ABSTRACT. In this paper we introduce three dynamic eigenvalue ratio estimators for the number of dynamic factors. The advantage of such estimators is that they do not require preliminary determination of discretionary parameters. The Dynamic Eigenvalue Ratio and the Dynamic Growth Ratio are the dynamic counterparts of the eigenvalue ratio estimators proposed by Ahn and Horenstein (2013). The Dynamic eigenvalue Difference Ratio (DDR) is closely related to the test statistic proposed by Onatsky (2009). In addition, we introduce a new eigenvalue ratio estimator for the number of static factors, the eigenvalue Difference Ratio estimator (DR). We prove consistency of such estimators and evaluate their performance under simulation. We conclude that both DDR and DR are valid alternatives to existing criteria.

ABSTRACT. I use Bayesian tools to develop a dynamic test of conditional factor models from the point of view of an investor who recognizes that parameters are uncertain, time-varying, and predictors are an imperfect proxy for macroeconomic and firms-specific news. Time-varying alphas, betas and idiosyncratic risks are jointly estimated in a single-step together with no-arbitrage restrictions. The test can be applied for a single asset or jointly across portfolios. As empirical application, I estimate over fifty years of postwar monthly data a conditional version of the CAPM and multi-factor models on size, book-to-market and momentum deciles portfolios. I show that once the dynamic and uncertain nature of the portfolios returns generating process is fully acknowledged, the null that conditional factor models hold is not sensibly rejected both in the time series and in the cross-section.

ABSTRACT. We introduce a frequency domain version of the EM algorithm for general dynamic factor models. We consider not only AR processes but also ARMA ones, for which we use iterative indirect inference procedures analogous to the algorithms in Hannan (1969). Our algorithm reduces the computational burden so much that researchers can estimate such models by maximum likelihood with a large number of series even without good initial values. However, near the optimum it slows down signicantly. Then, the best practical strategy is to switch to a gradient method that uses the EM principle to swiftly compute frequency domain analytical scores.

ABSTRACT. Quantitative assessments of the relationship between health and medical treatment are of great importance to policy makers, but simply looking at the raw correlation between health and medical care is unlikely to give the right answer because of simultaneity through the unobservable components of health deterioration. In this paper, we present a tractable dynamic factor model of health and medical treatment where individual observed health outcomes, including mortality, are driven by the individual's latent health stock and the dynamics of latent health reflects both exogeneous health depreciation and endogenous health investments. Nonlinearity is introduced to accommodate the discrete nature of some measured outcome. We estimate the model by the minimum distance method using a rich longitudinal data set that contains very detailed information on medical drug use, hospitalization, and mortality for a representative sample of hypertensive individuals, and has been constructed by combining a number of administrative archives. Our results confirm that better health is associated with lower medical treatment. More interestingly, we find an inverse U-shaped relation between lagged medical drug use and current health. This supports the view that taking the recommended amount of medication helps reduce health deterioration and avoid its negative consequences, but that taking more pills than needed cannot improve health. Our findings have important policy implications, suggesting that policies aimed at increasing awareness of hypertensive diseases and the importance of the treatment of high blood pressure may help reduce cardiovascular risks, and consequent hospitalization and mortality.

ABSTRACT. This paper proposes a test to verify whether the k-th moment of a random variable is finite. We use the fact that, under general assumptions, sample moments either converge to a finite number or diverge to infinity according as the corresponding population moment is finite or not. Building on this, we propose a test for the null that the k-th moment does not exist. Since, by construction, our test statistic diverges under the null and converges under the alternative, we propose a randomised testing procedure to discern between the two cases. We study the application of the test to raw data, and to regression residuals. Monte Carlo evidence shows that the test has the correct size and good power; the results are further illustrated through an application to financial data.

ABSTRACT. The contribution of this paper is two-fold. First we extend the available asymptotic theory for sieve estimators to cover the case of stationary and invertible linear processes driven by independent identically distributed (i.i.d.) infinite variance (IV) innovations. In particular, we show that the ordinary least squares sieve estimates, together with estimates of the impulse responses derived from these, obtained from an autoregression whose order is an increasing function of the sample size, are consistent and exhibit asymptotic properties analogous to those which obtain for a finite-order autoregressive process driven by i.i.d. IV errors. As such, their limit distributions cannot be directly employed for inference because they either may not exist or, where they do, depend on the unknown tail index of the IV process. As a consequence, our second contribution is to investigate the usefulness of bootstrap methods in this setting. Focusing on the theoretical properties of three sieve bootstraps: namely the wild bootstrap, the permutation bootstrap and a hybrid of the two, we show that, in contrast to the case of finite variance innovations, the wild bootstrap requires an infeasible correction to be consistent, whereas the other two bootstrap schemes are consistent under general conditions. We present Monte Carlo simulation results which suggest that, even without this infeasible correction, the basic wild bootstrap procedure displays very decent empirical finite sample size properties, only slightly inferior to those of the permutation and hybrid bootstraps which perform best overall. At the same time the basic wild bootstrap, and to a somewhat lesser extent the permutation and hybrid bootstraps, yield very large power gains, relative to other well-known bootstrap procedures such as the m out of n and i.i.d. residual bootstraps.

