Improving social harm indices with a modulated Hawkes process

Communities are affected adversely by a range of social harm events, such as crime, traffic crashes, medical emergencies, and drug use. The police, fire, health and social service departments are tasked with mitigating such social harm through various types of interventions. While various different social harm indices have been proposed for allocating resources to spatially fixed hotspots, the risk of social harm events is dynamic, and new algorithms and software systems that are capable of quickly identifying risks and triggering appropriate public safety responses are needed. We propose a novel modulated Hawkes process for this purpose that offers flexible approaches to both (i) the incorporation of spatial covariates and leading indicators for variance reduction in the case of rarer event categories, and (ii) the capture of dynamic hotspot formation through self-excitation. We present an efficient l1-penalized EM algorithm for estimating the model that performs feature selection for the spatial covariates of each incident type simultaneously. We provide simulation results using data from the Indianapolis Metropolitan Police Department in order to illustrate the advantages of the modulated Hawkes process model of social harm over various recently introduced social harm indices and property crime Hawkes processes.     

Do cryptocurrencies provide better hedging? Evidence from major equity markets during COVID-19 pandemic

Using the five-minute interval price data of two cryptocurrencies and eight stock market indices, we examine the risk spillover and hedging effectiveness between these two assets. Our approach provides a comparative assessment encompassing the pre-COVID-19 and COVID-19 sample periods. We employ copula models to assess the dependence and risk spillover from Bitcoin and Ethereum to stock market returns during both the pre-COVID-19 and COVID-19 periods. Notably, the COVID-19 pandemic has increased the risk spillover from Bitcoin and Ethereum to stock market returns. The findings vis-à-vis portfolio weights and hedge effectiveness highlight hedging gains; however, optimal investments in Bitcoin and Ethereum have reduced during the COVID-19 pandemic, while the cost of hedging has increased during this period. The findings also confirm that cryptocurrencies cannot provide incremental gains by hedging stock market risk during the COVID-19 pandemic.     

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.     

Economic fundamentals,policy responses,and state-level municipal bond sensitivity to COVID-19 prevalence

This paper conducts a state-by-state analysis of the financial impact of the COVID-19 pandemic on the U.S. municipal bond market. Using panel regressions and state-by-state regressions, we find that the prevalence rates of the COVID-19 virus negatively impacted the aggregate performance of municipal bonds. The study also explored whether the disparities in the economic fundamentals of U.S. states, as well as the COVID-19 mitigation policies employed by each state, can explain the sensitivity of the state’s municipal bond to its COVID-19 prevalence rates. States with more desirable economic fundamentals and robust COVID-19 mitigation policies appeared to have higher COVID-19 sensitivities than states that do not. This finding may be due to a baseline effect, in which the higher levels of economic activities in these states also make them more susceptible to the deleterious effects of the stronger mitigation policies enacted by them.     

Modeling frailty-correlated defaults using many macroeconomic covariates

We propose a novel time series panel data framework for estimating and forecasting time-varying corporate default rates subject to observed and unobserved risk factors. In an empirical application for a U.S. dataset, we find a large and significant role for a dynamic frailty component even after controlling for more than 80% of the variation in more than 100 macro-financial covariates and other standard risk factors. We emphasize the need for a latent component to prevent a downward bias in estimated default rate volatility and in estimated probabilities of extreme default losses on portfolios of U.S. debt. The latent factor does not substitute for a single omitted macroeconomic variable. We argue that it captures different omitted effects at different times. We also provide empirical evidence that default and business cycle conditions partly depend on different processes. In an out-of-sample forecasting study for point-in-time default probabilities, we obtain mean absolute error reductions of more than forty percent when compared to models with observed risk factors only. The forecasts are relatively more accurate when default conditions diverge from aggregate macroeconomic conditions.     

How have the dependence structures between stock markets and economic factors changed during the COVID-19 pandemic?

This study investigates how the dependence structures between stock markets and economic factors have changed during the COVID-19 pandemic using the dynamic model averaging approach. A series of economic factors such as commodity markets, cryptocurrency, monetary policy, international capital flows, and market uncertainty indices are considered. We find that the importance of economic variables and the sign and size of their coefficients are significantly different from those before the COVID-19 pandemic. The stock markets are most influenced by economic factors during the COVID-19 outbreak.     

