the moral hazard effect due to health insurance should be lower in women than in men. After adjusting for the sample selection in the estimation, we found a statistically significant evidence of moral hazard for the general population but statistically insignificant difference in moral hazard between men and women.

Selection and Incentive Effects of Gatekeeping on Healthcare Utilisation in Germany

The aim of this piece of research was to assess whether or not gatekeeping participation is exogenous with respect to healthcare utilisation. To do so, an endogenous treatment regression model was be employed, which required maximum simulated likelihood estimation. Within this model, a latent factor structure is applied. This procedure allows to separate the effect of selection on unobservables from the causal incentive effect. Results imply that unobservable characteristics are conducive to both healthcare utilisation and gatekeeping participation. Even more so, selection effects outweigh the incentive effects. Overall, gatekeeping participation causally strengthens the role of the general practitioner, decreases emergency admissions, while decreasing both ambulatory and total medical costs. A likelihood-ratio test supports the rejection of exogeneity in three out of four utilisation measures. A literature review to make an informed decision about the required number of simulation draws was conducted. Also, extensive robustness analysis suggests that simulation error did not bias the results.

Bayesian estimation of mixed multinomial logit models: Advances and simulation-based evaluations

Variational Bayes (VB) methods have emerged as a fast and computationally-efficient alternative to Markov chain Monte Carlo (MCMC) methods for scalable Bayesian estimation of mixed multinomial logit (MMNL) models. It has been established that VB is substantially faster than MCMC at practically no compromises in predictive accuracy. In this paper, we address two critical gaps concerning the usage and understanding of VB for MMNL. First, extant VB methods are limited to utility specifications involving only individual-specific taste parameters. Second, the finite-sample properties of VB estimators and the relative performance of VB, MCMC and maximum simulated likelihood estimation (MSLE) are not known. To address the former, this study extends several VB methods for MMNL to admit utility specifications including both fixed and random utility parameters. To address the latter, we conduct an extensive simulation-based evaluation to benchmark the extended VB methods against MCMC and MSLE in terms of estimation times, parameter recovery and predictive accuracy. The results suggest that all VB variants with the exception of the ones relying on an alternative variational lower bound constructed with the help of the modified Jensen’s inequality perform as well as MCMC and MSLE at prediction and parameter recovery. In particular, VB with nonconjugate variational message passing and the delta-method (VB-NCVMP-Δ) is up to 16 times faster than MCMC and MSLE. Thus, VB-NCVMP-Δ can be an attractive alternative to MCMC and MSLE for fast, scalable and accurate estimation of MMNL models.