Perspectives

Americans still have $2.6 trillion in “excess savings”

What is most notable about the “super-spike” in the personal savings rate is that it leaves American households with a massive financial cushion of $2.6 trillion in “excess savings.” That can be computed simply by noting that personal savings from March 2020 to May 2022 averaged $204 billion a month, or double the 2019 average level of $103 billion a month (Chart 2). On a cumulative basis, that monthly gap over a 26-month

Excess savings of that magnitude represents an important wild card for the economic outlook. Historically, American recessions have typically involved a peak-to-trough decline in GDP of about 2%. Those swings in GDP have rightly been the subject of much market excitement and anxiety. So to have “dry powder” of 11% of GDP of excess savings resident on consumer balance sheets should induce humility among investors and analysts about key questions like if, when, and how those savings will impact future consumer behavior.

Economists at Goldman Sachs observed that excess savings built up during World War II represent the closest available analog in modern history.¹ In both the US and the UK savings rates spiked during the war before undershooting their long-term averages in the late 1940s (Chart 4). The Goldman economists also estimated that 9% – 12% of the excess savings were spent in the first year after the war in the US and the UK, respectively.

Whether that represents a good model of how America’s pandemic-era excess savings will be spent is an open question. To be sure, the potential deployment of pent-up savings could provide a positive  tailwind to consumer spending in coming quarters. As a matter of simple arithmetic, even if 10% of the $2.6 trillion is spent in the next year, that would represent a contribution to growth of about 1.1% of GDP.

Indeed, Bloomberg’s survey of economists projects robust real growth in US consumption spending of 3.2% for this year. And strong household balance sheets, along with rapid wage growth, are seen as key supports for the consumer in the face of significant headwinds like higher food and energy prices and rising interest rates.

The boost from pent-up saving seems to be unfolding gradually. As shown in Chart 3, our estimate of America’s excess savings declined to $2.58 trillion as of May 2022 from a peak of $2.68 trillion last December. In our view, there are good reasons to believe that, despite its massive size, the savings cushion is unlikely to be a source of excess demand or persistent inflationary pressure going forward. This reflects both the composition of the savings cushion and the influence of the Federal Reserve’s policies that are likely to prevent further material declines in the savings rate.

How are the excess savings held?

If all of the excess savings accumulated during the pandemic has already been deployed in illiquid assets like real estate or into long-term holdings in stocks or bonds, the implications for consumer spending would be muted. That’s because statistical models on consumer spending generally find that households consume only a few cents per dollar out of their wealth.² Likewise, if excess savings resides mainly with  high-income households, the implications for consumer spending would be muted. That’s because high-income households tend to have a lower marginal propensity to consume than lower-income households.

We find that most of the excess savings have accumulated in highly liquid assets like currency, deposits, and money market funds, which make them inherently easy to spend. That said, the excess savings are highly concentrated among high-income households, which may imply a more muted impact on consumer spending.

Although there is no direct way to map the data on personal savings into how the savings were deployed, we can look at the Fed’s data on household balance sheets and see a huge buildup of liquid assets that occurred over the course of the pandemic. For example, households in the top 1% of the income distribution saw their liquid assets surge from $700 billion in December 2019 to $2.0 trillion by March 2022 (Chart 5). Measured against the long-term real trend, that represented excess liquid assets at the end of last year of $950 billion. Using the same method across other income groups, we observe the following data as of March 2022 (Chart 6):

• Total excess liquid assets came to $2.6 trillion ($20k per household)
• The top 1% of the income distribution held nearly 37% of the total ($725k per household)
• The bottom 20% held almost none of the total ($30 per household)

Using the Fed’s balance sheet data confirms that America’s savings cushion — defined in terms of “excess liquid assets” — was a massive $2.6 trillion at the end of last year. That is right in line with the $2.6 trillion estimate for May 2022 based on the monthly savings data. This data suggests that the savings cushion is almost entirely in highly liquid assets, which make them inherently easy to spend.

That said, the Fed’s data suggests that the cushion is an almost entirely non-existent $30 for low-income households who tend to have a higher marginal propensity to consume. In fact, US households are now spending an estimated $5,000 a year on gasoline, up from $2,800 a year ago.3 So, any savings cushion that existed for low-income households is likely long gone due to soaring prices at the pump. Of course, higher fuel prices are less of an issue for top 1% households with average excess cash balances of $725,000 — or for the next 19% of the top 20% with average excess cash balances of roughly $35,000. But as the main group of asset owners in the nation, the top 20% of income earners are likely to view their cash balances as a small part of their overall asset allocation.