ABSTRACT. This paper proposes a novel and flexible framework to estimate autoregressive models with time-varying parameters. Our setup nests various adaptive algorithms that are commonly used in the macroeconometric literature, such as learning-expectations and forgetting-factor algorithms. These are generalized along several directions: specifically, we allow for both Student-t distributed innovations as well as time-varying volatility. Meaningful restrictions are imposed to the model parameters, so as to attain local stationarity and bounded mean values. The model is applied to the analysis of inflation dynamics. Allowing for heavy-tails leads to a significant improvement in terms of fit and forecast. Moreover, it proves to be crucial in order to obtain well-calibrated density forecasts.

ABSTRACT. GARCH (1,1) models are widely used for modelling processes with time varying volatility. These include financial time series, which can be particularly heavy tailed. In this paper, we propose a log-transform-based least squares estimator (LSE) for the GARCH (1,1) model. The asymptotic properties of the LSE are studied under very mild moment conditions for the errors. We establish the consistency, asymptotic normality at the standard convergence rate of $\sqrt n $ for our estimator. The finite sample properties are assessed by means of an extensive simulation study. Our results show that LSE is more accurate than the quasi-maximum likelihood estimator (QMLE) for heavy tailed errors. Finally, we provide some empirical evidence on two financial time series considering daily and high frequency returns. The results of the empirical analysis suggest that in some settings, depending on the specific measure of volatility adopted, the LSE can allow for more accurate predictions of volatility than the usual Gaussian QMLE.

ABSTRACT. Tournaments motivate workers to provide effort. However, theory says that large differences in workers' relative abilities undermine the effectiveness of tournaments to motivate workers. To test this prediction I use a novel data set from professional football competitions which allows to measure effort as the total distance covered by teams and heterogeneity in abilities as the ratio between the winning probabilities derived from the betting odds of professional bookmakers. I show that the higher is the heterogeneity between teams the lower is their provision of effort. Using the winning probabilities of teams and a structural identification approach allows to control for all unobservable variables affecting the provision of effort.

ABSTRACT. In Southern Europe youngsters leave the parental home signifi…cantly later than in Northern Europe and United States. Policies have been implemented in Southern Europe to incentivize young adults to leave parental home earlier. Do peer effects among siblings modify the effects of these policies? Estimating peer effects is challenging because of problems of re‡flection, endogenous group formation, and correlated unobservables. We overcome these issues in the context of a Spanish rental subsidy, exploiting the subsidy eligibility age threshold to analyze peer effects among siblings. Instrumental variable estimates show that peer effects among siblings are negative, and that the effect is explained by the presence of old and single parents, and attenuated by the presence of additional siblings in the household. Findings indicate that policy makers should target the household rather than the individual, and combine policies for young adults together with policies for the elderly.

ABSTRACT. We derive a theoretical model of effort in the presence of career concern based on the multi-unit all-pay auction, and closely inspired by the Italian academic market. In this model, the number of applicants, the number of new posts, and the relative importance of the determinants of promotion determine academics’ effort. Because of the specific characteristics of Italian universities, where incentives operate only through promotion, and where all appointment panels are drawn from strictly separated and relatively narrow scientific sectors, the model fits well Italian academia, and we test it in a newly constructed dataset which collects the journal publications of all Italian academics working in universities. We find that individual researchers respond to incentives in the manner predicted by the theoretical model: more capable researchers respond to increases in the importance of the measurable determinants of promotion and in the competitiveness of the scientific sector by exerting more effort; less able researchers do the opposite.

ABSTRACT. This paper provides new insights into expectation-driven cycles by estimating a structural VAR with time-varying coefficients and stochastic volatility. We use survey-based expectations of the unemployment rate to measure expectations of future developments in economic activity. We find that the effect of expectation shocks on the realized unemployment rate have been particularly large during the most recent recession. Unanticipated changes in expectations contributed to the gradual increase in the persistence of the unemployment rate and to the decline in the correlation between the inflation and the unemployment rate over time. Our results are robust to the introduction of financial variables in the model.

ABSTRACT. This paper investigates the interaction between uncertainty and monetary policy by estimating a non-linear VAR with US post-WWII data. The uncertainty indicator is treated both as an endogenous variable in the VAR and as the transition indicator discriminating "high" vs. "low" uncertainty states. The impact of monetary policy shocks in different phases of the "uncertainty cycle" is assessed via the computation of Generalized Impulse Response Functions. Monetary policy shocks are found to be less effective when uncertainty is high, with the peak reactions of a battery of real variables being about two-thirds milder than those conditional on an initially low level of uncertainty. The framework is then put at work to investigate the effects of uncertainty shocks in the presence of the Zero Lower Bound. The "drop and rebound" response of real variables to uncertainty shocks documented by Bloom (2009) is found to be present only if the policy rate is a long way from its ZLB. Conversely, an uncertainty shock occurring when the economy is near the ZLB triggers longer-lasting recessions and does not lead to any significant “rebound”.

ABSTRACT. We evaluate the role of sentiment shocks on US economy focusing on the impact on both financial and real variables. To understand whether these effects are constant over time or are instead regime--dependent, we consider a Structural VAR model estimated with kernel based methods which allow for changes in the transmission of shocks. Identification relies on a standard recursive scheme in order to disentangle sentiment from financial shocks, the latter modeled either as stock prices or credit spread or house prices shocks. We show that sentiment shocks (i) have time--varying effects (ii) strengthening their effect during periods of crisis, in particular accounting for a large share of output and stock prices dynamics in the Global Financial Crisis, and (iii) are at least as relevant as financial shocks for the US business cycle.

