AutoRegressive Distributed Lag (ARDL) Estimation. Part 3 - Practice

In Part 1 and Part 2 of this series, we discussed the theory behind ARDL and the Bounds Test for cointegration. Here, we demonstrate just how easily everything can be done in EViews 9 or higher.

While our two previous posts in this series have been heavily theoretically motivated, here we present a step by step procedure on how to implement Part 1 and Part 2 in practice.

AutoRegressive Distributed Lag (ARDL) Estimation. Part 2 - Inference

This is the second part of our AutoRegressive Distributed Lag (ARDL) post. For Part 1, please go here, and for Part 3, please visit here.

In this post we outline the correct theoretical underpinning of the inference behind the Bounds test for cointegration in an ARDL model. Whilst the discussion is by its nature quite technical, it is important that practitioners of the Bounds test have a grasp of the background behind its inferences.

Dynamic Factor Models in EViews

One of the current buzz topics in macro-econometrics is that of dynamic factor models. 

Factor models allow researchers to work with a large number of variables by reducing them down to a handful (often two) components, allowing tractable results to be obtained from unwieldy data. 

A natural extension to factor models is to allow dynamics to enter the relationships.  These dynamic factor models have become extremely popular due to their ability to model business cycles, and perform both forecasting and nowcasting (predicting the current state of the economy).

Although EViews has built-in factor analysis, we do not (yet!) have dynamic factor models included. 

Luckily two researchers from the Ministry of Finance in Sweden have recently posted a paper, and corresponding code, that estimates dynamic factor models in EViews with a simple programming subroutine utilising EViews' state-space estimation object.

This paper looks fantastic - good job guys!

AutoRegressive Distributed Lag (ARDL) Estimation. Part 1 - Theory

One of our favorite bloggers, Dave Giles often writes about current trends in econometric theory and practice. One of his most popular topics is ARDL modeling, and he has a number of fantastic posts about it.

Since we have recently updated ARDL estimation in EViews 9.5, and are in the midst of adding some enhanced features to ARDL for the next version of EViews, EViews 10, we thought we would jot down our own thoughts on the theory and practice of ARDL models, particularly in regard to their use as a cointegration test.

This blog post will be in three parts. The first will discuss the theory behind ARDL models, the second will present the theory behind correct inference of the Bounds test, while the third will bring everything together with an example in EViews.

EViews Add-In: Importing Ken French’s Data Library

Background

The frenchdata add-in is designed to make it easier and faster to download data from Ken French's data library. The data in the library are in zipped *.txt or *.csv files, many with multiple data sets and mixed date formats that can be tedious to import. This add-in, in contrast, is straightforward and requires minimal input. After downloading and processing each file is put in a separate workfile, multiple datasets in a single file are separated and each one is put in a separate page, data columns are put into series, and date formats are read from the files and applied to the page(s) of the workfile.

L1 Trend Filtering

Author and guest post by Eren Ocakverdi.

Extracting the trend of a time series is an important analytical task as it simply depicts the underlying movement of the variable of interest. Had this so-called long term component known in advance, we would have been able to foresee its future course. In practice, however, there are several other factors (e.g. cycle, noise) in play that have influence on the dynamics of a time dependent variable.

Time path of a variable can either be deterministic (assuming the change in trend is constant) or stochastic (assuming the change in trend varies randomly around a constant). Estimation of a deterministic trend is straightforward, yet it often oversimplifies the data generating process. The assumption of stochastic trend seems to be a better fit to observed behavior of various time series as they tend to evolve with abrupt changes. Nevertheless, its estimation is difficult and can have serious implications due to accumulation of past errors.

Add-in Round Up for 2016 Q2/Q3

Add-in Round Up for 2016 Q2/3

In this section of the blog, we provide a summary of the Add-ins that have been released or updated within the previous few months, and we announce the winner of our “Add-in of the Quarter” prize!

As a reminder, EViews Add-ins are additions to the EViews interface or command language written by our users or the EViews Development Team and released to the public. You can install Add-ins to your EViews by using the Add-ins menu from within EViews, or by visiting our Add-ins webpage.

We have 9 new Add-ins within the last few months, including a number related to VAR analysis: 

1. ThSVAR
2. FanChart
3. Croston
4. LocalIRFs
5. Speccaus
6. SIRF
7. ConfCast
8. URAll
9. DMA

ThSVAR

The ThSVAR Add-in continues Davaajargal Luvsannyam's line of VAR based Add-ins, and estimates Threshold Structural Vector Auto Regression (FAVAR) models, such as those described by Balke (2000).

Unlike traditional structural VAR approaches, the ThSVAR allows a threshold variable that determines which of two regimes the structural contemporaneous relationship is in.

FanChart

The FanChart add-in creates Bank of England style fan charts from forecast distribution data.  More details can be found on our Fan Chart blog post.

Croston

The Croston method is a way of using exponential smoothing techniques to forecast intermittent series (series with long periods of zeros, intermingled with sparse positive integers).  This add-in performs the Croston method in a simple fashion.

LocalIRFs

The LocalIRFs Add-in, written by Eren Ocakverdi (Trubador on the EViews forums), performs impulse response analysis by local projection method of Jordà (2005, 2009) on a previously estimated VAR model. 

As well as providing the impulse response graphs and tables, Eren allows equality hypothesis tests on the responses.

Speccaus

Nicolas Ronderos' speccaus Add-in computes a frequency domain Granger causality test in the context of VAR models, as given in Breitung and Candelon (2006). 

SIRF

Another Davaajargal Luvsannyam Add-in related to VARs, SIRF computes scaled impulse responses of Structural Vector Auto Regressions. 

Although a rather simple Add-in, it provides powerful functionality to users who wish to create their own impulses for structural VARs.

ConfCast

One more from Davaajargal Luvsannyam (who has been busy!) to add to the extensive list of VAR based add-ins.  ConfCast performs conditional forecasting from a VAR model, allowing you to constrain the future values of the VAR's underlying series.

URALL

URALL, by Imadeddin Almosabbeh, solves a time-old issue of wanting to perform individual unit root tests on a large number of series at once.  The add-in allows you to specify the type of unit root test to run, then collates the output from each one into an easy to read table. Nifty!

DMA

The final Davaajargal Luvsannyam add-in, and one unrelated to VARs!, performs dynamic model averaging.  

Model averaging is an exploding field in econometrics, with a common consensus held that averaging over different models is a better approach than choosing the single best model. 

Although EViews (9 and above) has various model averaging techniques built, dynamic model averaging is not yet available built in.  This add-in addresses that short-coming.

Quarterly Prize

The EViews Development Team has decided that the DMA Add-in contributed most significantly to the usage of EViews this quarter. 

For more information on writing Add-ins, you can read the Add-in chapter of the online help or visit the Add-in writer’s forum. 

If you would like to submit an Add-in, need more information on the Quarterly Prize, or have quesitons about writing Add-in for EViews, please email support@eviews.com.