Operations Research
http://lists.repec.org/mailman/listinfo/nep-ore
Operations Research2014-10-17Walter FrischLarge Bayesian VARMAs
http://d.repec.org/n?u=RePEc:str:wpaper:1409&r=ore
Abstract: Vector Autoregressive Moving Average (VARMA) models have many theoretical properties which should make them popular among empirical macroeconomists. However, they are rarely used in practice due to over-parameterization concerns, difficult - ties in ensuring identification and computational challenges. With the growing interest in multivariate time series models of high dimension, these problems with VARMAs become even more acute, accounting for the dominance of VARs in this field. In this paper, we develop a Bayesian approach for inference in VARMAs which surmounts these problems. It jointly ensures identification and parsimony in the context of an efficient Markov chain Monte Carlo (MCMC) algorithm. We use this approach in a macroeconomic application involving up to twelve dependent variables. We find our algorithm to work successfully and provide insights beyond those provided by VARsJoshua C C Chan, Eric Eisenstat, Gary Koop2014-09VARMA identification, Markov Chain Monte Carlo, Bayesian, stochastic search variable selectionTesting for Neglected Nonlinearity Using Regularized Artificial Neural Networks
http://d.repec.org/n?u=RePEc:ucr:wpaper:201422&r=ore
The artificial neural network (ANN) test of Lee et al. (Journal of Econometrics 56, 269–290, 1993) uses the ability of the ANN activation functions in the hidden layer to detect neglected functionalmisspecification. As the estimation of the ANN model is often quite difficult, LWG suggested activate the ANN hidden units based on randomly drawn activation parameters. To be robust to the random activations, a large number of activations is desirable. This leads to a situation for which regularization of the dimensionality is needed by techniques such as principal component analysis (PCA), Lasso, Pretest, partial least squares (PLS), among others. However, some regularization methods can lead to selection bias in testing if the dimensionality reduction is conducted by supervising the relationship between the ANN hidden layer activations of inputs and the output variable. This paper demonstrates that while these supervised regularization methods such as Lasso, Pretest, PLS, may be useful for forecasting, they may not be used for testing because the supervised regularizationwould create the post-sample inference or post-selection inference (PoSI) problem. Our Monte Carlo simulation shows that the PoSI problem is especially severe with PLS and Pretest while it seems relatively mild or even negligible with Lasso. This paper also demonstrates that the use of unsupervised regularization does not lead to the PoSI problem. Lee et al. (Journal of Econometrics 56, 269–290, 1993) suggested a regularization by principal components, which is a unsupervised regularization.While the supervised regularizations may be useful in forecasting, regularization should not be supervised in inference.Tae-Hwy Lee, Zhou Xi, Ru Zhang2013-09Randomized ANN activations â€¢ Dimension reduction â€¢ Supervised regularization â€¢ Unsupervised regularization â€¢ PCA â€¢ Lasso â€¢ PLS â€¢ Pretest â€¢ PoSI problemThe wild tapered block bootstrap
http://d.repec.org/n?u=RePEc:aah:create:2014-32&r=ore
In this paper, a new resampling procedure, called the wild tapered block bootstrap, is introduced as a means of calculating standard errors of estimators and constructing confidence regions for parameters based on dependent heterogeneous data. The method consists in tapering each overlapping block of the series first, then applying the standard wild bootstrap for independent and heteroscedastic distributed observations to overlapping tapered blocks in an appropriate way. It preserves the favorable bias and mean squared error properties of the tapered block bootstrap, which is the state-of-the-art block-based method in terms of asymptotic accuracy of variance estimation and distribution approximation. For stationary time series, the asymptotic validity, and the favorable bias properties of the new bootstrap method are shown in two important cases: smooth functions of means, and M-estimators. The first-order asymptotic validity of the tapered block bootstrap as well as the wild tapered block bootstrap approximation to the actual distribution of the sample mean is also established when data are assumed to satisfy a near epoch dependent condition. The consistency of the bootstrap variance estimator for the sample mean is shown to be robust against heteroskedasticity and dependence of unknown form. Simulation studies illustrate the finite-sample performance of the wild tapered block bootstrap. This easy to implement alternative bootstrap method works very well even for moderate sample sizes.Ulrich Hounyo2014-09-24Block bootstrap, Near epoch dependence, Tapering, Variance estimationOn the infinite-dimensional representation of stochastic controlled systems with delayed control in the diffusion term
http://d.repec.org/n?u=RePEc:eve:wpaper:14-06&r=ore
