nep-for New Economics Papers
on Forecasting
Issue of 2020‒01‒20
nine papers chosen by
Rob J Hyndman
Monash University

  1. Fast Forecast Reconciliation Using Linear Models By Mahsa Ashouri; Rob J Hyndman; Galit Shmueli
  2. Adaptive Trees: a new approach to economic forecasting By Nicolas Woloszko
  3. Comparing forecast accuracy in small samples By Döhrn, Roland
  4. Forecasting energy commodity prices: a large global dataset sparse approach By Ferrari, Davide; Ravazzolo, Francesco; Vespignani, Joaquin
  5. Focused Bayesian Prediction By Ruben Loaiza-Maya; Gael M Martin; David T. Frazier
  6. Forecasting Unemployment Rates with International Factors By Pincheira, Pablo; Hernández, Ana María
  7. Forecasting with Unbalanced Panel Data By Badi Baltagi; Long Liu
  8. Autonomous Factor Forecast Quality: The Case of the Eurosystem By Romain M Veyrune; Shaoyu Guo
  9. Understanding SLL / US$ exchange rate dynamics in Sierra Leone using Box-Jenkins ARIMA approach By Jackson, Emerson Abraham

  1. By: Mahsa Ashouri; Rob J Hyndman; Galit Shmueli
    Abstract: Forecasting hierarchical or grouped time series usually involves two steps: computing base forecasts and reconciling the forecasts. Base forecasts can be computed by popular time series forecasting methods such as Exponential Smoothing (ETS) and Autoregressive Integrated Moving Average (ARIMA) models. The reconciliation step is a linear process that adjusts the base forecasts to ensure they are coherent. However using ETS or ARIMA for base forecasts can be computationally challenging when there are a large number of series to forecast, as each model must be numerically optimized for each series. We propose a linear model that avoids this computational problem and handles the forecasting and reconciliation in a single step. The proposed method is very flexible in incorporating external data, handling missing values and model selection. We illustrate our approach using two datasets: monthly Australian domestic tourism and daily Wikipedia pageviews. We compare our approach to reconciliation using ETS and ARIMA, and show that our approach is much faster while providing similar levels of forecast accuracy.
    Keywords: hierarchical forecasting, grouped forecasting, reconciling forecast, linear regression.
    JEL: C10 C14 C22
    Date: 2019
  2. By: Nicolas Woloszko
    Abstract: The present paper develops Adaptive Trees, a new machine learning approach specifically designed for economic forecasting. Economic forecasting is made difficult by economic complexity, which implies non-linearities (multiple interactions and discontinuities) and unknown structural changes (the continuous change in the distribution of economic variables). The forecast methodology aims at addressing these challenges. The algorithm is said to be “adaptive” insofar as it adapts to the quantity of structural change it detects in the economy by giving more weight to more recent observations. The performance of the algorithm in forecasting GDP growth 3- to 12-months ahead is assessed through simulations in pseudo-real-time for six major economies (USA, UK, Germany, France, Japan, Italy). The performance of Adaptive Trees is on average broadly similar to forecasts obtained from the OECD’s Indicator Model and generally performs better than a simple AR(1) benchmark model as well as Random Forests and Gradient Boosted Trees.
    Keywords: business cycles, concept drift, feature engineering, forecasting, GDP growth, interpretable AI, machine learning, short-term forecasts, structural change
    JEL: C01 C18 C23 C45 C53 C63 E37
    Date: 2020–01–16
  3. By: Döhrn, Roland
    Abstract: The Diebold-Mariano-Test has become a common tool to compare the accuracy of macroeconomic forecasts. Since these are typically model-free forecasts, distribution free tests might be a good alternative to the Diebold-Mariano-Test. This paper suggests a permutation test. Stochastic simulations show that permutation tests outperform the Diebold-Mariano-Test. Furthermore, a test statistic based on absolute errors seems to be more sensitive to differences in forecast accuracy than a statistic based on squared errors.
    Keywords: macroeconomic forecast,forecast accuracy,Diebold-Mariano test,permutation test
    JEL: C14 C15 C53
    Date: 2019
  4. By: Ferrari, Davide (Free University of Bozen-Bolzano, Faculty of Economics and Management, Italy); Ravazzolo, Francesco (Free University of Bozen-Bolzano, Faculty of Economics and Management, Italy); Vespignani, Joaquin (Tasmanian School of Business & Economics, University of Tasmania)
    Abstract: This paper focuses on forecasting quarterly energy prices of commodities, such as oil, gas and coal, using the Global VAR dataset proposed by Mohaddes and Raissi (2018). This dataset includes a number of potentially informative quarterly macroeconomic variables for the 33 largest economies, overall accounting for more than 80% of the global GDP. To deal with the information on this large database, we apply a dynamic factor model based on a penalized maximum likelihood approach that allows to shrink parameters to zero and to estimate sparse factor loadings. The estimated latent factors show considerable sparsity and heterogeneity in the selected loadings across variables. When the model is extended to predict energy commodity prices up to four periods ahead, results indicate larger predictability relative to the benchmark random walk model for 1-quarter ahead for all energy commodities. In our application, the largest improvement in terms of prediction accuracy is observed when predicting gas prices from 1 to 4 quarters ahead.
