Comparing trained and untrained probabilistic ensemble forecasts of COVID-19 cases and deaths in the United States

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.     

The dynamic effects of infectious disease outbreaks: The case of pandemic influenza and human coronavirus

Pandemic influenza is a regularly recurring form of infectious disease; this work analyses its economic effects. Like many other infectious diseases influenza pandemics are usually of short, sharp duration. Human coronavirus is a less regularly recurring infectious disease. The human coronavirus pandemic of 2019 (COVID-19) has presented with seemingly high transmissibility and led to extraordinary socioeconomic disruption due to severe preventative measures by governments. To understand and compare these events, epidemiological and economic models are linked to capture the transmission of a pandemic from regional populations to regional economies and then across regional economies. In contrast to past pandemics, COVID-19 is likely to be of longer duration and more severe in its economic effects given the greater uncertainty surrounding its nature. The analysis indicates how economies are likely to be affected due to the risk-modifying behaviour in the form of preventative measures taken in response to the latest novel pandemic virus.     

COVID-19: Forecasting confirmed cases and deaths with a simple time series model

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.     

Epidemiological versus meteorological forecasts: Best practice for linking models to policymaking

Weather forecasts, climate change projections, and epidemiological predictions all represent domains that are using forecast data to take early action for risk management. However, the methods and applications of the modeling efforts in each of these three fields have been developed and applied with little cross-fertilization. This perspective identifies best practices in each domain that can be adopted by the others, which can be used to inform each field separately as well as to facilitate a more effective combined use for the management of compound and evolving risks. In light of increased attention to predictive modeling during the COVID-19 pandemic, we identify three major areas that all three of these modeling fields should prioritize for future investment and improvement: (1) decision support, (2) conveying uncertainty, and (3) capturing vulnerability.     

Hawkes process modeling of COVID-19 with mobility leading indicators and spatial covariates

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.     

Coronavirus: Short-term pain but strong rebound likely

• ▪ Attention has understandably focused on limiting the damage from the short-term effects of the coronavirus outbreak. But it's likely that, once disruption and uncertainty fade, the rebound in activity will be strong. It's important for firms to position themselves for such a recovery.
• ▪ Historical evidence supports this view. In the past 200 years, short recessions have typically been followed by robust recovery. Long-term impacts from natural disasters have generally only been evident for specific hazards. Except for AIDS, longer-term pandemic effects also appear to have been contained.
• ▪ Surveys during the 2003 SARS and 2009 influenza outbreaks highlight one explanation for time-limited impacts. Public fears increased alongside rising infection rates, but they dissipated promptly as outbreaks came under control.
• ▪ Our modelling is consistent with these stylised facts. In our coronavirus pandemic scenario, global growth grinds to a halt in Q2 2020 but then rebounds to a rapid pace within a year. With much of the initial output loss recovered in a relatively short period of time, long-term impacts are limited.
• ▪ But there are risks to this view. The period of disruption could be longer than anticipated, depending on the potential spread and seasonality of COVID-19 and policy actions to mitigate the fallout. Opinion polls also highlight the potential risk of larger, more persistent effects for some countries.
• ▪ Moreover, coronavirus-related weakness and associated financial distress could expose other key vulnerabilities - for example related to deteriorating corporate sector balance sheets and fragile trade relations. These would be expected to have persistent effects on global activity over the coming years

     

Comparing the accuracy of several network-based COVID-19 prediction algorithms

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.     

A hybrid system-identification method for forecasting telecommunications product demands

A crucial challenge for telecommunications companies is how to forecast changes in demand for specific products over the next 6 to 18 months—the length of a typical short-range capacity-planning and capital-budgeting planning horizon. The problem is especially acute when only short histories of product sales are available. This paper presents a new two-level approach to forecasting demand from short-term data. The lower of the two levels consists of adaptive system-identification algorithms borrowed from signal processing, especially, Hidden Markov Model (HMM) methods [Hidden Markov Models: Estimation and Control (1995) Springer Verlag]. Although they have primarily been used in engineering applications such as automated speech recognition and seismic data processing, HMM techniques also appear to be very promising for predicting probabilities of individual customer behaviors from relatively short samples of recent product-purchasing histories. The upper level of our approach applies a classification tree algorithm to combine information from the lower-level forecasting algorithms. In contrast to other forecast-combination algorithms, such as weighted averaging or Bayesian aggregation formulas, the classification tree approach exploits high-order interactions among error patterns from different predictive systems. It creates a hybrid, forecasting algorithm that out-performs any of the individual algorithms on which it is based. This tree-based approach to hybridizing forecasts provides a new, general way to combine and improve individual forecasts, whether or not they are based on HMM algorithms. The paper concludes with the results of validation tests. These show the power of HMM methods to forecast what individual customers are likely to do next. They also show the gain from classification tree post-processing of the predictions from lower-level forecasts. In essence, these techniques enhance the limited techniques available for new product forecasting.     

Implementation of the SutteARIMA method to predict short-term cases of stock market and COVID-19 pandemic in USA

Quality & Quantity - The objective of this study is to compare the different methods which are effective in predicting data of the short-term effect of COVID-19 confirmed cases and DJI closed...     