ABSTRACT. Financial markets are central to the transmission of uncertainty shocks. This paper documents a new aspect of the interaction between the two by showing that uncertainty shocks have radically different macroeconomic implications depending on the state …financial markets are in when they occur. Using monthly US data, we estimate a nonlinear VAR where economic uncertainty is proxied by the (unobserved) volatility of the structural shocks, and a regime change occurs whenever credit conditions cross a critical threshold. An exogenous increase in uncertainty has recessionary effects in both good and bad credit regimes, but its impact on output is estimated to be fi…ve times larger when the economy is experiencing …financial distress. Accounting for this nonlinearity, uncertainty accounts for about 1% of the peak fall in industrial production observed in the 2007–2009 recession.

ABSTRACT. This paper tests the impact of family ownership on …firms' ’export decisions using a data set of 20,000 Italian manufacturers. We …find that family ownership increases the probability that …firms export, although the effect weakens as own- ership concentration rises. The bene…t of family owners is especially pronounced when they retain control rights (ownership is aligned with control) and seek the support of external managers (ownership is partially separated from manage- ment). The results suggest that families better internalize the long-run bene…fits of internationalization, but that their limited competencies attenuate this benefit in high-tech industries and in remote and unfamiliar export markets.

ABSTRACT. In this paper I examine the effect of trade liberalization on the productivity of domestic suppliers of exporting firms. Using a panel of Chilean firms during a period of trade liberalization with the European Union, the United States, and Korea, I show that when downstream firms expand into foreign markets, the increase in derived demand for intermediate inputs leads to productivity gains along the production chain. Export shocks increase market size for upstream firms through input-output linkages. This finding confirms the importance of demand in explaining firms’ productivity dynamics.

ABSTRACT. In this paper we explore the geographical dimension of insularity, measuring its effect on a comprehensive measure of trade costs (Novy 2012). Controlling for other geographical characteristics, connectedness (spatial proximity) and the role of historical events in shaping modern attitudes towards openness (measured through a quantification of routes descriptions in logbooks between 1750 and 1850), we give evidence that to be an island is not bad per se in terms of trade costs. Bad geography can be reversed by connectedness and open institutions.

ABSTRACT. This paper sheds additional light on the relationship between firm level innovative capacity and export intensity. By drawing from the recent literature on exporters' heterogeneity, we apply quantile regression techniques to a sample of Italian firms in order to verify whether the effect of innovative capacity – measured by R&D expenditures – varies along the conditional distribution of the export intensity, after controlling for censoring and potential endogeneity of the innovation variable. We confirm that R&D expenditures positively affect export intensity and we find that such effect has a bell shaped pattern along its conditional distribution: firms characterized by export intensity of about 60% can take highest advantage from investing in R&D activity. Overall results prove to be robust to several specification checks and suggest not only that firms innovative capacity helps to explain heterogeneity in export intensity performance, but also that its positive effect differs across the export to sales ratio distribution.

ABSTRACT. In this paper we study consumption network effects. Does the consumption of our peers affect our own consumption? How large is such effect? What is the economic mechanism that is behind it? We use long panel data on the entire Danish population to construct a measure of consumption based on administrative tax records on income and assets. We combine tax record data with matched employer-employee data so that we can construct peer groups based on workplace, which gives us a much tighter, precise, and credible definition of networks than used in previous literature. We use the available data to construct peer groups that do not perfectly overlap, and as such provide valid instruments derived from the network structure of one's peers group. The longitudinal nature of our data also allow us to estimate fixed effects models, which help us tackle reflection, self-selection, and common-shocks issues all at once. We estimate non- negligible and statistically significant endogenous and exogenous network effects. Estimated effects are quite relevant for policies as they generate non-negligible social multiplier responses. We also investigate what mechanisms generate such effects, distinguishing between "keeping up with the Joneses", a status model, and a more traditional risk sharing view.

ABSTRACT. We argue that US welfare would rise if unemployment insurance were increased for younger and decreased for older workers. This is because the young tend to lack the means to smooth consumption during unemployment and want jobs to accumulate high-return human capital. So unemployment insurance is most valuable to them, while moral hazard is mild. By calibrating a life cycle model with unemployment risk and endogenous search effort, we find that allowing unemployment replacement rates to decline with age yields sizeable welfare gains to US workers.

ABSTRACT. Modeling crude oil volatility is of substantial interest for both energy researchers and policy makers. Many authors emphasize the link between this volatility and some exogenous economic variables. This paper aims to investigate the impact of the U.S. Federal Reserve monetary policy on crude oil future price (COFP) volatility. By means of the recently proposed generalized autoregressive conditional heteroskedasticity-mixed data sampling (GARCH-MIDAS) model, the Effective Federal Fund Rate (EFFR) - as a proxy of the monetary policy - is plugged into the mean-reverting unit GARCH(1,1) model. Strong evidence of an inverse relation between the EFFR and COFP volatility is found. This means that an expansionary monetary policy is associated with an increase of the COFP volatility. The results are confirmed even controlling for the U.S. monthly Industrial Production percentage change as a proxy of the crude oil demand. Furthermore, since the unusual behavior (bubble) of the COFP in 2008 casts some doubts on the reliability of full sample results, we apply the same experimental set-up to two sub-samples before and after this anomalous time frame.

