Scoring rules and survey density forecasts

Incorporating Uncertainty Into USDA Commodity Price Forecasts

“…Rather than assigning an arbitrary log score (Boero, Smith, and Wallis ), therefore, we employ instead the CRPS—a quadratic scoring method that calculates the divergence of each forecasted density from an “ideal forecast” that places all probability mass at the realized price. An important advantage of the CRPS is that, unlike the log score, it awards forecasts that place more probability near (but not at) the realized value (Gneiting and Raftery ).…”

Does central bank communication really lead to better forecasts of policy decisions? New evidence based on a Taylor rule model for the ECB

“…The table shows the so-called Brier score (QPS) and the ranked probability score (RPS). Following Boero et al (2009), these measures can be explained as follows. The Brier score is calculated as: where p kt is a probability forecast of the outcome k (no change of the interest rate, an increase, or a decrease of the interest rate) at time t, while d kt , k = 1,…,K, takes the value 1 if the outcome x t is k, while otherwise d kt = 0.…”

Words that shake traders

“…(2) delivers the best forecasting performance among six different forecasting regression models according to the pseudo-R 2 criterion, and the quadratic and ranked probability scoring rules (see Boero et al, 2011). 11 Gurkaynak et al (2007) find that among a variety of financial market instruments (term federal funds loans, federal funds futures, term eurodollar deposits, eurodollar futures, Treasury bills and commercial papers) the federal funds futures dominate all the other securities in forecasting U.S. monetary policy at horizons out to 6 months.…”