Regional heterogeneity and U.S. presidential elections: Real-time 2020 forecasts and evaluation

This paper exploits cross-sectional variation at the level of U.S. counties to generate real-time forecasts for the 2020 U.S. presidential election. The forecasting models are trained on data covering the period 2000–2016, using high-dimensional variable selection techniques. Our county-based approach contrasts the literature that focuses on national and state level data but uses longer time periods to train their models. The paper reports forecasts of popular and electoral college vote outcomes and provides a detailed ex-post evaluation of the forecasts released in real time before the election. It is shown that all of these forecasts outperform autoregressive benchmarks. A pooled national model using One-Covariate-at-a-time-Multiple-Testing (OCMT) variable selection significantly outperformed all models in forecasting the U.S. mainland national vote share and electoral college outcomes (forecasting 236 electoral votes for the Republican party compared to 232 realized). This paper also shows that key determinants of voting outcomes at the county level include incumbency effects, unemployment, poverty, educational attainment, house price changes, and international competitiveness. The results are also supportive of myopic voting: economic fluctuations realized a few months before the election tend to be more powerful predictors of voting outcomes than their long-horizon analogs.     

An empirical analysis: Did green supply chain management alleviate the effects of COVID-19?

Supply chain management played a central role during the COVID-19 crisis, as the outbreak of the pandemic disrupted the majority of all global supply chains. This paper tests whether companies that use green supply chain management (GSCM) practices benefited from a buffer effect in the context of COVID-19. Our empirical analysis, conducted on a sample of U.S. companies, shows that GSCM companies experienced less negative abnormal stock returns during the crisis. This result contributes to the literature on financial impact of GSCM, finding that GSCM is perceived as an effective risk management tool and can serve as an effective drug against COVID-19 crisis. Our paper also contributes to the business debate on the role of green supply chains in the post-COVID19 world.     

Improving forecasting performance using covariate-dependent copula models

Copulas provide an attractive approach to the construction of multivariate distributions with flexible marginal distributions and different forms of dependences. Of particular importance in many areas is the possibility of forecasting the tail-dependences explicitly. Most of the available approaches are only able to estimate tail-dependences and correlations via nuisance parameters, and cannot be used for either interpretation or forecasting. We propose a general Bayesian approach for modeling and forecasting tail-dependences and correlations as explicit functions of covariates, with the aim of improving the copula forecasting performance. The proposed covariate-dependent copula model also allows for Bayesian variable selection from among the covariates of the marginal models, as well as the copula density. The copulas that we study include the Joe-Clayton copula, the Clayton copula, the Gumbel copula and the Student’s $t$-copula. Posterior inference is carried out using an efficient MCMC simulation method. Our approach is applied to both simulated data and the S&P 100 and S&P 600 stock indices. The forecasting performance of the proposed approach is compared with those of other modeling strategies based on log predictive scores. A value-at-risk evaluation is also performed for the model comparisons.     

How do stock price indices absorb the COVID-19 pandemic shocks?

To assess the resiliency of stock price indices during the COVID-19 crisis, this study provides a distinctive perspective; that is, we evaluate the ability of stock price indices to absorb COVID-19 shocks. We construct the measures of absorptive intensity and duration to identify a stock price’s absorptive capacity. We then employ the Granger causality test and a topology network approach to investigate the interactions of absorptivity among stock price indices. Our results show that stock price absorptivity varies over time and across countries and industries. The US and the Brazil stock indices have relatively high absorptive intensity while short duration. The health care industry shows distinctive trend in absorptive intensity from the other industries. The intensity of the non-cyclical industries such as utilities and consumer staples is high, while the cyclical industries such as banking, real estate, and energy have lower absorptive intensity. Moreover, the utilities, consumer staples, and financials industries are the main resiliency transmitters.     

Decomposing hospital productivity changes, 1985–1988: A nonparametric Malmquist approach

Annual data on U.S. hospitals from 1985–1988 are evaluated by ownership type—profit, nonprofit, state and local government, and U.S. Department of Veterans Affairs (VA)—for changes in hospital productivity over time. Distance functions are used to measure Malmquist indices of productivity change, which are then decomposed into indices of efficiency change and technology change. In contrast to previous studies using this approach, we allow for variable returns to scale and use both input and output orientations. We find that changes in technology dominate changes in inefficiency in determining changes in productivity.     

Calibrating longitudinal models of residential mobility and migration: an assessment of a non-parametric marginal likelihood approach

"This paper discusses the problems of controlling for omitted variables in estimating the structural parameters of longitudinal models and focuses upon an assessment of a non-parametric marginal maximum likelihood approach suggested by the results of Laird....The approach is shown to be statistically valid for a plausible discrete-time model of the incidence of residential or migration moves, at least for data in which no household moves in every time period. Empirical evaluation with two large [U.S.] datasets on residential mobility indicates that the approach is also computationally feasible and provides a promising alternative to more conventional methods for controlling for omitted variables."     

