Leading forecasting experts believe that big data and non-linear models are the primary challenges in macroeconomic forecasting in central banks right now.
Rafael Domenech, head of economic analysis of BBVA Research and professor of economics at the University of Valencia, retweeted an article shared by J Julian Cubero, lead economist of global scenarios at the BBVA Research, on macroeconomic forecasting being a challenging task after pandemics like the Covid-19 economic shock.
Researchers, as result, are discussing new advancements in methods and approaches, particularly inspired by machine learning (ML) techniques and big data, to forecast during abnormal times like the pandemic. The 11th Conference on Forecasting Techniques highlighted that the single most likely future outcome of GDP growth is not enough to understand the prospect and nature of extreme events, but that it will be key for central bankers consider factors such as its distribution, its tails, and how it could change extreme events like floods, pandemic or the Covid-19 crisis.
Forecasting experts also believe that pure model-based forecasts combined with off-model information by professional experts can also improve the forecasting performance in extreme scenarios like the Covid-19 shock, while policies should pay more attention to possible tail outcomes.
Resumen de la última conferencia bienal del #ECB sobre técnicas de previsión, “Forecasting in abnormal times”. No basta con la previsión máximo verosímil del PIB, hay que pensar en su distribución, sus colas, y en cómo cambiaría con eventos extremos https://t.co/7MVDulaaVc
— J. Julian Cubero (@juliancubero) October 31, 2021
Jonathan Portes, professor of economics and public policy at the School of Politics and Economics of King’s College, London and a senior fellow at the UK in a Changing Europe, shared an article on Brexit most likely to damage the economy long into the future. Portes believes that despite a waning Covid-19 pandemic, the impact of Brexit is going to weigh down UK’s economy for a longer period than expected.
The Office for National Statistics described the impact of the Covid-19 pandemic on the UK economy as the largest decline in economic output since 1709. However, the Office for Budget Responsibility (OBR) has confirmed that the long-term impact of Brexit on the economy is going to be twice as great as Covid-19. The agency believes that Brexit will reduce UK’s economic productivity, causing the country’s GDP to drop by 4%, while the impact of the pandemic on the GDP is likely to be just 2%.
Portes also highlights that the OBR’s model may have excluded the impact of the Covid-19 crisis on education, especially for poorer kids, which will leave permanent scars. Meanwhile, efforts to liberalise migration to non-European migrants is expected to offset some damage caused by Brexit.
As I explained here, the OBR *reduced* its estimate of the long-term impact of the pandemic, while maintaining its estimate (more than twice as large) of the long-term impact of covid. So it's mostly *Brexit* that requires higher taxes..https://t.co/I8gMagwm9O
— Jonathan Portes (@jdportes) October 31, 2021
Adam Tooze, professor at Columbia University and director of the European Institute, shared an article on sectoral inflation dispersion during the Covid-19 pandemic largely due to smaller and shorter-lived fluctuations in food, fuel, and housing prices.
Economists believe that the global recovery from the Covid-19 recession continues, but the pace has weakened and uncertainties have increased. The global toll for Covid-19 deaths caused by the highly transmissible Delta variant has reached up to 5 million, with growing risks of further mutations of the virus.
Additionally, supply chain disruptions have been further feeding inflation across countries, with policy trade-offs becoming more complex and economic prospects faltering amid supply and labour shortages and price rises. As a result, economists have revised the global growth forecast for 2021 to 5.9%, while that of 2022 remains unchanged at 4.9%.
Sectoral inflation dispersion during the pandemic does not stand out by historical
standards. This is largely due to smaller and shorter-lived swings in fuel, food, and housing prices.https://t.co/Pzzmp5qOdv pic.twitter.com/NAlgV794cR
— Adam Tooze (@adam_tooze) October 31, 2021