Why Are Illiquid Households Affected More by Inflation?

Inflation affects households by reducing the real value of their nominal assets and nominal liabilities. In a previous blog post, we showed that many U.S. households’ balance sheets are “nominal maturity mismatched,” which means that their nominal assets usually have shorter maturities than their nominal liabilities. As a result, unexpected inflation causes households to lose in the short term as financial assets (e.g., checking deposits) lose real value; some households, however, may offset part of those losses with long-run gains produced by a drop in the value of their liabilities (e.g., mortgages).

For households who are “liquid”—that is, those who have plenty of cash or can easily borrow—this maturity mismatch, and the resulting gains and losses, is irrelevant. They can either tap into their savings or borrow to cover the effects of short-term losses and then later replenish their savings or repay the debt with long-term gains. For example, a sudden burst of inflation might spur a household to use a credit card to help finance expenses in the short run, with the expectation that a nominal wage hike in the future will allow the household to pay off that debt.

However, in practice, most households face at least some degree of “illiquidity”; that is, they might be short on cash or find it difficult to borrow money. In this case, the short-run losses these households suffer can outweigh their long-run gains, leading to a liquidity problem associated with inflation, as documented in a 2024 St. Louis Fed working paper by Yu-Ting Chiang and Ezra Karger.See Yu-Ting Chiang and Ezra Karger, “Nominal Maturity Mismatch and the Liquidity Cost of Inflation,” Federal Reserve Bank of St. Louis Working Paper 2024-031A, September 2024. In this blog post, we summarize the working paper’s findings about how illiquid are U.S. households and how illiquidity changes our understanding of the effects of unexpected inflation.

How “Illiquid” Are U.S. Households?

To measure how illiquid households are, we separate out two sources of difficulty that households face when borrowing money: the cost of borrowing and a limit on borrowing.

First, households typically pay a “spread” when borrowing. This is the difference between the interest rate at which a household can borrow and the rate of return that they would receive on an investment with a low default risk, such as a U.S. government bond. This spread is positive, since lending to households involves more risk.

This spread can make borrowing expensive for households. To estimate the borrowing spread, we calculated each household’s borrowing cost on their credit card, since credit cards are a common source of short-term borrowing.Our data on household balance sheets come from the Federal Reserve’s triennial Survey of Consumer Finances for 2019. We looked at two questions from the SCF: “What interest rate do you pay on the card where you have the largest balance?” and “After the last payment was made, what was the total balance still owed?” For households that hold credit card debt and pay interest on it, we calculated the spread between the interest rate each household pays on their credit card debt and the U.S. Treasury yield. Households not paying interest on credit card debt are considered to have some other cheaper form of borrowing or the ability to tap into savings instead. We then took the weighted average of the spread among all households in a specific wealth decile.

Second, and perhaps more importantly, sometimes households may face a limit on what others are willing to lend to them. Even if they might be willing to pay a higher interest rate, no one wants to lend these households additional money; credit cards have credit score requirements and spending limits, and other types of loans often require proof of employment, income or car title, which not all households have.

Determining a “Shadow Rate”

Even if the household does not actually borrow, is important to know the rate at which these households would borrow if they could find a lender for additional funds. Since we cannot directly observe this, we refer to this implied rate as the “shadow rate.” We can, however, infer the shadow rate from the household’s consumption behavior. For example, imagine that a household is given some cash. How much they decide to spend (as opposed to save) represents how much more they would consume if they were able to borrow more. The more cash a household spends, the more they want to consume and the higher the rate they would be willing to pay to borrow.

Applying this concept to real-world data, we measured how much households spent upon receiving their COVID-19 stimulus payments.We used data from Facteus to estimate household consumption responses following the 2020-2021 federal stimulus payments. Controlling for differences in age and income, we quantify changes in households’ consumption following the stimulus check relative to their typical consumption level.

To convert these consumption changes to a measure of interest rates, we adopt a common assumption that there needs to be a 2 percentage point decrease in interest rate for the average household to increase consumption by 1 percentage point. Therefore, for a household that increases consumption by 5 percentage points upon a cash transfer, we infer that the household would have been willing to pay an interest rate 10 percentage points higher to borrow. This gives us the shadow rate.

The figure below shows both components of illiquidity (the spreads and shadow rates) from those with the least wealth (the first decile) to those with the most wealth (the 10th decile). Again, the shadow rate is based on our calculations of the change in consumption for each individual household as a result of the COVID-19 stimulus payments, and then averaged across deciles.