Indeed, looking beyond the top 20%, American households’ $2.6 trillion in excess savings represent less than 2% of their collective net worth, which stood at $149.3 trillion as of March 2022 (Chart 7). Will they deploy that cash by briskly buying goods and services, which could fuel persistent inflation? Or will they decide to put it to work in stock, bonds, or real estate assets? That comes down to millions of savings and asset allocation decisions that will be influenced crucially by interest rates and asset prices.

Key drivers of the savings rate: stock wealth, property wealth, and bond yields

This is where the Fed enters the picture. Given its current focus on fighting inflation, its job is to make sure that the massive excess liquidity that currently resides with American households does not get spent too rapidly. Put simply, when the Fed’s focus was on boosting the economy it wanted the savings rate to fall rapidly, fueling brisk consumer spending and creating jobs. Now, it needs consumer spending and the savings rate to stabilize so that the overall level of demand for goods and services is better calibrated with the economy’s productive capacity. It can do so by making sure that interest rates and asset prices are set at levels so that households keep their money in cash or other assets, rather than dipping into savings and fueling further demand for goods and services.

The nature of its challenge is evident from a simple statistical model of household savings behavior developed by the San Francisco Federal Reserve Bank in 2005. Their model showed that about 90% of the variance in the US personal savings rate from 1960 to 2005 could be tracked by three explanatory variables: (1) the ratio of household stock market wealth to personal disposable income, (2) the ratio of household residential property wealth to personal disposable income, and (3) the yield on 10-year US Treasury bonds.

Not surprisingly, they found that higher bond yields tend to raise the savings rate, as do lower levels of stock market and residential property wealth. The history of those explanatory factors is shown in Chart 8. It shows that even with this year’s bear market, our estimated mid-year ratios of stock market and property wealth remain toward the high end of their historic ranges (97th and 95th percentiles respectively since 1960). In addition, 10-year Treasury yields remain quite low (19th percentile) relative to history even factoring in the recent rise in yields to 3.0%.

An update of the San Francisco Fed’s model through the second quarter of 2022 is shown in Chart 9, which plots the actual US personal savings rate together with the fitted savings rate from the model. The model suggests that the recent configuration of bond yields and asset prices was consistent with a first-quarter savings rate of negative 1.2% compared to the actual rate of 5.6%. For the second quarter, the model estimates a savings rate of 0.1%, reflecting the influence of higher bond yields and substantially lower equity prices.

Clearly, the model did not predict the extremely elevated savings rate associated with the pandemic since that had unprecedented effects on household and government behavior. Indeed, the ongoing impact of the pandemic on household behavior may help explain why the saving rate remains higher than what the model predicts.

However — and take this with a large grain of salt — the model does suggest the following: As the health situation normalizes, there could be a tendency for the savings rate to decline to record low levels in response to still elevated asset price ratios and historically low bond yields. That would be an unwelcome development for a Fed focused on fighting inflation.

The $7 trillion question: will “cash on the sidelines” stay on the sidelines?

Referring back to Chart 7, we can see that American households as of March 2022 were sitting on $7.2 trillion of cash. Of that, we suggested earlier that $2.6 trillion — or 11% of GDP — could be considered “excess cash” relative to households’ normal need for real cash balances. The chart also shows that overall household cash as a percent of net worth hit a post-1960 record of 4.8%. That is an elevated level normally associated with recession periods when investor sentiment is highly negative. In such periods, cash is prized for its option value — i.e., as dry powder for taking advantage of future lower prices.

In short, there truly is a lot of “cash on the sidelines.” The Fed’s job now is to make sure it stays on the sidelines by offering investors attractive bond yields and asset prices to keep it from fueling a consumption boom. How far it has to raise rates and how fast it has to shrink its balance sheet to do so remains to be seen. But some sympathy for the complexity of their task is in order. There is no playbook for exiting from a pandemic-induced super-spike in savings, so data-driven and flexible policy improvisation is likely to continue.

William P. Sterling, Ph.D.