nep-for New Economics Papers
on Forecasting
Issue of 2020‒04‒27
nine papers chosen by
Rob J Hyndman
Monash University

  1. A Vector Autoregressive Model of Forecast Electricity Consumption in France By Stéphane AURAY; Vincent CAPONI
  2. Forecasting inflation in Bosnia and Herzegovina By Elma Hasanovic
  3. Using the Eye of the Storm to Predict the Wave of Covid-19 UI Claims By Daniel Aaronson; Scott A. Brave; R. Andrew Butters; Daniel W. Sacks; Boyoung Seo
  4. Time-varying Uncertainty of the Federal Reserve’s Output Gap Estimate By Travis J. Berge
  5. The value of publicly available, textual and non-textual information for startup performance prediction By Kaiser, Ulrich; Kuhn, Johan M.
  6. Boltzmann Entropy in Cryptocurrencies: A Statistical Ensemble Based Approach By Grilli, Luca; Santoro, Domenico
  7. COVID-Induced Economic Uncertainty By Scott R. Baker; Nicholas Bloom; Steven J. Davis; Stephen J. Terry
  8. Saudi Vision 2030 Dynamic Input-Output Table: Computing Macroeconomic Forecasts with RAS Method By David Havrlant; Mehmet Ali Soytas
  9. Further Estimations of the Likely Total Infections and Deaths Due to COVID19 in Select Countries (Version 2 dt. April 10, 2020) By Morris, Sebastian

  1. By: Stéphane AURAY (CREST-ENSAI and ULCO); Vincent CAPONI (CREST-ENSAI and IZA)
    Abstract: This provides a VARX approach for the estimation of electricity demand in metropolitan France. Our methodology takes into account the complex relation- ship between weather variables and electricity demand, especially in the short and medium run, and the correlation in the longer run, between electricity and macroeconomic variables. We are able to provide a reliable conditional forecasting that, within the VAR framework, takes into account the common dependency of electricity consumption and other variables. While the VAR approach is not novel within this literature, our main contributions lie on the use of exible functions that capture the role of weather to explain electricity consumption together with macroeconomic trend and cycle variables, and on the use of very detailed and comprehensive data on actual metered consumption of electricity in France. In- sample and out-sample forecasts provide evidence that our method is reliable for predicting future scenarios conditional on exogenous variables.
    Keywords: Electricity Forecast.
    JEL: Q43 Q47
    Date: 2020–02–12
  2. By: Elma Hasanovic (Central Bank of Bosnia and Herzegovina)
    Abstract: The purpose of this paper is to evaluate the performance of some leading univariate and multivariate models: ARIMA, the standard OLS VAR and Bayesian VAR models, in forecasting inflation in Bosnia and Herzegovina. Although the presented models are small and highly aggregated, they provide a convenient framework to illustrate practical forecast issues. Furthermore, they are a good starting point in the process of the forecast development. The empirical part of this paper estimates the domestic and international transmission effects on inflation and tries to find good predictors of the inflation. A variety of inflation indicators included in the VAR models are assessed as potential predictors of inflation. They have been suggested by economic theory and existing research. A pseudo out-of-sample forecast approach is employed to assess the models’ performance at different horizons using a recursive strategy. The study then evaluates the relative forecast performance of univariate model and various alternative specifications of the VAR models and offers conclusions. The results confirm the significant improvement in forecasting performance at all forecast horizons when Bayesian techniques, which incorporate information from the likelihood function and some informative prior distributions, are used.
    Keywords: Bayesian VAR, model selection, inflation forecasting
    Date: 2020–02–25
  3. By: Daniel Aaronson; Scott A. Brave; R. Andrew Butters; Daniel W. Sacks; Boyoung Seo
    Abstract: We leverage an event-study research design focused on the seven costliest hurricanes to hit the US mainland since 2004 to identify the elasticity of unemployment insurance filings with respect to search intensity. Applying our elasticity estimate to the state-level Google Trends indexes for the topic “unemployment,” we show that out-of-sample forecasts made ahead of the official data releases for March 21 and 28 predicted to a large degree the extent of the Covid-19 related surge in the demand for unemployment insurance. In addition, we provide a robust assessment of the uncertainty surrounding these estimates and demonstrate their use within a broader forecasting framework for US economic activity.
