Effects Worksheet

As a set of single-asset forecasts and contrast forecasts develops, each new forecast or contrast tends to have ripple effects across the entire set of assets, sometimes unpredictably so. The Components and Effects worksheets are diagnostic tools designed to assess the impact of each individual single-asset forecast or contrast, so that the impacts of changes can be better understood and unpredictable movements in the result estimates can be managed by adjusting their principal causes.

Both worksheets have important information that can be useful in diagnosing problems with a set of Bayes priors. The Components worksheet mainly deals with the impact of changes to the mean of a forecast, and the Effects worksheet shows the impact across all of the assets of including each forecast versus not including it, as a generalization of the “forecast effect” row, which shows only the effect of the forecast on its own value.

Before Running Bayes on a new case, or in a case saved in a previous version of the Estimator, the tables may be populated with N/A. Numerical values will appear after Running Bayes, since the calculation of these tables depend on the interaction of the Bayes calculations.

The Effects worksheet shows the impact of each forecast on the case, i. e. it is the result mean of the case minus the result mean of a case without that forecast.

Example:

In the illustrated example, a single-asset Forecast has been added for Short Term Bonds with a Forecast Mean of 4.5% and a Standard Deviation of 2%. Adding this contrast has the effect of raising the Result Mean for Short Term Bonds by 1.69%, but the other Treasurys (Intermediate and Long Term Bonds) are raised by 3.21% and 4.70%. The US Equities are negatively affected, due to the correlations between the assets, in all cases by over 1%. 

Whereas the Components Worksheet shows the impact of adjusting the Forecast Mean for any forecast, the Effects Worksheet shows the impact of including or excluding the entire forecast.

This example serves as an excellent illustration that sometime Forecasts may have unintended ripple effects on other assets' result means.