The 'fixed regressor' -or 'fixed design' -bootstrap is usually considered in the context of classic regression, or conditional mean (autoregressive) models, see for example, Goncalves and Kilian, 2004). We consider here inference for a general class of (non)linear ARCH models of order q, based on a 'Fixed Volatility' bootstrap. In the Fixed Volatility bootstrap, the lagged variables in the conditional variance equation are kept fixed at their values in the original series, while the bootstrap innovations are, as is standard, resampled with replacement from the estimated residuals based on quasi maximum likelihood estimation. We derive a full asymptotic theory to establish validity for the Fixed Volatility bootstrap applied to Wald statistics for general restrictions on the parameters. A key feature of the Fixed Volatility bootstrap is that the bootstrap sample, conditional on the original data, is an independent sequence. Inspection of the proof of bootstrap validity reveals that such conditional independence simplifies the asymptotic analysis considerably. In contrast to other bootstrap methods, one does not have to take into account the conditional dependence structure of the bootstrap process itself. We also investigate the finite sample performance of the Fixed Volatility bootstrap by means of a small scale Monte Carlo experiment. We find evidence that for small sample sizes, the Fixed Volatility bootstrap test is superior to the asymptotic test, and to the recursive bootstrap-based test. For large samples, both bootstrap schemes and the asymptotic test share properties, as expected from the asymptotic theory. Its appealing theoretical properties, together with its good finite sample performance, suggest that the proposed Fixed Volatility bootstrap may be an important tool for the analysis of the bootstrap in more general volatility models.
FIXED VOLATILITY BOOTSTRAP
921With respect to the fixed design bootstrap in regression models, which keep the conditional mean of the bootstrap data fixed among bootstrap repetitions, when dealing with volatility models the fixed design bootstrap maintains the conditional volatility fixed among bootstrap repetitions. This requires generating heteroskedasticity in the bootstrap data by recolouring the (conditionally i.i.d.) bootstrap innovations with some estimates of the volatility process which depend on the original data only. Hence, and in contrast to other bootstrap methods such as the recursive bootstrap, this method -which we label the 'Fixed Volatility' bootstrap in what follows -generates bootstrap samples which, conditional on the original data, features the same conditional volatility process.The recursive bootstrap for ARCH type model has been explored in the existing literature. For instance, Hidalgo and Zaffaroni (2007) propose a recursive bootstrap scheme for model specification tests in ARCH(∞) models. Corradi and Iglesias (2008) compare recursive bootstrap with block bootstrap methods in terms of asymptotic refinements. Pascual et al. (2006) exploit the rec...