The Oracle or the Crowd? : Experts versus the Stock Market in Forecasting Ceasefire Success in the Levant
2017, Schneider, Gerald, Hadar, Maya, Bosler, Naomi
The forecasting literature has come to mistrust the predictions made by experts who forecast political events in mass media. Distinguishing between judgements made by one or few individuals (‘oracles’) and assessments made by larger groups (‘crowds’), we contrast journalistic predictions with forecasts stemming from the financial industry. These two competing views were evaluated in a quantitative analysis of the ex ante success of 24 ceasefire agreements in various conflicts which took place in the Levant from 1993 to 2014. Our analysis compares the forecasts appearing in press commentaries (Haaretz, Jerusalem Post and New York Times) with the expectations that the Tel Aviv Stock Exchange had about the stability of these cooperative efforts. To evaluate the predictions of these very dissimilar sources, the effectiveness of the ceasefires was analysed through the number of violent events following the official start of the truce. The analysis shows that the financial industry performs better than the media industry in the comparative evaluation of ceasefire forecasts, but that neither source provides sufficiently accurate predictions. The partial support for the crowd thesis is discussed in light of recent literature that resuscitates the usage of well-trained experts for forecasting purposes, but warns against the dramatizing predictions of media pundits.