Search

Going back to the roots of modern business cycle modelling, the paper advances a macroeconomic model of a most elementary (and now largely forgotten) cycle mechanism within a growth framework. Essentially, it combines a nonlinear Harrodian instability for a business sentiment index with Harrod’s and Kaldor’s old idea that firms tend to reduce investment if they have already built up high capacities relative to their assessment of the normal market potential. For greater realism, a previous version is augmented by a delay in the capital growth rate and by in general three random shock processes. This makes it possible to estimate the model on quarterly US data of utilization and a certain capital ratio. The focus is on an inquiry into the model’s goodness-of-fit, for which the method of simulated moments is a fruitful technique. Here, 14 fairly ambitious summary statistics that the model should try to reproduce allow a detailed diagnosis of its merits and demerits. The results are sufficiently encouraging and form a sound basis for stepwise extensions of the model in future work, or for refinements in the estimation procedure.

We consider the problem of estimating a Markov process that is endogenously sampled. We observe a discrete-time Markov process { p t } at a set of random times { t 1 , … , t n } that depend on the outcome of a probabilistic sampling rule that depends on the state of the process and other observed covariates x t . We focus on a particular example where p t is the daily wholesale price of a standardized steel product. The endogenous sampling problem arises from the fact that we only observe p t on the days the firm purchases steel. We show how to solve this problem under two different assumptions about firm behavior: (1) optimality: the timing of steel purchases is governed by an optimal purchasing strategy that maximizes expected discounted profits, and (2) potential suboptimality: we allow the firm to use any randomized, Markovian purchasing strategy. In the latter case, the estimation problem becomes semi-parametric and we use the method of sieves to estimate a flexible parametric approximation to the firm’s purchasing behavior that best fits the data without imposing optimality. We show how estimation of this model becomes tractable under either of these assumptions using the method of simulated moments (MSM). We simulate realizations of wholesale steel prices and sample them in the same way as they are sampled in the actual data, i.e. only on days where purchases occur. We use the MSM estimator to estimate a truncated lognormal AR(1) model of the wholesale price processes for particular types of steel plate and test and reject the assumption that the firm is behaving optimally.

The elasticity of substitution between capital and labor plays an important role in the analysis of economic and policy issues such as factors' share in national income and tax policies on business capital formation. Rather than focusing on long-run relationships to estimate this elasticity, I exploit the short-run variations in the labor income share due to changes in capital-embodied technology. Using the simulated method of moments approach, I obtain an elasticity estimate that is clearly less than one. The study indicates that estimates based on the long-run relationship of factor's share may tend to be significantly larger.

This study extends the hybrid version of the baseline New-Keynesian model with heterogeneous agents who may adopt various forecast heuristics. With a focus on consumer expectations, we identify the most appropriate pairs of forecast heuristics that can lead to an equivalent fit to the data compared with the model specification under rational expectations. The competing specifications are estimated using the simulated method of moments. Our empirical results suggest that expectations under bounded rationality in the United States are grounded on consumers’ emotional state, while for the Euro Area they are technical in nature. This observation questions the need for a hybrid model specification under rational expectations.

We propose a new monitoring procedure based on moving sums (MOSUM) for detecting single or multiple structural breaks in factor copula models. The test compares parameter estimates from a rolling window to those from a historical data set and analyzes the behavior under the null hypothesis of no parameter change. The case of multiple breaks is also treated. In the model, the joint copula is given by the copula of random variables which arise from a factor model. This is particularly useful for analyzing high dimensional data. Parameters are estimated with the simulated method of moments (SMM). We analyze the behavior of the monitoring procedure in Monte Carlo simulations and a real data application. We consider an online procedure for predicting the day-ahead Value-at-risk based on the suggested monitoring procedure.

Within the seminal asset-pricing model by Brock and Hommes (Journal of Economic Dynamics Control 22, 1235–1274, 1998), heterogeneous boundedly rational agents choose between a fixed number of expectation rules to forecast asset prices. However, agents’ heterogeneity is limited in the sense that they typically switch between a representative technical and a representative fundamental expectation rule. Here, we generalize their framework by considering that all agents follow their own time-varying technical and fundamental expectation rules. Estimating our model using the method of simulated moments reveals that it is able to explain the statistical properties of the daily and monthly behavior of the S&P500 quite well. Moreover, our analysis reveals that heterogeneity is not only a realistic model property but clearly helps to explain the intricate dynamics of financial markets.

