Search

In this paper we investigate to what extent the bootstrap can be applied to conditional mean models, such as regression or time series models, when the volatility of the innovations is random and possibly non-stationary. In fact, the volatility of many economic and financial time series displays persistent changes and possible non-stationarity. However, the theory of the bootstrap for such models has focused on deterministic changes of the unconditional variance and little is known about the performance and the validity of the bootstrap when the volatility is driven by a non-stationary stochastic process. This includes near-integrated exogenous volatility processes as well as near-integrated GARCH processes, where the conditional variance has a diffusion limit; a further important example is the case where volatility exhibits infrequent jumps. This paper fills this gap in the literature by developing conditions for bootstrap validity in time series and regression models with non-stationary, stochastic volatility. We show that in such cases the distribution of bootstrap statistics (conditional on the data) is random in the limit. Consequently, the conventional approaches to proofs of bootstrap consistency, based on the notion of weak convergence in probability of the bootstrap statistic, fail to deliver the required validity results. Instead, we use the concept of ‘weak convergence in distribution’ to develop and establish novel conditions for validity of the wild bootstrap, conditional on the volatility process. We apply our results to several testing problems in the presence of non-stationary stochastic volatility, including testing in a location model, testing for structural change using CUSUM-type functionals, and testing for a unit root in autoregressive models. Importantly, we work under sufficient conditions for bootstrap validity that include the absence of statistical leverage effects, i.e., correlation between the error process and its future conditional variance. The results of the paper are illustrated using Monte Carlo simulations, which indicate that a wild bootstrap approach leads to size control even in small samples., This research was supported by the Danish Council for Independent Research, Denmark (DSF Grant 015-00028B), by the University of Bologna, Italy (ALMA IDEA 2017 Grant), and by the Italian Ministry of University and Research, Italy (PRIN 2017 Grant 2017TA7TYC).

The present special issue features a collection of papers presented at the 2017 International Panel Data Conference, hosted by the University of Macedonia in Thessaloniki, Greece. The conference marked the 40th anniversary of the inaugural International Panel Data Conference, which was held in 1977 at INSEE in Paris, under the auspices of the French National Centre for Scientific Research. As a collection, the papers appearing in this special issue of the Journal of Econometrics continue to advance the analysis of panel data, and paint a state-of-the-art picture of the field. (c) 2020 Elsevier B.V. All rights reserved.

Corresponding publicationsWorking Paper | Generating Univariate Fractional Integration within a Large VAR(1) 〈halshs-01944588〉; International audience; This paper shows that a large dimensional vector autoregressive model (VAR) of finite order can generate fractional integration in the marginalized univariate series. We derive high-level assumptions under which the final equation representation of a VAR(1) leads to univariate fractional white noises and verify the validity of these assumptions for two specific models.

This paper considers linear panel data models with a grouped pattern of heterogeneity when the latent group membership structure and/or the values of slope coefficients change at a break point. We propose a least squares approach to jointly estimate the break point, group membership structure, and coefficients. The proposed estimators are consistent, and the asymptotic distribution of the coefficient estimators is identical to that under known break point and group structure even when the cross-sectional sample size is much larger than the length of time series. Monte Carlo simulations and an empirical example illustrate the use of the approach and associated inference.

Many statistical and econometric learning methods rely on Bayesian ideas, often applied or reinterpreted in a frequentist setting. Two leading examples are shrinkage estimators and model averaging estimators, such as weighted-average least squares (WALS). In many instances, the accuracy of these learning methods in repeated samples is assessed using the variance of the posterior distribution of the parameters of interest given the data. This may be permissible when the sample size is large because, under the conditions of the Bernstein--von Mises theorem, the posterior variance agrees asymptotically with the frequentist variance. In finite samples, however, things are less clear. In this paper we explore this issue by first considering the frequentist properties (bias and variance) of the posterior mean in the important case of the normal location model, which consists of a single observation on a univariate Gaussian distribution with unknown mean and known variance. Based on these results, we derive new estimators of the frequentist bias and variance of the WALS estimator in finite samples. We then study the finite-sample performance of the proposed estimators by a Monte Carlo experiment with design derived from a real data application about the effect of abortion on crime rates.

