The Big Idea
Balance sheet and jobs lift consumer spending
Stephen Stanley | September 15, 2023
This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors. This material does not constitute research.
The latest snapshot of the US consumer balance sheet shows continuing strength. Debt burdens seem manageable. And households seem awash in liquidity. Combine that with a robust job market that continues to kick off substantial real advances in wage and salary income, and real consumer spending should remain on a solid uptrend.
Debt loads are light
A good place to start is the Federal Reserve’s latest Financial Accounts of the United States and the picture it paints of the overall level of household debt. As a percentage of GDP, household debt was at its lowest reading in over 20 years at the end of the second quarter (Exhibit 1). The series has fallen well below the trend line established from 1970 through 2000.
Exhibit 1: Household debt to-GDP-ratio
A similar way to assess the scale of household debt is to compare it to disposable income (Exhibit 2). This gauge is slightly less impressive, as the ratio has inched modestly above pre-pandemic levels. However, it has come down in each of the past four quarters and is currently at a lower level than at any time from 2002 to 2018.
Exhibit 2: Household debt-to-disposable-income ratio
Debt service burden
In a separate report, the Federal Reserve publishes estimates of the debt service burden, as defined by the percentage of disposable income needed to stay current on debt payments. The Fed makes two sets of calculations. The narrow one is called the debt service ratio and only covers traditional debt. The financial obligations ratio takes a broader view, including rent payments, auto lease payments, homeowners’ insurance, and property tax payments.
With interest rates surging since early 2022, one might imagine that even for a constant level of debt, the servicing burden of that borrowing would have risen sharply. However, the Fed data, which unfortunately only extends through the first quarter for now—a quarterly update is due any day—show that both measures have been roughly steady on balance over the past year and are historically low. In fact, aside from the pandemic period, both gauges are the lowest on record going back to 1980 (Exhibit 3). This is a testament to the prevalence of fixed-rate debt held by households, most notably for mortgages.
Exhibit 3: Debt service burden
Household assets
The Fed’s Financial Accounts data for the second quarter tell a very happy story for households. The value of household assets surged by $5.7 trillion in the quarter, as home prices rebounded and equity valuations increased noticeably. By comparison, over the prior four quarters combined, household assets had declined by a total of $2.8 trillion. Total household net worth jumped to an all-time high of $154.3 trillion through June.
Household liquid assets
Now we turn to the lynchpin of my analysis of household finances since the beginning of the pandemic, liquid assets. This series covers the portion of the balance sheet that represents cash equivalents and includes currency, bank deposits, and money market fund shares. I have emphasized this measure is a proxy for spendable funds.
Household liquid assets spiked during the pandemic, reflecting in large part the unprecedented waves of federal government largesse. Most economists presumed that consumers would spend down those balances quickly once the economy fully reopened last year. However, liquid assets remain elevated by historical standards (Exhibit 4). In the second quarter, the gauge fell by less than $100 billion. After surging by almost $5 trillion from the end of 2019 through the peak in early 2022, the series has come off by only about $550 billion over the most recent five quarters.
Exhibit 4: Household liquid assets
To be fair, a significant portion of the increase in spending power represented by these liquid assets has been eroded by inflation. The level of prices, as measured by the PCE deflator is up by roughly 15% from the end of 2019. Real household liquid assets, using the PCE deflator as the price index, has declined more than the nominal balance (Exhibit 5).
Exhibit 5: Real household liquid assets
This series, in my view, offers a more realistic view of the cushion that consumers still have left over from the pandemic. In the decade prior to the pandemic, the real liquid assets series rose by an average of about $290 billion a year. Applying that trend forward and comparing this trend line to the actual readings, the real Covid liquid assets bonanza peaked at the end of 2021 at about $2.6 trillion—that is, the actual reading was $2.6 trillion above the extrapolated trend line. As I noted a quarter ago, the cushion had shrunk by the end of the first quarter to about $1 trillion. The increment declined further in the second quarter to about $880 billion.
In my view, this is a far better way of assessing households’ liquidity cushion than the model-based estimates that the vast majority of economists have employed. These models are highly sensitive to the assumptions that underlie the calculations, which is why various economists have come up with wildly varying conclusions using quite similar methodologies. I fail to understand why economists are so insistent on crafting models based on the savings rate to approximate household liquidity when the Fed puts out hard data, erasing the need to concoct estimates. I would argue strongly that actual data, unless its accuracy is in question, always trump model-based estimates.
In any case, the real liquid assets series suggests that households had burned through about two-thirds of their cushion by the middle of this year. The good news is that the incremental decline in the second quarter was relatively small, only $120 billion. I believe this reflects the fact that household incomes have advanced robustly in recent quarters. Even after falling in July, real disposable income, which fell in 2021 and 2022, is up by an annualized 4.8% in the first seven months of this year, more than sufficient to support a robust gain in real consumer spending. Thus, households likely failed to draw heavily from their accumulated savings in recent months because they did not need to, relying on their solid real income gains to pay for their outlays.
To underscore that households remain flush with liquid assets, look at the component of household liquid assets that is the most liquid – checkable deposits and currency (Exhibit 6). This series ran mostly between $1.1 trillion and $1.3 trillion in the five years prior to the pandemic. It spiked to well over $4.5 trillion at its peak last year. Since then, it has receded only marginally, to $4.4 trillion as of June 30.
Exhibit 6: Household checkable deposits and currency
The Bank of America Institute offers corroborating evidence that bank deposits remain elevated. Their internal data on the Bank of America client base show that through August, median household savings and checking balances were still close to 50% higher than the 2019 benchmark levels. While the figure has been slowly coming down since early this year, the recent trend suggests that balances are going to remain far above pre-pandemic levels for the foreseeable future. It is also noteworthy that the Institute’s finding applies across all income levels, i.e., both low- and high-income households have median balances that are close to 50% higher than in 2019.
Nominal and real
As an aside, it is easy to get tripped up by comparing the levels over time of nominal data in a period of high inflation. After 30 years of low inflation, most analysts have gotten out of the habit of thinking in real terms. This has led to a number of faulty conclusions. One example that is relevant to this discussion is the level of credit card debt. When the New York Fed reported recently that credit card balances had moved above $1 trillion for the first time ever in the second quarter, many observers quickly asserted that this was evidence that households were up to their gills in debt. However, if we deflate the New York Fed credit card debt series by the PCE deflator, the real level of credit card balances was lower in the second quarter than at the end of 2019—and that is before we account for the higher levels of GDP, income and so on in 2023 compared to 2019—far from evidence of obvious trouble.
Conclusion
Household balance sheets remained in extraordinarily good shape through mid-2023. The Covid cushion has shrunk in real terms but remains sizable. In any case, households have not needed to fall back on their accumulated savings heavily so far this year because their real incomes have finally expanded after declining in 2021 and 2022. As usual, the key for the consumer going forward is likely to be the health of the labor market. As long as a robust job market continues to kick off substantial real advances in wage and salary income, real consumer spending should remain on a solid uptrend.