|
on Forecasting |
| By: | John Geweke (University of Iowa. USA); Gianni Amisano (University of Brescia - Italy, European Central Bank and The RImini Centre for Economic Analisys - Italy) |
| Abstract: | A prediction model is any statement of a probability distribution for an outcome not yet observed. This study considers the properties of weighted linear combinations of n prediction models, or linear pools, evaluated using the conventional log predictive scoring rule. The log score is a concave function of the weights and, in general, an optimal linear combination will include several models with positive weights despite the fact that exactly one model has limiting posterior probability one. The paper derives several interesting formal results: for example, a prediction model with positive weight in a pool may have zero weight if some other models are deleted from that pool. The results are illustrated using S&P 500 returns with prediction models from the ARCH, stochastic volatility and Markov mixture families. In this example models that are clearly inferior by the usual scoring criteria have positive weights in optimal linear pools, and these pools substantially outperform their best components. |
| Keywords: | forecasting; GARCH; log scoring; Markov mixture; model combination; S&P 500 returns; stochastic volatility |
| Date: | 2008–01 |
| URL: | http://d.repec.org/n?u=RePEc:rim:rimwps:22-08&r=for |
| By: | Tatevik Sekhposyan; Barbara Rossi |
| Abstract: | We evaluate various models’ relative performance in forecasting future US output growth and inflation on a monthly basis. Our approach takes into account the possibility that the models’ relative performance can be varying over time. We show that the models’ relative performance has, in fact, changed dramatically over time, both for revised and real-time data, and investigate possible factors that might explain such changes. In addition, this paper establishes two empirical stylized facts. Namely, most predictors for output growth lost their predictive ability in the mid-1970s, and became essentially useless in the last two decades. When forecasting inflation, instead, fewer predictors are significant (among which, notably, capacity utilization and unemployment), and their predictive ability significantly worsened around the time of the Great Moderation. |
| Keywords: | Output Forecasts, Inflation Forecasts, Model Selection, Structural Change, Forecast Evaluation, Real-time data. Evaluation |
| JEL: | C22 C52 C53 |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:duk:dukeec:08-5&r=for |
| By: | Brzoza-Brzezina, Michal; Kot, Adam |
| Abstract: | In a New Keynesian model with asymmetric information we show that publication of macroeconomic projections and of the future interest rate path by the central bank can improve macroeconomic outcomes. However, the gains from publishing interest rate paths are small relative to those from publishing macroeconomic projections. Given that most inflation targeting central banks are already publishing macroeconomic projections this means that most gains from increasing transparency in this area may already have been reaped. This, together with the potential costs, may explain the relative reluctance of central banks to publish interest rate paths. |
| Keywords: | interest rate path; monetary policy; adaptive learning |
| JEL: | E43 E58 E52 |
| Date: | 2008–07–20 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:10296&r=for |