ABSTRACT. We establish the strong consistency and asymptotic normality of the Whittle estimator of a class of multivariate exponential volatility models.Both properties are established under a variety of parametrization. The variables of interest might not have finite fractional moment of any order and so in particular finite variance is nowhere imposed. The results cover both the case of exponentially and hyperbolically decaying coefficients, allowing for different degrees of persistence of shocks to the conditional variances. It is shown that the rate of convergence and the asymptotic normality of the Whittle estimates do not depend on the degree of persistence implied by the parametrization as the Whittle function automatically compensates for the possible lack of square integrability of the model spectral density

ABSTRACT. The dynamics of financial volatility shows a behavior characterized by alternating periods of turbulence and relative quiet. We suggest to model it as a mixture memory model where time-varying mixing weights are a function of some forcing variable capable of sudden changes. In choosing a mixture approach we rely on previous evidence on the presence of a short-- and a long--memory component in the observed series. We apply our model to the main Spanish stock index (IBEX) using the spread between the sovereign national and German bond rates as the forcing variable. The results show a good performance in sample, pointing to the fact that fixed weights may be a limitation to an accurate description of volatility behavior.

ABSTRACT. The OGARCH specification is the leading model for a class of multivariate GARCH (MGARCH) specifications based on linear combinations of univariate GARCH estimates. The majority of those MGARCH models adopts a spectral decomposition of the covariance matrix, allowing for heteroskedasticity on (some of) the principal components, while the loading matrix (that is, the matrix mapping the conditional principal components to the asset returns), is constant over time. In this paper we extend the OGARCH model class to allow for time-varying loadings. Our modelling approach closely parallels the DCC modelling approach introduced as an extension of the CCC model to allow for dynamic correlations. We first introduce an auxiliary process which is capable of capturing the relevant features of the unobservable loading dynamics. The time-varying loading matrix is then computed from the auxiliary process subject to the necessary orthonormality constraints. The resulting model (Dynamic Principal Components, or DPC, model) preserves the ease of interpretation and the feasibility of the OGARCH model. In particular, we prove that the eigenvectors of the intercept term of the time-varying loadings can be consistently estimated by the eigenvectors of the sample covariance matrix. This property extends to the dynamic framework the well-known analogous property of the OGARCH model. Empirical examples show evidences of the benefits provided by the introduction of time-varying properties for the loading matrix.

ABSTRACT. Ordinary least squares (OLS) is well-known to produce an inconsistent estimator of the spatial parameter in pure spatial autoregression (SAR). This paper explores the potential of indirect inference to correct the inconsistency of OLS. Under broad conditions, it is shown that indirect inference (II) based on OLS produces consistent and asymptotically normal estimates in pure SAR regression. The II estimator is robust to departures from normal disturbances and is computationally straightforward compared with pseudo Gaussian maximum likelihood (PML). Monte Carlo experiments based on various specifications of the weighting matrix confirm that the indirect inference estimator displays little bias even in very small samples and gives overall performance that is comparable to the Gaussian PML.

ABSTRACT. As spatial econometrics has become popular in empirical economic research, two main estimation procedures have been proposed in the literature to properly estimate parameters: Maximum Likelihood (ML) and Instrumental Variables (IV). From a theoretical point of view, these approaches are asymptotically equivalent, implying that results should be equal across estimators using large samples. However, due to the reduced dimensionality of the data at hands, the results of these estimators are typically different in small samples. In this work we show how, under certain hypothesis, this two approaches could lead to similar results.

ABSTRACT. The paper assesses spillovers from total factor productivity (TFP) in the Italian manufacturing industry and the existence of potential TFP premiums originating from innovation, properly accounting for spatial distances in place between firms. We resort to firm-level geo-referenced data to estimate geographical TFP spillovers and to input-output matrices of inter-sectorial trade to detect sectorial TFP spillovers (both demand and supply driven). A complete spatial autoregressive model with spatial autoregressive disturbances (SARAR) is estimated, with the purpose to analyze the spatial diffusion of productivity shocks as well. To the best of our knowledge the present work comes to represent one of the first attempts to estimate a model of this type based on a relatively large panel of micro-data and resorting to dense matrices of firms’ distances: customized versions of the available R routines were developed. Results show how firms’ TFP levels benefit from spillovers originated within the neighborhood while spatial diffusion of productivity shocks might differ, depending on how spatial influences are modeled: shocks spread negatively within geographical-based neighborhoods (competition framework) and positively within sectorial-based neighborhoods (cooperation framework). Furthermore, firms located in patent-intensive areas are suitable to show local productivity premiums: the result emerges clearly even in correspondence to the subsample of small firms, thus corroborating the findings of a comprehensive active role played by innovation in enhancing productivity.

ABSTRACT. It is well known that the maximum likelihood estimator (MLE) for the autoregressive parameter in a spatial autoregressive model can suffer from substantial bias. One approach to reducing the bias is to recenter the profile score. We study properties of the resulting estimator, which we name the adjusted MLE. A comparison of the MLE and the adjusted MLE is not straightforward, because the distributions of the two estimators have different supports. We use analytical and simulation methods to perform such a comparison.

ABSTRACT. This study evaluates the macroeconomic effects of Outright Monetary Transaction (OMT) announcements by the European Central Bank (ECB). Using high-frequency data, we find that OMT announcements decreased the Italian and Spanish 2-year government bond yields by about 2 percentage points, while leaving unchanged the bond yields of the same maturity in Germany and France. These results are used to calibrate a scenario in a multi-country model describing the macrofinancial linkages in France, Germany, Italy, and Spain. The scenario analysis suggests that the reduction in bond yields due to OMT announcements is associated with a significant increase in real activity, credit, and prices in Italy and Spain with relatively muted spillovers in France and Germany.

