Click here for search results

Newsletter

Site Tools

C. Hevia - Other Publications

Research Topics

Economic Growth

  • "Argentina's Divergence. Failed Reforms and Incredible Policies in Argentine Exceptionalism", with P.A. Neumeyer and H. Hopenhayn, E. Glaeser and R. Di Tella (eds.), JFK Harvard University . Forthcoming.
  • Saving and Growth in Egypt, with N. Loayza, World Bank Policy Research Working Paper 5529, 2011.
  • "Saving and Growth in Sri Lanka", with N. Loayza, 2011
  • "Saving in Turkey: An International Comparison", 2010.
  • "Saving in Egypt: Trend and Explanations", with N. Loayza and Y. Ikeda, June 2010.
  • Privatization-Nationalization Cycles, with N. V. Loayza and R. Chang, World Bank Policy Research Working Paper 5029, 2009.

International Economics

Monetary and Exchange Rate Policy

  • "Estimating and Forecasting the Term Structure of Government Bonds with Regime Changes", with M. Sola, M. Gonzalez-Rozada and F. Spagnolo, 2011
  • "Is the Euro a Constraint on Policy?" with J.P. Nicolini, 2011.
  • Optimal Devaluation, with J. P. Nicolini, World Bank Policy Research Working Paper 4926, 2009.

Software (Matlab routines)
The files below can be read in textpad/notepad, and then saved to the user's PC.  To use the files as intended, the user will need their own copy of Matlab.

Econometrics routines

  • arma_mle.m is function for performing the Maximum Likelihood estimation of an autoregressive moving average model ARMA( p, q ).
  • harvey_pierse.m is a function that estimates an ARMA(p,q) model with missing observations or time-aggregated data, and computes an estimates of these missing observations or disaggregated data.
  • var_decp.m is a function that computes the variance decomposition of the forecast errors of a vector autoregressive process of arbitrary order `p’. Shocks are orthogonalized using the Cholesky decomposition, so that the ordering of the variables in the VAR matter for the results.
     
  • Routines to compute robust GMM-based standard errors of several statistics: correl_gmm.m computes the standard error of the correlation coefficient between two time series, ratsdev_gmm.m computes the standard error of the ratio of standard deviations of two time series, and sdev_gmm.m  computes the standard error of the standard deviation of a single time series.  These standard errors are robust to heteroskedasticity and to serial correlation in the series. (The programs can be easily extended to compute the standard error of virtually any estimator.)
    • Documentation (this file also explains how to compute the gmm standard errors of, virtually, anything).

Filtering and smoothing routines

  • hpfilter.m computes the Hodrick-Prescott filter of a time series.
     
  • hpfiltbreak.m  computes the Hodrick-Prescott filter of a time series with structural breaks at known points.
     
  • kalman_filter.m  computes the Kalman filter and the log-likelihood of a discrete time linear dynamical system written in state-space form.
     
  • kalman_smoother.m computes the Kalman filter estimates and Kalman smoother estimates of a discrete time linear dynamical system  written in state space form.
     
  • kfilter_udu.m  is a function that computes the stationary Kalman filter using the UDU’ decomposition of the covariance matrix. This approach increases substantially the numerical precision of the algorithm because it propagates the square root of the covariance matrix instead of the covariance matrix itself. This improves numerical precision because the condition number of the square root of the covariance matrix is half of that of the covariance matrix.

 Additional routines (numerical integration and linear algebra)

  • The function gauss_hermite.m computes the abscissas and weights of the n-point Gauss-Hermite quadrature formula. This routine is useful to compute the numerical integral of a function over the domain (-infty,+infty).

*Typing “help +(the name of the routine)” in Matlab provides a detailed explanation of the routine, inputs, and output.




Permanent URL for this page: http://go.worldbank.org/7IC10OPOV0


© 2016 The World Bank Group, All Rights Reserved. Legal