|
on Forecasting |
| By: | David T. Frazier; Ryan Covey; Gael M. Martin; Donald Poskitt |
| Abstract: | We demonstrate that the forecasting combination puzzle is a consequence of the methodology commonly used to produce forecast combinations. By the combination puzzle, we refer to the empirical finding that predictions formed by combining multiple forecasts in ways that seek to optimize forecast performance often do not out-perform more naive, e.g. equally-weighted, approaches. In particular, we demonstrate that, due to the manner in which such forecasts are typically produced, tests that aim to discriminate between the predictive accuracy of competing combination strategies can have low power, and can lack size control, leading to an outcome that favours the naive approach. We show that this poor performance is due to the behavior of the corresponding test statistic, which has a non-standard asymptotic distribution under the null hypothesis of no inferior predictive accuracy, rather than the {standard normal distribution that is} {typically adopted}. In addition, we demonstrate that the low power of such predictive accuracy tests in the forecast combination setting can be completely avoided if more efficient estimation strategies are used in the production of the combinations, when feasible. We illustrate these findings both in the context of forecasting a functional of interest and in terms of predictive densities. A short empirical example {using daily financial returns} exemplifies how researchers can avoid the puzzle in practical settings. |
| Date: | 2023–08 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2308.05263&r=for |
| By: | Bartzsch, Nikolaus; Brandi, Marco; de Pastor, Raymond; Devigne, Lucas; Maddaloni, Gianluca; Posada Restrepo, Diana; Sene, Gabriele |
| Abstract: | As part of the Eurosystem's annual banknote production planning, the national central banks draw up forecasts estimating the volumes of national-issued banknotes in circulation for the three years ahead. As at the end of 2021, more than 80 per cent of euro banknotes in circulation (cumulated net issuance) had been issued by the national central banks of France, Germany, Italy and Spain ('4 NCBs'). To date, the 4 NCBs have been using ARIMAX models to forecast the banknotes issued nationally in circulation by denomination ('benchmark models'). This paper presents the structural time series models developed by the 4 NCBs as an additional forecasting tool. The forecast accuracy measures used in this study show that the structural time series models outperform the benchmark models currently in use at each of the 4 NCBs for most of the denominations. However, it should be borne in mind that the statistical informative value of this comparison is limited by the fact the projection period is only twelve months. |
| Keywords: | euro, demand for banknotes, forecast of banknotes in circulation, structural time series models, ARIMA models, intervention variables |
| JEL: | C22 E41 E47 E51 |
| Date: | 2023 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:bubdps:202023&r=for |
| By: | Bartzsch Nikolaus; Brandi Marco; Devigne Lucas; De Pastor Raymond; Maddaloni Gianluca; Posada Restrepo Diana; Sene Gabriele |
| Abstract: | As part of the Eurosystem’s annual banknote production planning, the national central banks draw up forecasts estimating the volumes of national-issued banknotes in circulation for the three years ahead. As at the end of 2021, more than 80 per cent of euro banknotes in circulation (cumulated net issuance) had been issued by the national central banks of France, Germany, Italy and Spain (collectively referred to as the “4 NCBs”). To date, the 4 NCBs have been using ARIMAX models to forecast the banknotes issued nationally in circulation by denomination (“benchmark models”). This paper presents the structural time series models developed by the 4 NCBs as an additional forecasting tool. According to the forecast accuracy measures employed, the Structural Time series Models (“STSMs”) outperform the benchmark models at each of the 4 NCBs and for most of the denominations. However, it should be borne in mind that the statistical informative value of this comparison is limited by the short projection period of just 12 months. |
| Keywords: | Euro, Demand for Banknotes, Forecast of Banknotes in Circulation, Structural Time Series Models, ARIMA Models, Intervention Variables |
| JEL: | C22 E41 E47 E51 |
| Date: | 2023 |
| URL: | http://d.repec.org/n?u=RePEc:bfr:banfra:919&r=for |
| By: | Shahnaz Parsaeian (Department of Economics, University of Kansas, Lawrence, KS 66045) |
| Abstract: | This paper develops an efficient Stein-like shrinkage estimator for estimating slope parameters under structural breaks in seemingly unrelated regression models, which is then used for forecasting. The proposed method is a weighted average of two estimators: a restricted estimator that estimates the parameters under the restriction of no break in the coefficients, and an unrestricted estimator that considers break points and estimates the parameters using the observations within each regime. It is established that the asymptotic risk of the Stein-like shrinkage estimator is smaller than that of the unrestricted estimator, which is the method typically used to estimate the slope coefficients under structural breaks. Furthermore, this paper proposes an averaging minimal mean squared error estimator in which the averaging weight is derived by minimizing its asymptotic risk. The superiority of the two proposed estimators over the unrestricted estimator in terms of the mean squared forecast errors are also derived. Further, analytical comparison between the asymptotic risks of the proposed estimators is provided. Insights from the theoretical analysis are demonstrated in Monte Carlo simulations, and through an empirical example of forecasting output growth rates of G7 countries. |
| Keywords: | Forecasting; Seemingly unrelated regression; Structural breaks; Stein-like shrinkage estimator; Minimal mean squared error estimator |
| JEL: | C13 C23 C52 C53 |
| Date: | 2023–08 |
| URL: | http://d.repec.org/n?u=RePEc:kan:wpaper:202308&r=for |