Powerful Integrated Solutions for Econometric Analysis

An object-oriented matrix programming language. It is an important tool for statistical and econometric programming with a syntax similar to C++ and a comprehensive range of commands for matrix and statistical operations. Ox is at the core of OxMetrics. Most of the other modules of OxMetrics (such as PcGive, STAMP, G@RCH) are implemented with the Ox language. Ox Professional belongs to theOxMetrics Enterprise Edition.

An essential tool for modern econometric modelling. PcGive Professional is also part of OxMetrics Enterprise Edition. It provides the latest econometric techniques, from single equation methods to advanced cointegration, volatility models static and dynamic panel data models, discrete choice models and time-series models.

PcGive Professional includes Autometrics

Autometrics is the automatic econometric model selection procedure that is available in PcGive. Autometrics is a revolutionary new approach to model building, based on recent advances in the understanding of model selection procedures. Experiments show that Autometrics outperforms even the most experienced econometrician. Starting from an initial model, Autometrics will find the best simplified model. Thus removing the drudgery of model selection, allowing you to concentrate on the variable choice and interpretation of the model(s).

Modelling and forecasting time series, based on structural time series models. These models use advanced techniques, such as Kalman filtering, but are set to be easy to use. The hard work is done by the program, leaving the user free to concentrate on formulating models, then using them to make forecasts. STAMP 9 includes both univariate and multivariate models and automatic outlier detection. STAMP is also part of OxMetrics Enterprise Edition.

G@RCH is an OxMetrics module dedicated to the estimation and forecast of univariate and multivariate ARCH-type models. It also allows the estimation of univariate and multivariate non-parametric estimators of the quadratic variation and the integrated volatility. G@RCH provides a menu-driven easy-to-use interface, as well as some graphical features. For repeating tasks, the models can be estimated via the Batch Editor of OxMetrics or using the Ox language together with the ‘Garch’, ‘MGarch’ and ‘Realized’ classes. Version 9.0 is a major update that features many improvements. G@RCH is also part of OxMetrics Enterprise Edition.

SsfPack is a suite of C routines for carrying out computations involving the statistical analysis of univariate and multivariate models in state space form and requires Ox 4 or above to run. SsfPack is not a member of OxMetrics Enterprise Edition.

CATS uses OxMetrics for data input and graphical and text output, and is part of the OxMetrics family. The third generation of CATS is a complete rewrite in more than one way. It is now written in Ox for use within OxMetrics, either using the graphical user interface or programmatically. Furthermore, many algorithms have been improved or newly invented, in particular for I(2) models. The new CATs module with I(2) cointegration and many new I(1) cointegration features includes corrections and is considerably faster.

CATS uses OxMetrics for data input and graphical and text output, and is part of the OxMetrics family.

The third generation of CATS is a complete rewrite in more than one way. It is now written in Ox for use within OxMetrics, either using the graphical user interface or programmatically. Furthermore, many algorithms have been improved or newly invented, in particular for I(2) models. The new CATs module with I(2) cointegration and many new I(1) cointegration features includes corrections and is considerably faster.

Here is a brief summary of new features in the I(1) part of CATS

• Much more efficient computations (can be several orders of magnitude faster) in Bartlett correction and recursive estimation;
• Bartlett correction always included when valid;
• Improved beta-switching algorithm;
• New alpha-beta-switching algorithm allowing linear restrictions on alpha and not requiring identification;
• Bootstrap of rank test;
• Bootstrap of restrictions;
• More Monte Carlo facilities: draw from estimated model, either with estimated or with specified coefficients;
• General-to-specific CATSmining;
• Automatic generation of Ox code;
• New convenient way to express restrictions;
• Most algorithms QR based.

And for the I(2) part of CATS:

• Improved tau-switching algorithm;
• New delta-switching algorithm;
• New triangular-switching algorithm allowing linear restrictions on alpha, beta, tau and not requiring identification;
• Estimation with delta=0;
• Bootstrap of rank test;
• Simulation of asymptotic distribution of rank test;
• Bootstrap of restrictions;
• More Monte Carlo facilities: draw from estimated model, either with estimated or with specified coefficients;
• Automatic tests of unit vectors and variables;
• CATSmining;
• Improved computation of standard errors;
• Automatic generation of Ox code;
• All algorithms QR based.

The Special Issue reprint book "Recent Developments in Cointegration" has been published online and is freely accessible on the MDPI Books platform here

New Packages in Ox:

• oxurl - download resources from the internet through simple interface to libcurl.

