Narrative should support data - Data should not just support narrative

 

A problem with fundamental discretionary macro is that data is often used to support the narrative when the reasoning should be reversed. Narrative should support the data analysis. From data comes narrative, not the other way. 

There is an issue if a manager has a narrative in his mind and then searches for data to confirm what he already believes he knows. That may work at times, but it is not an effective strategy.

If you believe inflation is transitory, you will look for data to support the transitory story. If you believe the Fed will pivot, you will look for data to support the pivot story. The causality should be reversed.

Following data first is more difficult. Inflation is higher than average over the last decade and inflation is generally mean reverting; however, we may have little evidence on the speed of adjustment. Saying inflation will fall is not an easy conclusion. We may believe that the Fed does not want to push the economy into a recession, but we may have little evidence on their current reaction function. Most times we may have to fall back on the idea that we don't know or cannot make a good prediction. We often rely on a narrative as a substitute to an answer that we don't know. 

The focus should be on systematically reviewing data and forming some likelihood towards an event. This is the essence of systematic global macro. The narrative are the words around the data analysis.