High-Frequency Data & COVID-19: Investing for the Next Phase of the Cycle
Investors and financial markets are forward-looking. They are constantly trying to anticipate the future and how to position for it. In a COVID-19 world, where economic uncertainty and financial market conditions are at extremes, that’s no easy task. A key question at times like this is: which models and data sets may be best equipped to inform our decision making?
Macroeconomic data alone isn’t a prescient guide because it is lagged. Relying on backward-looking data to make forward-looking decisions is like driving by looking through the rearview mirror. In recent years, there has been tremendous progress in harnessing important alternative data insights through machine learning models. Insights from high-frequency data such as live traffic, hotel and restaurant reservations, and mobility can help investors analyze the road ahead.
For example, in order to understand how the economic and sector impacts of the virus might play out in the United States, it can be important to track developments in countries whose cases peaked ahead of the US.
Chart 1 below shows that Wuhan, the first epicenter of the COVID-19 outbreak, has been limping back to normal after the city’s lockdown was eased in the first week of April. Chart 2 shows that work and commerce have returned to pre-COVID-19 levels in Beijing and other major Chinese cities following more than three months of restrictions. However, there’s a noticeable pattern to the Beijing chart: a sharp drop in activity during lunch hours and after 6:00 PM. The country remains closed for entertainment. It may be months before people feel comfortable stepping out for non-essential activities. This doesn't bode well for hotels, restaurants, cafes, or sports venues, which are all likely to feel pain for a long time.