• oxdoc - generate documentation from source code markup (written in Ox). Documentation for oxdoc is created by running it on itself.

• gwg2ox - generates Ox code for a gwg file (used when calling Save Ox code from the graph window contact menu).

• Python - call Python from Ox code.

• Rox - call R from Ox code.

Ox 9.0

• parallel if (test-expression) for, parallel if (test-expression) foreach allows the code to decide to run in parallel or not

• Multiplicative concatenation of arrays and strings

• Indexing array by string: searches for an odd index wich has that string value, then accesses the next element. This emulates dictionary style lookup. If an item is not present, .Null is returned (instead of a index out of range error).

• G@RCH 9.0 now includes univariate Multiplicative Error Models (MEM). MEM models are primarily aimed at modeling non-negative time series. The most popular MEM type model is the Autoregressive Conditional Duration (ACD) model of Engle and Russell (1998). G@RCH 9.0 provides no fewer than seven different specifications.

• Mulcom 3.0 is now fully integrated into OxMetrics 9.0. MulCom is designed for forecasting evaluation. It provides an analysis of whether some among a set of competing models are significantly better than the others, in terms of predictive accuracy. MulCom can also be applied in any setting where one what to compare the means of two or more populations.

• G@RCH 9.0 can be used with DataFetch to download data from the internet and then estimate a model on the downloaded data.

The improvements in PcGive mainly relate to (1) model formulation, (2) saturation estimation using Autometrics, (3) automatic model selection in simultaneous equations models (SEM).

Fixed and Improved in PcGive 15

 

• Easier desktop layout with additional direct access to recursive graphs, forecasts and tests.
• Trend indicator saturation (TIS) (we have undertaken research using TIS on GPs speed of takeup of generics, which yielded useful results, but still in progress, as is the technical paper)—note this as PcGive being at the frontier implementing powerful new approaches.
• Recursive graphs can be implemented after conventional estimation rather than needing a separate implementation, and can be undertaken with or without indicator saturation (IS).
• Additional diagnostic plots are available
• Easy choice of form of estimated parameter standard errors (SEs) including HCSE and HACSE.
• Multivariate robust Hedgehog plots had variables scrambled.
• Hedgehog plots use a different color in the forecast period.
• Hedgehog Levels forecasts beyond estimation sample: not integrated.
• Levels forecasts with gap: created cGap extra forecasts Also used wrong levels if cGap > 0.
• Recursive hedgehog would omit early part when there are dummies.
• Fixed Ox issue with find, affecting forecasting.
• Autometrics options are presented differently, offering more flexibility in the choice of saturation.
• The AR test after FIML has been changed, using the 3SLS likelihood instead of estimating the extended model by FIML.
• When the lag length exceeds 12, the presearch uses lag blocking and a more e cient search for contrasting terminals.

Unique to PcGive 15

 

• Hedgehog graphs where a sequence of forecasts for say 1 through 8 steps ahead starting at successive horizons T, T+1, etc. are plotted against outcomes looking like the spins of a hedgehog.
• Easily computed forecasts by a robust device as an additional choice to avoid that problem: an illustration is attached showing how much better the robust device is after a shift.

• STAMP 8.3 works under OxMetrics 6.1.
• The Ox code generator is introduced and fully supported by STAMP. This new facility can generate Ox code for the model that is estimated in STAMP. It complements the Batch code generator in STAMP. It is particularly useful for those who use Ox for time series analysis in a production environment.
• The online help facility of STAMP is updated. In particular, the online help for the Batch language and the new Ox code generator are rewritten.
• AIC and BIC added to default output.
• The confidence bounds of seasonal, cycle, and AR components can be centered around zero or following the components.

• All weights and related computations in the Test/Weights dialog can be carried out, also for time series with missing data.
• The Write forecasts option is combined with a Store forecasts option in the Test/Forecasting dialog. The observations forecasts are stored after confirmation as a new variable with the forecasts attached at the end of the sample. When necessary, the database sample is automatically extended such that the forecast window is included. The in-sample values of the new variable are the same as in the original series.
• The Edit/Save forecasts option in the Test/Forecasting dialog is reactivated for model without explanatory variables.
• The Batch code options for Forecasting is extended; see Batch documentation.
• Variables and components in the Batch code need to be written between accolades. Specifically, in the setcmp batch command we have "level", "slope", "seasonal", "cycle", "ar" and "irregular".
• Inclusion of lagged dependent variables is discouraged. A new facility will be built in for the next version. In this version it is best to treat and to have it as an exogenous variable.