We study the determinants of the takeover processes duration. In our model, we focus on bidder initiated, one-to-one negotiations. An initial bidder submits an offer to the target. The target either accepts, and the transaction is completed, or rejects. The bidder has then the choice between going away or continuing the negotiation. As time goes on, the bidder and the target learn about true synergies thanks to the negotiation process. But rival bidders can show up and compete to acquire the target. Using a discrete time finite horizon dynamic programming approach, we study the equilibrium relations between the negotiation duration, pressure of potential competition and the learning process. We turn next to a calibration exercise based on a large sample of merger negotiations, collecting initiation and negotiation duration in SEC filings. We parallel the simulated method of moments to match the frequency distribution of private negotiation duration. Our results suggest that the M&A market is much more competitive than previously thought.

Customs record data reveal a number of patterns in relationships Colombian firms have with their U.S. buyers. We interpret these patterns in terms of a continuous-time model in which heterogeneous sellers search for buyers in a market. Success in selling to a buyer reveals information to the seller about the appeal of her product in the market, affecting her incentive to search for more buyers. Fit using the method of simulated moments, the model replicates key patterns in the customs records and allows us quantify several types of trade costs, including the search costs of identifying potential clients and the costs of maintaining business relationships with existing clients. It also allows us to estimate the effect of previous exporting activity on the costs of meeting new clients, and to characterize the cumulative effects of learning on firms' search intensities. Finally, we use our fitted model to explore the effects of these trade costs and learning effects on aggregate export dynamics

Consumers may face demand uncertainty when choosing a service plan under three‐part tariffs, and preferences for multiple services may be inter‐dependent. To examine such a demand system, we construct a two‐stage discrete/continuous choice model for service bundles, allowing for interactive utility and preference correlations. Implementing a piecewise maximization approach to consumers’ non‐differentiable utility maximization problem, we estimate the model via simulated method of moments. We empirically illustrate the model using data from a Chinese wireless service provider. Our counterfactual analysis shows that the three‐part tariffs with interchangeable units show no significant loss of revenue, compared to existing tariffs.

This paper focuses on investigating financial asset returns' extreme risks, which are defined as the negative log-returns over a certain threshold. A simple agent-based model is constructed to explain the behavior of the market traders when extreme risks occur. We consider both the volatility clustering and the heavy tail characteristics when constructing the model. Empirical study uses the China securities index 300 daily level data and applies the method of simulated moments to estimate the model parameters. The stationarity and ergodicity tests provide evidence that the proposed model is good for estimation and prediction. The goodness-of-fit measures show that our proposed model fits the empirical data well. Our estimated model performs well in out-of-sample Value-at-Risk prediction, which contributes to the risk management.

This paper studies the allocation of heterogeneous agents to levels of educational attainment. The goal is to understand the magnitudes and sources of mismatch in this assignment, both in theory and in the data. The paper presents evidence of substantial mismatch between ability and educational attainment across 21 OECD countries, with a focus on Germany, Italy, Japan and the US. In the model, mismatch originates from: (i) taste shocks, (ii) binding borrowing constraints and (iii) noisy measures of ability in test scores. The model is estimated using a simulated method of moments approach. The main finding is that measured mismatch arises largely from noise in test scores and does not reflect borrowing constraints. Differences in tastes for education across households play a minor role in explaining mismatch. Further, the estimation allows us to decompose the college wage premium, isolating cross-country differences in selection effects from the return to education.

We propose new fluctuation tests for detecting structural breaks in factor copula models and analyse the behaviour under the null hypothesis of no change. In the model, the joint copula is given by the copula of random variables which arise from a factor model. This is particularly useful for analysing data with high dimensions. Parameters are estimated with the simulated method of moments (SMM). The discontinuity of the SMM objective function complicates the derivation of a functional limit theorem for the parameters. We analyse the behaviour of the tests in Monte Carlo simulations and a real data application. It turns out that our test is more powerful than nonparametric tests for copula constancy in high dimensions.

