Klusowski, J. Prediction Horizons Influence Expectations of Trend Continuation vs. Reversal. Manuscript in preparation.


People commonly make forecasts across short and long horizons (e.g., next month vs. next quarter), yet the effects of these variations remain understudied. This research shows that such horizons influence the degree to which people expect existing trends to continue or reverse, even for the immediate future. For example, people expect greater trend continuation in the next period when predicting the next period only rather than the next three periods. These effects manifest consistently for increasing, decreasing, and repeating sequences as well as for skill and chance domains, which typically show opposing patterns. These tendencies seem to emerge because, compared to short horizons, long horizons elicit beliefs that imbalances in trends cannot persist. Theoretically, this research contributes to the literature on post-trend predictions (positive vs. negative recency). Previous research has largely considered forecasting as a retrospective process dependent on past trend characteristics, e.g., streak origin and length. The present research suggests that this process can also be prospective, i.e., dependent on future period attributes. In addition, it also extends the work on temporal distance (near vs. distant forecasts) or categorical scope (narrow vs. broad brackets) by suggesting distinct and even opposite predictions. Practically, these findings suggest that business managers and policymakers should be aware of these effects to understand and influence forecasts: When trend characteristics themselves are not malleable, prediction horizons can be a useful method to manage expectations in financial markets, performance outcomes, and many other important metrics across domains.