forecast.midas_r: Forecast MIDAS regression

In midasr: Mixed Data Sampling Regression

Description

Forecasts MIDAS regression given the future values of regressors. For dynamic models (with lagged response variable) there is an option to calculate dynamic forecast, when forecasted values of response variable are substituted into the lags of response variable.

Usage

## S3 method for class 'midas_r'
forecast(
  object,
  newdata = NULL,
  se = FALSE,
  level = c(80, 95),
  fan = FALSE,
  npaths = 999,
  method = c("static", "dynamic"),
  insample = get_estimation_sample(object),
  show_progress = TRUE,
  add_ts_info = FALSE,
  ...
)

Arguments

object	midas_r object
newdata	a named list containing future values of mixed frequency regressors. The default is NULL, meaning that only in-sample data is used.
se	logical, if TRUE, the prediction intervals are calculated
level	confidence level for prediction intervals
fan	if TRUE, level is set to seq(50,99,by=1). This is suitable for fan plots
npaths	the number of samples for simulating prediction intervals
method	the forecasting method, either "static" or "dynamic"
insample	a list containing the historic mixed frequency data
show_progress	logical, if TRUE, the progress bar is shown if se = TRUE
add_ts_info	logical, if TRUE, the forecast is cast as ts object. Some attempts are made to guess the correct start, by assuming that the response variable is a ts object of frequency 1. If FALSE, then the result is simply a numeric vector.
...	additional arguments to simulate.midas_r

Details

Given future values of regressors this function combines the historical values used in the fitting the MIDAS regression model and calculates the forecasts.

Value

an object of class "forecast", a list containing following elements:

method	the name of forecasting method: MIDAS regression, static or dynamic
model	original object of class midas_r
mean	point forecasts
lower	lower limits for prediction intervals
upper	upper limits for prediction intervals
fitted	fitted values, one-step forecasts
residuals	residuals from the fitted model
x	the original response variable

The methods print, summary and plot from package forecast can be used on the object.

Author(s)

Vaidotas Zemlys

Examples

data("USrealgdp")
data("USunempr")

y <- diff(log(USrealgdp))
x <- window(diff(USunempr), start = 1949)
trend <- 1:length(y)

##24 high frequency lags of x included
mr <- midas_r(y ~ trend + fmls(x, 23, 12, nealmon), start = list(x = rep(0, 3)))

##Forecast horizon
h <- 3
##Declining unemployment
xn <- rep(-0.1, 12*h)
##New trend values
trendn <- length(y) + 1:h

##Static forecasts combining historic and new high frequency data
forecast(mr, list(trend = trendn, x = xn), method = "static")

##Dynamic AR* model
mr.dyn <- midas_r(y ~ trend + mls(y, 1:2, 1, "*")
                   + fmls(x, 11, 12, nealmon),
                  start = list(x = rep(0, 3)))

forecast(mr.dyn, list(trend = trendn, x = xn), method = "dynamic")

##Use print, summary and plot methods from package forecast

fmr <- forecast(mr, list(trend = trendn, x = xn), method = "static")
fmr
summary(fmr)
plot(fmr)

Example output

Loading required package: sandwich
Loading required package: optimx
     Point Forecast
2012     0.04798328
2013     0.04638829
2014     0.04609471
     Point Forecast
2012     0.04598409
2013     0.04446394
2014     0.04355180
     Point Forecast
2012     0.04798328
2013     0.04638829
2014     0.04609471

Forecast method: MIDAS regression forecast (static)

Model Information:
MIDAS regression model
 model: y ~ trend + fmls(x, 23, 12, nealmon) 
(Intercept)       trend          x1          x2          x3 
  0.0421355  -0.0002936  -0.2333520   0.4611676  -0.0250372 

Function optim was used for fitting

Error measures:
                       ME        RMSE         MAE      MPE     MAPE      MASE
Training set 2.266363e-07 0.008224211 0.006657488 5.453882 43.69982 0.2984747
                   ACF1
Training set 0.06705183

Forecasts:
     Point Forecast
2012     0.04798328
2013     0.04638829
2014     0.04609471

midasr documentation built on Feb. 23, 2021, 5:11 p.m.