Computing moments of ratios of quadratic forms in normal variables

The accuracy and speed of numerical methods for computing the moments of a ratio of quadratic forms in normal variables is examined, with particular application to the sample auto-correlation function. Methods based on a saddlepoint approximation are demonstrated to be not only superior to existing approximations, but are numerically reliable and virtually as accurate as the method suitable for exact computations, while taking only a fraction of the time to compute. The new method also maintains its accuracy for time series models near the nonstationary border, which is of significant interest for unit-root inference and also a case for which first-order mean and variance approximations break down. As a wide variety of test statistics and their power functions arising in econometric models are expressible in the general form considered, the method should prove very useful for data analysis and model building.

[1]  T. Sawa The exact moments of the least squares estimator for the autoregressive model , 1978 .

[2]  Offer Lieberman A Laplace approximation to the moments of a ratio of quadratic forms , 1994 .

[3]  Aman Ullah,et al.  Moments of the ratio of quadratic forms in non-normal variables with econometric examples , 1994 .

[4]  P. A. P. Moran,et al.  SOME THEOREMS ON TIME SERIES II. THE SIGNIFICANCE OF THE SERIAL CORRELATION COEFFICIENT , 1948 .

[5]  Hung T. Nguyen,et al.  Quadratic Forms in Normal Random Variables , 1989 .

[6]  Mark A. McComb A Practical Guide to Heavy Tails , 2000, Technometrics.

[7]  Leigh A. Roberts,et al.  On the Existence of Moments of Ratios of Quadratic Forms , 1995, Econometric Theory.

[8]  S. Provost,et al.  The exact distribution of indefinite quadratic forms in noncentral normal vectors , 1996 .

[9]  H. Daniels Saddlepoint Approximations in Statistics , 1954 .

[10]  Oliver D. Anderson,et al.  EXACT GENERAL-LAG SERIAL CORRELATION MOMENTS AND APPROXIMATE LOW-LAG PARTIAL CORRELATION MOMENTS FOR GAUSSIAN WHITE NOISE , 1993 .

[11]  S. Rice,et al.  Saddle point approximation for the distribution of the sum of independent random variables , 1980, Advances in Applied Probability.

[12]  R. W. Farebrother,et al.  The Distribution of a Quadratic Form in Normal Variables , 1990 .

[13]  Jean-Marie Dufour,et al.  Some Robust Exact Results on Sample Autocorrelations and Tests of Randomness , 1985 .

[14]  J. S. Mehta,et al.  THE EXISTENCE OF MOMENTS OF SOME SIMPLE BAYES ESTIMATORS OF COEFFICIENTS IN A SIMULTANEOUS EQUATION MODEL. , 1978 .

[15]  Offer Lieberman,et al.  Saddlepoint Approximation for the Distribution of a Ratio of Quadratic Forms in Normal Variables , 1994 .

[16]  Murray D. Smith,et al.  On the expectation of a ratio of quadratic forms in normal variables , 1989 .

[17]  J. Gooijer Exact moments of the sample autocorrelations from series generated by general arima processes of order (p, d, q), d=0 or 1 , 1980 .

[18]  J. Magnus,et al.  Evaluation of moments of quadratic forms and ratios of quadratic forms in normal variables: background, motivation and examples , 1993 .

[19]  Ronald W. Butler,et al.  Approximate distributions for the various serial correlograms , 1998 .

[20]  Murray D. Smith,et al.  EXPECTATIONS OF RATIOS OF QUADRATIC FORMS IN NORMAL VARIABLES: EVALUATING SOME TOP‐ORDER INVARIANT POLYNOMIALS , 1993 .

[21]  F. Graybill,et al.  Matrices with Applications in Statistics. , 1984 .

[22]  J. Magnus The Exact Moments of a Ratio of Quadratic Forms in Normal Variables , 1986 .

[23]  T. Sawa Finite-Sample Properties of the k-Class Estimators , 1972 .

[24]  Aman Ullah,et al.  Moments of the function of non-normal random vector with applications to econometric estimators and test statistics 1 , 1995 .

[25]  Mukhtar M. Ali Distributions of the Sample Autocorrelations When Observations Are From a Stationary Autoregressive-Moving-Average Process , 1984 .

[26]  William H. Press,et al.  Numerical recipes : the art of scientific computing : FORTRAN version , 1989 .

[27]  P. Moran Testing for serial correlation with exponentially distributed variates. , 1967, Biometrika.

[28]  G. A. Ghazal Moments of the ratio of two dependent quadratic forms , 1994 .

[29]  Antti J. Kanto Covariances between estimated autocorrelations of an ARMA process , 1988 .

[30]  J. Imhof Computing the distribution of quadratic forms in normal variables , 1961 .

[31]  Georgi N. Boshnakov,et al.  Bartlett's formulae—Closed forms and recurrent equations , 1996 .

[32]  William H. Press,et al.  Book-Review - Numerical Recipes in Pascal - the Art of Scientific Computing , 1989 .

[33]  W. Krämer The Linear Regression Model Under Test , 1986 .

[34]  P. Moran,et al.  Some theorems on time series. , 1947, Biometrika.