nep-for New Economics Papers
on Forecasting
Issue of 2014‒08‒25
six papers chosen by
Rob J Hyndman
Monash University

  1. Exponential Smoothing, Long Memory and Volatility Prediction By Tommaso Proietti
  2. Uncertainty of Macroeconomic Forecasters and the Prediction of Stock Market Bubbles By Helmut Herwartz; Konstantin A. Kholodilin
  3. Asian Development Outlook Forecast Skill By Ferrarini, Benno
  4. Big Data: Google Searches Predict Unemployment in Finland By Tuhkuri, Joonas
  5. Mobility in an enlarging European Union: Projections of potential flows from EU's eastern neighbors and Croatia By Michael Fertig; Martin Kahanec
  6. Forecasting the occurrence of electricity price spikes in the UK power market By Pawel Maryniak; Rafal Weron

  1. By: Tommaso Proietti (DEF and CEIS, Università di Roma "Tor Vergata")
    Abstract: Extracting and forecasting the volatility of financial markets is an important empirical problem. Time series of realized volatility or other volatility proxies, such as squared returns, display long range dependence. Exponential smoothing (ES) is a very popular and successful forecasting and signal extraction scheme, but it can be suboptimal for long memory time series. This paper discusses possible long memory extensions of ES and finally implements a generalization based on a fractional equal root integrated moving average (FerIMA) model, proposed originally by Hosking in his seminal 1981 article on fractional differencing. We provide a decomposition of the process into the sum of fractional noise processes with decreasing orders of integration, encompassing simple and double exponential smoothing, and introduce a lowpass real time filter arising in the long memory case. Signal extraction and prediction depend on two parameters: the memory (fractional integration) parameter and a mean reversion parameter. They can be estimated by pseudo maximum likelihood in the frequency domain. We then address the prediction of volatility by a FerIMA model and carry out a recursive forecasting experiment, which proves that the proposed generalized exponential smoothing predictor improves significantly upon commonly used methods for forecasting realized volatility.
    Keywords: Realized Volatility. Signal Extraction. Permanent-Transitory Decomposition. Fractional equal-root IMA model
    JEL: C22 C53 G17
    Date: 2014–07–30
  2. By: Helmut Herwartz; Konstantin A. Kholodilin
    Abstract: We assess the contribution of macroeconomic uncertainty -- approximated by the dispersion of the real GDP survey forecasts -- to the ex post and ex ante prediction of stock price bubbles. For a panel of six OECD economies covering 24 years, two alternative binary chronologies of bubble periods are determined and subjected to panel logit regressions conditioning on macroeconomic indicators and expectation uncertainty. Measures of macroeconomic uncertainty improve the ex ante signalling of stock price booms and bubbles.
    Keywords: Stock market bubbles, out-of-sample forecasting, consensus forecasts, macroeconomic uncertainty, OECD countries
    JEL: G01 G17 E27
    Date: 2014
  3. By: Ferrarini, Benno (Asian Development Bank)
    Abstract: The Asian Development Outlook (ADO) provides growth and inflation forecasts for more than 40 economies in the region. This paper assesses the accuracy of those forecasts against actual outcomes for the years from 2008 to 2011. The World Economic Outlook (WEO) forecasts by the International Monetary Fund are used as a benchmark against which to derive a comparative measure of the accuracy of ADO forecasts, or skill. ADO is found to be ‘more skillful’ than WEO in estimating both current-year gross domestic product (GDP) growth and consumer price index (CPI) inflation of Asian economies. WEO may have an edge over ADO when it comes to year-ahead GDP forecasts, while ADO’s inflation forecasts tend to be more accurate. By and large, and notwithstanding much heterogeneity across economies and years, both sets of forecasts display a high degree of inaccuracy during the crisis years.
    Keywords: economic forecasts; forecast skill; Asian Development Outlook; Asian Development Bank; World Economic Outlook; International Monetary Fund
    JEL: E17 E37
    Date: 2014–02–01
  4. By: Tuhkuri, Joonas
    Abstract: There are over 3 billion searches globally on Google every day. This report examines whether Google search queries can be used to predict the present and the near future unemployment rate in Finland. Predicting the present and the near future is of interest, as the official records of the state of the economy are published with a delay. To assess the information contained in Google search queries, the report compares a simple predictive model of unemployment to a model that contains a variable, Google Index, formed from Google data. In addition, cross-correlation analysis and Granger-causality tests are performed. Compared to a simple benchmark, Google search queries improve the prediction of the present by 10 % measured by mean absolute error. Moreover, predictions using search terms perform 39 % better over the benchmark for near future unemployment 3 months ahead. Google search queries also tend to improve the prediction accuracy around turning points. The results suggest that Google searches contain useful information of the present and the near future unemployment rate in Finland.
    Keywords: Big Data, Google, Internet, nowcasting, forecasting, unemployment, time-series analysis
    JEL: C1 C22 C43 C53 C82 E27
    Date: 2014–08–14
  5. By: Michael Fertig; Martin Kahanec
    Abstract: This study evaluates potential migration flows to the European Union from its eastern neighbors and Croatia. We perform out-of-sample forecasts using an adaption of the model of Hatton (1995) to time series cross-sectional data about post-enlargement migration flows following the EU’s 2004 enlar­gement. We consider two baseline policy scenarios, with and without accession of sending countries to the EU. Our results show that migration flows are driven by migration costs and economic conditions, but the largest effects accrue to policy variables. In terms of the predicted flows: (i) we can expect modest migration flows in case of no liberalization of labor markets and only moderately increased migration flows under liberalization; (ii) after an initial increase following liberalization, migration flows will subside to long run steady state; (iii) Ukraine will send the most migrants; and (iv) the largest inflows in absolute terms are predicted for Germany, Italy and Austria, whereas Ireland, Denmark, Finland and again Austria are the main receiving countries relative to their population.
    Keywords: Migration, free movement of workers, European Union, Eastern Partnership, EU enlargement, migration potential, out-of-sample forecasting
    JEL: F22 C23 C53
    Date: 2013–10–23
  6. By: Pawel Maryniak; Rafal Weron
    Abstract: We study the forward looking information that is available to all participants in the UK power market and measure its predictive value with respect to forecasting the occurrence of electricity price spikes. We focus on information that measures the extent to which the capacity of the UK generation park will be constrained over the next two weeks. Following Cartea et al. (2009) we use a measure of 'tight market conditions', based on predicted capacity constraints, which identifies the periods when price spikes are more likely to occur. Considering a longer and a more recent time period (1.1.2006-31.12.2012), we find that irrespective of the spike identification algorithm the probability of observing a spike is roughly an exponentially increasing function of the demand-to-capacity ratio.
    Keywords: Electricity price spike; Forecasting; Capacity constraints; Indicated demand; Indicated generation
    JEL: C14 C51 C53 Q47
    Date: 2014–08–16

This nep-for issue is ©2014 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.