Creative reconstruction: a structured literature review of the early empirical research on the COVID-19 crisis and entrepreneurship

Management Review Quarterly - The COVID-19 crisis triggered by the novel coronavirus (SARS-CoV-2) and the infection control measures taken have extended beyond affecting health issues to impact...     

Impact of the COVID-19 outbreak and its related announcements on the Chinese conventional and Islamic stocks’ connectedness

In this paper, we assess the impacts of the COVID-19 counts (infected cases, deaths and recovered) and related announcements on the Islamic and conventional stocks interplays in the Chinese market. We test whether Islamic stocks are perceived as assets providing diversification benefits in time of COVID-19 pandemic. Doing so, we implement a multivariate GJR-GARCH model under dynamic conditional correlation (DCC) as well as multiple and partial wavelet coherence methods to recent Chinese daily data ranging from 2 December 2019 to 8 May 2020 and COVID-19 related announcement for the period. Our results from multivariate GJR-GARCH models reveal that COVID-19 infected cases and deaths do impact mean DCCs between Islamic and conventional stocks, number of recovered do not have such impact, while none of the above have any significant impact on the DCCs fluctuations. However, when we analyze the impact of COVID-19 related announcement on the variation of conditional correlation between two stocks (i.e. DCC volatility) our findings show that 7 out of 10 such announcements (mainly those with serious health treats or economic implications) do effect those volatilities in Chinese equity market. The empirical findings from partial and multiple wavelet coherences provide robust evidence of instability in the co-movement between Islamic and conventional indexes for different scales and over dissimilar sub-periods. Indeed, the weakening of co-movements is especially notable in the very short and short-run where operating the short-term investors. Our empirical findings offer several key propositions for policy makers and portfolio managers in China with broad implications applicable to other markets.     

Analyzing the emotional impact of COVID-19 with Twitter data: Lessons from a B-VAR analysis on Italy

The novel coronavirus 2019 revolutionized the way of living and the communication of people making social media a popular tool to express concerns and perceptions. Starting from this context we built an original database based on the Twitter users’ emotions shown in the early weeks of the pandemic in Italy. Specifically, using a single index we measured the feelings of four groups of stakeholders (journalists, people, doctors, and politicians), in three groups of Italian regions (0,1,2), grouped according to the impact of the COVID-19 crises as defined by the Conte Government Ministerial Decree (8^th March 2020). We then applied B-VAR techniques to analyze the sentiment relationships between the groups of stakeholders in every Region Groups. Results show a high influence of doctors at the beginning of the epidemic in the Group that includes most of Italian regions (Group 0), and in Lombardy that has been the region of Italy hit the most by the pandemic (Group 2). Our outcomes suggest that, given the role played by stakeholders and the COVID-19 magnitude, health policy interventions based on communication strategies may be used as best practices to develop regional mitigation plans for the containment and contrast of epidemiological emergencies.     

新冠肺炎疫情下管理层业绩预告归因行为研究

基于A股上市公司2020年1—4季度业绩预告归因的研究发现,公司将业绩变化归因于新冠疫情的概率呈现随季度下降的趋势,坏消息归因于新冠疫情的概率显著高于好消息,且这种差异未随季度发生显著变化。业绩预告归因反映了管理层自利性归因与新冠疫情的叠加影响,且在特定情境下投资者能够识别自利性归因。研究结论从微观经济主体和利益相关者感知的角度展示了中国抗击新冠疫情的成效,也为监管部门规范重大外生冲击下的信息披露操纵行为提供了借鉴。     

Guest Editorial: A purchasing and supply management view of supply resilience for better crisis response

Purchasing and supply management (PSM) has been under great pressure since the COVID-19 pandemic first shook the world. Companies and public organizations faced new kinds of supply disruptions, and at a scale never seen before. New response abilities were required from PSM to address these challenges and disruptions. This Editorial introduces four articles in the Special Issue on “PSM learning from the pandemic: transforming for better crisis management.” These empirical contributions show how companies could build resilience to survive and be competitive during the COVID-19 pandemic. This Editorial discusses how supply resilience should be conceptualized in post-pandemic supply chains adopting a PSM perspective. We suggest that supply resilience practices should be developed and planned according to whether they strengthen existing supply chain relationships (bridging) or establish new ones (buffering) and whether they are short-term (temporary) or long-term (permanent) orientated. Furthermore, three supply resilience capabilities, absorbing, responding and capitalizing, should be prioritized in supply chains for responding to and recovering from global crises and disruptions. Supply resilience is key to crisis response and recovery, and PSM has an essential role in building and sustaining that resilience.     

Herding behavior in the cryptocurrency market during COVID-19 pandemic: The role of media coverage

This paper examines herding behavior in the cryptocurrency market during the COVID-19 pandemic using daily data and based on static and regime-switching models. Furthermore, we investigate whether herding behavior is affected by the coronavirus media coverage. Based on a sample of the top-43 cryptocurrencies in terms of market capitalization between 2013 and 2020, we find significant evidence of herding for the entire sample period only during high volatility state. Moreover, during the COVID-19 crisis, results suggest that investors in the cryptocurrency market follow the consensus. Finally, the impact of coronavirus media coverage is significant on herding among investors, explaining such behavior in the cryptocurrency market during the COVID-19 crisis. Our findings explain herding determinants that may help investors avoid such comportment, mainly during the crisis.     

