What Are the Characteristics of Banks with Large Unrealized Losses?

The period from March 2022 to July 2023 marks one of the fastest instances of monetary policy tightening in recent decades. The effective federal funds rate increased from less than 0.1% to over 5.3% during this period, reflecting the Federal Reserve’s response to surging inflation during the COVID-19 pandemic.

One of the consequences of the tightening was the fall in the price of outstanding fixed-income securities; these include things like corporate bonds and U.S. Treasuries. Such securities are issued on a regular basis and pay investors fixed interest payments until they mature. Why does the price of these securities fall when interest rates rise? When rates increase, newly issued securities offer investors better returns and thus become more attractive than previously issued ones. Thus, the price of these outstanding securities must fall to induce investors to purchase them.

Book Values and Market Values

Who are the investors holding and trading these securities? There are many types of such investors, and commercial banks are an important one. Banks hold fixed-income securities, both government- and privately issued, for many reasons, including liquidity and capital management and to comply with regulatory requirements. Banks often hold these securities to maturity, meaning that they record them on their balance sheets at purchase value, or “book value.” As interest rates rose in 2022-23 and the price of previously issued securities fell, market values started diverging from these book values. This should not matter much for a bank that holds the security to maturity. However, if the bank had to sell these securities, it would have to do so at a price lower than the one recorded in its balance sheet, which would force it to post or realize a loss.

This generalized decrease in prices due to rising interest rates has given rise to a large increase in “unrealized losses” in the U.S. banking system since 2022. This is a measure of how much banks stand to lose if they were to liquidate their entire security portfolio, taking into account the difference between prevailing market values and the prices at which those securities were acquired (book values). Note that we can think of unrealized gains in the same way: When interest rates fall, security prices rise, but if banks choose not to sell their securities, then those gains are not realized.

Unrealized Losses in the U.S. Banking System

The figure below shows net unrealized gains in billions of dollars for all banks from the first quarter of 2001 to the second quarter of 2025.The St. Louis Fed’s Supervisory Policy and Risk Analysis team tracks this measure, along with other key banking metrics. See the Nov. 6 blog post, “Banking Analytics: Unrealized Losses Decrease Again at U.S. Banks.” The data come from the Federal Financial Institutions Examination Council (FFIEC) via their Central Data Repository’s Public Data Distribution. When this measure becomes negative, it corresponds to unrealized losses instead of gains. Readers will notice the sharp drop in unrealized gains in 2022-23 corresponding to the tightening of monetary policy.

Aggregate Net Unrealized Gains in the U.S. Banking System

SOURCES: Federal Financial Institutions Examination Council and authors’ calculations.

NOTE: Data are quarterly.

Are Unrealized Losses Widespread among U.S. Banks?

The U.S. banking system has become increasingly concentrated, with a few banks dominating the industry. A natural follow-up question is this: Are these unrealized losses concentrated in just a few large banks, or is this a widespread phenomenon that affects large and small banks alike?

The next figure tells us that the vast majority of banks that report to the FFIEC experienced unrealized losses; the blue line shows the median net unrealized gains among said banks, while the gray band captures the middle 80% of the net unrealized gains distribution for each quarter (the lower limit of the band is thus the 10th percentile, while the upper limit is the 90th).

Notice that we are now looking at net unrealized gains as a percentage of Tier 1 capital. (Tier 1 capital is the core capital a bank holds.) We do this to put all banks on a level playing field size-wise; you might expect that larger banks will have more capital and can buy more assets, meaning they probably have larger unrealized gains than smaller banks. Dividing the unrealized gains for each bank by that bank’s Tier 1 capital stock negates this size difference, meaning for our figure below that both large and small banks saw unrealized losses in 2022-23.

Median Value of Net Unrealized Gains as a Share of Tier 1 Capital

SOURCES: Federal Financial Institutions Examination Council and authors’ calculations.

NOTES: Data are quarterly. The gray area represents the range of values for the 10th and 90th percentiles of unrealized gains as a share of Tier 1 capital.

What Are the Characteristics of Banks with Large Unrealized Losses?

This revelation prompts a new question: If a bank’s size does not affect its unrealized gains (when expressed as a percentage of the bank’s Tier 1 capital), what does? The next few figures present binned scatterplots of a bank-level measure of unrealized losses (net unrealized gains as a share of Tier 1 capital) against a series of bank-level characteristics: (1) deposits as a share of liabilities, (2) liquid assets as a share of total assets, and (3) Tier 1 capital as a share of total assets.We bin our scatterplots by first dividing the x-axis into 20 equally sized, mutually exclusive groups and placing banks into those groups based on their x-axis values; the group’s x-axis value is the midpoint of the range of grouped bank values. We then average the y-axis values within each group. The number of groups in each figure will differ because of differences in the availability of data.

The Relationship between Net Unrealized Gains as a Share of Tier 1 Capital and Deposits as a Share of Liabilities: U.S. Banks in 2025:Q2

The Relationship between Net Unrealized Gains as a Share of Tier 1 Capital and Liquid Assets as a Share of Total Assets: U.S. Banks in 2025:Q2

The Relationship between Net Unrealized Gains as a Share of Tier 1 Capital and Tier 1 Capital as a Share of Total Assets: U.S. Banks in 2025:Q2

SOURCES FOR ALL THREE FIGURES: Federal Financial Institutions Examination Council and authors’ calculations.

NOTE FOR ALL THREE FIGURES: The results of the scatterplots have been grouped into bins for ease of viewing. See Endnote 2 for details.

From these three figures, we learn that banks with more unrealized losses, relative to their Tier 1 capital, tend to have the following:

• A greater deposits-to-liabilities ratio, meaning that they rely more on deposits as a source of funding and less on other types of financing instruments, such as bonds.
• A lower liquid asset-to-asset ratio, meaning that their asset portfolios are less liquid: They may take longer to convert to cash if the bank experiences a negative funding shock (and may lose more value during conversion).
• A lower Tier 1 capital-to-assets ratio, meaning that they have a smaller capital buffer to absorb losses in their asset portfolio.

Taking Stock

Tier 1 capital as a share of assets is a standard measure of bank solvency: Banks that have more capital relative to the size of their balance sheet are perceived as being able to withstand larger negative shocks to their portfolio without risking bankruptcy. Large unrealized losses effectively reduce the size of this buffer.

Additionally, deposits can potentially be a volatile funding source, and banks that rely more on this type of funding are exposed to the risk of runs. A bank that is extremely reliant on deposits might find itself with frequent funding shortfalls, which may need to be satisfied by security sales. Thus, banks with more deposits are more likely to be forced to realize losses.

Our admittedly simple analysis shows that, according to these two metrics, banks with greater unrealized losses are also the ones who tend to exhibit worse metrics in terms of solvency and run risk. This constellation of risks can give rise to systemic risk in the banking industry, which is one reason why unrealized losses are an important metric for policymakers and researchers to track.

Notes

1. The St. Louis Fed’s Supervisory Policy and Risk Analysis team tracks this measure, along with other key banking metrics. See the Nov. 6 blog post, “Banking Analytics: Unrealized Losses Decrease Again at U.S. Banks.”
2. We bin our scatterplots by first dividing the x-axis into 20 equally sized, mutually exclusive groups and placing banks into those groups based on their x-axis values; the group’s x-axis value is the midpoint of the range of grouped bank values. We then average the y-axis values within each group. The number of groups in each figure will differ because of differences in the availability of data.