Maximum Likelihood Estimators for ARMA and ARFIMA Models: A Monte Carlo Study

Hauser, Michael A.

doi:10.2139/ssrn.137955

Cited by 10 publications

(7 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the sake of clarity, we only present a subset of the results. 7 The rows show the results from the ARFIMA (0,d,0 There is a considerable amount of uncertainty in the estimates of d. Consistent with Hauser (1999), the estimator produces downwardly biased estimates of d across all models. In particular, performance is poor when t ¼ 100 and t ¼ 250 and is worse when the data-generating process contains short-run dynamics.…”

Section: Fractional Integrationmentioning

confidence: 72%

“…Other estimators often used are the Whittle likelihood(Robinson 1995) and the modified profile likelihood(An and Bloomfield 1993). Evidence suggests the exact MLE is not a panacea(Hauser 1999). Specifically, the modified profile likelihood dominates the exact MLE, which is biased downward, in small samples, especially when long-and short-run dynamics both characterize the data-generating process.LukeKeele et al…”

mentioning

confidence: 99%

See 1 more Smart Citation

Treating Time with All Due Seriousness

Keele

Linn²,

Webb³

2016

Polit. anal.

View full text Add to dashboard Cite

In this article, we highlight three points. First, we counter Grant and Lebo's claim that the error correction model (ECM) cannot be applied to stationary data. We maintain that when data are properly stationary, the ECM is an entirely appropriate model. We clarify that for a model to be properly stationary, it must be balanced. Second, we contend that while fractional integration techniques can be useful, they also have important weaknesses, especially when applied to many time series typical in political science. We also highlight two related but often ignored complications in time series: low power and overfitting. We argue that the statistical tests used in time-series analyses have little power to detect differences in many of the sample sizes typical in political science. Moreover, given the small sample sizes, many analysts overfit their time-series models. Overfitting occurs when a statical model describes random error or noise instead of the underlying relationship. We argue that the results in the Grant and Lebo replications could easily be a function of overfitting.

show abstract

Section: Fractional Integrationmentioning

confidence: 72%

mentioning

confidence: 99%

Treating Time with All Due Seriousness

Keele

Linn²,

Webb³

2016

Polit. anal.

View full text Add to dashboard Cite

show abstract

“…13 The autoregressive fractionally integrated moving average (ARFIMA) model estimate for the gender gap is (0, d, 0) with d = 0.465, shown in Table 1. 14 The series is fractionally integrated since 13 The modified profile likelihood has been shown to have superior small-sample properties over the exact maximum likelihood estimator derived by Hauser 1999 andSowell 1992. In particular, bias in estimates of d are, on average, smaller using the MPL.…”

Section: The Persistence Of Gender Differencesmentioning

confidence: 99%

The Dynamics of the Partisan Gender Gap

2004

View full text Add to dashboard Cite

Gender differences in vote choice, opinion, and party identification have become a common feature of the American political landscape. We examine the nature and causes of gender differences in partisanship using a time series approach. We show that gender differences are pervasive—existing outside of the context of specific elections or issues—and that they are a product of the interaction of societal conditions and politics. We find that from 1979 to 2000, the partisan gender gap has grown when the political climate moved in a conservative direction, the economy deteriorated, and the percentage of economically vulnerable, single women increased. The gender gap is likely to be a continual feature of the American political landscape: one that shapes everything from elite political behavior to election outcomes.

show abstract

“…In response, we note that there is more than one way to estimate these parameters. The method that KL&W choose, the time-domain exact maximum likelihood (EML) of Sowell (1992), may be the estimator favored by Stata, but it suffers from a well-known and persistent negative bias (see, e.g., Li and McLeod 1986;Cheung and Diebold 1994;Hauser 1999).…”

Section: Fractional Integration Methodsmentioning

confidence: 99%

Equation Balance and Dynamic Political Modeling

Lebo

Grant

2016

Polit. anal.

View full text Add to dashboard Cite

The papers in this symposium agree on several points. In this article, we sort through some remaining areas of disagreement and discuss some of the practical issues of time series modeling we think deserve further explanation. In particular, we have five points: (1) clarifying our stance on the general error correction model in light of the comments in this issue; (2) clarifying equation balance and discussing how bounded series affects our thinking about stationarity, balance, and modeling choices; (3) answering lingering questions about our Monte Carlo simulations and exploring potential problems in the inferences drawn from long-run multipliers; (4) reviewing and defending fractional integration methods in light of the questions raised in this symposium and elsewhere; and (5) providing a short practical guide to estimating a multivariate autoregressive fractionally integrated moving average model with or without an error correction term.

show abstract

Maximum Likelihood Estimators for ARMA and ARFIMA Models: A Monte Carlo Study

Cited by 10 publications

References 30 publications

Treating Time with All Due Seriousness

Treating Time with All Due Seriousness

The Dynamics of the Partisan Gender Gap

Equation Balance and Dynamic Political Modeling

Contact Info

Product

Resources

About