In the deterministic context a series of well established results allow to reformu- late delay differential equations (DDEs) as evolution equations in infinite dimensional spaces. Several models in the theoretical economic literature have been studied using this reformulation. On the other hand, in the stochastic case only few results of this kind are available and only for specific problems. The contribution of the present letter is to present a way to reformulate in infinite dimension a prototype controlled stochastic DDE, where the control variable appears delayed in the diffusion term. As application, we present a model for quadratic risk minimization hedging of European options with execution delay and a time-to-build model with shock. Some comments concerning the possible employment of the dynamic programming after the reformulation in infinite dimension conclude the letter.Giorgio Fabbri, Salvatore Federico2014Stochastic delay differential equations, Evolution equations in Hilbert space, Dynamic programmingIdentification of DSGE Models - the Effect of Higher-Order Approximation and Pruning
http://d.repec.org/n?u=RePEc:cqe:wpaper:3314&r=ore
Several formal methods have been proposed to check local identification in linearized DSGE models using rank criteria. Recently there has been huge progress in the estimation of non-linear DSGE models, yet formal identification criteria are missing. The contribution of the paper is threefold: First, we extend the existent methods to higher-order approximations and establish rank criteria for local identification given the pruned state-space representation. It is shown that this may improve overall identification of a DSGE model via imposing additional restrictions on the moments and spectrum. Second, we derive analytical derivatives of the reduced-form matrices, unconditional moments and spectral density for the pruned state-space system. Third, using a second-order approximation, we are able to identify previously non-identifiable parameters: namely the parameters governing the investment adjustment costs in the Kim (2003) model and all parameters in the An and Schorfheide (2007) model, including the coeffcients of the Taylor-rule.Willi Mutschler2014-10non-linear DSGE, rank condition, analytical derivatives, pruned state-spaceIncreases In Risk and Demand for Risky Asset
http://d.repec.org/n?u=RePEc:ipg:wpaper:2014-600&r=ore
In this paper, we examine the eect of a decrease in risk on the demand for risky asset in the standard portfolio problem. We introduce a new class of dominance, that we name relative order and we prove that this class of dominance is consistent both with central dominance introduced by Gollier (1995) and with mean preserving in- crease in risk. Finally, we show that some known classes of dominance are particular cases of our new class of dominance.A.Chateauneuf, G.Lakhnati2014-09-29Central Dominance, EU Model, Mean Preserving Increase in Risk, Port- folio Choice, Relative Simple Dominance, Relative Dominance.The zero lower bound and parameter bias in an estimated DSGE model
http://d.repec.org/n?u=RePEc:van:wpaper:vuecon-sub-14-00009&r=ore
This paper examines how and to what extent parameter estimates can be biased in a dynamic stochastic general equilibrium (DSGE) model that omits the zero lower bound (ZLB) constraint on the nominal interest rate. Our Monte Carlo experiments using a standard sticky-price DSGE model show that no significant bias is detected in parameter estimates and that the estimated impulse response functions are quite similar to the true ones. However, as the probability of hitting the ZLB increases, the parameter bias becomes larger and therefore leads to substantial differences between the estimated and true impulse responses. It is also demonstrated that the model missing the ZLB causes biased estimates of structural shocks even with the virtually unbiased parameters.Yasuo Hirose, Atsushi Inoue2014-09-09A Network Analysis of the Evolution of the German Interbank Market
http://d.repec.org/n?u=RePEc:rza:wpaper:461&r=ore
In this paper, we report a descriptive investigation of the structural evolution of two of the most important over-the-counter markets for liquidity in Germany: the interbank market for credit and for derivatives. We use end-of-quarter data from the German large credit register between 2002 and 2012 and characterize the underlying networks. Surprisingly, the data show little or no impact of the 2008 crisis on the structure of credit market. The derivative market however exhibits a peak of concentration in the run up to the crisis. Globally, both markets exhibit high stability for most of the networks metrics and high correlation amongst them.Tarik Roukny, Co-Pierre Georg and Stefano Battiston2014financial networks, interbank market, credit default swaps, liquidityDynamic Strategies for Successful Online Crowdfunding
http://d.repec.org/n?u=RePEc:net:wpaper:1409&r=ore
