EXPO/ECONOMETRICS™

EXPO/Econometrics is a powerful add-in module to EXPO that enables users to design, evaluate, and apply a series of econometric tests, models and analytics to market data from a wide variety of sources. Traders and Quantitative Analysts now have "point-and-click" access to a full collection of econometric tests, model estimation and forecasting techniques. Through EXPO, econometrics users gain complete integration with all major data distribution platforms and major market data sources including the BLOOMBERG service, Reuters, Bridge, FAME, and LIM.

Research is demonstrating that econometric tests and techniques can provide significant improvements in forecasting and estimation accuracy for financial markets. The problem has been that most econometric software packages were designed for use by economists and academics working with low frequency economic data. Furthermore, existing econometric software packages did not provide easy access to market data sources and were not designed for integration with data distribution platforms like Bloomberg, Triarch, TIB, OTS, and Kapiti FIST. This made it difficult or impossible to use econometrics in real-time at the trading desk.

EXPO/Econometrics solves these problems. Designed with a simple, point-and-click interface, EXPO/Econometrics offers built-in access to all major market data sources and complete integration with all major data distribution platforms.

• Data Transformations: Box-Cox transformations, Differencing, Logit, Seasonal adjustment, and Periodicity conversion.
• Statistical Analysis: Autocorrelation and Partial autocorrelation Analysis, Q-statistics, Restricted Histogram, Correlation, and Variance/Covariance Matrix.
• Econometric Tests: Additional variables, superfluous variables, Dickey-Fuller unit root, Engle-Granger cointegration, Granger Causality, Multicollinearity, Normality, LM Serial Correlation, GARCH and White heteroskedasticity, Chow, and Ramsey.
• Model Estimation and Forecasting: OLS, GARCH, ARIMA, Ridge, Rolling/moving regression, Instrumental variables, and Autoregressive errors.
• Random Number Generation: Using Binomial, Chi-squared, Exponential, F, Student-t, Normal, Lognormal, and Poisson distributions.

The main EXPO package includes several analyses that are also directly relevant to econometric analysis:

• Frequency Analysis: Convolution, Discrete Fourier Transform, Fast Fourier Transform, Inverse Fourier transforms, impulse filters, Power spectral density, trigonometric functions generator, Digital filter functions.
• Polynomial Analysis: Cubic spline interpolation, polynomial estimation and statistics.
• Statistics: Summary statistics, rolling correlation and statistics; Student-t, F, ANOVA, and Chi-square tests.
• Mathematical Functions: An extensive set of advanced functions for matrix math and calculus are provided in EXPO's Analyze menu.

  Econometric applications in finance can be organized into the following categories:

  • Parameter Estimation
  • Forecasting
  • Classification
  • Hypothesis Testing

  Specific applications of EXPO/Econometrics in finance could include:

  • Forecasting daily Forex rates using ARIMA;
  • Forecasting monthly interest rates based on a multivariate regression macro model;
  • Forecasting daily and hourly volatility using GARCH;
  • Assessing asset linkages using cointegration analysis;
  • Exploiting Forex market inefficiencies via error-correction models;
  • Combining trading signals from several indicators using regression;
  • Predicting bond risk of default using logit models;
  • Testing the performance of trading systems using GARCH and Cointegration;
  • Option pricing via simulation using different probability distributions;
  • Estimating the weekly correlation between international stock market returns;
  • Estimating multi-index beta coefficients;
  • Quantifying seasonalities in asset returns.