I think that is an assumption too far. The behavioral finance literature shows that overconfidence often leads to errors; is there a reason why “conviction” recommendations from research providers should be exempt from that? I don’t think so, because my observation of asset managers and research providers over the years has provided too many concrete examples of stronger conviction recommendations or positions being detrimental to performance. Too often “experts” are even more susceptible to the pitfalls of overconfidence, because they believe they are immune due to their expertise.

I agree that this all is worth further study, and I believe that you will find that one dividing line in the analysis that will be of interest is “quantitative” versus “fundamental.” The patterns that emerge regarding those more confident recommendations (or, in the case of an asset manager, larger weights) vary considerably by the nature of the methodology employed.