ABSTRACT. In this paper, we model business investment distinguishing between Information and Communication Technologies (ICT) and traditional (Non-ICT) components taking into account asset specific characteristics potentially affecting the reactivity of capital accumulation over the business cycle. Business investment and ICT and Non-ICT assets are estimated within a Vector Error Correction Model (VECM) model to test, in a unified empirical framework, the assumptions of the flexible accelerator model (Clark, 1944, and Koyck, 1954) and of the neoclassical model of Hall and Jorgenson (1967), as well as how financial constraints and uncertainty influence investment behaviour (Hall and Lerner, 2010, and Bloom, 2007). Our findings suggest that the existence of a long-run relationship with standard macro determinants (output and user cost) is validated for aggregate business capital stock as well as for individual Non-ICT assets but not for ICT. In the short run, liquidity is a key determinant of investment behaviour independently of the asset type. In the long-run, uncertainty significantly affects ICT. Finally, the results of the counterfactual exercises over the latest Italian recession support the idea that ICT is a key policy variable to foster the economic recovery.

ABSTRACT. This paper empirically tests the Expectation Hypothesis of the term structure of the US repurchasing agreements (repo) rates, considered in a Vector Auto Regression (VAR) model. A multiple hypotheses approach is adopted, in order to jointly test all the statistical hypotheses implied by the EH, i.e. the long-run and short-run implications of the theory. Furthermore, the testing procedures are carried out by taking into account non-stationary volatility of the series and of the error terms through bootstrap inference, White correction and rolling windows analysis. Differently from previous results, overall evidences in favor of the statistical non-rejection of the EH are found. In particular, the rolling window analysis clarifies that the EH has been rejected only during periods of turmoil of the financial/repo markets.

ABSTRACT. Understanding the dynamics of the leverage ratio is at the heart of the empirical research about firms' capital structure, and is well grounded in alternative theoretical models. The results of this strand of empirical research are always based on the maintained assumptions of poolability and stationarity, despite this choice is very much a matter of faith, based more on theoretical considerations rather than on available empirical evidence. The aim of this paper is to re-examine and to interpret the outcomes gathered under the assumptions of poolability and stationarity in the light of the results of statistical tools and models devoted to the assessment of stationarity in the context of heterogeneity. We provide robust evidence of non-stationarity for a significantly large share of US firms' debt ratios and, consequently, of strong heterogeneity in the speeds at which firms adjust towards their targets. These results should stimulate new directions of the empirical research to further deepen the understanding of debt ratio dynamics.

ABSTRACT. We link administrative public data on regional screening policies to individual Survey of Health Ageing and Retirement in Europe (SHARE) data to estimate the causal effect of home invitation on mammography uptake. Exploiting regional variation in the availability of screening policies and in age eligibility criteria, we find that home invitation increases mammography uptakes by around 24%. Significant effects are found when at least 50% of the population is invited. The stock of health information and the ability to process it play a role, as the effects of invitation are higher among low educated and lower among cognitive impaired women.

ABSTRACT. One of the main objectives of empirical analysis of experiments and quasi-experiments is to inform policy decisions that determine how treatments are allocated to individuals with different observable covariates. We propose the Empirical Welfare Maximization (EWM) method, which estimates a treatment assignment policy by maximizing the sample analog of average social welfare over a class of candidate treatment policies. We show that, when propensity score is known, the average social welfare attained by EWM rules converges at least at n^(-1/2) rate to the maximum obtainable welfare. This holds uniformly over a minimally constrained class of data distributions, and this uniform convergence rate is minimax optimal. In comparison with this benchmark rate, we examine how the uniform convergence rate of the average welfare improves or deteriorates depending on the richness of the class of candidate decision rules, on the distribution of conditional treatment effects, on the lack of knowledge for the propensity score, and on additional smoothness assumptions for the regression functions or propensity scores. We also discuss practically implementable computation for the EWM rule. As an empirical application, we derive an EWM rule for the a training program using the experiment data analyzed in LaLonde (1986).

ABSTRACT. We consider empirical measurement of exact equivalent/compensating variation resulting from price-change of a discrete good using individual-level data, when there is unobserved heterogeneity in preferences. We show that for binary and multinomial choice, the marginal distributions of EV/CV are nonparametrically point-identified solely from the conditional choice-probabilities, under extremely general preference-distributions. These results hold even when the distribution/dimension of unobserved heterogeneity are neither specified, nor identified and utilities are neither quasi-linear nor parametrically specified. Welfare-distributions can be expressed as closed-form functionals of choice-probabilities, thus enabling easy computation in applications. Average EV for price-rise equals the change in average consumer-surplus and is smaller than average CV for a normal good. Point-identification fails for ordered choice if the unit-price is identical for all alternatives, thereby providing a connection to Hausman-Newey's (2013) partial identification results for the limiting case of continuous choice.

ABSTRACT. In 2005, the Indian Government launched a conditional cash-incentive program to encourage institutional delivery. This paper studies the effects of the program on neonatal mortality using district-level household survey data. We model mortality using survival analysis, paying special attention to the substantial heaping present in the data. The main objective of this paper is to provide a set of sufficient conditions for identification and consistent estimation of the baseline hazard accounting for heaping and unobserved heterogeneity. Our identification strategy requires neither administrative data nor multiple measurements, but a correctly reported duration and the presence of some flat segments in the baseline hazard which includes this correctly reported duration point. We establish the asymptotic properties of the maximum likelihood estimator and provide a simple procedure to test whether the policy had (uniformly) reduced mortality. While our empirical findings do not confirm the latter, they do indicate that accounting for heaping matters for the estimation of the baseline hazard.