We fit stochastic spatial-temporal models to high-resolution rainfall radar data using Approximate Bayesian Computation (ABC). We consider models constructed from cluster point-processes, starting with the model of Cox, Isham and Northrop, which is the current state of the art. We then generalise this model to allow for more realistic rainfall intensity gradients and for a richer covariance structure that can capture negative correlation between the intensity and size of localised rain cells. The use of ABC is of central importance, as it is not possible to fit models of this complexity using previous approaches. We also introduce the use of Simulated Method of Moments (SMM) to initialise the ABC fit.

We propose a new efficient method of calibrating agent-based models using the Method of Simulated Moments. It utilizes the Filtered Neighborhoods optimization algorithm, gradually narrowing down the search area by examining local neighborhoods of promising solutions. The new method obtains better calibration accuracy for a benchmark financial agent-based model in comparison to a broad selection of other methods, while using just a tiny fraction of their computational budget.

Altres ajuts: RES-589-28-0001 A cross validation method for selection of statistics for Approximate Bayesian Computing, and for related estimation methods such as the Method of Simulated Moments, is presented. The method uses simulated annealing to minimize the cross validation criterion over a combinatorial search space that may contain an extremely large number of elements. A first simple example, for which optimal statistics are known from theory, shows that the method is able to select these optimal statistics out of a large set of candidate statistics. A second example of selection of statistics for a stochastic volatility model illustrates the method in a more complex case. Code to replicate the results, or to use the method for other applications, is provided at http://www.runmycode.org/companion/view/1116.

This paper studies method of simulated moments (MSM) estimators that are implemented using Bayesian methods, specifically Markov chain Monte Carlo (MCMC). Motivation and theory for the methods is provided by Chernozhukov and Hong (2003). The paper shows, experimentally, that confidence intervals using these methods may have coverage which is far from the nominal level, a result which has parallels in the literature that studies overidentified GMM estimators. A neural network may be used to reduce the dimension of an initial set of moments to the minimum number that maintains identification, as in Creel (2017). When MSM-MCMC estimation and inference is based on such moments, and using a continuously updating criteria function, confidence intervals have statistically correct coverage in all cases studied. The methods are illustrated by application to several test models, including a small DSGE model, and to a jump-diffusion model for returns of the S&amp, P 500 index.

We argue that the waning of nepotism in academia bolstered scientific production in pre-industrial Europe. We build a database of families of scholars (1088–1800), measure their scientific output, and develop a general method to disentangle nepotism from inherited human capital—two determinants of occupational persistence. This requires jointly addressing measurement error in human capital proxies and sample selection bias arising from nepotism. Our method exploits multi-generation correlations together with parent-child distributional differences to identify the structural parameters of a first-order Markov process of human capital transmission with nepotism. We find an intergenerational human capital elasticity of 0.59, higher than that suggested by parent-child elasticities, yet lower than multi-generation estimates ignoring nepotism. In early academia, 40 percent of scholars’ sons achieved their position because of nepotism. Nepotism was lower in science than in law and in Protestant than in Catholic institutions, and declined substantially during the Scientific Revolution and the Enlightenment—two periods of buoyant scientific advancement.

We present a dynamic lifecycle model of women's choices with respect to partnership status, labour supply and fertility when they cannot directly observe whether a given male partner is of a violent type or not. The model is estimated by the method of simulated moments using longitudinal data from the Avon Longitudinal Study of Parents and Children. The results indicate that uncertainty about a partner's abusive type creates incentives for women to delay fertility, reduce fertility overall, divorce more often and increase labour supply. We also study the impact of higher female wages, income support to single mothers, and subsidized childcare when the mother is working. While higher wages reduce women's overall exposure to abuse, both income support and subsidized childcare fail to do so because they encourage early fertility. Income support also leads to less accumulated labour market experience and hence higher abuse rates.

