Lack of precision in market views - the need for quant models

It isn't even wrong.
-Wolfgang Pauli referring to a worthless research paper

Most market commentary and opinions are problematic because they just do not say enough with precision. Hence, the Pauli opinion applies. Most commentary isn't wrong but rather the non-opinion opinion.  There is not enough content to be wrong. The commentators often want it that way. They prefer not going on limb with a clear view. The old adage that most forecasters may either get the direction or timing right but not both usually applies.

Most commentary follows a set framework. There is nothing wrong with this framework and it follows a normal story-line. The problem is not with the set-up of the story but with the conclusion. It is the resolution at the end that makes the commentary useful.

A commentary first engages in telling facts; a recap of what is known about the markets, or a selected set of information. There is the view that more facts will add creditability to anything that is being said. The expert provides details. This is useful for providing information to investors but it is not a recommendation. This is a review of the past.

Next comes the analysis of whether new facts or information is good or bad for the markets. This a review of the present and is usually a simple correlation of information and events. The Fed said "x" and the markets moved "y". If "y" moved up, then the information was market positive. This is the approach taken by most reporters. Events are matched with market behavior. Seldom do reporters say that an economic announcement seemed inconsistent with the market behavior.

Finally comes the market view or forecast what will happen next.  This is usually in the form of "I believe", "I think", "It is likely".  There will be some contingent forecasting. If the Fed raises rates, the market will sell-off. There is little precision with this forecast. It is an opinion with little weight or a timeframe for which it will apply. There is often not a clear tie between what has happened in the past, what is occurring in the present, and what will happen in the future.

This is why quant models are so important. The quant model takes data from the past to find relationships. Uses the new data to update the model and the extrapolates the model into the future with some standard error to place bounds on what is possible. This model can be wrong, but it is not worthless. It is not qualitative but allows for some level of precision.