Leveraging multi-tier healthcare facility network simulations for capacity planning in a pandemic

The COVID-19 pandemic has placed severe demands on healthcare facilities across the world, and in several countries, makeshift COVID-19 centres have been operationalised to handle patient overflow. In developing countries such as India, the public healthcare system (PHS) is organised as a hierarchical network with patient flows from lower-tier primary health centres (PHC) to mid-tier community health centres (CHC) and downstream to district hospitals (DH). In this study, we demonstrate how a network-based modelling and simulation approach utilising generic modelling principles can (a) quantify the extent to which the existing facilities in the PHS can effectively cope with the forecasted COVID-19 caseload; and (b) inform decisions on capacity at makeshift COVID-19 Care Centres (CCC) to handle patient overflows. We apply the approach to an empirical study of a local PHS comprising ten PHCs, three CHCs, one DH and one makeshift CCC. Our work demonstrates how the generic modelling approach finds extensive use in the development of simulations of multi-tier facility networks that may contain multiple instances of generic simulation models of facilities at each network tier. Further, our work demonstrates how multi-tier healthcare facility network simulations can be leveraged for capacity planning in health crises.     

A comparison of monthly global indicators for forecasting growth

This paper evaluates the predictive content of a set of alternative monthly indicators of global economic activity for nowcasting and forecasting quarterly world real GDP growth using mixed-frequency models. It shows that a recently proposed indicator that covers multiple dimensions of the global economy consistently produces substantial improvements in forecasting accuracy, while other monthly measures have more mixed success. Specifically, the best-performing model yields impressive gains with MSPE reductions of up to 34% at short horizons and up to 13% at long horizons relative to an autoregressive benchmark. The global economic conditions indicator also contains valuable information for assessing the current and future state of the economy for a set of individual countries and groups of countries. This indicator is used to track the evolution of the nowcasts for the U.S., the OECD area, and the world economy during the COVID-19 pandemic and the main factors that drive the nowcasts are quantified.     

Mixed-frequency machine learning: Nowcasting and backcasting weekly initial claims with daily internet search volume data

We propose an out-of-sample prediction approach that combines unrestricted mixed-data sampling with machine learning (mixed-frequency machine learning, MFML). We use the MFML approach to generate a sequence of nowcasts and backcasts of weekly unemployment insurance initial claims based on a rich trove of daily Google Trends search volume data for terms related to unemployment. The predictions are based on linear models estimated via the LASSO and elastic net, nonlinear models based on artificial neural networks, and ensembles of linear and nonlinear models. Nowcasts and backcasts of weekly initial claims based on models that incorporate the information in the daily Google Trends search volume data substantially outperform those based on models that ignore the information. Predictive accuracy increases as the nowcasts and backcasts include more recent daily Google Trends data. The relevance of daily Google Trends data for predicting weekly initial claims is strongly linked to the COVID-19 crisis.     

Latent social distancing: Identification,causes and consequences

It is not directly observable how effectively a society practices social distancing during the COVID-19 pandemic. This paper proposes a novel and robust methodology to identify latent social distancing at the country level. We extend the Susceptible-Exposed-Infectious-Recovered-Deceased (SEIRD) model with a time-varying, country-specific distancing term, and derive the Model-Inferred DIStancing index (MIDIS) for 120 countries using readily available epidemiological data. The index is not sensitive to measurement errors in epidemiological data and to the values assigned to model parameters. The evolution of MIDIS shows that countries exhibit diverse patterns of distancing during the first wave of the COVID-19 pandemic—a persistent increase, a trendless fluctuation, and an inverted U are among these patterns. We then implement regression analyses using MIDIS and obtain the following results: First, MIDIS is strongly correlated with available mobility statistics, at least for high income countries. Second, MIDIS is also strongly associated with (i) the stringency of lockdown measures (governmental response), (ii) the cumulative number of deceased persons (behavioral response), and (iii) the time that passed since the first confirmed case (temporal response). Third, there is statistically significant regional variation in MIDIS, and more developed societies achieve higher distancing levels. Finally, MIDIS is used to explain output losses experienced during the pandemic, and it is shown that there is a robust positive relationship between the two, with sizable economic effects.     

