The disconnect between hard and soft data

There is a strong disconnect between hard and soft data in the US. Soft data is telling us that we are surely heading for a recession. The hard data is still looking strong albeit well off the highs from the post-pandemic surge. Usually soft and hard data are generally consistent with the soft data usually leading and more volatile. This does not seem to be the case currently. The divergence tells that there is a high degree of uncertainty regardless of market forecasts.

The St Louis Fed first wrote about hard and soft data and provided measures for each index. See "Hard Data, Soft Data and Forecasting". The measure for hard and soft can change. I like survey data for soft and actual collected data for hard. A diffusion index can serve as an alternative to the principal component approach. Soft data usually has a shorter lag and will be a better representation for the current environment.