The paper develops a novel realized matrix-exponential stochastic volatility model of multivariate returns and realized covariances that incorporates asymmetry and long memory (hereafter the RMESV-ALM model), and higher-moment spillovers. The matrix exponential transformation guarantees the positive definiteness of the dynamic covariance matrix. We decompose the likelihood function of the RMESV-ALM model into two components: one based on the conventional Kalman filter, and the other evaluated by a Monte Carlo likelihood technique. We consider a two-step quasi-maximum likelihood estimator for maximizing the likelihood function, and examine the finite sample properties of the estimator. The specification enables us to analyze asymmetric and higher-moment spillover effects in the covariance dynamics via news impact curves and impulse response functions. Using high frequency data for three US financial assets, the new model is estimated and evaluated. The forecasting performance of the new model is compared with a novel dynamic realized matrix-exponential conditional covariance model. Our empirical results suggest the RMESV-ALE specification to be superior, and spillover effects are found from returns or volatility to the remaining volatilities.

Large economic and financial panels often contain time series that influence the entire cross-section. We name such series granular. In this paper we introduce a panel data model that allows to formalize the notion of granular time series. We then propose a methodology, which is inspired by the network literature in statistics and econometrics, to detect the set of granulars when such set is unknown. The influence of the i-th series in the panel is measured by the norm of the i-th column of the inverse covariance matrix. We show that a detection procedure based on the column norms allows to consistently select granular series when the cross-section and time series dimensions are large. Importantly, the methodology allows to consistently detect granulars also when the series in the panel are influenced by common factors. A simulation study shows that the proposed procedures perform satisfactorily in finite samples. Our empirical studies demonstrate, among other findings, the granular influence of the automobile sector in US industrial production.

A novel approach to inference for a specific region of the predictive distribution is introduced. An important domain of application is accurate prediction of financial risk measures, where the area of interest is the left tail of the predictive density of logreturns. Our proposed approach originates from the Bayesian approach to parameter estimation and time series forecasting, however it is robust in the sense that it provides a more accurate estimation of the predictive density in the region of interest in case of misspecification. The first main contribution of the paper is the novel concept of the Partially Censored Posterior (PCP), where the set of model parameters is partitioned into two subsets: for the first subset of parameters we consider the standard marginal posterior, for the second subset of parameters (that are particularly related to the region of interest) we consider the conditional censored posterior. The censoring means that observations outside the region of interest are censored: for those observations only the probability of being outside the region of interest matters. This quasi-Bayesian approach yields more precise parameter estimation than a fully censored posterior for all parameters, and has more focus on the region of interest than a standard Bayesian approach. The second main contribution is that we introduce two novel methods for computationally efficient simulation: Conditional MitISEM, a Markov chain Monte Carlo method to simulate model parameters from the Partially Censored Posterior, and PCP-QERMit, an Importance Sampling method that is introduced to further decrease the numerical standard errors of the Value-at-Risk and Expected Shortfall estimators. The third main contribution is that we consider the effect of using a time-varying boundary of the region of interest. Extensive simulation and empirical studies show the ability of the introduced method to outperform standard approaches.

In this paper, we develop econometric tools to analyze the integrated volatility of the efficient price and the dynamic properties of microstructure noise in high-frequency data under general dependent noise. We first develop consistent estimators of the variance and autocovariances of noise using a variant of realized volatility. Next, we employ these estimators to adapt the pre-averaging method and derive a consistent estimator of the integrated volatility, which converges stably to a mixed Gaussian distribution at the optimal rate $n^{1/4}$. To refine the finite sample performance, we propose a two-step approach that corrects the finite sample bias, which turns out to be crucial in applications. Our extensive simulation studies demonstrate the excellent performance of our two-step estimators. In an empirical study, we characterize the dependence structures of microstructure noise in several popular sampling schemes and provide intuitive economic interpretations; we also illustrate the importance of accounting for both the serial dependence in noise and the finite sample bias when estimating integrated volatility.

This paper is concerned with inference on the cumulative distribution function (cdf) F X ∗ in the classical measurement error model X = X ∗ + ϵ . We consider the case where the density of the measurement error ϵ is unknown and estimated by repeated measurements, and show validity of a bootstrap approximation for the distribution of the deviation in the sup-norm between the deconvolution cdf estimator and F X ∗ . We allow the density of ϵ to be ordinary or super smooth. We also provide several theoretical results on the bootstrap and asymptotic Gumbel approximations of the sup-norm deviation for the case where the density of ϵ is known. Our approximation results are applicable to various contexts, such as confidence bands for F X ∗ and its quantiles, and for performing various cdf-based tests such as goodness-of-fit tests for parametric models of X ∗ , two sample homogeneity tests, and tests for stochastic dominance. Simulation and real data examples illustrate satisfactory performance of the proposed methods.