    Keywords: energy prices, forecasting, Dynamic Factor model, sparse esti- mation, penalized maximum likelihood
    JEL: C1 C5 C8 E3 Q4
    Date: 2019
  5. By: Ruben Loaiza-Maya; Gael M Martin; David T. Frazier
    Abstract: We propose a new method for conducting Bayesian prediction that delivers accurate predictions without correctly specifying the unknown true data generating process. A prior is defined over a class of plausible predictive models. After observing data, we update the prior to a posterior over these models, via a criterion that captures a user-specified measure of predictive accuracy. Under regularity, this update yields posterior concentration onto the element of the predictive class that maximizes the expectation of the accuracy measure. In a series of simulation experiments and empirical examples we find notable gains in predictive accuracy relative to conventional likelihood-based prediction..
    Keywords: loss-based prediction, Bayesian forecasting, proper scoring rules, stochastic volatility model, expected shortfall, M4 forecasting competition.
    JEL: C11 C53 C58
    Date: 2020
  6. By: Pincheira, Pablo; Hernández, Ana María
    Abstract: In this paper we study international linkages when forecasting unemployment rates in a sample of 24 OECD economies. We propose a Global Unemployment Factor (GUF) and test its predictive ability considering in-sample and out-of-sample exercises. Our main results indicate that the predictive ability of the GUF is heterogeneous across countries. In-sample results are statistically significant for Austria, Belgium, Czech Republic, Finland, France, Ireland, The Netherlands, Portugal, Slovenia, Sweden and United States. Robust statistically significant out-of-sample results are found for Belgium, Czech Republic, France, The Netherlands, Slovenia, Sweden and the United States. This means that the inclusion of the GUF adds valuable information to predict domestic unemployment rates, at least for these last seven countries.
    Keywords: Forecasting, unemployment, international factors, time-series models, out-of-sample comparison, nested models.
    JEL: C1 C12 C2 C22 C4 C49 C5 C52 C53 C6 E2 E24 E27 E3 E37 E6 E63 F0 F00 F3 F36 F37 F6 F62 F66 J0 J00 J01 J08 J6 J60 J64
    Date: 2019–12–28
  7. By: Badi Baltagi (Center for Policy Research, Maxwell School, Syracuse University, 426 Eggers Hall, Syracuse, NY 13244); Long Liu (College of Business, University of Texas at San Antonio)
    Abstract: This paper derives the best linear unbiased prediction (BLUP) for an unbalanced panel data model. Starting with a simple error component regression model with unbalanced panel data and random effects, it generalizes the BLUP derived by Taub (1979) to unbalanced panels. Next it derives the BLUP for an unequally spaced panel data model with serial correlation of the AR(1) type in the remainder disturbances considered by Baltagi and Wu (1999). This in turn extends the BLUP for a panel data model with AR(1) type remainder disturbances derived by Baltagi and Li (1992) from the balanced to the unequally spaced panel data case. The derivations are easily implemented and reduce to tractable expressions using an extension of the Fuller and Battese (1974) transformation from the balanced to the unbalanced panel data case.
    Keywords: Forecasting, BLUP, Unbalanced Panel Data, Unequally Spaced Panels, Serial Correlation
    JEL: C33
    Date: 2020–01
  8. By: Romain M Veyrune; Shaoyu Guo
    Abstract: The publication of liquidity forecasts can be understood as part of central banks’ push toward greater transparency regarding monetary policy implementation. However, the advantages of transparency can only be realized if the information provided is accurate and reliable. This paper (1) provides an overview of the international practice of publishing the forecasts; (2) proposes and implements a framework to evaluate the accuracy and reliability of forecasts using the long history of Eurosystem forecasts as a case study; and (3) analyzes the Eurosystem forecast errors to determine the factors influencing forecast quality. A supporting factor for a high-quality forecast is the contemporaneousness of the information used, whereas money market segmentation can weigh on forecast quality.
    Date: 2019–12–27
  9. By: Jackson, Emerson Abraham
    Abstract: This study was carried out with the purpose of producing twelve out-of-sample forecast for a univariate exchange rate variable as a way of addressing challenges faced around dollarization issues in the domestic economy. In pursuit of this, the ARIMA model was utilised, with the best model [1,4,7] indicating that the Sierra Leone - Leone [SLL] currency will continue to depreciate against the United States Dollar [US$] throughout most part of the year 2020. This was done on the assumption of Ceteris Paribus condition, and most importantly on the view that past events of the univariate exchange rate variable is a determinant of future outcomes or performances. In a bid to moving forward, policy recommendations have suggested high level collaboration between relevant policy institutions like the Bank of Sierra Leone and the Ministry of Finance to address issues of concern, for example, a boost to the real sector and many more.
    Keywords: Box-Jenkins ARIMA, Exchange Rate, Forecast, Sierra Leone
    JEL: C52 C53 E47 F31 F47
    Date: 2020–01–01

This nep-for issue is ©2020 by Rob J Hyndman. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.