Analysis of the impact of COVID-19 pandemic on G20 stock markets

We examine the impact of the COVID-19 pandemic on G20 stock markets from multiple perspectives. To measure the impact of COVID-19 on cross-market linkages and deeply explore the dynamic evolution of risk transmission relations and paths among G20 stock markets, we statically and dynamically measure total, net, and pairwise volatility connectedness among G20 stock markets based on the DY approach by Diebold and Yilmaz (2012, 2014). The results indicate that the total volatility connectedness among G20 stock markets increases significantly during the COVID-19 crisis, moreover, the volatility connectedness display dynamic evolution characteristics during different periods of the COVID-19 pandemic. Besides, we also find that the developed markets are the main spillover transmitters while the emerging markets are the main spillover receivers. Furthermore, to capture the impact of COVID-19 on the volatility spillovers of G20 stock markets, we individually apply the spatial econometrics methods to analyze both the direct and indirect effects of COVID-19 on the stock markets’ volatility spillovers based on the “volatility spillover network matrix” innovatively constructed in this paper. The empirical results suggest that stock markets react more strongly to the COVID-19 confirmed cases and cured cases than the death cases. In general, our study offers some reference for both the investors and policymakers to understand the impact of COVID-19 on global stock markets.     

Forecasting unemployment insurance claims in realtime with Google Trends

Leveraging the increasing availability of ”big data” to inform forecasts of labor market activity is an active, yet challenging, area of research. Often, the primary difficulty is finding credible ways with which to consistently identify key elasticities necessary for prediction. To illustrate, we utilize a state-level event-study focused on the costliest hurricanes to hit the U.S. mainland since 2004 in order to estimate the elasticity of initial unemployment insurance (UI) claims with respect to search intensity, as measured by Google Trends. We show that our hurricane-driven Google Trends elasticity leads to superior real-time forecasts of initial UI claims relative to other commonly used models. Our approach is also amenable to forecasting both at the state and national levels, and is shown to be well-calibrated in its assessment of the level of uncertainty for its out-of-sample predictions during the Covid-19 pandemic.     

Happiness sentiments and the prediction of cross-border country exchange-traded fund returns

This research examines whether social media (Twitter) happiness sentiment and country-level happiness sentiment indices predict cross-border ETF returns. To account for complicated associations between happiness sentiment and ETF returns, we use a quantile regression approach and find that Twitter and trading market (U.S.) happiness sentiments are strong predictors of future ETF returns, for which both have far greater predictive power than those of their home countries. Home country happiness indices exhibit asymmetric impacts across quantiles, suggesting the importance of trading country (U.S.) and Twitter happiness sentiments. Higher U.S. and home countries’ freedom to make life choices, absence of corruption perception, and confidence in national government precede higher ETF returns, while U.S. GDP, social support, health life expectancy, positive affect, and negative affect precede lower (abnormal) returns. We find that higher return quantile country ETFs provide a safe haven for U.S. investors during a U.S. bear market.     

The ‘COVID’ crash of the 2020 U.S. Stock market

We employed the log-periodic power law singularity (LPPLS) methodology to systematically investigate the 2020 stock market crash in the U.S. equities sectors with different levels of total market capitalizations through four major U.S. stock market indexes, including the Wilshire 5000 Total Market index, the S&P 500 index, the S&P MidCap 400 index, and the Russell 2000 index, representing the stocks overall, the large capitalization stocks, the middle capitalization stocks and the small capitalization stocks, respectively. During the 2020 U.S. stock market crash, all four indexes lost more than a third of their values within five weeks, while both the middle capitalization stocks and the small capitalization stocks have suffered much greater losses than the large capitalization stocks and stocks overall. Our results indicate that the price trajectories of these four stock market indexes prior to the 2020 stock market crash have clearly featured the obvious LPPLS bubble pattern and were indeed in a positive bubble regime. Contrary to the popular belief that the 2020 US stock market crash was mainly due to the COVID-19 pandemic, we have shown that COVID merely served as sparks and the 2020 U.S. stock market crash had stemmed from the increasingly systemic instability of the stock market itself. We also performed the complementary post-mortem analysis of the 2020 U.S. stock market crash. Our analyses indicate that the probability density distributions of the critical time for these four indexes are positively skewed; the 2020 U.S. stock market crash originated from a bubble that had begun to form as early as September 2018; and the bubble profiles for stocks with different levels of total market capitalizations have distinct temporal patterns. This study not only sheds new light on the makings of the 2020 U.S. stock market crash but also creates a novel pipeline for future real-time crash detection and mechanism dissection of any financial market and/or economic index.