Impacts of COVID-19 on global stock sectors: Evidence from time-varying connectedness and asymmetric nexus analysis

In this paper, we analyze the impact of the COVID-19 crisis on global stock sectors from two perspectives. First, to measure the effect of the COVID-19 on the volatility connectedness among global stock sectors in the time–frequency domain, we combine the time-varying connectedness and frequency connectedness method and focus on the total, directional, and net connectedness. The empirical results indicate a dramatic rise in the total connectedness among the global stock sectors following the outbreak of COVID-19. However, the high level of the total connectedness lasted only about two months, representing that the impact of COVID-19 is significant but not durable. Furthermore, we observe that the directional and net connectedness changes of different stock sectors during the COVID-19 pandemic are heterogeneous, and the diverse possible driving factors. In addition, the transmission of spillovers among sectors is driven mainly by the high-frequency component (short-term spillovers) during the full sample time. However, the effects of the COVID-19 outbreak also persisted in the long term. Second, we explore how the changing COVID-19 pandemic intensity (represented by the daily new COVID-19 confirmed cases and the daily new COVID-19 death cases worldwide) affect the daily returns of the global stock sectors by using the Quantile-on-Quantile Regression (QQR) methodology of Sim and Zhou (2015). The results indicate the different characteristics in responses of the stock sectors to the pandemic intensity. Specifically, most sectors are severely impacted by the COVID-19. In contrast, some sectors (Necessary Consume and Medical & Health) that are least affected by the COVID-19 pandemic (especially in the milder stage of the COVID-19 pandemic) are those that are related to the provision of goods and services which can be considered as necessities and substitutes. These results also hold after several robustness checks. Our findings may help understand the sectoral dynamics in the global stock market and provide significant implications for portfolio managers, investors, and government agencies in times of highly stressful events like the COVID-19 crisis.     

Learning,disagreement and inflation forecasting

This paper studies inflation forecasting based on the Bayesian learning algorithm which simultaneously learns about parameters and state variables. The Bayesian learning method updates posterior beliefs with accumulating information from inflation and disagreement about expected inflation from the Survey of Professional Forecasters (SPF). The empirical results show that Bayesian learning helps refine inflation forecasts at all horizons over time. Incorporating a Student’s t innovation improves the accuracy of long-term inflation forecasts. Including disagreement has an effect on refining short-term inflation density forecasts. Furthermore, there is strong evidence supporting a positive correlation between disagreement and trend inflation uncertainty. Our findings are helpful for policymakers when they forecast the future and make forward-looking decisions.     

Forecasting for COVID-19 has failed

Epidemic forecasting has a dubious track-record, and its failures became more prominent with COVID-19. Poor data input, wrong modeling assumptions, high sensitivity of estimates, lack of incorporation of epidemiological features, poor past evidence on effects of available interventions, lack of transparency, errors, lack of determinacy, consideration of only one or a few dimensions of the problem at hand, lack of expertise in crucial disciplines, groupthink and bandwagon effects, and selective reporting are some of the causes of these failures. Nevertheless, epidemic forecasting is unlikely to be abandoned. Some (but not all) of these problems can be fixed. Careful modeling of predictive distributions rather than focusing on point estimates, considering multiple dimensions of impact, and continuously reappraising models based on their validated performance may help. If extreme values are considered, extremes should be considered for the consequences of multiple dimensions of impact so as to continuously calibrate predictive insights and decision-making. When major decisions (e.g. draconian lockdowns) are based on forecasts, the harms (in terms of health, economy, and society at large) and the asymmetry of risks need to be approached in a holistic fashion, considering the totality of the evidence.     

The COVID-19 shock and challenges for inflation modelling

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.     

On the forecasting performance of a small-scale DSGE model

Dynamic stochastic general equilibrium (DSGE) models have recently become standard tools for policy analysis. Nevertheless, their forecasting properties have still barely been explored. In this article, we address this problem by examining the quality of forecasts of the key U.S. economic variables: the three-month Treasury bill yield, the GDP growth rate and GDP price index inflation, from a small-size DSGE model, trivariate vector autoregression (VAR) models and the Philadelphia Fed Survey of Professional Forecasters (SPF). The ex post forecast errors are evaluated on the basis of the data from the period 1994–2006. We apply the Philadelphia Fed “Real-Time Data Set for Macroeconomists” to ensure that the data used in estimating the DSGE and VAR models was comparable to the information available to the SPF.Overall, the results are mixed. When comparing the root mean squared errors for some forecast horizons, it appears that the DSGE model outperforms the other methods in forecasting the GDP growth rate. However, this characteristic turned out to be statistically insignificant. Most of the SPF's forecasts of GDP price index inflation and the short-term interest rate are better than those from the DSGE and VAR models.