1,721,101 research outputs found
Structural changes in large economic datasets: A nonparametric homogeneity test
No full textThis paper proposes a Bayesian nonparametric homogeneity test for distributional changes. We provide an asymptotic approximation of the Bayes factor and show that it is related to the Shannon entropy. The proposed test is suitable for large high-dimensional datasets which otherwise require time-consuming computation for posterior approximation. An analysis on the FRED-QD macroeconomic dataset shows the ability of the test to detect relevant structural changes in the US economy
Monte carlo within simulated annealing for integral constrained optimizations
Full text linkFor years, Value-at-Risk and Expected Shortfall have been well established measures of market risk and the Basel Committee on Banking Supervision recommends their use when controlling risk. But their computations might be intractable if we do not rely on simplifying assumptions, in particular on distributions of returns. One of the difficulties is linked to the need for Integral Constrained Optimizations. In this article, two new stochastic optimization-based Simulated Annealing algorithms are proposed for addressing problems associated with the use of statistical methods that rely on extremizing a non-necessarily differentiable criterion function, therefore facing the problem of the computation of a non-analytically reducible integral constraint. We first provide an illustrative example when maximizing an integral constrained likelihood for the stress-strength reliability that confirms the effectiveness of the algorithms. Our results indicate no clear difference in convergence, but we favor the use of the problem approximation strategy styled algorithm as it is less expensive in terms of computing time. Second, we run a classical financial problem such as portfolio optimization, showing the potential of our proposed methods in financial applications
Sticky proposal densities for adaptive MCMC methods
No full textMonte Carlo (MC) methods are commonly used in Bayesian signal processing to address complex inference problems. The performance of any MC scheme depends on the similarity between the proposal (chosen by the user) and the target (which depends on the problem). In order to address this issue, many adaptive MC approaches have been developed to construct the proposal density iteratively. In this paper, we focus on adaptive Markov chain MC (MCMC) algorithms, introducing a novel class of adaptive proposal functions that progressively 'stick' to the target. This proposed class of sticky MCMC methods converge very fast to the target, thus being able to generate virtually independent samples after a few iterations. Numerical simulations illustrate the excellent performance of the sticky proposals when compared to other adaptive and non-adaptive schemes
On the role of dependence in sticky price and sticky information Phillips curve: Modelling and forecasting
No full textUnderstanding the role of sticky price and sticky information for inflation dynamics is a key issue in economics. The literature has treated the two forms of stickiness as independent. This paper proposes a new dual stickiness Phillips curve based on dependence among the events of setting prices and updating information. Using US data over the period 1947Q1–2020Q1, the new model is scrutinized against a dual stickiness model without dependence, a pure sticky price model, and a pure sticky information model, through in- and out-of-sample analyses. The results show: (i) the new model outperforms the model without dependence in-sample; (ii) the dual stickiness models perform similarly out-of-sample; and (iii) the pure sticky models yield the worst forecasts. The results have some implications for policy makers and practitioners. A policy maker may consider the new model given its performance in- and out-of-sample, while a practitioner may prefer the model without dependence, given its lesser complexity and its competitive forecasting performance
Generalized Poisson difference autoregressive processes
Full text linkThis paper introduces a novel stochastic process with signed integer values. Its autoregressive dynamics effectively captures persistence in conditional moments, rendering it a valuable feature for forecasting applications. The increments follow a Generalized Poisson distribution, capable of accommodating over- and under-dispersion in the conditional distribution, thereby extending standard Poisson difference models. We derive key properties of the process, including stationarity conditions, the stationary distribution, and conditional and unconditional moments, which prove essential for accurate forecasting. We provide a Bayesian inference framework with an efficient posterior approximation based on Markov Chain Monte Carlo. This approach seamlessly incorporates inherent parameter uncertainty into predictive distributions. The effectiveness of the proposed model is demonstrated through applications to benchmark datasets on car accidents and an original dataset on cyber threats, highlighting its superior fitting and forecasting capabilities compared to standard Poisson model
