Among Federal Reserve officials and many economists, it is fashionable to argue that any losses the Federal Reserve should suffer, no matter how large, will have no operational consequence. Is this true? If so, how does the Fed account for its losses and stay solvent? And who ends up paying for these losses? As the Fed executes its strategy to reign in run-away inflation, the answers to these questions take center stage as the Fed has already experienced mark-to-market losses of epic proportions and will soon post large operating losses, something it has never faced in its 108-year history.
We estimate that, between December 31, 2021 and the end of May 31, 2022, the Federal Reserve lost $540 billion in market value on its huge portfolio of investments in Treasury bonds and mortgage securities. To put this loss in perspective, $540 billion is equivalent to 60 percent of the value of the Federal Reserve System’s entire asset holdings on September 1, 2008, just prior to the onset of the financial crisis. $540 billion is more than 13 times the Federal Reserve System’s recently reported consolidated capital of $41 billion meaning that the market value of the Fed’s outstanding liabilities—primarily member bank reserves and Federal Reserve notes—exceed the market value of the assets the Fed owns by about half a trillion dollars. As interest rates go higher, this loss increases. Moreover, if the Fed’s inflation-fighting campaign eventually requires short-term interest rates to rise above 2.7 percent, we project the Federal Reserve will experience net operating losses, in addition to its mark-to-market losses.
Unlike banks and other financial institutions, no matter how big the losses it may face and how negative its true capital position, the Federal Reserve will not fail. But if losses, however large, can’t end the Fed, who pays for these losses? Will the Fed’s shareholders be hit in some fashion or will the losses be monetized and contribute to spiraling inflation? Should member banks be paid interest when the interest payments cause Federal Reserve losses? In recent years, questions like these have been irrelevant because the Fed has made very large profits. But this year is different.
Federal Reserve officials try to downplay the gravity of these issues. For example, at a recent Federal Reserve Bank of Atlanta Financial Markets Conference, Federal Reserve Bank of Cleveland president Loretta Mester said that losses would have no impact on the Fed’s ability to conduct monetary policy, but admitted they could raise “communication challenges” for the system. This rather cavalier treatment of massive Federal Reserve losses is curious since it is potentially at odds with the way Federal Reserve losses should be treated according to the Federal Reserve Act. Moreover, given the large interest income banks earn on their reserve balances, the issue of burden sharing of Federal Reserve System losses may become much more contentious as the Fed executes its inflation-fighting policies in the coming months.
The real story of how the Fed accounts for losses, how the losses impact monetary policy, and who ultimately pays for these losses is a complicated one. The details are in some little-known provisions of the Federal Reserve Act of 1913, in more recent Federal Reserve Board policy decisions regarding the Fed’s accounting standards, in legislation changing its dividend and capital surplus policies, and in its post-financial crisis decision to pay interest on banks’ reserve accounts.
The Federal Reserve Act stipulates that Federal Reserve shareholders—the member banks– should bear at least some Federal Reserve System losses, but to date, this has never happened. “Innovations” in accounting policies adopted by the Federal Reserve Board in 2011 suggest that the Board intends to ignore the law and monetize Federal Reserve losses, thereby transferring them indirectly through inflation to anyone holding Federal Reserve notes, dollar denominated cash balances and fixed-rate assets.
Federal Reserve System Current and Prospective Losses
In the Federal Reserve System’s most recent financial statements for the quarter ending March 31, 2022, the fair value note to the statements shows a total unrealized capital loss of $458 billion during the quarter on the Fed’s $8.8 trillion book value of the Fed’s System Open Market Account (SOMA) securities holdings. This loss took the fair value of the portfolio from a mark-to-market gain of $128 billion on December 31, 2021, to a mark-to-market loss of $330 billion on March 31.
With interest rates continuing to increase, we estimate that the Fed’s unrealized capital loss grew by an additional $210 billion, bringing the Fed’s total unrecognized capital loss to an estimated $540 billion as of May 31, 2022. Losses continued to grow through mid-June and should the Fed maintain its plan to continue raising interest rates to fight inflation, these losses will only increase. If the Fed was a bank or other regulated financial institution, it would be closed because it is already deeply economically insolvent.
In addition to the deleterious impact of rising interest rates on the market value of its SOMA portfolio, rising interest rates will sharply reduce the Fed’s net interest income. In the first quarter of 2022, the Fed reported net interest earnings of $35 billion which, when netted against expenses, yielded a reported operating income of $32 billion, a figure that excludes the mark-to-market loss on its securities portfolio.
