Luiggi Donayre

The Asymmetric Effects of Oil Price Shocks on the Canadian Economy

A threshold vector autoregression (TVAR) is estimated to study the effects of oil price shocks on Canadian output and price level. While much of the literature has investigated potential asymmetric effects of positive and negative oil price shocks within a linear vector autoregression (VAR), we do so within a nonlinear VAR. Further, we extend the analysis to consider the correlation between asymmetries associated with the business cycle phase and size/sign asymmetries. Positive oil price shocks are found to have a stronger effect on output than negative oil price shocks. This asymmetry is significant in recessions, but lessened during expansions. The results also suggest that the reduction in inflation due to a negative oil price shock is larger than the increase in inflation following a positive oil price shock, especially during periods of low output growth. Yet, neither inflation nor output growth seems to vary disproportionately with the size of the oil price shock. In general, the results are robust to the ordering of the variables in the VAR process and to the time window over which the net oil price change is computed

ESTIMATED THRESHOLDS IN THE RESPONSE OF OUTPUT TO MONETARY POLICY: ARE LARGE POLICY CHANGES LESS EFFECTIVE?

This paper investigates the potential sources of the mixed evidence found in the empirical literature studying asymmetries in the response of output to monetary policy shocks of different magnitudes. Further, it argues that such mixed evidence is a consequence of the exogenous imposition of the threshold that classifies monetary shocks as small or large. To address this issue, I propose an unobserved-components model of output, augmented by a monetary policy variable, which allows the threshold to be endogenously estimated. The results show strong statistical evidence that the effect of monetary policy on output varies disproportionately with the size of the monetary shock once the threshold is estimated. Meanwhile, the estimates of the model are consistent with a key implication of menu-cost models: smaller monetary shocks trigger a larger response on output.

Outliers and persistence in threshold autoregressive processes

Abstract This paper uses Monte Carlo simulations to investigate the effects of outlier observations on the properties of linearity tests against threshold autoregressive (TAR) processes. By considering different specifications and levels of persistence for the data-generating processes, we find that additive outliers distort the size of the test and that the distortion increases with the level of persistence. In addition, we also find that larger additive outliers can help to improve the power of the test in the case of persistent TAR processes.

Improving likelihood-ratio-based confidence intervals for threshold parameters in finite samples

Within the context of threshold regressions, we show that asymptotically-valid likelihood-ratio-based confidence intervals for threshold parameters perform poorly in finite samples when the threshold effect is large. A large threshold effect leads to a poor approximation of the profile likelihood in finite samples such that the conventional approach to constructing confidence intervals excludes the true threshold parameter value too often, resulting in low coverage rates. We propose a conservative modification to the standard likelihood-ratio-based confidence interval that has coverage rates at least as high as the nominal level, while still being informative in the sense of including relatively few observations of the threshold variable. An application to thresholds for U.S. industrial production growth at a disaggregated level shows the empirical relevance of applying the proposed approach.

Causality between Public Debt and Real Growth in the OECD: A Country-by-Country Analysis

We analyze the direction of causality between public debt and real economic growth in a sample of 20 OECD countries for a period of 40 years starting in 1970. Given the persistence of real growth rates, we estimate canonical cointegrating regressions to allow for the possibility of stochastic cointegrating vectors. We then make inferences about the direction of causality by means of both Granger tests and VAR-based tests that do not depend on whether the series are integrated or cointegrated. We find that while modern welfare states tend to face low real growth following increases in public debt, more traditional welfare states and those with larger governments typically exhibit either causality from low growth to debt accumulation or bidirectional causality. However, the heterogeneity of the results suggests caution when making general statements about the relationship between these variables. In particular, the causal link is intrinsic to each country and it cannot be inferred that higher debt always leads to lower economic growth.

Do monetary policy shocks generate TAR or STAR dynamics in output

Abstract This paper studies whether the relationship between monetary policy shocks of different size and output is better described by threshold autoregressive (TAR) or smooth transition autoregressive (STAR) dynamics. Using a Bayesian framework, a TAR process and a STAR process are formally compared within an unobserved components model of output, augmented with a monetary policy variable. The Bayesian model comparison favors the notion that the dynamics are nonlinear and better described by a smooth transition between regimes, which suggests that aggregation plays a role in the dynamics between monetary policy and output. This evidence is further supported by the results of a model that uses output data at the sectoral level: when more disaggregated data are employed, the transition between regimes is more abrupt. Moreover, the results show that, when the transition between regimes is smooth, large monetary policy shocks identified as the residuals of a VAR are neutral, consistent with the implications of menu-costs models.