Average Spreads and Shadow Rates by Wealth Decile

SOURCE: Chiang and Karger (2024).

This measure of illiquidity decreases with wealth (i.e., net worth), starting from 21% at the bottom wealth decile and going up to 5.5% at the top decile. Additionally, the relative size of the spread and shadow rate changes across the distribution. If we just looked at spreads, we might conclude that the least-wealthy households face less illiquidity. However, many households with the lowest net worth do not have credit cards and cannot easily borrow using standard financial services. This inability to borrow is represented by high shadow rates that they face.

The Liquidity Problem of Unexpected Inflation

As we showed in our previous blog post, an unexpected inflation shock like the one experienced in 2021-2022 results in short-term losses (as assets like labor income lose real value) and long-term gains (as liabilities like mortgages lose real value). For many households, this may mean that current labor income can no longer cover their daily grocery expenses, even though they may see long-term gains when the real value of their mortgage payment falls and their nominal wages go up as the labor market responds to the inflation. Some households could simply borrow money to cover their short-term loss in income and then repay in the future using their long-term gain from lower real liabilities. However, this may not be possible for households that experience illiquidity, producing short-run losses that outweigh the long-term gains.

To understand how a liquidity problem can alter households’ experience during an inflation episode, consider a household that sees a $1,000 loss in the real value of nominal assets this year and a $1,000 gain next year because of the reduced real value of their nominal liabilities.

For a liquid household, these gains and losses cancel out, and the household experiences no net loss due to the inflation episode, as they can simply reduce their savings today to make up for the loss. However, consider an illiquid household in the bottom wealth decile who is, in fact, willing to borrow at a 20% rate. For this household, the $1,000 next year is worth only about $800 today. As a result, this household experiences a loss of $200 due to illiquidity during the inflation episode.

The Effect of Illiquidity on Welfare

To show the effect of illiquidity, we combined the illiquidity results from the first figure with the direct change in wealth from the inflation shock, which makes both assets and liabilities lose value. (The calculation of the direct changes in wealth are outlined in our previous blog post.)

The result is the second figure below, which shows combinations of all the different welfare effects from an unexpected inflation shock by wealth decile. We first separated out the effects operating through the “wealth channel,” which is just the change in overall wealth from the inflation shock, not factoring in any information about household illiquidity. We then combined the wealth channel with the effects from different types of illiquidity: spreads, shadow rates, and both spreads and shadow rates.

Effect on Welfare from Unexpected Inflation, Averaged within Each Wealth Decile

SOURCE: Chiang and Karger (2024).

When accounting for both types of illiquidity, the least-wealthy households suffer the greatest welfare loss, equivalent to around 0.7% of their lifetime wealth. The wealthiest households see a smaller loss of about 0.2%. Crucially, when combining the effects of both types of illiquidity, every decile loses from unexpected inflation.

However, the sources of welfare losses vary across the wealth distribution. For the wealthiest households, the losses almost entirely come from the wealth channel; for the least-wealthy households, however, the illiquidity effect is particularly important, accounting for more than three-quarters of the losses. This means that when inflation strikes unexpectedly, those with the smallest net worth sustain the sharpest loss because of their inability to borrow money.

In conclusion, while some households gain and some lose wealth because of unexpected inflation alone, the average household in each wealth decile loses when one accounts for illiquidity, which means that short-term losses cannot be offset by long-term gains. In particular, the least-wealthy households face the most difficulty borrowing and thus are affected the most from surprise inflation.

Notes

1. See Yu-Ting Chiang and Ezra Karger, “Nominal Maturity Mismatch and the Liquidity Cost of Inflation,” Federal Reserve Bank of St. Louis Working Paper 2024-031A, September 2024.
2. Our data on household balance sheets come from the Federal Reserve’s triennial Survey of Consumer Finances for 2019. We looked at two questions from the SCF: “What interest rate do you pay on the card where you have the largest balance?” and “After the last payment was made, what was the total balance still owed?” For households that hold credit card debt and pay interest on it, we calculated the spread between the interest rate each household pays on their credit card debt and the U.S. Treasury yield. Households not paying interest on credit card debt are considered to have some other cheaper form of borrowing or the ability to tap into savings instead. We then took the weighted average of the spread among all households in a specific wealth decile.
3. We used data from Facteus to estimate household consumption responses following the 2020-2021 federal stimulus payments.