ABSTRACT. Twitter text data may be very useful to predict financial tangibles, such as share prices, as well as intangible assets, such as company reputation. While twitter data are becoming widely available to researchers, methods aimed at selecting which twitter data are reliable are, to our knowledge, not yet available. To overcome this problem, and allow to employ twitter data for nowcasting and forecasting purposes, in this contribution we propose an effective statistical method that formalises and extends a quality index employed in the context of the evaluation of academic research: the h-index. Our proposal will be tested on a list of twitterers described by the Financial Times as "the top financial tweeters to follow", for the year 2013. Using our methodology we rank these twitterers and provide confidence intervals to decide whether they are significantly different.

ABSTRACT. Time series characterize a large part of the data stored in economic, financial, medical and scientific databases. The automatic statistical modelling of such data may be a very hard problem when the time series show ``complex'' features, such as nonlinearity, local nonstationarity, high frequency, long memory and periodic components. In such a context, the aim of this paper is to propose an effective clustering technique in the frequency domain, where the need of computational and memory resources is much reduced in order to make the algorithm efficient for large and complex temporal data bases. The problem of the selection of the optimal partition is also discussed along with a proposal which takes into account the stability of the clusters and the efficiency of the procedure in classifying the time series among the different clusters. An application to a large time-series database provided by an Italian electric company is finally discussed showing the good performance of the proposed clustering technique.

ABSTRACT. When a large number of moment restrictions is available there may be restrictions that are more important or credible than others. In these situations it might be desirable to weight each restriction based on our beliefs. This is automatically implemented by a Bayesian procedure. We develop, in this paper, a Bayesian approach to moment estimation and study how to impose moment restrictions on the data distribution through a semiparametric prior distribution for the data generating process F and the structural parameter. We show that a Gaussian process prior for the density function associated with F is particularly convenient in order to impose over-identifying restrictions and allows to have a posterior distribution in closed-form. The posterior distribution resulting from our prior specification is shown to be consistent and asymptotically normal.

ABSTRACT. Patient mobility is a crucial phenomenon in contexts of hospital competition based on quality and driven by patient choice. This study examines inter-regional patient mobility in the Italian National Health Service, a regionally decentralised tax-funded system where in-patient hospital services are provided free at any point of use in the whole country, using administrative data on hospital discharges occurred from 2001 to 2010 in all public and private accredited hospitals. The aim is to understand whether mobility patterns might have consequences on the efficiency and effectiveness of the healthcare provided at the regional level, as well as universalism and equity in healthcare. We specify a gravity model for Origin-Destination (OD) flow data that distinguishes between emissiveness (at Origin) and attractiveness (at Destination) factors affecting bilateral flows. We exploit the longitudinal dimension of the data and estimate a negative binomial conditionally correlated random effects (CCRE) dynamic model that allows for region-pair-specific unobservable heterogeneity. Total and specific types of flow (surgical, medical, acute and cancer-related admissions) are modelled, accounting for correlation between unobserved region-pairs effects and time-variant covariates and their spatial lags. Our main findings indicate that RHSs in the richest regions attract more patients from other regions and that the most effective pull factors are the number of beds and the technology endowment. We also find that the ability of a RHS to attract patients who reside in other regions decreases with the concentration of the organizational structure. Finally, we have detected a mild explosive dynamics in inter-regional patient mobility over time, which could have implications for the long-run sustainability of the overall national-regional health system.

ABSTRACT. This study explores the long-run effects of World War II on health conditions and eating habits in later life. Using microdata for adults from the Italian Multipurpose Survey on Households and exploring regional and cohort variation, we show that individuals that experienced World War II during childhood in more affected regions tend to suffer more from obesity and have a higher probability of myocardial infarction during late adulthood. We provide evidence that the mechanism behind this finding is the excessive consumption of meat later in life as a response to the scarcity of meat at an early age. The estimated effect is particularly strong for individuals that experienced the war during early childhood. Several robustness and placebo tests using different waves and cohorts as well as a triple-differences estimation confirm the causal interpretation of the proposed mechanism. Our results shed light on the hypothesis of a behavioral channel from early-life shocks to adult health.

ABSTRACT. One aim of the giving English NHS patients the right to a wider choice of hospitals in 2006 was to encourage hospitals to compete by raising quality. Previous studies of the effect of Choice policy have found that there was a more rapid reduction in AMI mortality after 2006 in hospitals facing more competitive market structures. We extend these analyses to hip fracture and stroke, and also examine the 2008 extension of Choice policy. We use HES individual patient data from 2002/3 to 2010/11 and measure quality by mortality for AMI, stroke, and hip fracture. We measure competition by (a) the reciprocal of the Herfindahl index computed from predicted patient flows and (b) the number of sites (NHS or private) providing elective care within 30km of a site providing care to AMI, hip fracture or stroke patients. We estimate models with hospital site fixed effects, year effects, patient covariates including morbidity, age, gender, deprivation, distance to provider used and hospital covariates. We find that in the pre-Choice period hospital sites facing more competition had higher AMI and hip fracture mortality. In the post-Choice period the deleterious effect of competition on mortality was smaller and there was no statistically significant relationship between competition and mortality. We found no additional effect of the extension of Choice in 2008 on the relationship between competition and mortality.

