Konsten att förutspå konjunkturen - Hur användbara är enkätbaserade snabbindikatorer?

Foreseeing future changes in economic activity is of up most importance to a wide array of actors. The time lag in the presentation of official GDP statistics means that to somewhat accurately capture a tendency of where the business cycle is heading one need turn to qualitative sentiment indicators such as Purchasing Managers’ Index (PMI) and official sentiment indicators for valuable clues. This thesis uses in-sample and out-of-sample methods to evaluate how well PMI and official Business Tendency Surveys (BTS) in Sweden and Denmark fair against the year-on-year growth rate of industrial production. Moreover, various Swedish regional sentiment indicators get examined and a case study of the indicator for the Öresund region is performed. The lead/lag relationships of the indicators are examined through three tests. Cross correlations, Granger causality tests and turning point analysis suggest that all observed indicators have a lead over industrial production. In general PMI outperforms BTS and the Swedish indicators have a longer lead on average compared to the Danish ones. Out-of-sample forecast evaluations point partly in the same direction. The Swedish indicators provide the best forecasts for industrial production. In an international outlook, Swedish and Danish PMI compared with other European PMI perform poorly. The high export dependency in Sweden and Denmark is presented as a possible explanation among others