Model Uncertainty and Serial Correlation in Forecast Errors

2009 ◽  
Author(s):  
Juhani T. Linnainmaa ◽  
Walter N. Torous
Keyword(s):  
Model Uncertainty ◽  
Serial Correlation ◽  
Forecast Errors

scholarly journals Can Cattle Basis Forecasts Be Improved? A Bayesian Model Averaging Approach

2019 ◽  
Vol 51 (02) ◽  
pp. 249-266
Author(s):  
Nicholas D. Payne ◽  
Berna Karali ◽  
Jeffrey H. Dorfman
Keyword(s):  
Bayesian Approach ◽  
Model Uncertainty ◽  
Bayesian Model Averaging ◽  
Model Performance ◽  
Model Averaging ◽  
Forecasting Model ◽  
Forecast Errors ◽  
Forecast Horizon ◽  
Combining Forecasts ◽  
The Bayesian Approach

AbstractBasis forecasting is important for producers and consumers of agricultural commodities in their risk management decisions. However, the best performing forecasting model found in previous studies varies substantially. Given this inconsistency, we take a Bayesian approach, which addresses model uncertainty by combining forecasts from different models. Results show model performance differs by location and forecast horizon, but the forecast from the Bayesian approach often performs favorably. In some cases, however, the simple moving averages have lower forecast errors. Besides the nearby basis, we also examine basis in a specific month and find that regression-based models outperform others in longer horizons.

scholarly journals Forecast Comparison Based on Random Walks

2016 ◽  
Vol 144 (2) ◽  
pp. 615-626 ◽  
Author(s):  
Timothy DelSole ◽  
Michael K. Tippett
Keyword(s):  
Random Walks ◽  
Serial Correlation ◽  
Model Development ◽  
Sign Test ◽  
Forecast Errors ◽  
Dynamical Models ◽  
Multimodel Ensemble ◽  
The North ◽  
Forecast Comparison ◽  
North American Multimodel Ensemble

Abstract This paper proposes a procedure based on random walks for testing and visualizing differences in forecast skill. The test is formally equivalent to the sign test and has numerous attractive statistical properties, including being independent of distributional assumptions about the forecast errors and being applicable to a wide class of measures of forecast quality. While the test is best suited for independent outcomes, it provides useful information even when serial correlation exists. The procedure is applied to deterministic ENSO forecasts from the North American Multimodel Ensemble and yields several revealing results, including 1) the Canadian models are the most skillful dynamical models, even when compared to the multimodel mean; 2) a regression model is significantly more skillful than all but one dynamical model (to which it is equally skillful); and 3) in some cases, there are significant differences in skill between ensemble members from the same model, potentially reflecting differences in initialization. The method requires only a few years of data to detect significant differences in the skill of models with known errors/biases, suggesting that the procedure may be useful for model development and monitoring of real-time forecasts.

Nonlinear Adjustments of Volatility Expectations to Forecast Errors: Evidence from Markov-Regime Switches in Implied Volatility

2012 ◽  
Vol 15 (03) ◽  
pp. 1250007 ◽  
Author(s):  
Kazuhiko Nishina ◽  
Nabil Maghrebi ◽  
Mark J. Holmes
Keyword(s):  
Financial Crises ◽  
Regime Switching ◽  
Serial Correlation ◽  
Implied Volatility ◽  
Nonlinear Behavior ◽  
Heterogeneous Beliefs ◽  
Forecast Errors ◽  
Market Participants ◽  
Model Free ◽  
Switching Models

This paper tests for nonlinearities in the behavior of volatility expectations based on model-free implied volatility indices. Using Markov regime-switching models, the empirical evidence from the German, Japanese and U.S. markets suggests that there are indeed regime-specific levels of volatility expectations. Whereas the regimes seem to be governed by the degree of serial correlation and adjustment to forecast errors, there is no evidence of significant leverage effects. The frequency of regime shifts in volatility expectations is affected by the onset of financial crises, which have the effect of increasing the likelihood of regimes driven by lower autoregressive effects and faster speeds of adjustment. The evidence suggests that despite the heterogeneous beliefs of market participants, implied volatility indices provide a measure of consensus expectations that can be useful in understanding the nonlinear behavior of volatility expectations during periods of financial instability.

scholarly journals On the Rationality of Medium-Term Tax Revenue Forecasts: Evidence from Germany

2015 ◽  
Vol 235 (1) ◽  
pp. 22-40 ◽  
Author(s):  
Christian Breuer
Keyword(s):  
Serial Correlation ◽  
Institutional Setting ◽  
Tax Revenue ◽  
Forecast Errors ◽  
Structural Level ◽  
Medium Term ◽  
Short Term ◽  
Percentage Points ◽  
Rational Behaviour ◽  
Revenue Projections

Summary This paper examines tax revenue projections in Germany for the period 1968 to 2012 with a focus on forecasting rationality. It is shown that tax revenue forecasts for the medium-term are upward biased. Overoptimistic revenue projections are particularly pronounced after the German reunification and reflect upward-biased GDP projections in this period. The predicted tax-GDP-ratio appears to be upward biased, as well. The forecasts are likely to overestimate tax revenues if the predicted tax-GDP-ratio exceeds its structural level of approximately 22½ percentage points. The results also indicate that forecast errors of short-term projections for the current year exhibit serial correlation. It is conceivable that the non-rational behaviour can be traced back to the specific institutional setting of revenue forecasting and budgetary planning in Germany.

New Evidence on Serial Correlation in Analyst Forecast Errors

1999 ◽  
Vol 28 (4) ◽  
pp. 106 ◽  
Author(s):  
Stacey R. Nutt ◽  
John C. Easterwood ◽  
Cintia M. Easterwood
Keyword(s):  
Serial Correlation ◽  
Analyst Forecast ◽  
Forecast Errors ◽  
New Evidence
Sign in / Sign up

Export Citation Format

Share Document