ABSTRACT. Family income shapes college opportunities for US students, even when its correlation with academic ability is taken into account. I propose a general equilibrium model to estimate the productivity costs deriving from the fact that human capital investment is not always allocated where its marginal product would be highest. Using the equilibrium conditions of the model, I back out the value of barriers to college investment for disadvantaged students from data on family income, ability, schooling and wages. Counterfactual experiments suggest that a more meritocratic access to college education could boost output by approximately 10%, and wages by between 8% and 11%. I conclude that returns from policies aimed to expand college opportunities are potentially very large.

ABSTRACT. The aim of this paper is to study the effect of teacher digital knowledge and skills on student achievement. We match INVALSI data with a unique student-teacher data-set for a representative sample of Italian students in their second year of upper secondary school containing a wide range of ICT-related variables on both ICT knowledge and ICT using teaching practices. Within-student between-subject estimates show heterogeneous effects of ICT-related teaching practices on student performance. Positive effects are associated with practices which change the lecture style or how the teacher communicates with students, parents and colleagues. On the contrary, more traditional practices have a negative and statistically significant impact. Teacher’s ICT knowledge per se is not significantly related with student performance, but it influences the effectiveness of some ICT teaching practices.

ABSTRACT. It is widely recognized that educational attainment of individuals is affected not only by parental education, but also by the skill composition of the local parental peer group. However, estimating the influence of neighbors on human capital investment decisions is complicated by the endogeneity of location choice. We exploit a rare immigrant settlement policy in Germany to identify the causal impact of parental peer-heterogeneity on the educational outcomes of their children. Children belonging to low educated parents benefit significantly from the presence of high-educated neighbors. In contrast, we do not find any negative influence of the low educated neighbors. To address the reflection problem we restrict to adult peers with no children in the relevant age group. These results are similar to our baseline, reflecting increase in parental aspirations rather than direct child-to-child peer effect. Our results are robust to a range of flexible peer definitions.

ABSTRACT. We study to what extent childhood skills and signals determine socioeconomic adult outcomes for disadvantaged youths who were institutionalized in a juvenile treatment facility in their teenage years, while allowing for multiplier effects from dynamic contagion among the subsequent adult outcomes. We combine data from their treatment files with retrospective interview data and official registry data to decompose adult outcome variables for crime, employment, social welfare, drug use and intimate relationships from age 16 until 32, into payoffs from childhood skills and signals, and payoffs from dynamic contagion in adult outcomes using a factor augmented nonlinear dynamic panel data model. The results suggest that cognitive and non-cognitive childhood skills have lasting effects on male adult outcomes for crime and employment. For females non-cognitive skills are more important in explaining crime, employment, drug use and intimate relationships, but cognitive skills have little persistent effects. The payoffs from the childhood education signal are important for employment and social welfare, whereas the criminal records signal has predictive power for crime and drug use. Further, simulation results suggest the efficiency from interventions targeted to improving non-cognitive skills, as multiplier effects from dynamic contagion in adult outcomes imply large returns from these investments on multiple adult outcomes which exceed the marginal returns.

ABSTRACT. We derive the optimal hedging ratios for a portfolio of assets driven by a Cointegrated Vector Autoregressive model (CVAR) with general cointegration rank. Our hedge is optimal in the sense of minimum variance portfolio. We consider a model that allows for the hedges to be cointegrated with the hedged asset and among themselves. We find that the minimum variance hedge for assets driven by the CVAR, depends strongly on the portfolio holding period. The hedge is defined as a function of correlation and cointegration parameters. For short holding periods the correlation impact is predominant. For long horizons, the hedge ratio should overweight the cointegration parameters rather then short-run correlation information. In the infinite horizon, the hedge ratios shall be equal to the cointegrating vector. The hedge ratios for any intermediate portfolio holding period should be based on the weighted average of correlation and cointegration parameters. The results are general and can be applied for any portfolio of assets that can be modeled by the CVAR of any rank and order.

ABSTRACT. Recent research has emphasized that permanent changes in the innovation variance (caused by structural shifts or an integrated volatility process) lead to size distortions in conventional unit root tests. Cavaliere and Taylor (2004) and Beare (2004) propose nonparametrically corrected versions of unit root tests that have the same asymptotic null distribution as the uncorrected versions in case of homoskedasticity. In this paper, we first derive the asymptotic power envelope for the unit root testing problem when the nonstationary volatility process is known. Next, we show that under suitable conditions, adaptation with respect to the volatility process is possible, in the sense that nonparametric estimation of the volatility process leads to the same asymptotic power envelope. A Monte Carlo experiment shows that these asymptotic results are reflected in finite sample properties, although fairly large sample sizes are needed to fully obtain the asymptotic local power gains.

ABSTRACT. We propose a monitoring procedure to detect a structural change from stationary to integrated behavior. When the procedure is applied to the residuals of a relationship between integrated series it thus monitors a structural change from a cointegrating relationship to a spurious relationship. The cointegration monitoring procedure is based on residuals from modified least squares estimation, using either Fully Modified, Dynamic or Integrated Modified OLS. The procedure is inspired by Chu et al. (1996) in that it is based on parameter estimation on a pre-break calibration period only rather than being based on sequential estimation over the full sample. We investigate the asymptotic behavior of the procedures under the null, for (fixed and local) alternatives and in case of parameter changes. We also study the finite sample performance via simulations. An application to credit default swap spreads illustrates the potential usefulness of the procedure.