    Keywords: Covid-19; Google trends; hurricanes; unemployment; unemployment insurance
    JEL: C53 H12 J65
    Date: 2020–04–07
  4. By: Travis J. Berge
    Abstract: What is the output gap and when do we know it? A factor stochastic volatility model estimates the common component to forecasts of the output gap produced by the staff of the Federal Reserve, its time-varying volatility, and time-varying, horizon-specific forecast uncertainty. The common factor to these forecasts is highly procyclical, and unexpected increases to the common factor are associated with persistent responses in other macroeconomic variables. However, output gap estimates are very uncertain, even well after the fact. Output gap uncertainty increases around business cycle turning points. Lastly, increased macroeconomic uncertainty, as measured by the output gap's time-varying volatility, produces pronounced negative responses to other macroeconomic variables.
    Keywords: Output gap; Unobserved variables; Real-time data; Factor model; Stochastic volatility; Macroeconomic uncertainty
    JEL: C53 E32
    Date: 2020–02–03
  5. By: Kaiser, Ulrich; Kuhn, Johan M.
    Abstract: Can publicly available, web-scraped data be used to identify promising business startups at an early stage? To answer this question, we use such textual and non-textual information about the names of Danish firms and their addresses as well as their business purpose statements (BPSs) supplemented by core accounting information along with founder and initial startup characteristics to forecast the performance of newly started enterprises over a five years' time horizon. The performance outcomes we consider are involuntary exit, above{average employment growth, a return on assets of above 20 percent, new patent applications and participation in an innovation subsidy program. Our first key finding is that our models predict startup performance with either high or very high accuracy with the exception of high returns on assets where predictive power remains poor. Our second key finding is that the data requirements for predicting performance outcomes with such accuracy are low. To forecast the two innovation-related performance outcomes well, we only need to include a set of variables derived from the BPS texts while an accurate prediction of startup survival and high employment growth needs the combination of (i) information derived from the names of the startups, (ii) data on elementary founder-related characteristics and (iii) either variables describing the initial characteristics of the startup (to predict startup survival) or business purpose statement information (to predict high employment growth). These sets of variables are easily obtainable since the underlying information is mandatory to report upon business registration. The substantial accuracy of our predictions for survival, employment growth, new patents and participation in innovation subsidy programs indicates ample scope for algorithmic scoring models as an additional pillar of funding and innovation support decisions.
    Keywords: startup,performance,prediction,text as data
    JEL: L26 C53
    Date: 2020
  6. By: Grilli, Luca; Santoro, Domenico
    Abstract: In this paper we try to build a statistical ensemble to describe a cryptocurrency-based system, emphasizing an "affinity" between the system of agents trading in these currencies and statistical mechanics. We focus our study on the concept of entropy in the sense of Boltzmann and we try to extend such a definition to a model in which the particles are replaced by N agents completely described by their ability to buy and to sell a certain quantity of cryptocurrencies. After providing some numerical examples, we show that entropy can be used as an indicator to forecast the price trend of cryptocurrencies.
    Keywords: Cryptocurrency, Entropy, Prices Forecast, Boltzmann, Blockchain
    JEL: C02 C69 E44 E47 G12 G17 G19
    Date: 2020–04–11
  7. By: Scott R. Baker; Nicholas Bloom; Steven J. Davis; Stephen J. Terry
    Abstract: Assessing the economic impact of the COVID-19 pandemic is essential for policymakers, but challenging because the crisis has unfolded with extreme speed. We identify three indicators – stock market volatility, newspaper-based economic uncertainty, and subjective uncertainty in business expectation surveys – that provide real-time forward-looking uncertainty measures. We use these indicators to document and quantify the enormous increase in economic uncertainty in the past several weeks. We also illustrate how these forward-looking measures can be used to assess the macroeconomic impact of the COVID-19 crisis. Specifically, we feed COVID-induced first-moment and uncertainty shocks into an estimated model of disaster effects developed by Baker, Bloom and Terry (2020). Our illustrative exercise implies a year-on-year contraction in U.S. real GDP of nearly 11 percent as of 2020 Q4, with a 90 percent confidence interval extending to a nearly 20 percent contraction. The exercise says that about 60 percent of the forecasted output contraction reflects a negative effect of COVID-induced uncertainty.
