HAR Inference: Recommendations for Practice

Lazarus, Eben; Lewis, Daniel; Stock, James H.; Watson, Mark W.

doi:10.1080/07350015.2018.1506926

Cited by 173 publications

(137 citation statements)

References 33 publications

(48 reference statements)

Supporting

Mentioning

115

Contrasting

Order By: Relevance

“…Monte Carlo simulations in Hualde and Iacone () and Lazarus et al () show that fixed‐ m asymptotics has the same size–power tradeoff documented for fixed‐ b asymptotics: the smaller the value for m , the better the empirical size, but also the weaker the power.…”

Section: Fixed‐smoothing Asymptoticsmentioning

confidence: 89%

“…With this type of asymptotics, the assumption on the bandwidth parameter implies that the estimate of the long‐run variance is not consistent. However, inference is more precise than with heteroskedasticity‐ and autocorrelation‐consistent (HAC) standard asymptotics, and therefore it is often referred to as “heteroskedasticity–autocorrelation robust” (HAR; see, e.g., Lazarus et al, ).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Comparing predictive accuracy in small samples using fixed‐smoothing asymptotics

Coroneo

Iacone

2020

J of Applied Econometrics

View full text Add to dashboard Cite

We consider fixed-smoothing asymptotics for the Diebold and Mariano (Journal of Business and Economic Statistics, 1995, 13(3), 253-263) test of predictive accuracy. We show that this approach delivers predictive accuracy tests that are correctly sized even when only a small number of out-of-sample observations is available. We apply the fixed-smoothing asymptotics to the Diebold and Mariano test to evaluate the predictive accuracy of the Survey of Professional Forecasters (SPF) and of the European Central Bank Survey of Professional Forecasters (ECB SPF) against a simple random walk. Our results show that the predictive abilities of the SPF and of the ECB SPF were partially spurious. INTRODUCTIONGood forecasts are key to good decision making, and being able to compare predictive accuracy is key to discriminating between good and bad forecasts. To this end, one of the most used tests to compare the predictive accuracy of two competing forecasts is the Diebold and Mariano (1995; DM) test.The DM test is based on a loss function associated with the forecast errors of each forecast, testing the null hypothesis of zero expected loss differential between two competing forecasts. This framework allows us to test for equal predictive accuracy using any loss function, and the test statistic is valid for contemporaneously correlated, serially correlated, and nonnormal forecast errors.The DM approach takes forecast errors as model free, and the test is valid also when the forecasts are produced from unknown models, as for example from forecast survey data. When the forecasts are produced by estimated models, nested or nonnested, it is in general necessary to account for the impact of the model parameter estimation uncertainty on the distribution of the forecast accuracy test (see Clark & McCracken, 2001;West, 1996). In this case, the limiting distribution of the test statistics depends on the specific modeling assumptions made for obtaining the forecast errors (see Clark & McCracken, 2013;West, 2006). West (1996) showed that in some cases the DM approach was asymptotically valid even when forecasts were obtained from estimated models. This happens when the number of in-sample observations is large relative to the number of out-of-sample observations or when in-sample and out-of-sample loss is the same, for example using a quadratic loss as an evaluation criterion for models that are estimated by ordinary least squares (OLS). However, in practice, it is not uncommon to compare forecasts produced by models for which accounting for the model parameter estimation uncertainty is not tractable. In addition, if the objective is to compare forecasting methods as opposed to forecasting models, then Giacomini and White (2006) showed that in an environment with asymptotically nonvanishing estimation uncertainty the DM test can still be applied. For these reasons, the DM test is still widely applied also when forecasts are obtained by estimated models (see Diebold, 2015).One additional reason for the success of the DM test is that the t...

show abstract

Section: Fixed‐smoothing Asymptoticsmentioning

confidence: 89%

Section: Introductionmentioning

confidence: 99%

Comparing predictive accuracy in small samples using fixed‐smoothing asymptotics

Coroneo

Iacone

2020

J of Applied Econometrics

View full text Add to dashboard Cite

show abstract

“…Specifically, for each model j = 1, ..., 254 that includes at least one factor, the test statistic takes the form P −1/2 ∑ T t=R (û 2 t+h,AR −û 2 t+h,j )/ω j whereω 2 j is an estimate of the long-run variance of u 2 t+h,AR −û 2 t+h,j . This is estimated using the Bartlett kernel and bandwidth ⌊1.3 √ P ⌋ + 1 as advocated in Lazarus, Lewis, Stock, and Watson (2018). Critical values for the asymptotic distribution are approximated using the formula provided in Table 1 of Kiefer and Vogelsang (2005;p.…”

Section: Predictability Of Factor-based Modelsmentioning

confidence: 99%

FRED-QD: A Quarterly Database for Macroeconomic Research

McCracken,

2020

View full text Add to dashboard Cite

In this paper we present and describe a large quarterly frequency, macroeconomic database. The data provided are closely modeled to that used in Stock and Watson (2012a). As in our previous work on FRED-MD, our goal is simply to provide a publicly available source of macroeconomic "big data" that is updated in real time using the FRED database. We show that factors extracted from this data set exhibit similar behavior to those extracted from the original Stock and Watson data set. The dominant factors are shown to be insensitive to outliers, but outliers do affect the relative influence of the series as indicated by leverage scores. We then investigate the role unit root tests play in the choice of transformation codes with an emphasis on identifying instances in which the unit root-based codes differ from those already used in the literature. Finally, we show that factors extracted from our data set are useful for forecasting a range of macroeconomic series and that the choice of transformation codes can contribute substantially to the accuracy of these forecasts. JEL Classification: C30, C33, C82

show abstract

“…Both panels show the ECI index conditional on a one standard deviation negative surprise in the official unemployment rate (meaning a higher-than-expected unemployment rate) and in the number of nonfarm payrolls. Panel 3a provides the local projection estimates together with 68% and 90% confidence bands based on the equal-weighted cosines long-run variance estimator and optimal bandwidth recommended by Lazarus et al (2018). For comparison, the broken lines in Panel 3a show the smoothed estimates and the associated Lazarus et al (2018) 90% bands.…”

Section: Confidence Changes Following Surprises In Macroeconomic Relementioning

confidence: 99%

Do Monetary Policy Announcements Shift Household Expectations?

Lewis¹,

Makridis²,

Mertens³

2019

Self Cite

View full text Add to dashboard Cite

We use daily survey data from Gallup to assess whether households' beliefs about economic conditions are influenced by surprises in monetary policy announcements. We first provide more general evidence that public confidence in the state of the economy reacts to certain types of macroeconomic news very quickly. Next, we show that surprises to the Federal Funds target rate are among the news that have statistically significant and instantaneous effects on economic confidence. In contrast, surprises about forward guidance and asset purchases do not have similar effects on household beliefs, perhaps because they are less well understood. We document heterogeneity in the responsiveness of sentiment across demographics.

show abstract

HAR Inference: Recommendations for Practice

Cited by 173 publications

References 33 publications

Comparing predictive accuracy in small samples using fixed‐smoothing asymptotics

Comparing predictive accuracy in small samples using fixed‐smoothing asymptotics

FRED-QD: A Quarterly Database for Macroeconomic Research

Do Monetary Policy Announcements Shift Household Expectations?

Contact Info

Product

Resources

About