As interest rates continue to increase, the Fed net interest revenue and operating income will decline as the Fed pays higher interest rates on $3.3 trillion in member bank reserve balances and its nearly $2.3 trillion (as of June 1) in reverse repurchase agreements while it earns interest on the largely fixed-rate securities its SOMA portfolio. According to our estimates, if short-term interest rates were to reach 2.7 percent, the Fed’s net interest income would no longer be sufficient to cover its approximately $9 billion in annual operating costs, and the Fed would post an annual operating loss. This fact is especially relevant given that the FOMC forecast has the federal funds rate at 3.4 percent by year-end 2022.
With annual inflation currently running at 8.6 percent, 3.4 percent may not be a high enough short-term interest rate to tame inflationary pressures. Federal Funds futures and several bank economists project that policy rates will need to rise to 4 percent or higher in 2023. Ignoring market-to-market losses on its SOMA portfolio, and absent any realized losses from SOMA asset sales, we project that the Fed will post an annual operating loss of $62 billion if short-term rates rise to 4 percent. A $62 billion loss is 150 percent of the Federal Reserve system’s current capital.
This unenviable financial situation in which the Fed has placed itself—huge mark-to-market investment losses and declining, and eventually negative operating income—is the predictable consequence of the balance sheet it has created when the stance of Fed monetary policy transitions to fighting inflation. The balance sheet now combines paying rising rates of interest on bank reserves and reverse repurchase transactions after more than a decade of Fed quantitative easing and zero interest rate policies that stuffed the Fed’s balance sheet with low-yielding long-term fixed rate securities. In short, the Fed’s earning dynamics now resemble those of a typical failing 1980s savings and loan.
Does the Federal Reserve Need a Positive Capital Cushion?
In 1913, the members of the 63rd Congress which passed the Federal Reserve Act, decreed that the Federal Reserve’s 12 district banks should be capitalized by their member banks. The Act also specifies that member banks must absorb the first tranche of losses should Fed revenues fail to cover expenses. Indeed, as discussed below, Section 2.4 of the Act specifically says that member banks are liable for an amount up to double the value of their subscribed stock to cover Federal Reserve district bank losses.
The Federal Reserve system comprises a Board of Governors and the Federal Open Market Committee in Washington DC, and 12 district Federal Reserve banks. All the financial assets and liabilities of the Federal Reserve are held by the district banks. Each district bank is owned by the commercial and mutual savings banks of that district that applied for, and were granted, Federal Reserve membership.
District banks issue equity shares with a par value of $100. Member banks must subscribe to the shares issued by their district bank in a dollar value equal to 6 percent of a member institution’s paid in capital and surplus. Member banks only pay in half the subscribed share value “while the remaining half of the subscription shall be subject to call by the Board.” Each member bank must true up its district bank stock subscription annually to reflect changes in the member bank’s capital and surplus.
The 1913 Federal Reserve Act required that district banks have positive capital. In particular, the 1913 Act stated that, “no Federal Reserve bank shall commence business with a subscribed capital less than $4 million.” $4 million in capital was the minimum amount needed to open a district reserve bank, but those organizing each district bank could require a higher initial capital threshold should prudence dictate.
If a district bank failed to generate the capital needed to commence operations from member bank contributions, the Act authorized the sales of district bank shares to the public, and should public subscriptions prove insufficient, sales of shares to the U.S. Treasury. Clearly, Congress placed a high priority on ensuring that each district reserve bank had adequate capital before commencing operations. A reasonable interpretation of this legislative language is that Congress established $4 million as the minimum required capitalization of a Federal Reserve district bank. Accounting for inflation, the minimum capital needed operate a Federal Reserve district bank would exceed $110 million today.
Member banks earn dividends on their Federal Reserve district bank stock holdings and the Federal Reserve system dividend policy impacts the Fed’s capital surplus account, which according to the GAO (p.9) is “intended to cushion against the possibility that total Reserve Bank capital would be depleted by losses incurred through Federal Reserve operations.”
In return for providing the district bank’s capital base, all member banks were initially entitled to receive a generous 6 percent dividend on the par value of their paid-in shares. The dividend was cumulative in the event a district bank had insufficient operating revenues to cover expenses and dividends in any given year. More recently, the dividend rate was reduced for large banks, currently defined to be banks with assets in excess of $11.2 billion. The annual dividend rate for these banks is the lesser of, “the high yield of the 10-year Treasury note auctioned at the last auction held prior to the payment of such dividend, or 6 percent.” Since the change, large bank dividends have been less than, sometimes substantially less than, half of the 6 percent rate promised in the original Federal Reserve Act.