Crowdfunding is a fast emerging internet fundraising mechanism for soliciting capital from the crowd to support entrepreneurial ventures. This paper empirically investigates the dynamics of investors’ backing behaviors in the presence of network externalities and a finite time window. The proposed model captures how investors dynamically update their expectations on the prospect of a project based on its current funding status and time progress. Model estimation shows that investors are more likely to back a project that has already attracted a critical mass of funding (positive network externalities). For the same amount of achieved funding, the backing propensity declines over time (negative time effects). These two opposing forces give rise to a critical mass of funding the project must attain on time to achieve successful funding by the deadline. Counterfactual simulations show that projects may fail to attain the critical mass because of unfavorable shocks in investor visits at the early stage of the funding cycle. We derive dynamic seeding strategies for project owners to maximize the likelihood of funding success.Zhuoxin Li, Jason A. Duan2014-09crowdfunding; group buying; entrepreneurship; network externality; hazards model; Bayesian inferenceInitial-Condition Free Estimation of Fixed Effects Dynamic Panel Data Models
http://d.repec.org/n?u=RePEc:siu:wpaper:16-2014&r=ore
It is well known that (quasi) MLE of dynamic panel data (DPD) models with short panels depends on the assumptions on the initial values; ignoring them or a wrong treatment of them will result in inconsistency or serious bias. This paper introduces a initial-condition free method for estimating the fixed-effects DPD models, through a simple modification of the quasi-score. An outer-product-of-gradients (OPG) method is also proposed for robust inference. The MLE of Hsiao, Pesaran and Tahmiscioglu (2002, Journal of Econometrics), where the initial observations are modeled, is extended to quasi MLE and an OPG method is proposed for robust inference. Consistency and asymptotic normality for both estimation strategies are established, and the two methods are compared through Monte Carlo simulations. The proposed method performs well in general, whether the panel is short or not. The quasi MLE performs comparably, except when model does not contain time-varying regressor, or the panel is not short and the dynamic parameter is small. The proposed method is much simpler and easier to apply.Zhenlin Yang2014-09Bias reduction; Consistency; Asymptotic normality; Dynamic panel; Fixed effects; Modified quasi-score; Robust standard error; Short panelQuality Provision in the Presence of a Biased Intermediary
http://d.repec.org/n?u=RePEc:net:wpaper:1406&r=ore
In many industries, consumers rely on recommendations by an intermediary when choosing between competing products. In this paper, we look at how the existence of contracts between firms and intermediaries affects the quality of the advice received by consumers, and firms' incentives to invest in improving the quality of their products. We consider a model with one intermediary and two firms who decide how much to invest. Under a variety of contractual environments (vertical integration, ex post endorsement) we show that, even though the intermediary tends to endorse the best firm, contractual endorsement distorts firms' incentives to invest. Quality can then decrease or increase compared to an objective benchmark. We contrast our approach to a setup with fixed qualities and endogenous prices, under which contractual endorsement hurts consumers.Alexandre de Cornière, Greg Taylor2014-09intermediary, quality, biasMoment-based estimation of nonlinear regression models with boundary outcomes and endogeneity, with applications to nonnegative and fractional responses
http://d.repec.org/n?u=RePEc:cfe:wpcefa:2014_09&r=ore
In this paper we suggest simple moment-based estimators to deal with unobserved heterogeneity in a special class of nonlinear regression models that includes as main particular cases exponential models for nonnegative responses and logit and complementary loglog models for fractional responses. The proposed estimators: (i) treat observed and omitted covariates in a similar manner; (ii) can deal with boundary outcomes; (iii) accommodate endogenous explanatory variables without requiring knowledge on the reduced form model, although such information may be easily incorporated in the estimation process; (iv) do not require distributional assumptions on the unobservables, a conditional mean assumption being enough for consistent estimation of the structural parameters; and (v) under the additional assumption that the dependence between observables and unobservables is restricted to the conditional mean, produce consistent estimators of partial effects conditional only on observables.Esmeralda A. Ramalho, Joaquim J.S. Ramalho2014Unobserved heterogeneity; Endogeneity; Boundary outcomes; Fractional regression; Exponential regression; Transformation regression models.Impact Estimates for Static Spatial Panel Data Models in R
http://d.repec.org/n?u=RePEc:rri:wpaper:2013wp05&r=ore
In the present note we demonstrate how to implement the Lee and Yu (2010) procedure for fixed effects spatial panel data models available from the R (R Development Core Team 2012) package splm (Millo and Piras 2012). Additionally, we also show how to compute the impact estimates introduced by Kelejian, Tavlas, and Hondroyiannis (2006) and formalized in LeSage and Pace (2009). Unlike Matlab (MATLAB 2011), there was no R function specific to static panel data models for the calculation of the impact measures. After receiving numerous requests from the users of splm, we decided to extend the cross sectional functions available from spdep (Bivand 2013) to spatial panel data models.Gianfranco Piras2013-05spatial panel data models, R, computational methods, impact measures