In a speech last April1, Fed Chair Nominee, Kevin Warsh, lamented the mission creep of the Federal Reserve as well as the Fed’s expanded balance sheet. Many academics agree with Warsh – the Fed has often tried to address problems that are not within its remit and with tools not suited to their purpose. Some of these endeavors have inflated the balance sheet which, rather like a Swiss army knife, is a tool used for many tasks, none of which it does particularly well.
I generally agree with Warsh’s criticism. While I don’t believe that the much larger Fed balance sheet of recent years has been the cause of inflation or government overspending, it has distorted capital markets. That being said, the size and composition of the balance sheet is mostly a reaction to decisions made elsewhere in government – over-regulation in some areas, under-regulation in others and a general lack of fiscal discipline.
For Fed policy-makers, it is important to work with the other side of Washington to address these issues before shrinking the balance sheet. For investors, it is important to recognize that this probably won’t happen. Consequently, regardless of the aspirations of its new Chairman, assuming he is confirmed, the Fed is unlikely to make much progress in reducing its balance sheet, leaving in place its distortive effects on long-term interest rates, the slope of the yield curve and credit spreads.
The Balance Sheet before the Global Financial Crisis
The first thing to appreciate about the Fed’s balance sheet is that both the asset and liability sides have important economic impacts. One way to understand these impacts is to look at how the balance sheet has changed since the Global Financial Crisis (GFC).
Twenty years ago, in March of 2006, the assets of the Federal Reserve system amounted to $870 billion, or just over 6% of GDP. There was nothing unusual about this number – despite many economic cycles and financial convulsions, Fed assets had largely remained in a narrow band of 4% to 6% of GDP for decades.
The details of the balance sheet were also unremarkable and reflected the Fed’s primary focus on setting interest rates to achieve low and stable inflation and maximum employment while facilitating the smooth functioning of commercial banking and transactions within the economy.
In March of 2006, on the asset side of the ledger, they held roughly $760 billion in U.S. Treasuries, almost 60% of which had a maturity of less than a year. Among their other assets was gold valued at $11 billion. However, it’s worth noting that this was a book entry, pricing gold at $42.22 per ounce as mandated by Congress in 1973. Even 20 years ago, with gold priced at $580 an ounce, this stockpile was actually worth $150 billion. It is worth $1.2 trillion today.
On the other side of the balance sheet, the single biggest liability was $790 billion in paper currency in circulation. This is, at first glance an extraordinary number – roughly $2,600 for every person living in the United States. However, it should be noted that much of this cash was held outside the United States and presumably the vast majority of these banknotes were held in safety-deposit boxes rather than wallets.
In addition, in March of 2006, the Treasury Department held just $4 billion in its general account at the Federal Reserve, employing relatively tight cash management techniques to align daily outlays with revenues as much as possible.
Finally, on the liability side, commercial banks held just $11 billion in reserves. These reserve balances, combined with cash the banks held in their own vaults, constituted total reserves from a regulatory perspective, and, given reserve requirements, limited the creation of deposits. Each week, commercial banks would have to ensure they held sufficient reserves to meet these requirements and so would lend and borrow reserves amongst themselves at the federal funds rate, thereby establishing a basic short-term interest rate for the economy. This was a “scarce reserves” regime with banks holding as little as possible at the Fed since reserves paid zero interest.
Under this system, the Federal Reserve controlled the federal funds rate through open market operations, sometimes selling or buying T-bills, but more normally using repos to increase or reduce banking system reserves.
Lastly, it’s worth noting that the Fed made a healthy profit on its balance sheet, since it paid no interest on its largest liabilities and earned interest on its Treasury portfolio. These profits were remitted back to the Treasury department.
The Growth in the Balance Sheet
The metamorphosis of this modest balance sheet into today’s giant started in 2006 when Congress approved a measure allowing the Fed to pay interest on reserves, with an effective date of October 2011. When the financial crisis exploded, Congress brought this date forward to October 2008.
This turned out to be important in the months and years that followed.
As the financial crisis enveloped global markets, the Fed acted to shore up the shell-shocked mortgage market and the economy by adding billions of dollars in Treasuries and mortgage-backed securities to the asset side of its balance sheet. In theory, the reserves created to purchase these assets could have led to explosive growth in the money supply and inflation. The key to avoiding this, was for the banking system to hold these reserves at the Fed rather than lending them out. Doing so was also desirable since there was a perceived need to increase bank capitalization without adding further stress to the economy or the banking system.
Initially, in the deeply recessionary environment that followed the financial crisis, there was little danger of inflation getting out of hand and the commercial banks, which, in truth, were not at the heart of the crisis, were well able to weather loan losses. Moreover, as the banks became subject to new and overlapping requirements relating to capital ratios, leverage ratios and liquidity ratios, as well as new stress tests, they had a greater need to hold reserves.
The recovery from the GFC was painfully slow, leading the Fed to engage in multiple rounds of quantitative easing, buying both Treasuries and mortgage-backed securities in an attempt to lower long-term interest rates and mortgage spreads. It could do this without ballooning the money supply by continuing to pay interest on reserves, essentially converting a supply of short-term assets from the banking system into a long-term demand for Treasuries and mortgages.
In addition, after the GFC and, particularly, during and after the pandemic, the Treasury Department started carrying much larger balances in its general account at the Fed. This provided Treasury with a buffer of cash for the numerous occasions in recent years when Congress has failed to raise the debt ceiling. It also had the impact of quietly funneling money back into the Fed once rising short-term rates had turned its annual profit into a deficit, since the Fed paid no interest to the Treasury on its general account balances but could generate interest through the securities it bought to offset these balances.
Finally, even in an era of increasingly electronic transactions, the demand for cash continued to rise.