Random Projection Methods in Economics and Finance
No full textDimension reduction techniques have been proposed to cope with some modelling and forecasting issues with large and complex datasets such as overfitting, over-parametrization and inefficiency. Among all dimension reduction techniques, the most widespread are principal component analysis and factor analysis. Recently, Random Projection (RP) techniques became popular in many fields due to their simplicity and effectiveness and found applications to machine learning, statistics and econometrics. The basis of the RP technique relies on the remarkable result in the Johnson-Lindenstrauss lemma, which provides some conditions to achieve an effective reduction of the size of the data, without altering their information content. This chapter reviews the most used dimensionality reduction techniques, introduces random projection methods and shows their effectiveness in time series analysis through a simulation study and some original applications to tracking and forecasting financial indexes and to predicting electricity trading volumes. Our empirical results suggest that random projection preprocessing of the data does not jeopardize the validity of inference and prediction procedures and possibly improves their efficiency.Dimension reduction techniques have been proposed to cope with some modelling and forecasting issues with large and complex datasets such as overfitting, over-parametrization and inefficiency. Among all dimension reduction techniques, the most widespread are principal component analysis and factor analysis. Recently, Random Projection (RP) techniques became popular in many fields due to their simplicity and effectiveness and found applications to machine learning, statistics and econometrics. The basis of the RP technique relies on the remarkable result in the Johnson-Lindenstrauss lemma, which provides some conditions to achieve an effective reduction of the size of the data, without altering their information content. This chapter reviews the most used dimensionality reduction techniques, introduces random projection methods and shows their effectiveness in time series analysis through a simulation study and some original applications to tracking and forecasting financial indexes and to predicting electricity trading volumes. Our empirical results suggest that random projection preprocessing of the data does not jeopardize the validity of inference and prediction procedures and possibly improves their efficiency
Density calibration with consistent scoring functions
No full textThis contribution studies a calibration approach for predictive densities based on generalized scoring rules. We consider a set of simulated experiments in order to study the effectiveness of the metho
Endogeneity in Interlocks and Performance Analysis: A Firm Size Perspective
Full text linkThis paper contributes to the literature on interlocking directorates (ID) by providing a new solution to the two econometric issues arising in the joint analysis of interlocks and firm performance which are the endogenous nature of ID and sample selection bias due to the exclusion of isolated firms. Some key determinants of ID network formation are identified and used to check for endogeneity. We analyze the impact of the positioning in the network on firms’ performance and inspect how the impact varies across firms of different sizes drawing on information relating to 37,324 firms in the interlocking network which, to our knowledge, is the widest dataset ever used in approaching the study of ID. Our results, made robust for endogeneity and sample selection bias, suggest that eigenvector centrality and the clustering coefficient have a positive and significant impact on all the performance measures and that this effect is more pronounced for small firms
A Dynamic Stochastic Block Model with infinite communities
No full textThis contribution proposes the use of bayesian non–parametric techniques
to make inference on the number of communities in a Dynamic Stochastic Block
Model which is then applied to real network data on international financial flow
Fiscal Policy Regimes in Resource-Rich Economies
Full text linkWe analyse fiscal policy in resource-rich economies using a novel Bayesian regime-switching panel model. The identified regimes capture pro- or countercyclical fiscal behaviour, while the switches between the regimes have the interpretation of changes in fiscal policy. Applying the model to sixteen oil-producing economies, we show that fiscal policy has alternated between a procyclical and countercyclical regime multiple times over the sample. Furthermore, we find fiscal policy to be the most volatile in the procyclical regime and that the probability of being in the procyclical regime is higher for OPEC countries rather than non OPEC countries. We also show that following either an increase or decrease in oil revenues, the growth in government expenditure mostly increases, suggesting there is an upward bias in expenditures in oil-producing countries. These are new findings in the literature
- …
