nep-for New Economics Papers
on Forecasting
Issue of 2018‒10‒01
eight papers chosen by
Rob J Hyndman
Monash University

  1. A Probabilistic Cohort-Component Model for Population Forecasting - The Case of Germany By Vanella, Patrizio; Deschermeier, Philipp
  2. Nowcasting Canadian Economic Activity in an Uncertain Environment By Tony Chernis; Rodrigo Sekkel
  3. Improving Underlying Scenarios for Aggregate Forecasts: A Multi-level Combination Approach By Cobb, Marcus P A
  4. The Use of Business Expectations in the Short-Term Forecasting of Economic Activity in Ukrainian By Roman Lysenko; Nataliia Kolesnichenko
  5. Forecasting India's Economic Growth: A Time-Varying Parameter Regression Approach. By Bhattacharya, Rudrani; Chakravarti, Parma; Mundle, Sudipto
  6. Forecasting with second-order approximations and Markov-switching DSGE models By Sergey Ivashchenko; Semih Emre Cekin; Kevin Kotze; Rangan Gupta
  7. How CBO Produces Its 10-Year Economic Forecast: Working Paper 2018-02 By Robert W. Arnold
  8. TIIE-28 Swaps as Risk-Adjusted Forecasts of Monetary Policy in Mexico By García-Verdú Santiago; Ramos Francia Manuel; Sánchez-Martínez Manuel