    JEL: D80 E17 E32 E66 L50
    Date: 2020–04
  8. By: David Havrlant; Mehmet Ali Soytas (King Abdullah Petroleum Studies and Research Center)
    Abstract: When an economy is in the midst of a transformation and diversification process, it is hard to assume that its sectoral composition and inter-industry transactions will remain unchanged. This is especially the case since substantial adjustments to a country’s economic structure are at the heart of any restructuring plan. This paper introduces an approach that combines macroeconomic forecasts with the RAS method to produce long-term projections of input-output tables (IOTs), with an emphasis on key targets of Saudi Vision 2030, Saudi Arabia’s blueprint for economic diversification. A significant advantage of the input-output framework is its high sectoral granularity, allowing it to capture the impacts of adjustments to final demand or government policies with respect to individual sectors. Our hybrid approach enables the introduction of different growth paths for the main variables, so that Vision 2030’s transformation plan is reflected appropriately in the projected IOTs. The framework is flexible enough to accommodate sudden adjustments with relative ease, such as the introduction of new technologies or entire sectors into the economy. Saudi Vison 2030 includes a set of targets relating to economic diversification, improved energy efficiency, the introduction of new technologies, social transformation and the support of selected emerging sectors. These policies are expected to have a substantial impact on the Saudi economy, underlining the need for an adequate and flexible tool for projecting and evaluating structural adjustments in the economy.
    Keywords: Economic Diversification, Economic projection, Input-Output table, Saudi Vision 2030
    Date: 2020–04–12
  9. By: Morris, Sebastian
    Abstract: We had earlier estimated the likely cases and deaths over the course of the pandemic for a number of countries. This was an early attempt and gave somewhat tentative results. With some 7 more days of data being now available, better estimates are possible which we bring out in this paper. As in the previous paper we use a logistic model of cumulative cases and deaths, to estimate the zero growth level of cases and deaths. We also provide an upper bound to these estimates. The earlier estimates are further reinforced, and new estimates are made for a select set of countries where the growth rates in the numbers of cases, and in deaths have begun to decline. We also give estimates of the current growth rates in cases and deaths that these countries are likely to witness. The study as before presumes that the spread of infection is one-stage logistic process, once significant numbers of infections have taken place. This may not be true of countries which witnessed low deaths and cases. In countries that have witnessed much spread and deaths relative to their populations and with more sustainable approaches to containment may not witness significantly more deaths than what has happened thus far. This would be the case of Iran, Italy. China and Korea too with their rather highly coordinated approach despite low spread of cases and low number of deaths relative to their population would along with Iran, Italy and Denmark and Turkey would most likely not see a secondary wave of infections. Argentina and South Africa show very high growth rate in deaths even the increase in cases have slowed down considerable. Spain has stabilized its growth in deaths to nearly zero levels bit since the cases are continuing to grow at around 5.7% the death rates could again turn positive after a while. Germany and Indonesia show continuing rise in deaths and cases at moderately high rates. Japan, Malaysia, Brazil and Singapore show low to moderate death rates, but since the rise in cases continues to be between 5 and 8%, these low(Japan) moderate growth rate in deaths are likely to continue for a while before they fall to zero. France, Sweden Australia and Thailand would see continuing growth in cases at moderate rates even though the growth in deaths continue to be at high rates. The US most notably shows very high growth rates in both deaths and in cases indicating that the deaths at high rates are likely to continue for a while. While estimates are made for Canada, India, Bangladesh, Russia, Mexico, UK and the Philippines, they are of limited value since it is too early for the logistic model to fit. However, all of these except Russia show high death rates and high case rates. These countries could all see continuing rise in cases before the decline in rates happen, so that their current decline in death rates even when statistically significant could change for the worse. We have as in the previous paper used a logistic model to estimate the current growth rates, and made forecasts of the ultimate stable cases and deaths before these stop rising any further. For 26 countries (with a combined population of 3.8 billion) the total cases as on date 9th /10th April was where the logistic trend has been realized for cases, was 1.36 million. We expect the cases to rise to a maximum in the countries covered to 2.9 million. The death trends in only 22 of the 29 countries considered had stabilized to a logistic model. In these 22 countries (with a combined population of 3.7billion) the deaths as on date were 87,472. These would surely rise to between 121,000 to 355,000 before stabilizing. In the estimates above India most notably has not been included, since its trends have not yet stablised to a logistic unfoldment. At present it is engaged in a titanic struggle through near complete lock downs to restrict the cases and deaths to low levels. Whether this would work to quell the spread to very levels, or whether the problem explodes later is still an open question.
    Date: 2020–04–13

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.