Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Statistics›Block Bootstrap (Moving Block and Stationary)
Regression model

Block Bootstrap (Moving Block and Stationary)

Block Bootstrap (Moving Block and Stationary Bootstrap) · Also known as: moving block bootstrap, stationary bootstrap, blok bootstrap (moving block / stationary)

Block bootstrap is a resampling method for dependent, autocorrelated time-series data: instead of resampling single observations, it resamples whole blocks of consecutive observations so the serial-correlation structure is preserved. The moving block variant was introduced by Künsch (1989) and the stationary variant by Politis and Romano (1994).

ScholarGate
  1. Regression model
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Block Bootstrap
Bootstrap InferenceJackknifeOLS RegressionPermutation TestQuantile RegressionBayesian BootstrapDouble BootstrapWild Bootstrap

When to use it

Use block bootstrap when you need standard errors or confidence intervals for a statistic computed on autocorrelated time-series data and you do not want to assume a parametric error model. It expects a genuine time-series structure and a reasonable length — at least about 50 observations — because the block length must be chosen to match the autocorrelation. With short series (n below 30) too few blocks are available and the dependence cannot be reproduced; below 20 observations the block-length choice becomes unreliable and a plain bootstrap is preferable.

Strengths & limitations

Strengths
  • Preserves the serial-correlation structure of the data, so inference stays valid under dependence where the ordinary bootstrap fails.
  • Distribution-free: it does not require normality or a parametric model for the errors.
  • Offers both fixed-length (moving block) and stationarity-preserving (stationary) variants to suit the series at hand.
Limitations
  • Results are sensitive to the block length, which must be tuned to the autocorrelation structure and has no universally optimal choice.
  • Needs a reasonably long series (around 50+ observations); short series leave too few blocks to capture the dependence.
  • Assumes (approximate) stationarity of the series; strong trends or structural breaks undermine the resampling.

Frequently asked

How is block bootstrap different from the ordinary bootstrap?

The ordinary bootstrap resamples individual observations independently, which destroys any time dependence. Block bootstrap resamples consecutive blocks of observations, keeping the within-block correlation intact so inference remains valid for autocorrelated data.

What is the difference between the moving block and stationary versions?

The moving block bootstrap uses fixed-length overlapping blocks. The stationary bootstrap of Politis and Romano draws blocks of random, geometrically distributed length, which makes the resampled series itself stationary and reduces sensitivity to a single block-length choice.

How do I choose the block length?

The block length should match the autocorrelation structure: long enough to capture the meaningful dependence but short enough to leave many distinct blocks. There is no single optimal value, so it is the key tuning decision and should be guided by the autocorrelation of the series.

What if my time series is short?

With fewer than about 30 observations there are too few blocks to preserve the dependence, and below 20 the block-length choice is unreliable. In those cases a standard bootstrap or a permutation test is the safer alternative.

Sources

  1. Künsch, H. R. (1989). The Jackknife and the Bootstrap for General Stationary Observations. Annals of Statistics, 17(3), 1217-1241. DOI: 10.1214/aos/1176347265 ↗
  2. Politis, D. N., & Romano, J. P. (1994). The Stationary Bootstrap. Journal of the American Statistical Association, 89(428), 1303-1313. DOI: 10.1080/01621459.1994.10476870 ↗

How to cite this page

ScholarGate. (2026, June 1). Block Bootstrap (Moving Block and Stationary Bootstrap). ScholarGate. https://scholargate.app/en/statistics/block-bootstrap

Related methods

Bootstrap InferenceJackknifeOLS RegressionPermutation TestQuantile Regression

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Bootstrap InferenceStatistics↔ compare
  • JackknifeStatistics↔ compare
  • OLS RegressionEconometrics↔ compare
  • Permutation TestStatistics↔ compare
  • Quantile RegressionEconometrics↔ compare
Compare side by side →

Referenced by

Bayesian BootstrapDouble BootstrapWild Bootstrap

Similar methods

Spatial Bootstrap SimulationBootstrap SimulationParametric BootstrapBootstrap InferenceDouble BootstrapMultilevel Bootstrap SimulationHierarchical Bootstrap SimulationWild Bootstrap

Related reference concepts

Bootstrap MethodsBootstrap and ResamplingResampling MethodsJackknife ResamplingPermutation TestsCross-Validation and Resampling

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Block Bootstrap (Block Bootstrap (Moving Block and Stationary Bootstrap)). Retrieved 2026-07-21 from https://scholargate.app/en/statistics/block-bootstrap · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Künsch (moving block, 1989); Politis & Romano (stationary, 1994)
Year
1989
Type
Resampling inference for dependent data
Estimator
Block resampling (moving or stationary blocks)
Outcome
continuous time series
MinSample
50
Related methods
Bootstrap InferenceJackknifeOLS RegressionPermutation TestQuantile Regression
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account