In this paper, we consider the least squares estimation of a panel structure threshold regression (PSTR) model where both the slope coefficients and threshold parameters may exhibit latent group structures. We study the asymptotic properties of the estimators of the latent group structure and the slope and threshold coefficients. We show that we can estimate the latent group structure correctly with probability approaching 1 and the estimators of the slope and threshold coefficients are asymptotically equivalent to the infeasible estimators that are obtained as if the true group structures were known. We study likelihood-ratio-based inferences on the group-specific threshold parameters under the shrinking-threshold-effect framework. We also propose two specification tests: one tests whether the threshold parameters are homogeneous across groups, and the other tests whether the threshold effects are present. When the number of latent groups is unknown, we propose a BIC-type information criterion to determine the number of groups in the data. Simulations demonstrate that our estimators and tests perform reasonably well in finite samples. We apply our model to revisit the relationship between capital market imperfection and the investment behavior of firms and to examine the impact of bank deregulation on income inequality. We document a large degree of heterogeneous effects in both applications that cannot be captured by conventional panel threshold regressions.

What transpires from recent research is that temperatures and radiative forcing seem to be characterized by a linear trend with two changes in the rate of growth. The first occurs in the early 60s and indicates a very large increase in the rate of growth of both temperature and radiative forcing series. This was termed as the "onset of sustained global warming". The second is related to the more recent so-called hiatus period, which suggests that temperatures and total radiative forcing have increased less rapidly since the mid-90s compared to the larger rate of increase from 1960 to 1990. There are two issues that remain unresolved. The first is whether the breaks in the slope of the trend functions of temperatures and radiative forcing are common. This is important because common breaks coupled with the basic science of climate change would strongly suggest a causal effect from anthropogenic factors to temperatures. The second issue relates to establishing formally via a proper testing procedure that takes into account the noise in the series, whether there was indeed a `hiatus period' for temperatures since the mid 90s. This is important because such a test would counter the widely held view that the hiatus is the product of natural internal variability. Our paper provides tests related to both issues. The results show that the breaks in temperatures and radiative forcing are common and that the hiatus is characterized by a significant decrease in their rate of growth. The statistical results are of independent interest and applicable more generally., Comment: 42 pages, 8 tables, 2 figures

In this paper we propose an autoregressive wild bootstrap method to construct confidence bands around a smooth deterministic trend. The bootstrap method is easy to implement and does not require any adjustments in the presence of missing data, which makes it particularly suitable for climatological applications. We establish the asymptotic validity of the bootstrap method for both pointwise and simultaneous confidence bands under general conditions, allowing for general patterns of missing data, serial dependence and heteroskedasticity. The finite sample properties of the method are studied in a simulation study. We use the method to study the evolution of trends in daily measurements of atmospheric ethane obtained from a weather station in the Swiss Alps, where the method can easily deal with the many missing observations due to adverse weather conditions. (C) 2019 Elsevier B.V. All rights reserved.

This paper studies the SEIRD epidemic model for COVID-19. First, I show that the model is poorly identified from the observed number of deaths and confirmed cases. There are many sets of parameters that are observationally equivalent in the short run but lead to markedly different long run forecasts. Second, I show that the basic reproduction number R 0 can be identified from the data, conditional on epidemiologic parameters, and propose several nonlinear SUR approaches to estimate R 0 . I examine the performance of these methods using Monte Carlo studies and demonstrate that they yield fairly accurate estimates of R 0 . Next, I apply these methods to estimate R 0 for the US, California, and Japan, and document heterogeneity in the value of R 0 across regions. My estimation approach accounts for possible underreporting of the number of cases. I demonstrate that if one fails to take underreporting into account and estimates R 0 from the reported cases data, the resulting estimate of R 0 may be biased downward and the resulting forecasts may exaggerate the long run number of deaths. Finally, I discuss how auxiliary information from random tests can be used to calibrate the initial parameters of the model and narrow down the range of possible forecasts of the future number of deaths.

