Approaches to Economic Forecasting
There are at least three approaches to economic forecasting that economists and analysts use. These methods are econometric modelling, economic indicators, and the checklist approach. We discuss these economic forecasting in detail. We also highlight the advantages and disadvantages of each method. Generally speaking, ti is best to rely on more than one method to make sure the forecasts are robust
The first method is econometric analysis. This method uses statistical methods to explain economic relationships and formulate forecasting models. There are two kinds of econometric models.
- Structural econometric models are based on economic theory
- Reduced-form econometric models are compact versions of structural approaches
Econometric models can range from very simple to very complex, involving several hundred relationships. The most common approach to estimate econometric models is ordinary least squares (OLS), although more sophisticated models can also be used.
- can incorporate many variables
- once specified, it can be reused
- output is quantified and based on a consistent set of relationships
- complex and time-consuming to construct
- the data may be difficult to forecast and the relationships can change
- output may require interpretation
- does not work well to forecast turning points
The second method is economic indicators. This kind of data is available from governments, international organizations, and private organisations. There are several kinds of indicators:
- leading indicators that move ahead of the business cycle. They can be combined to form a composite.
- coincident indicators that move with changes in the business cycle
- lagging indicators that move after changes in the business cycle.
Composites can also be interpreted as a diffusion index. This means that we look at the number of indicators pointing towards expansion versus contraction in the economy.
- simple, intuitive, and easy to interpret
- data is often readily available
- lists can be tailored to meet specific forecasting needs
- forecasting results tend to be inconsistent
- tend to give false signals
- are revised frequently which can make them appear to fit data better than they actually do
The third approach is the checklist approach. This approach is more subjective than the other two approaches. In this case, the analyst considers a series of questions and uses judgement and perhaps some statistical modelling to make a forecast.
- less complex than econometrics
- flexible in mixing objective statistical analysis with judgement
- complexity must be limited because the process is manual
We discussed three popular approaches to economic forecasting used by analysts and economists. Each have their advantages and disadvantages. The best approach is to combine them. That way, the forecast will be more robust.