INTERNATIONAL JOURNAL OF FORECASTING

Testing big data in a big crisis: Nowcasting under COVID-19
Barbaglia L, Frattarolo L, Onorante L, Pericoli FM, Ratto M and Pezzoli LT
During the COVID-19 pandemic, economists have struggled to obtain reliable economic predictions, with standard models becoming outdated and their forecasting performance deteriorating rapidly. This paper presents two novelties that could be adopted by forecasting institutions in unconventional times. The first innovation is the construction of an extensive data set for macroeconomic forecasting in Europe. We collect more than a thousand time series from conventional and unconventional sources, complementing traditional macroeconomic variables with timely big data indicators and assessing their added value at nowcasting. The second novelty consists of a methodology to merge an enormous amount of non-encompassing data with a large battery of classical and more sophisticated forecasting methods in a seamlessly dynamic Bayesian framework. Specifically, we introduce an innovative "selection prior" that is used not as a way to influence model outcomes, but as a selecting device among competing models. By applying this methodology to the COVID-19 crisis, we show which variables are good predictors for nowcasting Gross Domestic Product and draw lessons for dealing with possible future crises.
On single point forecasts for fat-tailed variables
Taleb NN, Bar-Yam Y and Cirillo P
We discuss common errors and fallacies when using naive "evidence based" empiricism and point forecasts for fat-tailed variables, as well as the insufficiency of using naive first-order scientific methods for tail risk management. We use the COVID-19 pandemic as the background for the discussion and as an example of a phenomenon characterized by a multiplicative nature, and what mitigating policies must result from the statistical properties and associated risks. In doing so, we also respond to the points raised by Ioannidis et al. (2020).
Guest editorial: Economic forecasting in times of COVID-19
Ferrara L and Sheng XS
Why was economic forecasting so difficult during COVID-19? To answer this question, we organized an online workshop in July 2020, sponsored by the International Institute of Forecasters (IIF) and hosted by American University. Below you will find some of the lessons that can be drawn from the special issue we edited.
A flexible framework for intervention analysis applied to credit-card usage during the coronavirus pandemic
Ho ATY, Morin L, Paarsch HJ and Huynh KP
We develop a variant of intervention analysis designed to measure a change in the law of motion for the distribution of individuals in a cross-section, rather than modeling the moments of the distribution. To calculate a counterfactual forecast, we discretize the distribution and employ a Markov model in which the transition probabilities are modeled as a multinomial logit distribution. Our approach is scalable and is designed to be applied to micro-level data. A wide panel often carries with it several imperfections that complicate the analysis when using traditional time-series methods; our framework accommodates these imperfections. The result is a framework rich enough to detect intervention effects that not only shift the mean, but also those that shift higher moments, while leaving lower moments unchanged. We apply this framework to document the changes in credit usage of consumers during the COVID-19 pandemic. We consider multinomial logit models of the dependence of credit-card balances, with categorical variables representing monthly seasonality, homeownership status, and credit scores. We find that, relative to our forecasts, consumers have greatly reduced their use of credit. This result holds for homeowners and renters as well as consumers with both high and low credit scores.
The COVID-19 shock and challenges for inflation modelling
Bobeica E and Hartwig B
We document the impact of COVID-19 on inflation modelling within a vector autoregression (VAR) model and provide guidance for forecasting euro area inflation during the pandemic. We show that estimated parameters are strongly affected, leading to different and sometimes implausible projections. As a solution, we propose to augment the VAR by allowing the residuals to have a fat-tailed distribution instead of a Gaussian one. This also outperforms with respect to unconditional forecasts. Yet, what brings sizeable forecast gains during the pandemic is adding meaningful off-model information, such as that entailed in the Survey of Professional Forecasters. The fat-tailed VAR loses part, but not all of its relative advantage compared to the Gaussian version when producing conditional inflation forecasts in a real-time setup. It is the joint fat-tailed errors and multi-equation modelling that manage to robustify models against extreme observations; in a single-equation model the same solution is less effective.
Editorial: Epidemics and forecasting with a focus on COVID-19
Pinson P
Forecasting the Covid-19 recession and recovery: Lessons from the financial crisis
Foroni C, Marcellino M and Stevanovic D
We consider simple methods to improve the growth nowcasts and forecasts obtained by mixed-frequency MIDAS and UMIDAS models with a variety of indicators during the Covid-19 crisis and recovery period, such as combining forecasts across various specifications for the same model and/or across different models, extending the model specification by adding MA terms, enhancing the estimation method by taking a similarity approach, and adjusting the forecasts to put them back on track using a specific form of intercept correction. Among these methods, adjusting the original nowcasts and forecasts by an amount similar to the nowcast and forecast errors made during the financial crisis and subsequent recovery seems to produce the best results for the US, notwithstanding the different source and characteristics of the financial crisis. In particular, the adjusted growth nowcasts for 2020Q1 get closer to the actual value, and the adjusted forecasts based on alternative indicators become much more similar, all unfortunately indicating a much slower recovery than without adjustment, and very persistent negative effects on trend growth. Similar findings also emerge for forecasts by institutions, for survey forecasts, and for the other G7 countries.