This paper extends the notion of self-excitation in jumps to a rich class of continuous time semimartingale models, proposes statistical tests to detect its presence in a discretely observed sample path at high frequency, and derives the tests’ asymptotic properties. Our statistical setting is semiparametric: except for necessary parametric assumptions on the jump size measure, the other components of the semimartingale model are left essentially unrestricted. We analyze the finite sample performance of our tests in Monte Carlo simulations.

This paper provides pseudo-Gaussian and locally optimal rank-based tests for the cointegration rank in linear cointegrated error-correction models with common trends and i.i.d. elliptical innovations. The proposed tests are asymptotically distribution-free, hence their validity does not depend on the actual distribution of the innovations. The proposed rank-based tests depend on the choice of scores, associated with a reference density that can freely be chosen. Under appropriate choices they are achieving the semiparametric efficiency bounds; when based on Gaussian scores, they moreover uniformly dominate their pseudo-Gaussian counterparts. Simulations show that the asymptotic analysis provides an accurate approximation to finite-sample behavior. The theoretical results are based on a complete picture of the asymptotic statistical structure of the model under consideration.

It is well established that the shocks driving many key macro-economic and financial variables display time-varying volatility. In this paper we consider estimation and hypothesis testing on the coefficients of the co-integrating relations and the adjustment coefficients in vector autoregressions driven by both conditional and unconditional heteroskedasticity of a quite general and unknown form in the shocks. We show that the conventional results in Johansen (1996) for the maximum likelihood estimators and associated likelihood ratio tests derived under homoskedasticity do not in general hold in the presence of heteroskedasticity. As a consequence, standard confidence intervals and tests of hypothesis on these coefficients are potentially unreliable. Solutions to this inference problem based on Wald tests (using a "sandwich" estimator of the variance matrix) and on the use of the wild bootstrap are discussed. These do not require the practitioner to specify a parametric model for volatility, or to assume that the pattern of volatility is common to, or independent across, the vector of series under analysis. We formally establish the conditions under which these methods are asymptotically valid. A Monte Carlo simulation study demonstrates that significant improvements in finite sample size can be obtained by the bootstrap over the corresponding asymptotic tests in both heteroskedastic and homoskedastic environments. An application to the term structure of interest rates in the US illustrates the difference between standard and bootstrap inferences regarding hypotheses on the co-integrating vectors and adjustment coefficients.

This paper derives the asymptotic distribution for a number of rank-based and classical residual specification tests in AR-GARCH type models. We consider tests for the null hypotheses of no linear and quadratic serial residual autocorrelation, residual symmetry, and no structural breaks. We also apply our method to backtesting Value-at-Risk. For these tests we show that, generally, no size correction is needed in the asymptotic test distribution when applied to AR-GARCH residuals obtained through Gaussian quasi maximum likelihood estimation. To be precise, we give exact expressions for the limiting null distribution of the test statistics applied to (standardized) residuals, and find that standard critical values often, though not always, lead to conservative tests. For this result, we give simple necessary and sufficient conditions. Simulations show that our asymptotic approximations work well for a large number of AR-GARCH models and parameter values. We also show that the rank-based tests often, though not always, have superior power properties over the classical tests, even if they are conservative. An empirical application illustrates the relevance of these tests to the AR-GARCH models for weekly stock market return indices of some major and emerging countries. © 2015 Published by Elsevier B.V. 185 305 331 305-331

In order to derive closed-form expressions of the prices of credit derivatives, standard credit-risk models typically price the default intensities, but not the default events themselves. The default indicator is replaced by an appropriate prediction and the prediction error, that is the default-event surprise, is neglected. Our paper develops an approach to get closed-form expressions for the prices of credit derivatives written on multiple names without neglecting default-event surprises. This approach differs from the standard one, since the default counts necessarily cause the factor process under the risk-neutral probability, even if this is not the case under the historical probability. This implies that the standard exponential pricing formula of default does not apply. Using the US bond data, we show that allowing for the pricing of default events has important implications in terms of both data-fitting and model-implied physical probabilities of default. In particular, it may provide a solution to the credit spread puzzle. Besides, we show how our approach can be used to account for the propagation of defaults on the prices of credit derivatives.

