Abstract:The frequentist variability of Bayesian posterior expectations can provide meaningful measures of uncertainty even when models are misspecified. Classical methods to asymptotically approximate the frequentist covariance of Bayesian estimators such as the Laplace approximation and the nonparametric bootstrap can be practically inconvenient, since the Laplace approximation may require an intractable integral to compute the marginal log posterior, and the bootstrap requires computing the posterior for many different bootstrap datasets. We develop and explore the infinitesimal jackknife (IJ), an alternative method for computing asymptotic frequentist covariance of smooth functionals of exchangeable data, which is based on the "influence function" of robust statistics. We show that the influence function for posterior expectations has the form of a simple posterior covariance, and that the IJ covariance estimate is, in turn, easily computed from a single set of posterior samples. Under conditions similar to those required for a Bayesian central limit theorem to apply, we prove that the corresponding IJ covariance estimate is asymptotically equivalent to the Laplace approximation and the bootstrap. In the presence of nuisance parameters that may not obey a central limit theorem, we argue using a von Mises expansion that the IJ covariance is inconsistent, but can remain a good approximation to the limiting frequentist variance. We demonstrate the accuracy and computational benefits of the IJ covariance estimates with simulated and real-world experiments.

Bayesian Nonparametric Estimation of Ex-post Variance

Bayesian Nonparametric Covariance Estimation with Noisy and Nonsynchronous Asset Prices

Bayesian variable selection in linear regression models with instrumental variables

Bayesian Variance Changepoint Detection in Linear Models with Symmetric Heavy-Tailed Errors

Inferring Volatility Dynamics Using Stock Prices and Variance Swap Rates: an Efficient Bayesian Approach

Bayesian Nonparametric Covariance Regression

Nonparametric Estimation of Large Spot Volatility Matrices for High-Frequency Financial Data

Bayesian Hypothesis Tests Using Nonparametric Statistics

Bayesian Estimation of Dynamic Asset Pricing Models with Informative Observations

The Bayesian Infinitesimal Jackknife for Variance

Improving Markov Switching Models Using Realized Variance

Bayesian Inference for Non-Parametric Extreme Value Theory

Variance function estimation in high-dimensions

A Bayesian Semiparametric Realized Stochastic Volatility Model

High-Dimensional Volatility Matrix Estimation with Cross-Sectional Dependent and Heavy-Tailed Microstructural Noise

Bayesian Nonparametric Adaptive Spectral Density Estimation for Financial Time Series

A general Bayesian model for heteroskedastic data with fully conjugate full-conditional distributions

Variational Inference for Large Bayesian Vector Autoregressions

The Observed Asymptotic Variance: Hard edges, and a regression approach

NOISY HIGH FREQUENCY DATA-BASED ESTIMATION OF VOLATILITY FUNCTION WITH APPLICATIONS

Discussion of “Inferring Volatility Dynamics and Variance Risk Premia: An Efficient Bayesian Approach”