Monitoring Recessions: A Bayesian Sequential Quickest Detection Method
Li H, Sheng XS and Yang J
Monitoring business cycles faces two potentially conflicting objectives: accuracy and timeliness. To strike a balance between the dual objectives, we develop a Bayesian sequential quickest detection method to identify turning points in real time and propose a sequential stopping time as a solution. Using four monthly indexes of real economic activity in the US, we evaluate the method's real-time ability to date the past five recessions. The proposed method identifies similar turning point dates as the National Bureau of Economic Research (NBER), with no false alarms, but on average dates peaks 4 months faster and troughs 10 months faster relative to the NBER announcement. The timeliness of our method is also notable compared to the dynamic factor Markov-switching model - the average lead time is about 5 months in dating peaks and 2 months in dating troughs.
Nowcasting unemployment insurance claims in the time of COVID-19
Larson WD and Sinclair TM
Near-term forecasts, also called nowcasts, are most challenging but also most important when the economy experiences an abrupt change. In this paper, we explore the performance of models with different information sets and data structures in order to best nowcast US initial unemployment claims in spring of 2020 in the midst of the COVID-19 pandemic. We show that the best model, particularly near the structural break in claims, is a state-level panel model that includes dummy variables to capture the variation in timing of state-of-emergency declarations. Autoregressive models perform poorly at first but catch up relatively quickly. The state-level panel model, exploiting the variation in timing of state-of-emergency declarations, also performs better than models including Google Trends. Our results suggest that in times of structural change there is a bias-variance tradeoff. Early on, simple approaches to exploit relevant information in the cross sectional dimension improve forecasts, but in later periods the efficiency of autoregressive models dominates.
Commentary on "Transparent modeling of influenza incidence": Because the model said so
Moss R
Introduction to the M5 forecasting competition Special Issue
Makridakis S, Petropoulos F and Spiliotis E
Comparing trained and untrained probabilistic ensemble forecasts of COVID-19 cases and deaths in the United States
Ray EL, Brooks LC, Bien J, Biggerstaff M, Bosse NI, Bracher J, Cramer EY, Funk S, Gerding A, Johansson MA, Rumack A, Wang Y, Zorn M, Tibshirani RJ and Reich NG
The U.S. COVID-19 Forecast Hub aggregates forecasts of the short-term burden of COVID-19 in the United States from many contributing teams. We study methods for building an ensemble that combines forecasts from these teams. These experiments have informed the ensemble methods used by the Hub. To be most useful to policymakers, ensemble forecasts must have stable performance in the presence of two key characteristics of the component forecasts: (1) occasional misalignment with the reported data, and (2) instability in the relative performance of component forecasters over time. Our results indicate that in the presence of these challenges, an untrained and robust approach to ensembling using an equally weighted median of all component forecasts is a good choice to support public health decision-makers. In settings where some contributing forecasters have a stable record of good performance, trained ensembles that give those forecasters higher weight can also be helpful.
Post-script-Retail forecasting: Research and practice
Fildes R, Kolassa S and Ma S
This note updates the 2019 review article "Retail forecasting: Research and practice" in the context of the COVID-19 pandemic and the substantial new research on machine-learning algorithms, when applied to retail. It offers new conclusions and challenges for both research and practice in retail demand forecasting.
Short-term Covid-19 forecast for latecomers
Medeiros MC, Street A, Valladão D, Vasconcelos G and Zilberman E
The number of new Covid-19 cases is still high in several countries, despite vaccination efforts. A number of countries are experiencing new and severe waves of infection. Therefore, the availability of reliable forecasts for the number of cases and deaths in the coming days is of fundamental importance. We propose a simple statistical method for short-term real-time forecasting of the number of Covid-19 cases and fatalities in countries that are latecomers-i.e., countries where cases of the disease started to appear some time after others. In particular, we propose a penalized LASSO regression model with an error correction mechanism to construct a model of a latecomer country in terms of other countries that were at a similar stage of the pandemic some days before. By tracking the number of cases in those countries, we use an adaptive rolling-window scheme to forecast the number of cases and deaths in the latecomer. We apply this methodology to 45 countries and we provide detailed results for four of them: Brazil, Chile, Mexico, and Portugal. We show that the methodology performs very well when compared to alternative methods. These forecasts aim to foster better short-run management of the healthcare system and can be applied not only to countries but also to different regions within a country. Finally, the modeling framework derived in the paper can be applied to other infectious diseases.