This paper develops shrinkage methods for addressing the “many instruments” problem in the context of instrumental variable estimation. It has been observed that instrumental variable estimators may behave poorly if the number of instruments is large. This problem can be addressed by shrinking the influence of a subset of instrumental variables. The procedure can be understood as a two-step process of shrinking some of the OLS coefficient estimates from the regression of the endogenous variables on the instruments, then using the predicted values of the endogenous variables (based on the shrunk coefficient estimates) as the instruments. The shrinkage parameter is chosen to minimize the asymptotic mean square error. The optimal shrinkage parameter has a closed form, which makes it easy to implement. A Monte Carlo study shows that the shrinkage method works well and performs better in many situations than do existing instrument selection procedures.

The objective of this paper is to extend the results on Pseudo Maximum Likelihood (PML) theory derived in Gourieroux, Monfort, and Trognon (GMT) (1984) to a situation where the first four conditional moments are specified. Such an extension is relevant in light of pervasive evidence that conditional distributions are non-Gaussian in many economic situations. The key statistical tool here is the quartic exponential family, which allows us to generalize the PML2 and QGPML1 methods proposed in GMT(1984) to PML4 and QGPML2 methods, respectively. An asymptotic theory is developed which shows, in particular, that the QGPML2 method reaches the semi-parametric bound. The key numerical tool that we use is the Gauss-Freud integration scheme which solves a computational problem that has previously been raised in several econometric fields. Simulation exercises show the feasibility and robustness of the methods.

Cointegration ideas as introduced by Granger in 1981 are commonly embodied in empirical macroeconomic modelling through the vector error correction model (VECM). It has become common practice in these models to treat some variables as weakly exogenous, resulting in conditional VECMs. This paper studies the consequences of different approaches to weak exogeneity for the dynamic properties of such models, in the context of two models of the UK economy, one a national-economy model, the other the UK submodel of a global model. Impulse response and common trend analyses are shown to be sensitive to these assumptions and other specification choices. (C) 2010 Elsevier B.V. All rights reserved.

This paper proposes two types of stochastic correlation structures for Multivariate Stochastic Volatility (MSV) models, namely the constant correlation (CC) MSV and dynamic correlation (DC) MSV models, from which the stochastic covariance structures can easily be obtained. Both structures can be used for purposes of determining optimal portfolio and risk management strategies through the use of correlation matrices, and for calculating Value-at-Risk (VaR) forecasts and optimal capital charges under the Basel Accord through the use of covariance matrices. A technique is developed to estimate the DC MSV model using the Markov Chain Monte Carlo (MCMC) procedure, and simulated data show that the estimation method works well. Various multivariate conditional volatility and MSV models are compared via simulation, including an evaluation of alternative VaR estimators. The DC MSV model is also estimated using three sets of empirical data, namely Nikkei 225 Index, Hang Seng Index and Straits Times Index returns, and significant dynamic correlations are found. The Dynamic Conditional Correlation (DCC) model is also estimated, and is found to be far less sensitive to the covariation in the shocks to the indexes. The correlation process for the DCC model also appears to have a unit root, and hence constant conditional correlations in the long run. In contrast, the estimates arising from the DC MSV model indicate that the dynamic correlation process is stationary.

Importance sampling is used in many areas of modern econometrics to approximate unsolvable integrals. Its reliable use requires the sampler to possess a variance, for this guarantees a square root speed of convergence and asymptotic normality of the estimator of the integral. However, this assumption is seldom checked. In this paper we use extreme value theory to empirically assess the appropriateness of this assumption. Our main application is the stochastic volatility model, where importance sampling is commonly used for maximum likelihood estimation of the parameters of the model. © 2008 Elsevier B.V. All rights reserved.

