Yesterday I gave a virtual talk at JSM 2021 on Visualizing Economic and Financial Market Trends in Volatile Times.
Here is a summary of my remarks.
The COVID-19 pandemic and associated economic slowdown has led to unprecedented volatility in many economic and financial time series. With swings well outside of historical ranges, many forecasting models break down. Data visualization techniques are a powerful tool to help the analyst understand evolving economic and financial market trends.
In this talk I will discuss various data visualization methods that help track economic and financial market trends in real time. I will give examples of economic and financial market data, showing how traditional visualizations may falter when faced with extreme values, and explore how alternative visualizations may give new insights.
We will also discuss how these visualizations can assist in modeling time series. Using the example of weekly home purchase mortgage applications in United States, we will consider how activity contracted sharply in the spring of 2020 but then rebounded over summer and fall. We’ll then consider how a traditional univariate forecasting model can interact with alternative data visualization techniques to help the analyst understand evolving trends in real time.
Below are the slides I used, here is a link (pdf). The animated gif on slide 3 doens’t work in the pdf, but you can find a link to my original tweet about it in the pdf and more on this blog at my post on 30 sigma.
Also in the session:
Check out Nicholas Tierney’s presentation on the R broglar package.
Earo Wang on the tsibble pacakge.
Priyanga Dilini Talagala’s talk on the R mask package.