This paper examines the asset pricing implications of sector-specific shocks in a multi-sector economy where heterogeneous firms interact in the markets for material inputs, investment goods, and final goods. The model is solved using a third-order perturbation and is estimated by the simulated method of moments. The sparsity of the Capital Flow table implies that sectoral shocks tend to affect the most sectors that produce investment goods. However, overall, the input-output structure diffuses the sectoral shocks so that there is limited heterogeneity in ex-post mean stock returns across sectors. Sectoral productivity shocks jointly account for more than 50% of equity risk premia in all sectors with substantial heterogeneity in the contribution of sectoral shocks to the risk premia but little heterogeneity in the composition of risk premia across sectors. The model endogenously generates volatility clusters in sectoral stock returns even though all shocks are conditionally homoskedastic.

This paper estimates the importance of temptation (Gul and Pesendorfer, 2001) for consumption smoothing and asset accumulation in a structural life-cycle model. We use two complementary estimation strategies: ?rst, we estimate the Euler equation of this model; and second we match liquid and illiquid wealth accumulation using the Method of Simulated Moments. We ?nd that the utility cost of temptation is one-quarter of the utility bene?t of consumption. Further, we show that allowing for temptation is crucial for correctly estimating the elasticity of intertemporal substitution: estimates of the EIS are substantially higher than without temptation. Finally, our Method of Simulated Moments estimation is able to match well the life-cycle accumulation pro?les for both liquid and illiquid wealth only if temptation is part of the preference speci?cation. Our ?ndings on the importance of temptation are robust to the di?erent estimation strategies. Revised July 2020

We propose a new simulation-based estimation method, adversarial estimation, for structural models. The estimator is formulated as the solution to a minimax problem between a generator (which generates simulated observations using the structural model) and a discriminator (which classifies whether an observation is simulated). The discriminator maximizes the accuracy of its classification while the generator minimizes it. We show that, with a sufficiently rich discriminator, the adversarial estimator attains parametric efficiency under correct specification and the parametric rate under misspecification. We advocate the use of a neural network as a discriminator that can exploit adaptivity properties and attain fast rates of convergence. We apply our method to the elderly's saving decision model and show that our estimator uncovers the bequest motive as an important source of saving across the wealth distribution, not only for the rich., 56 pages, 3 tables, 11 figures

This study evaluates the policy that brings non-marginal changes in an urban environmental attribute, fire risk, by using real estate auction data. Our evaluation is based on the extrapolation of the value distribution for auctioned real estate. In the estimation, we propose a simulated method of moments that is based on the moment conditions on order statistics of bids. We consider the possibility of affiliation by applying the Archimedean copula. Our estimation result rejects the hypothesis that bidders' valuation for apartments is independent when we assume that the marginal distribution of value follows log-normal or exponential distributions.

We study the estimation of dynamic economic models for which some of the state variables are observed only occasionally by the econometrician—a common problem in many fields, ranging from industrial organization to marketing to finance. If such occasional state observations are serially correlated, the likelihood function of the model becomes a potentially high-dimensional integral over a nonstandard domain. We propose a method that generalizes the recursive likelihood function integration procedure (RLI; Reich, 2018) to numerically approximate this integral, and demonstrate its statistical efficiency in several well-understood examples from finance and industrial organization. In extensive Monte Carlo studies, we compare the performance of our approach to a recently suggested method of simulated moments. In all our demonstrations, we consistently and efficiently identify all model parameters, and we find that the additional variance of our estimator when going from full to occasional state observations is small for the parameters of interest.

In a neoclassical dynamic model of the firm with labor market frictions, optimal hiring is a forward-looking decision that depends on both discount rates and expected cash flows. Empirically, we show that: a) the aggregate hiring rate of publicly traded firms in the U.S. economy negatively predicts stock market excess returns and long-term cash flows both in-sample and out-of-sample, and positively predicts short-term cash flows; and b) through a variance decomposition, the time series variation in the aggregate hiring rate is mainly driven by changes in discount rates and short-term expected cash flows, each contributing roughly to 50% of the variation, with no contribution from variation in long-term expected cash flows. Through a structural estimation of the model, we show that labor adjustment costs and, to a lesser extent, time-variation in the price of aggregate productivity risk, are essential for the model to replicate the empirical patterns.