This paper focuses on the analysis of size distributions of innovations, which are known to be highly skewed. We use patent citations as one indicator of innovation significance, constructing two large datasets from the European and US Patent Offices at a high level of aggregation, and the Trajtenberg [1990, A penny for your quotes: patent citations and the value of innovations. Rand Journal of Economics 21(1), 172–187] dataset on CT scanners at a very low one. We also study self-assessed reports of patented innovation values using two very recent patent valuation datasets from the Netherlands and the UK, as well as a small dataset of patent licence revenues of Harvard University. Statistical methods are applied to analyse the properties of the empirical size distributions, where we put special emphasis on testing for the existence of ‘heavy tails’, i.e., whether or not the probability of very large innovations declines more slowly than exponentially. While overall the distributions appear to resemble a lognormal, we argue that the tails are indeed fat. We invoke some recent results from extreme value statistics and apply the Hill [1975. A simple general approach to inference about the tails of a distribution. The Annals of Statistics 3, 1163–1174] estimator with data-driven cut-offs to determine the tail index for the right tails of all datasets except the NL and UK patent valuations. On these latter datasets we use a maximum likelihood estimator for grouped data to estimate the tail index for varying definitions of the right tail. We find significantly and consistently lower tail estimates for the returns data than the citation data (around 0.6–1 vs. 3–5). The EPO and US patent citation tail indices are roughly constant over time, but the latter estimates are significantly lower than the former. The heaviness of the tails, particularly as measured by value indicators, we argue, has significant implications for technology policy and growth theory, since the second and possibly even the first moments of these distributions may not exist.

In this paper, we introduce path diagrams for multivariate time series which visualize the dynamic relationships among the variables. In these path diagrams, the vertices represent the components of the time series and are connected by arrows or lines according to the nonvanishing parameters in the autoregressive representation of the time series. We show that these path diagrams provide a framework for the analysis of the dependence structure of the time series. In particular, we give sufficient graphical conditions for granger-noncausality and granger-noncausality up to a certain horizon.

In econometrics the same data set is typically used to select the model and to estimate the parameters in the selected model. In applied econometrics practice, however, one typically acts as if the model had been given a priori, thus ignoring the fact that the estimators are in fact pretest estimators. Hence one assumes incorrectly that the estimator is unbiased, and that the reported variance, conditional on the selected model, is equal to its unconditional variance. In this paper, we find the unconditional first and second moments of the pretest estimator (in fact, of a more general estimator, the WALS estimator), taking full account of the fact that model selection and estimation are an integrated procedure, and show that the error in not reporting the correct moments can be large. We also show that this error can vary substantially between different model selection procedures. Finally, we ask how the error increases when the number of auxiliary regressors increases.

This paper proposes panel data tests of gaussian affine term structure models. Yield curve data for different moments in time are pooled with the factors treated as fixed effects. With fixed effects the time series properties of the price of risk can be ignored. Results of tests with us interest rate data show that the gaussian model is able to capture the cross sectional structure of yields as well as unrestricted factor loadings from principal components analysis. However, estimates of the mean reversion parameters in a 3-factor model differ significantly when the model is estimated from yield levels or forward differences, which is inconsistent with the gaussian model.

We combine micro and macro unemployment duration data to study the effects of the business cycle on the outflow from unemployment. We allow the cycle to affect individual exit probabilities of unemployed workers as well as the composition of the total inflow into unemployment. We estimate the model using (micro) survey data and (macro) administrative data from France. The distribution of the inflow composition is estimated along with the other parameters. The estimation method deals with differences between the micro and macro unemployment definitions. The results also show to what extent the unemployment duration distributions corresponding to the two definitions can be described by the same model.This discussion paper resulted in a publication in the Journal of Econometrics.

In this paper, we investigate the effect of mean-nonstationarity on the first-difference generalized method of moments (FD-GMM) estimator in dynamic panel data models. We find that when data is mean-nonstationary and the variance of individual effects is significantly larger than that of disturbances, the FD-GMM estimator performs quite well. We demonstrate that this is because the correlation between the lagged dependent variable and instruments gets larger owing to the unremoved individual effects, i.e., instruments become strong. This implies that, under mean-nonstationarity, the FD-GMM estimator does not always suffer from the weak instruments problem even when data is persistent.

This paper considers unit root tests based on robust estimators with a high breakdown point and high efficiency. The asymptotic distribution of these tests is derived. Critical values for the test are obtained via simulation. It is found that the size of the classical OLS based Dickey-Fuller test breaks down if the time series contains additive outliers For innovative outliers the size of the robust test is less stable, while its size-adjusted power properties are better. An example is provided by applying the robust tests to the extended Nelson-Plosser data. For four series the null hypothesis of nonstationarity is rejected.

This paper concerns the estimation of an intertemporal model for labour supply and consumption that recognises the presence of nonworkers and which is cast in a structural optimising framework that allows for uncertainty. Through the utilisation of a micro-data source that measures consumption and labour supply we are able to estimate a flexible preference model which incorporates corner solutions and uncertainty.