The impact of the COVID-19 pandemic on business expectations
Meyer BH, Prescott B and Sheng XS
We document and evaluate how businesses are reacting to the COVID-19 crisis through August 2020. First, on net, firms see the shock (thus far) largely as a demand rather than supply shock. A greater share of firms report significant or severe disruptions to sales activity than to supply chains. We compare these measures of disruption to their expected changes in selling prices and find that, even for firms that report supply chain disruptions, they expect to lower near-term selling prices on average. We also show that firms are engaging in wage cuts and expect to trim wages further before the end of 2020. These cuts stem from firms that have been disproportionally negatively impacted by the pandemic. Second, firms (like professional forecasters) have responded to the COVID-19 pandemic by lowering their one-year-ahead inflation expectations. These responses stand in stark contrast to that of household inflation expectations (as measured by the University of Michigan or the New York Fed). Indeed, firms' one-year-ahead inflation expectations fell precipitously (to a series low) following the onset of the pandemic, while household measures of inflation expectations jumped markedly. Third, despite the dramatic decline in firms' near-term inflation expectations, their longer-run inflation expectations have remained relatively stable.
Hawkes process modeling of COVID-19 with mobility leading indicators and spatial covariates
Chiang WH, Liu X and Mohler G
Hawkes processes are used in statistical modeling for event clustering and causal inference, while they also can be viewed as stochastic versions of popular compartmental models used in epidemiology. Here we show how to develop accurate models of COVID-19 transmission using Hawkes processes with spatial-temporal covariates. We model the conditional intensity of new COVID-19 cases and deaths in the U.S. at the county level, estimating the dynamic reproduction number of the virus within an EM algorithm through a regression on Google mobility indices and demographic covariates in the maximization step. We validate the approach on both short-term and long-term forecasting tasks, showing that the Hawkes process outperforms several models currently used to track the pandemic, including an ensemble approach and an SEIR-variant. We also investigate which covariates and mobility indices are most important for building forecasts of COVID-19 in the U.S.
Pandemics and forecasting: The way forward through the Taleb-Ioannidis debate
Pinson P and Makridakis S
COVID-19: Forecasting confirmed cases and deaths with a simple time series model
Petropoulos F, Makridakis S and Stylianou N
Forecasting the outcome of outbreaks as early and as accurately as possible is crucial for decision-making and policy implementations. A significant challenge faced by forecasters is that not all outbreaks and epidemics turn into pandemics, making the prediction of their severity difficult. At the same time, the decisions made to enforce lockdowns and other mitigating interventions versus their socioeconomic consequences are not only hard to make, but also highly uncertain. The majority of modeling approaches to outbreaks, epidemics, and pandemics take an epidemiological approach that considers biological and disease processes. In this paper, we accept the limitations of forecasting to predict the long-term trajectory of an outbreak, and instead, we propose a statistical, time series approach to modelling and predicting the short-term behavior of COVID-19. Our model assumes a multiplicative trend, aiming to capture the continuation of the two variables we predict (global confirmed cases and deaths) as well as their uncertainty. We present the timeline of producing and evaluating 10-day-ahead forecasts over a period of four months. Our simple model offers competitive forecast accuracy and estimates of uncertainty that are useful and practically relevant.
Testing the predictive accuracy of COVID-19 forecasts
Coroneo L, Iacone F, Paccagnini A and Santos Monteiro P
We test the predictive accuracy of forecasts of the number of COVID-19 fatalities produced by several forecasting teams and collected by the United States Centers for Disease Control and Prevention for the epidemic in the United States. We find three main results. First, at the short horizon (1 week ahead) no forecasting team outperforms a simple time-series benchmark. Second, at longer horizons (3 and 4 week ahead) forecasters are more successful and sometimes outperform the benchmark. Third, one of the best performing forecasts is the Ensemble forecast, that combines all available predictions using uniform weights. In view of these results, collecting a wide range of forecasts and combining them in an ensemble forecast may be a superior approach for health authorities, rather than relying on a small number of forecasts.
Comparing the accuracy of several network-based COVID-19 prediction algorithms
Achterberg MA, Prasse B, Ma L, Trajanovski S, Kitsak M and Van Mieghem P
Researchers from various scientific disciplines have attempted to forecast the spread of coronavirus disease 2019 (COVID-19). The proposed epidemic prediction methods range from basic curve fitting methods and traffic interaction models to machine-learning approaches. If we combine all these approaches, we obtain the Network Inference-based Prediction Algorithm (NIPA). In this paper, we analyse a diverse set of COVID-19 forecast algorithms, including several modifications of NIPA. Among the algorithms that we evaluated, the original NIPA performed best at forecasting the spread of COVID-19 in Hubei, China and in the Netherlands. In particular, we show that network-based forecasting is superior to any other forecasting algorithm.
Probabilistic population forecasting: Short to very long-term
Raftery AE and Ševčíková H
Population forecasts are used by governments and the private sector for planning, with horizons up to about three generations (around 2100) for different purposes. The traditional methods are deterministic using scenarios, but probabilistic forecasts are desired to get an idea of accuracy, assess changes, and make decisions involving risks. In a significant breakthrough, since 2015, the United Nations has issued probabilistic population forecasts for all countries using a Bayesian methodology that we review here. Assessment of the social cost of carbon relies on long-term forecasts of carbon emissions, which in turn depend on even longer-range population and economic forecasts, to 2300. We extend the UN method to very-long range population forecasts by combining the statistical approach with expert review and elicitation. While the world population is projected to grow for the rest of this century, it will likely stabilize in the 22nd century and decline in the 23rd century.