Time series

» Home » Products » Features » Time series

Handle all the statistical challenges inherent to time-series data—autocorrelations, common factors, autoregressive conditional heteroskedasticity, unit roots, cointegration, and much more. From graphing and filtering to fitting complex multivariate models, let Stata reveal the structure in your time-series data.

ARIMA

ARMA
ARMAX
Standard and robust variance estimates
Static and dynamic forecasts
Linear constraints
Multiplicative seasonal ARIMA
Spectral densities
Impulse–response functions (IRFs)
Parametric autocorrelation estimates and graphs
Check stability conditions

Watch Introduction to ARMA/ARIMA models.

ARCH/GARCH

GARCH
APARCH
EGARCH
NARCH
AARCH
GJR and more
ARCH in mean
Standard and robust variance estimates
Normal, Student's t, or generalized error distribution
Multiplicative deterministic heteroskedasticity
Static and dynamic forecasts
Linear constraints

Multivariate GARCH

Diagonal VECH models
Conditional correlation models

Constant conditional correlation
Dynamic conditional correlation
Varying conditional correlation

Multivariate normal or multivariate Student's t errors
Standard and robust variance estimates
Static and dynamic forecasts
Linear constraints

Markov-switching models

Dynamic regression
Autoregression
Tables of transition probabilities
Tables of expected durations
Standard and robust variance estimates

Watch Markov-switching models.

ARFIMA

Long-memory processes
Fractional integration
Standard and robust variance estimates
Static and dynamic forecasts
Linear constraints
Spectral densities
Impulse–response functions (IRFs)
Parametric autocorrelation estimates and graphs

Regression with AR(1) disturbances

Heteroskedasticity-and-autocorrelation-consistent covariance matrices
Cochrane–Orcutt/Prais–Winsten methods
ARMA/ARIMA estimators
ARCH estimators

Unobserved components model (UCM)

Trend-cycle decomposition
Stochastic cycles
Estimation by state-space methods
Standard and robust variance estimates
Static and dynamic forecasts
Linear constraints
Spectral densities

FRED data New

Over 470,000 U.S. and international economic and financial time series
Search or browse by subject, title, or source
Download directly into Stata
Put series on a common periodicity
Easily update datasets containing dozens, or even hundreds, of series
Easy-to-use interface for searching and browsing
Commands for updating datasets and replicability

Watch Import FRED (Import Federal Reserve Economic Data).

Business calendars

Define your own calendars
Create calendar from dataset
Format variables using business calendar format
Convert between business dates and regular dates
Lags and leads calculated according to calendar

Graphs and tables

Autocorrelations and partial correlations
Cross-correlations
Cumulative sample spectral density
Periodograms
Line plots
Range plot with lines
Patterns of missing data

Watch Line graphs and tin().
Watch Correlograms and partial correlograms.

Time-series functions

String conversion to date: daily, weekly, monthly, quarterly, half-yearly, yearly
Dates and times from numeric arguments
Date and time literal support
Periodicity conversion, e.g., daily date to quarterly
Date and time ranges

Time-series operators

L, lag
F, leads
D, differences
S#, seasonal lag

Watch Time-series operators.

Time-series time and date formats

Default formats for clock-time daily, weekly, monthly, quarterly, half-yearly, yearly
High-frequency data with millisecond resolution
User-specified formats

Watch Formatting dates, tsset, tsreport, and tsfill.

Support for Haver Analytics database

New import haver command makes using Haver datasets even easier
Quickly access worldwide economics and financial datasets

Read about the new DSGE features.

Forecast models

Combine results from multiple estimation commands
Specify identities and declare exogenous variables
Obtain dynamic and static forecasts
Use simulation methods to obtain prediction intervals
Specify alternative scenarios and perform "what-if" analyses

Watch A tour of forecasting.

VAR/SVAR/VECM

Vector autoregression (VAR)
Structural vector autoregression (SVAR)
Vector error-correction models (VECM)
Impulse–response functions (IRFs)

Simple IRFs
Orthogonalized IRFs
Structural IRFs
Cumulative IRFs

Dynamic multipliers
Forecast-error variance decompositions (FEVD)
Static and dynamic forecasts
Diagnostics and tests
- Cointegration tests
- Granger causality tests
- LM tests for residual autocorrelation
- Tests for normality of residuals
- Lag-order selection statistics
- Stability analysis using eigenvalues
- Wald lag-exclusion statistics
Graphical and tabular presentations and comparisons of IRFs and FEVDs

State-space models

VARMA models
Structural time-series models
Stochastic general-equilibrium models
Stationary and nonstationary models
Standard and robust variance estimates
Static and dynamic forecasts
Linear constraints

Dynamic-factor models

Unobserved factors with vector autoregressive structure
Exogenous covariates
Autocorrelated disturbances in dependent variables’ equations
Standard and robust variance estimates
Static and dynamic forecasts
Linear constraints

Threshold regression New

One threshold or multiple thresholds
Specify number of thresholds
Automatically choose the number of thresholds, using
- BIC
- AIC
- Hannan-Quinn information criterion
Thresholds may be:
- Points in time
- Values of covariates in the regression model
- Values of variables not in the regression model