This paper studies the positive and normative implication of extreme shocks for monetary policy. The analysis is based on a small-scale new Keynesian model with sticky prices and wages where shocks are drawn from asymmetric generalized extreme value (GEV) distributions. A nonlinear perturbation of the model is estimated by the simulated method of moments. Under both the Taylor and Ramsey policies, the central bank responds nonlinearly and asymmetrically to shocks. The trade-off between targeting a gross inaation rate above 1 as insurance against extreme shocks and strict price stability is unambiguously decided in favour of strict price stability.

A model of how personality traits affect household time and resource allocation decisions and wages is developed and estimated. In the model, households choose between two behavioral modes: cooperative or noncooperative. Spouses receive wage offers and allocate time to supplying labor market hours and to producing a public good. Personality traits, measured by the so-called Big Five traits, can affect household bargaining weights and wage offers. Model parameters are estimated by Simulated Method of Moments using the Household Income and Labor Dynamics in Australia (HILDA) data. Personality traits are found to be important determinants of household bargaining weights and of wage offers and to have substantial implications for understanding the sources of gender wage disparities.

This paper develops a general equilibrium model to study the link between the amount of capital invested in housing assets and the term structure of interest rates. In the model, the production of housing assets is irreversible and housing assets can be used as collateral for borrowing funds. Agents’ decisions about consumption and investments in housing and non-housing assets generate a time-varying market price of risk that drives the dynamics of the term structure. The calibration to U.S. data using the simulated method of moments technique captures the dynamics of consumption, and the short- and long-term interest rates.

Because the data show that market tightness is not orthogonal to unemployment, this paper identifies the many empirical difficulties caused by adopting the free entry of vacancies assumption in the Diamond-Mortensen-Pissarides (DMP) framework. Relaxing the free entry assumption and using Simulated Method of Moments (SMM) finds the vacancy creation process is less than infinitely elastic. Because a recession-leading job separation shock then causes vacancies to fall as unemployment increases, the ad hoc restriction to zero job separation shocks (to generate Beveridge curve dynamics) becomes redundant. In contrast to standard arguments, the calibrated model finds the job separation process drives unemployment volatility over the cycle. (JEL E24, E32, J24, J63, J64)

Estimation of agent-based models is currently an intense area of research. Recent contributions have to a large extent resorted to simulation-based methods mostly using some form of simulated method of moments estimation (SMM). There is, however, an entire branch of statistical methods that should appear promising, but has to our knowledge never been applied so far to estimate agent-based models in economics and finance: Markov chain Monte Carlo methods designed for state space models or models with latent variables. This latter class of models seems particularly relevant as agent-based models typically consist of some latent and some observable variables since not all the characteristics of agents would mostly be observable. Indeed, one might often not only be interested in estimating the parameters of a model, but also to infer the time development of some latent variable. However, agent-based models when interpreted as latent variable models would be typically characterized by non-linear dynamics and non-Gaussian fluctuations and, thus, would require a computational approach to statistical inference. Here we resort to Sequential Monte Carlo (SMC) estimation based on a particle filter. This approach is used here to numerically approximate the conditional densities that enter into the likelihood function of the problem. With this approximation we simultaneously obtain parameter estimates and filtered state probabilities for the unobservable variable(s) that drive(s) the dynamics of the observable time series. In our examples, the observable series will be asset returns (or prices) while the unobservable variables will be some measure of agents’ aggregate sentiment. We apply SMC to two selected agent-based models of speculative dynamics with somewhat different flavor. The empirical application to a selection of financial data includes an explicit comparison of the goodness-of-fit of both models.

We analyze the predictive power of value-at-risk forecasts generated by agent-based models. Specifically, we choose variants of the agent-based models proposed by Brock & Hommes (1998) and Franke & Westerhoff (2012) and calibrate them on the S&P 500 price and return series using a two-step process that enables the models to describe time series dynamics. To obtain a general approximation of the model parameters, our first estimation is conducted with the method of simulated moments. Following this, we apply a rolling window maximum likelihood estimation to obtain the state of the agent-based models at the current time step. The value-at-risk forecasts are then generated by iterating the models forward in time. Our results reveal that agent-based models are not only suitable for value-at-risk forecasting but are also capable of outperforming common benchmark models such as GARCH models. Most notably, we find that agent-based models outperform GARCH models in highly volatile recession periods, making agent-based models the superior choice for value-at-risk forecasting in periods with a high risk of loss.