On Thursday the Bureau of Labor Statistics reported that initial claims for unemployment insurance in the third week of March came in at an astounding seasonally adjusted 3.283 million workers, far eclipsing the previous peaks of 660 thousand workers seen in September 1982 and March 2009. Over the past two years, in contrast, we had been enjoying a healthy labor market averaging only 220 thousand claims each week.

What does this mean for the broader economy and for deposits in particular?

As figure 1 shows, there is a strong correlation between the more familiar monthly unemployment rate and the weekly initial claims for unemployment insurance. A regression analysis indicates that if initial claims in the final week of March fall to “only” 2 million workers the unemployment rate would exceed 11%. If we have a repeat of the third week’s performance, the unemployment rate would hit 13%. For context, in the Great Recession unemployment peaked at only 10.0%. Thursday’s report was for the week ended March 21, near the beginning of the nationwide lockdown and just as businesses began laying off employees in earnest. Thus it is not unreasonable to expect next week’s initial claims report to be as bad as this weeks.

Screen Shot 2020-03-28 at 10.39.10 AM

Our economic model suggests that such large spikes in the unemployment rate could cause second quarter GDP to fall on the order of 13% to 15% or more at an annualized rate, far worse than the -8.4% growth seen in 2008Q4 at the depth of the Great Recession. Our analysis is in line with most other economists. Friday morning Federal Reserve Bank of Dallas president Robert Kaplan was quoted in the Wall Street Journal (link) saying that second quarter GDP could shrink by 20% with the unemployment rate peaking in the low- to mid-teens but coming down rapidly.

Whether the economy recovers quickly or takes more time to heal remains a hotly debated topic among economists. Figure 2 shows two possible scenarios. In the mild scenario the unemployment rate quickly recovers as Kaplan suggests, ending this year at 8% and falling below 5% by the end of next year. In the more severe scenario the unemployment rate peaks at nearly 16% as the initial layoffs slow the economy and force even more layoffs later in the year. In this scenario the unemployment rate recovers more slowly, in line with past recessions, and stays over 10% through the end of 2022.

Screen Shot 2020-03-27 at 4.26.38 PM

Right now we do not know which scenario will play out. Therefore it is imperative to plan for the possibility that the economy remains weak and unemployment remains high for years to come.

Figure 3 shows how checking account balances behaved around the last four recessions, based on data reported in the Federal Reserve H.6 release.

Screen Shot 2020-03-27 at 4.32.10 PM

In general, balances hold steady after the onset of a recession, with monthly growth rates bobbing above and below zero. Most people recognize that as people are laid off from their jobs, those people will deplete their balances as they search for new work. However, people who are still employed tend to reduce consumption, reduce exposure to the stock market, and increase their own savings. Figure 3 suggests those two forces tend to cancel each other out. The large spike seen after the Great Recession began in December 2007 is apparently due to cash payments and subsequent household spending as a result of the American Recovery and Reinvestment Act signed into law in February 2009.

Figure 4 shows how consumer CD balances behaved. Here there is some evidence that CD balances decline after the onset of a recession.

Screen Shot 2020-03-27 at 4.32.49 PM

Keep in mind that deposit rates invariably fall as a recession progresses. Whether the decline in CD balances is due to household depletion of savings to get through the recession or simply due to depositors not renewing CDs because of low rates offered is difficult to determine. However, because non-maturity deposit balances do not show a fall off, we suspect the interest-rate channel dominates.

So how should your bank be thinking about deposits for the next 12 months? 

At the moment things are moving so quickly it’s hard to think about deposits beyond 30 days- but here’s how to think about your playbook right now:

Short term (<30 days): Focus on right-sizing your rate sheet. Anticipate an influx of liquid deposits as government stimulus checks begin to arrive.  Wind down specials and promotional pricing.

Mid-term (30-60 days): Develop a rate planning calendar. Don’t change all your rates all at once. Focus on front-line education and strategies to ensure clients aren’t thrown into chaos. Do a full rate review of both front-book and back-book pricing to look for opportunities to eliminate arbitrage and reduce high-cost deposits.

Long-term (60+ days): Your bank will likely be flush with deposits for the next 6 to 12 months, so now is the time to focus on margin management and on building capabilities for the next cycle. Add tools, tracking, and better customer profiling capability to track what is happening now, then use that data to your advantage when deposits are once again competitive.

If you have any questions or would like to talk further, contact us, we are here to help. For more weekly deposits insights, check out our latest Three Minute Banker post.


About Brian

Brian Poi is an economist with nearly 20 years of experience in macroeconomics, econometrics, model development and validation, and forecasting. At Nomis, Brian focuses on all things deposits, especially new product and new model development. Previously, Brian worked with many CCAR banks to develop stress-testing models and adapt those models to satisfy CECL requirements as well.