Framework to Analyze An Economy or The Macroeconomic Environment

I thought of sharing the analytical framework we usually use to analyze the state of an economy and its short and medium-term economic outlook or the overall macroeconomic environment. Hope it proves to be a useful framework for those who are interested in tracking/monitoring/analyzing the economic performance and outlook of an economy - developed or emerging - or the overall macroeconomic environment. 

This framework includes tracking the external economic environment and financial markets too, given their rising influence on the economic performance and outlook of economies across the globe:

A. Analytical Framework (includes statistical and econometric tools we use):

I. Domestic Economy Analysis (Developed or Emerging):

• Political Environment and Developments;
• Macroeconomic Policy (Monetary and Fiscal) Environment & Trends;
• Economic Performance, Growth and Macroeconomic Stability- Key Economic Indicators: GDP & its key constituents, inflation (consumer prices and producer prices), unemployment rate, labour market participation rate, wage and income growth, money supply indicators and credit growth, fiscal deficit, trade deficit, current account deficit, foreign exchange reserves, nominal and real exchange rates, FDI & investment climate, investment to GDP ratio, savings to GDP ratio, industrial production, capacity utilization, manufacturing and services PMI, retail & automobile sales, consumer confidence, business confidence, short term interest rates, bond yields, savings rate and other specific economic indicators pertaining to some economies (for example, electricity generation, foodgrains production, housing market indicators (existing and new home sales, housing starts etc.), Tankan Survey, oil inventories, durable goods orders etc.);
• Household, Government and Corporate Debt (as % of GDP); 
• Key Economic Reforms; 
• Financial Market Trends, Capital Flows and Developments; 
• Business Environment, Investment Climate and Corporate Profitability;
• Banking Sector and Financial Stability;
• Real Estate Sector;
• Infrastructure;
• Susceptibility of An Economy to Exogenous or External Shocks (for example, due to substantial or unsustainable levels of corporate borrowing in foreign currency); 
• Political, Economic (including Exchange Rate), Institutional and Financial Risk;
• Quarterly and Annual Economic Forecasts;

II. External Economic Environment Analysis:

• Global Economic Environment & Developments;
• International Monetary Policy (US, Eurozone, UK, Japan, Canada and China) Developments;
• Geo-Political Risks;
• Oil & Commodity Prices;
• Global Financial Markets; and,
• Key Currency Movements (USD, GBP, Euro etc). 

B. Analysis of Economic Indicators - Statistical and Time Series Econometric Tools:

Types of Exploratory Data Analysis - graphic analysis, descriptive statistics, histograms, pie charts, bar-charts, coefficient of variation, X-Y plots, correlation & lead-lag analysis, box-and-whisker plots, trend, seasonality, stationarity and normality analysis, outlier detection, structural break analysis, missing data and other data related analysis, and appropriate transformations to the data - logarithmic, differencing, growth rates etc. 

Statistical and Econometric Models (Time Series):

Given below are a range of models that can be used for estimation of elasticity, ‘‘what-if’’ scenarios, impact or multiplier analysis, estimation of parameters, historical simulations or economic forecasting - after carrying out various types of exploratory data analysis (it might be noted that choosing the right kind of model from the following range of models requires a significant amount of judgement too - which really comes from experience with real world economic data).

Range of Models:

• Naive Models;
• Simple Averaging Methods and Extrapolation Models; 
• Moving Averages Models;
• Exponential Smoothing Models;
• Classical Decomposition Model;
• Combining Forecasts Model;
• Single Equation Regression Models – Linear, Quadratic and Inverse Equation Models; 
• Log-Log Models;
• Log-Linear Models;
• Linear - Log Models; 
• ARIMA Models;
• Regression Models with Dummy Variables;
• Distributed Lag Models;
• Autoregressive Distributed Lag (ARDL) Models;
• Autoregressive Distributed Lag Models (ARDL) with Dummy Variables;
• Vector Autoregression (VAR) and Impulse Response Function;
• Cointegration and Error Correction Model (ECM); 
• Cointegration and Vector Error Correction Model (VECM);
• Fully Recursive Model;
• Indirect Least Square Model;
• Two Stage Least Square Model (2SLS);
• Two Stage Least Square Model plus Autoregressive (2SLS/AR);
• Limited Information Maximum Likelihood (LIML) Model;
• Three Stage Least Squares Model (3SLS);
• Full Information Maximum Likelihood Method (FIML);
• Full Information Maximum Likelihood Method (FIML) plus AR;
• Seemingly Unrelated Regression Estimation (SURE) Model; and
• Seemingly Unrelated Regression Estimator (SURE) plus AR Model.

* The aforesaid models have to pass various diagnostic tests, before they can to used for various purposes (stated above).

This post has been written by Sher Mehta, Director of Macroeconomic Research, Virtuoso Economics.

END