Small Data, Big Time—A Retrospect of the First Weeks of COVID-19

Abstract

Knowledge of the current state of economies, how they respond to COVID-19 mitigations and indicators, and what the future might hold for them is important. We use recently developed generalised network autoregressive (GNAR) models, using trade-determined networks, to model and forecast the Purchasing Managers' Indices for a number of countries. We use networks that link countries where the links themselves, or their weights, are determined by the degree of export trade between the countries. We extend these models to include node-specific time series exogenous variables (GNARX models), using this to incorporate COVID-19 mitigation stringency indices and COVID-19 death rates into our analysis. The highly parsimonious GNAR models considerably outperform vector autoregressive models in terms of mean-squared forecasting error and our GNARX models themselves outperform [Read before The Royal Statistical Society at the first meeting on 'Statistical Aspects of the Covid-19 Pandemic' held at the

Christine P. Chai's Contribution to the ‘First Discussion Meeting on Statistical Aspects of the Covid-19 Pandemic’

“…I totally agree with the second lesson in Zhao (2020): “Practitioners are still unfamiliar with basic statistical concepts.” Dr. Anthony Fauci mentioned statistical significance in his testimony to the US government in July 2020. It is appropriate to report experiment results, but over‐emphasising $$$ p< $$$ 0.05 is not a good idea.…”

“…Last but not least, I really like the conclusion in Zhao (2020): ‘What is important is not to find fault with others, but to learn from each others’ failures'. I hope we can all learn from the mistakes in fighting against the epidemic, so the same disaster would not repeat in the future.…”

“…The third paper by Zhao (2020) brings an interesting perspective regarding COVID‐19 from a statistician. I would take this with a grain of salt because the article was written in retrospective, and I am also concerned about potential regulations from the Chinese government.…”

Quantifying the Economic Response to COVID-19 Mitigations and Death Rates Via Forecasting Purchasing Managers' Indices Using Generalised Network Autoregressive Models with Exogenous Variables

“…Most of the epidemiological analyses of COVID-19 estimated the basic reproductive number R 0 instead of the growth exponent r. This is because R 0 can be expressed in terms of the parameters in standard compartmental epidemic models and is more relevant in forecasting the epidemic with various kinds of interventions. But from a statistical perspective, it is much easier to estimate r. Moreover, an initial doubling time of 2-3 days, which is what most countries in Europe and elsewhere experienced when the pandemic first hit (Pellis et al, 2021;Zhao, 2021) also highlights the urgency of the matter.…”

Tao Wang's Contribution to the ‘First Discussion Meeting on Statistical Aspects of the Covid-19 Pandemic’

“…Note that exogenous information, such as availability of vaccines mentioned by Qingyuan Zhao, is not used to improve efficiency of the estimate/forecast of daily infection curves. Zhao's paper (Zhao, 2021) highlights the importance of statistical thinking in the early murk of an emerging outbreak, when uncertainty abounds not just on quantitative measures of spread, but even on the mechanisms of transmission. Not stressed in the paper, but equally important, are uncertainties on the prevalence and role of a-, pre-and pauci-symptomatic transmission, of the role of children, and the severity profile of the emergent disease.…”