Notably, the payment of interest on reserves effectively changed the management of short-term interest rates from a “scarce reserves” regime to an “ample reserves” regime, with the interest rate paid on reserves serving, in theory, as a floor for the federal funds rate, regardless of the reserves in the system. In practice, the federal funds rate has normally hovered just below this, being set, effectively, by financial institutions, such as money market funds and government sponsored enterprises that are ineligible for interest on reserves and can only receive interest from the Fed at the overnight reverse repo rate.
This new procedure ran into problems in the fall of 2019 when the Fed tried to reduce the size of its balance sheet and overnight repo rates spiked, despite total bank reserves at the Fed still equaling $1.5 trillion. This highlighted the reality that, given the complexity of bank regulation, the Fed now cannot be sure when “ample” reserves cease to be ample in the eyes of commercial banks.
This episode ended the first round of quantitative tightening and the Fed’s balance sheet saw tremendous expansion during and after the pandemic recession, peaking in April of 2022 at $9.015 trillion in assets. Among the assets were $5.8 trillion in Treasuries and $2.7 trillion in mortgage-backed securities. The liability side included $2.3 trillion in currency, $543 billion held by the Treasury in its general account and $3.8 trillion in reserves.
Since then, the Federal Reserve has once again tried to shrink its balance sheet by allowing some maturing Treasuries and mortgage-backed securities to roll off, with a particular emphasis on moving away from holdings of mortgage securities. However, by late 2025, it was clear that less than abundant reserves were once again causing volatility in overnight repo markets, leading the Fed to wind down quantitative tightening.
As a result, by March of 2026, the balance sheet is smaller than it was four years ago but still contains $4.4 trillion in Treasuries and $2.0 trillion in mortgage-backed securities on the asset side and $2.4 trillion in cash, $875 billion in the Treasury general account and $3.0 trillion in reserves on the liability side.
The Economic Effects of an Expanded Balance Sheet
Contrary to the assertion of some critics, the Fed’s surging balance sheet hasn’t caused runaway government deficits or inflation over the years.
First, on deficits, while the Fed holds a substantial $4.4 trillion in Treasuries today, this represents 14% of federal debt in the hands of the public which is actually lower than the Fed share of federal debt 20 years ago. Moreover, if the Fed had not increased its Treasury holdings, its liabilities would also have been lower and the private sector would likely have funded more of the deficit outside of the Federal Reserve system at similar overall interest rates. The reason for high and rising U.S. federal deficits is that American voters elect politicians who are willing to pass legislation to cut taxes and raise spending without paying for it.
Similarly, a larger balance sheet has not raised inflation. Over the past 20 years, the monetary base has increased at an average pace of 10% per year. However, because banks have held much higher reserves at the Fed, the M2 money supply has risen by just 6.2% per year. Furthermore, the velocity of money, (that is, how often each dollar is used in the economy), has fallen sharply, so that over the same period, nominal GDP rose by just 4.3% per year. Finally, once real economic growth is taken into account, the broadest measure of inflation in the economy, the GDP deflator, has risen at a moderate annual pace of 2.3%. The truth is that this has generally been a period of low inflation, with the exception of the early 2020s when inflation spiked due to the pandemic, stimulus checks and the war in Ukraine.
That being said, the balance sheet has had some distortive effects. Of the Fed’s $4.4 trillion in Treasuries, $2.0 trillion have a maturity of over 5 years which probably exerts some downward pressure on long-term interest rates relative to short rates. It’s not clear that this serves any real public service – after all, much of American business needs to be financed on a shorter-term basis and savers perhaps deserve a better reward for locking their money away for the long-run than the compensation provided by the very shallow upward slope on the yield curve.
More seriously, the $2 trillion the Fed holds in mortgage-backed securities represents an implicit subsidy of home loans relative to all other forms of borrowing in the private sector. This is surely sub-optimal, since subsidizing education, R&D and business fixed investment would provide better gains in productivity and living standards in the long run.
Moreover, the advantage of lower mortgage rates accrues mainly to existing homeowners, while low rates over the last decade led directly to the explosion in home prices which put homeownership out of reach for many younger households, thereby worsening inequality.
Beyond these issues, the size of the Fed’s stockpile of assets is relatively harmless and, while it’s probably a good idea to shrink it over time, this can’t be achieved without making some tough choices regarding the liability side of the balance sheet.
First, there probably isn’t much to be gained from reducing the stockpile of circulating currency. While some of this is undoubtedly used for money laundering, much of it is just used as a convenient store of value in countries with runaway domestic inflation. Moreover, the willingness of foreigners to hold U.S. banknotes, interest free, implicitly subsidizes the Federal Reserve and, by extension, the U.S. federal government.
The Treasury department ought to be able to run a much smaller and more stable general account at the Fed. However, real stability would require Congress to permanently eliminate the debt ceiling, refrain from government shutdowns, and change tax laws far less frequently.
And commercial banks shouldn’t feel the need to hold quite so much in reserves at the Fed. However, getting there would require Congress to simplify the extraordinarily complicated overlapping regulations with which the commercial banking system is now governed.
In reality, none of this is likely to happen.
In a television interview last summer, Kevin Warsh argued for both a reduction in short-term interest rates and a shrinking of the Fed’s balance sheet. Given the inflationary impact of rising oil prices, even with a new Fed chair, lower short-term rates now look unlikely before the end of the year. However, given the political forces supporting cheap mortgage finance, heavy bank regulation and an unwillingness to bring stability to budget decisions, the wait for a smaller Fed balance sheet will likely be even longer.
This should hold both short-term and long-term interest rates in a narrow range going forward, suggesting that, while fixed income can still provide portfolios with income and stability, investors will need to venture elsewhere to achieve strong total returns.