Nowcasting for Decision Making in Crisis Times
(5 minute read)
Recent crises in the global socio-economic landscape are throwing fresh light on next generation nowcasting techniques. This article explains how nowcasting techniques can be adapted for decision making during times of crises.
During times of socio-economic crises, the ability to gather vast amounts of information quickly and make rapid decisions to respond to evolving situations, is critical for decision makers. In the past, overreliance on typical economic data, often published with months of time lag, proved unreliable to respond to fast changing situations. Dotcom crashes, financial recession of 2008 all led to institutions adding nowcasting to their decision-making toolboxes.
Nowcasting uses complex econometric techniques, contemporaneous data sets from a broad range of sources, and enables dynamic planning and decision-making during times of crises. However, the rapidity, scale, and enormity of global crises in the past few years – Brexit, US-China trade standoffs, and Covid-19 – have all challenged the stability of macroeconomic structures, and with them, the reliability of nowcasting.
Nowcasting models have become extremely complex with the variety and scale of data sources, assumptions, variables, and relationships. To ensure that nowcasting remains more reliable, McKinsey research suggests reducing number of variables, examining more high-frequency variables, setting robust KPIs that would provide consistent views and reliability in statistical models. This enables easy interpretation of estimates, understanding structural breaks, and provision of real-time information.
Real-Time Views
Nowcasting, beyond traditional forecasting, helps institutions to understand current, recent past, and future situations even without the help of formal published economic indicators. They help overcome challenges of time lag in published data, providing companies with timely intelligence, and facilitating accurate predictions of recoveries when traditional models have failed.
Despite this, early generation nowcasting is challenged by growing complexity of crises and the number of variables accounted for when relationships between multiple variables break down in times of great stress. Crises impact on countries and regions differently and comparable effects can be hard to calculate with accuracy in an increasingly interconnected world.
Next-Generation Nowcasting
McKinsey recommends a modified approach to nowcasting that builds on country- and industry-specific expertise to whittle down variables to a few select, high-frequency ones, which interrelationships will remain relatively stable even in times of major crisis. The revised model will be more flexible and robust in response to exogenous shocks, will support economically intuitive outcomes, consider high-frequency data, and offer timely economic insights.
Organisations can use up-to-date country and sector wide information produced by this type of model to support decision making and recovery strategies. Nowcasting provides government institutions with real-time insights into economic trends, informs revenue planning and cash flow and proactively supports recovery packages for affected sectors. Financial institutions can benefit from economic impact assessment underpinning investment opportunities, sectoral trends, lending strategies etc. from nowcasting models. Industrial businesses can build on nowcasting to ascertain timely consumer demand views and sector evolution, impacting on their production and sales strategies.
With the right set of variables, KPIs, and timely data, next-generation nowcasting provides a variety of benefits to a range of organisations helping them to cope with rapid, macro-structural crises efficiently.