ABSTRACT. This paper presents a recursive procedure that characterizes the order of the pole and the coefficients of the Laurent series representation of the inverse of a regular analytic matrix function. The algorithm is based on the `local rank factorization' of a matrix function, which links the local rank-structure of a regular analytic matrix function and the one of its inverse. These results can be used to derive the explicit form of the coefficients of the common trends representation of integrated processes of generic order.

ABSTRACT. We study how tail risk relates to macroeconomic fundamentals and uncertainty. We propose a novel time series model to study the dynamics of tail risk in financial returns. Using daily data from the U.S. stock market, we empirically show that tail risk is countercyclical and exhibits remarkable size effects: tail risk goes up when fundamentals deteriorate and a negative shock hits macroeconomic uncertainty; and larger firms respond more than smaller ones. We further show that connectedness, namely the comovement in tail risk among portfolios, is positively correlated with portfolio-level tail risk. Further evidence from international equity markets robustifies our empirical findings.

ABSTRACT. This paper compares the performance of a quadratic utility function and discusses how to change its characteristic parameter, ARA, so that rating is consistent with return and risk measurements. In particular, this parameter is modified in such a way that a positive return Fund has always a rating higher than one with a negative yield. This modification confirms the possibility to build a new ranking procedure which is more coherent with the actual behaviour of investors.

ABSTRACT. We find that uncertainty in the oil and gold market affects the cross-section of stock returns. We compare and benchmark the role of these alternative asset market uncertainties vis à vis the more traditional equity market uncertainty. Inspired by recent empirical evidence, uncertainty in those asset markets is proxied by the variance risk premiums derived from futures and options traded on the S&P500, oil and gold. We find evidence of both, systematic and asset-specific uncertainty. We document a negative relationship between the various types of uncertainty and firm’s stock returns. An independent increase in S&P, oil and gold market uncertainty coincides with lower return for an important proportion of the stock universe. On the opposite, we show that only S&P uncertainty is a market-wide priced factor in the cross-section of expected stock returns. The other uncertainty factors are sector-specific and are only priced within certain industries. Market industry segmentation explains why a specific factor such as oil uncertainty is only priced for a subset of the stocks.

ABSTRACT. This paper aims to assess dynamic tail risk exposure in the hedge fund sector using daily data. We use a copula function to model both lower and upper tail depen- dence between hedge-fund and broad-market returns as a function of market uncertainty. We proxy the latter by means of a single index that combines the options-implied market volatility, the volatility risk premium, and the swap and term spreads. We nd substantial time-variation in both lower- and upper-tail dependence, even for hedge fund styles that exhibit little unconditional tail dependence. In particular, dependence between hedge fund and equity market returns decreases in both tails signicantly with market uncertainty. There are only two styles that feature neither unconditional nor conditional tail depen- dence, namely, convertible arbitrage and equity market neutral. We also fail to observe any tail dependence with bond and currency markets, though we nd strong evidence that the lower-tail risk exposure of macro hedge funds to commodity markets increases with liquidity risk. Finally, further analysis shows mixed evidence on how much hedge funds contribute to systemic risk.

ABSTRACT. The effects of voluntary work on earnings have recently been studied for some developed countries such as Canada, France and Austria. This paper extends this line of research to Italy, using data from the European Union Statistics on Income and Living Conditions (EU-SILC) dataset. A double methodological approach is used in order to control for unobserved heterogeneity: Heckman and IV methods are employed to account for unobserved worker heterogeneity and endogeneity bias. Empirical results show that, when the unobserved heterogeneity is taken into account, a wage premium of 2.7 percent emerges, quite small if compared to previous investigations on Canada and Austria. The investigation into the channels of influence of volunteering on wages gives support to the hypotheses that volunteering enables the access to fruitful informal networks, avoids the human capital deterioration and provides a signal for intrinsically motivated individuals.

ABSTRACT. We estimate and compare the wage premium to supervision at different quantiles of the distribution of wages for 26 European economies. We build upon the work by Di Nardo, Fortin and Lemieux (1996) and Blau and Khan (1996a) and propose a framework allowing for international comparisons of the distribution of wages. Results from using the EU-SILC 2009 suggest that the premium is higher at the right tail of the distribution, in favour of the hypothesis that it increases inequality; that the premium depends on the economy where the supervisor is employed; and that the labour market institutions shapes the wage premium to supervision.

ABSTRACT. We analyze the determinants of unemployment duration adopting a finite mixture modeling. There are several factors and characteristics that affect unemployment duration and our approach allows identifying clusters of individuals which differs in the way that those factors affect unemployment duration. The idea we are pursuing in our analysis is that similar characteristics and variables can have different effects depending on the specific individual. Our methodology, allows to directly estimating the existence of such clusters, without an a priori definition of them. We apply our methodology to German data on unemployed workers from 2002 and we identify three groups of unemployed individuals that react differently to the same inputs. The groups differ in terms of baseline hazard and on the effect that unemployment benefits have on duration. In particular, some groups have mildly increasing baseline hazard suggesting they put a lot of effort in the search from the very start while others appear to have increasing hazards suggesting that during the first part of the spell they are less active. Moreover benefits have different effect on these groups.

ABSTRACT. We propose a new VAR identication scheme that enables us to iden- tify labor supply shocks from other labor market shocks. According to our analysis on U.S data over the period 1985-2013, labor supply shocks account for a small but significant share of fluctuations in output and unemployment at business-cycle frequencies. We also find that wage bar- gaining shocks are important but not exclusive drivers of unemployment both in the short and in the long run. These results suggest that tradi- tional identication strategies used in estimated New Keynesian models may be misguided.