Beware the comfort in the numbers—a reminder

Every year I need to remind myself that the expanding data sets upon which we rely are wonderful in the aura of transparency and comfort that we can study the numbers and analysis and relax about the details.  Back tests?  Challenges to granularity?  Scrutiny of the methodology?  That is a lot of work.

With the recent increase in volatility, I need a reminder. Each year I take a peek, in a granular way, and note that at the roots and drivers of the data we find:

                             Sentiment

                             Confidence

                             Animal Spirits

                             Reputation

                             Culture

All the usual qualitative inputs to the models, trends analysis, stress tests, and polls.  Now (as they say more than ever) these qualitative elements are imbedded and rarely documented at the root of the inputs to the analysis. Cultural bias to polls.  Reputation to appraisal. Animal spirits (pick your animal) for the markets. Confidence and reputation everywhere.

In isolation, these qualitative components are small if not trivial; but compounded, even additive; they have the potential for significance and materiality (even if they are masked through netting). Thus this is a note to myself reminding me to dig a bit into the data before relaxing in the results.