Is Knowledge Cursed When Forecasting the Forecasts of Others

Financial decision makers (lenders, insurers, advisees) often need to estimate how well others make decisions. Is knowledge a blessing or a curse when forecasting others' forecast accuracy? The authors show that this depends on its type. Within a single experimental setting, they identify and test 4 distinct information types that have different effects on forecast accuracy. First, the authors revisit the well-known "curse of knowledge" and show that it may have resulted from entirely arbitrary, uninformative anchors. Second, we show that in contrast, genuinely informative cues purged of anchoring potential enhance estimation accuracy. Third, richer, more detailed financial information has no effect even for participants better able to interpret it. Fourth, domain experts do not overimpute others' forecast ability. The authors conclude that in financial settings knowledge may be a blessing or a curse, or have no effect depending on its type.