  1. By: Vanella, Patrizio; Deschermeier, Philipp
    Abstract: The future development of population size and structure is of importance since planning in many areas of politics and business is conducted based on expectations about the future makeup of the population. Countries with both decreasing mortality and low fertility rates, which is the case for most countries in Europe, urgently need adequate population forecasts to identify future problems regarding social security systems as one determinant of overall macroeconomic development. This contribution proposes a stochastic cohort-component model that uses simulation techniques based on stochastic models for fertility, migration and mortality to forecast the population by age and sex. We specifically focus on quantifying the uncertainty of future development as previous studies have tended to underestimate future risk. The results provide detailed insight into the future population structure, disaggregated into both sexes and 116 age groups. Moreover, the uncertainty in the forecast is quantified as prediction intervals for each subgroup. The underlying models for forecasting the demographic components have been developed in earlier studies and rely on principal component time series models. Since the proposed model is fully probabilistic, it offers a wide range of information, not only identifying the most probable course of the population but also a vast number of possible scenarios for future development of the population and quantifying their respective likelihoods. The model is applied to forecast the population of Germany until 2040. The results indicate a larger future population for Germany compared to the population predicted in studies conducted before 2015. The driving factors are lower mortality, higher fertility and higher net migration as derived by us statistically in contrast to widely used qualitative assumptions. The present study shows that the increase in population is mainly due to a larger proportion of older individuals.
    Keywords: Population Forecasting, Stochastic Simulation, Cohort-Component Methods, Principal Component Analysis, Time Series Analysis
    JEL: C22 C53 J11
    Date: 2018–09
  2. By: Tony Chernis; Rodrigo Sekkel
    Abstract: This paper studies short-term forecasting of Canadian real GDP and its expenditure components using combinations of nowcasts from different models. Starting with a medium-sized data set, we use a suite of common nowcasting tools for quarterly real GDP and its expenditure components. Using a two-step combination procedure, the nowcasts are first combined within model classes and then merged into a single point forecast using simple performance-based weighting methods. We find that no single model clearly dominates over all forecast horizons, subsamples and target variables. This highlights that when operating in an uncertain environment, where the choice of model is not clear, combining forecasts is a prudent strategy.
    Keywords: Econometric and statistical methods
    JEL: C53 E52 E37
    Date: 2018
  3. By: Cobb, Marcus P A
    Abstract: Abstract In some situations forecasts for a number of sub-aggregations are required for analysis in addition to the aggregate itself. In this context, practitioners typically rely on bottom-up methods to produce a set of consistent forecasts in order to avoid conflicting messages. However, using this approach exclusively can mean that forecasting accuracy is negatively affected when compared to using other methods. This paper presents a method for increasing overall accuracy by jointly combining the forecasts for an aggregate, any sub-aggregations, and the components from any number of models and measurement approaches. The framework seeks to benefit from the strengths of each of the forecasting approaches by accounting for their reliability in the combination process and exploiting the constraints that the aggregation structure imposes on the set of forecasts as a whole. The results from the empirical application suggest that the method is successful in allowing the strengths of the better-performing approaches to contribute to increasing the performance of the rest.
    Keywords: Bottom-up forecasting; Forecast combination; Hierarchical forecasting; Reconciling forecasts
    JEL: C53 E27 E37
    Date: 2018–08
  4. By: Roman Lysenko (National Bank of Ukraine); Nataliia Kolesnichenko (National Bank of Ukraine)
    Abstract: This paper focuses on the predictive capability of business outlook survey findings in forecasting changes in Ukraine’s real GDP, and in its consumption and investment components. Survey findings have been generalized through the use of principal component analysis. The business outlook index compiled by the National Bank of Ukraine is used as an alternative measure. To forecast GDP and its components, we employ ARDL and VAR models, which factor in the estimated principal components, the business outlook index, and the business outlook survey findings for construction investment over the next 12 months. In estimating the predictive capability of the models, we use pseudo-out-of-sample forecasting. A comparison with actual data shows that annual GDP and consumption growth are best forecast in the current period by applying business outlook survey findings that have been generalized using a principal component analysis, and the first difference of the business outlook index.
    Keywords: business expectations, GDP, short-term forecasting
    JEL: E27 E58
    Date: 2018–06
  5. By: Bhattacharya, Rudrani (National Institute of Public Finance and Policy); Chakravarti, Parma (Ambedkar University); Mundle, Sudipto (National Institute of Public Finance and Policy)
    Abstract: Forecasting GDP growth is essential for effective and timely implementation of macroeconomic policies. This paper uses a Principal Component augmented Time Varying Parameter Regression (TVPR) approach to forecast real aggregate and sectoral growth rates for India. We estimate the model using a mix of fiscal, monetary, trade and production side-specific variables. To assess the importance of different growth drivers, three variants of the model are used. In `Demand-side' model, the set of variables exclude production-specific indicators, while in the `Supply-side' model, information is extracted only from the latter set. The `Combined' model consists of both sets of variables. We find that TVPR model consistently outperforms constant parameter factor-augmented regression model and Dynamic Factor Model in terms of forecasting performance for all the three specifications. Based on the TVPR model, we find that demand side variant minimises the error forecast for total GDP and the industrial sector GDP, while the supply side variant minimises the error forecast for services sector GDP. We also find that forecast error is minimised using both the supply side variant and the combined variant for agriculture sector GDP.
    Keywords: Real GDP growth ; Forecasting ; Time varying ; Parameter Regression Model ; Dynamic Factor Model ; India
    JEL: C32 C5 O4
    Date: 2018–09
  6. By: Sergey Ivashchenko (St. Petersburg Institute for Economics and Mathematics Russian Academy of Sciences (RAS)); Semih Emre Cekin (Department of Economics, Turkish-German University); Kevin Kotze (School of Economics, University of Cape Town); Rangan Gupta (Department of Economics, University of Pretoria)
    Abstract: This paper compares the out-of-sample forecasting performance of first- and second-order perturbation approximations for DSGE models that incorporate Markov-switching behaviour in the policy reaction function and the volatility of shocks. These results are compared to those of a model that does not incorporate any regime-switching. The results suggest that second-order approximations provide an improved forecasting performance in models that do not allow for regime-switching, while for the MS-DSGE models, a first-order approximation would appear to provide better out-of-sample properties. In addition, we find that over short-horizons, the MS-DSGE models provide superior forecasting results when compared to those models that do not allow for regime-switching (at both perturbation orders).
    Date: 2018
  7. By: Robert W. Arnold
    Abstract: As part of its mandate to provide nonpartisan analyses to assist economic and budgetary decisions by the Congress, CBO prepares an economic forecast twice per year. Those forecasts are used in the agency’s projections for the federal budget and cost estimates for proposed federal legislation. Forecasts of gross domestic product, inflation, interest rates, and income play a particularly significant role in the agency’s budgetary analysis; for example, projections of wages and salaries are used to forecast individual income tax receipts. This paper describes the process used
    JEL: E17
    Date: 2018–02–02
  8. By: García-Verdú Santiago; Ramos Francia Manuel; Sánchez-Martínez Manuel
    Abstract: Information extracted from financial derivatives on interest rates is commonly used to forecast movements in such rates. Yet, such an extraction generally assumes that agents are risk-neutral. Thus, it might be useful to account for their risk-aversion when doing forecasts. This can be done adding a risk-correction. In this context, we use TIIE-28 swaps to forecast changes in monetary policy in Mexico, using financial variables to account for the risk-correction. We assess whether models with a risk-correction outperform the TIIE-28 swaps rates. Their in-sample explained variability improves when using a risk-correction. Our main model’s out-of-sample forecasts are similar for short horizons (3-month), and statistically better for longer horizons (9 to 24-month), compared to the direct use of TIIE-28 swaps interest rates.
    Keywords: TIIE-28;Swaps;Interest Rates;Expected Monetary Policy
    JEL: E52 G12
    Date: 2018–08

This nep-for issue is ©2018 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.