Federal Reserve Board policies concerning member bank dividends and Federal Reserve System capital surplus confirm the view that district reserve banks need to maintain positive capital. In a 1922 memorandum, the General Counsel of the Federal Reserve Board clarified the cumulative nature of the dividend on bank share subscriptions and established a target value for the Federal Reserve System’s capital surplus account:
“[T ]he earnings of the Federal reserve banks shall be used for the following purposes in the order named:
(1) For the payment of or provision for expenses.
(2) For the payment to stockholders (who are member banks exclusively) of cumulative dividends at the rate of six percent per annum on paid-in capital.
(3) For creating and adding to a surplus fund until such fund equals 100 percent of subscribed capital.
(4) The balance to be paid 90 percent to the United States as a franchise tax and 10 percent into surplus.
…No payment can be made into the surplus fund unless the earnings for the current year are sufficient to pay in full the dividends for that year and any dividends for past years that may remain unpaid.
Effective January 1, 2021, revisions to the Federal Reserve Act limit the aggregate Federal Reserve system surplus account to $6.785 billion. Federal Reserve district banks now remit earnings to the U.S. Treasury to the extent that these earnings exceed member bank dividend commitments and any contributions necessary to maintain the Federal Reserve system’s surplus at $6.785 billion.
The original Federal Reserve Act, revisions to the Act, and subsequent Federal Reserve Board policy statements regarding dividends and the Federal Reserve system surplus account all suggest that the Federal Reserve system faces a minimum capitalization requirement codified in law and in the Federal Reserve Board policies that govern member bank dividend payments. But in the remaining sections will we explain how the Fed’s accounting policies belie its Congressional mandate and Federal Reserve Board policies designed to ensure that the Fed maintains a positive capital cushion.
Unrealized Losses, Realized Losses and Operating Losses
Unlike other financial institutions that must comply with GAAP accounting standards, the Federal Reserve Board decides on the accounting standards it uses to report the Federal Reserve system’s income and balance sheet positions. Under the Fed’s accounting rules, the implications of a Federal Reserve system loss depend on how the loss is generated. The Fed’s accounting standards distinguish among three types of losses: unrealized losses, realized losses, and operating losses.
Today, the Fed’s SOMA portfolio includes $8.8 trillion in interest-bearing assets, $7.8 trillion of which have a maturity of over 1 year. The Fed accounts for its SOMA securities using held-to-maturity accounting conventions. Securities are valued at par value with amortization of any price premium paid, or price discount received, at the time the Fed purchased the security. Premiums or discounts are amortized over the remaining life of the security.
The SOMA portfolio’s assets are fixed-rate instruments with market values that depend on the current interest rate environment. In general, the book value of the Fed’s securities holdings will not equal the current market value of the portfolio. The Fed does not recognize mark-to-market gains or losses on its SOMA securities portfolio when it calculates its earnings or losses. The Fed only recognizes realized gains or losses on these securities. If the Fed sells a security from the SOMA portfolio for greater (lower) than its amortized cost, it records a realized gain (loss) in income. Realized gains or losses are included in the Fed’s reported operating earnings but, as we explain below, under the Federal Reserve Board’s current accounting policies, negative earnings will not negatively impact the Fed’s reported capital or surplus accounts.
The third category of income (losses) important for Federal Reserve System account statements are total reserve bank income (losses) from operations [a.k.a. operating income (losses)]. Operating income (losses) are defined as: (1) net interest income (interest earnings less interest expense); plus (2) other income (loss) items that include realized losses on SOMA securities, foreign exchange translation gains (losses) and income from services provided, including those reimbursed by the government and other income; (3) less Federal Reserve district bank operating expenses; less (4) Federal Reserve Board operating expenses and currency printing costs; less (5) the assessment to pay the expenses of the Bureau of Consumer Financial Protection.
The Federal Reserve Act and Federal Reserve System Operating Losses
The authors of the 1913 Federal Reserve Act never envisioned that Federal Reserve district banks would suffer large capital losses on their investments since Section 14 of the Act restricts their asset holdings to gold, gold coins, gold certificates, short-term banker’s acceptances, real bills eligible for rediscount, US government notes and bonds, and short-term tax anticipation notes or revenue bonds issued by eligible state and local governments. Unlike the Fed’s current portfolio, the market value of district bank asset holdings was not sensitive to changes in market interest rates because most Fed assets matured very quickly, or in the case of gold-based assets, had values that were fixed by the international gold standard. Further, member bank reserves which are deposit liabilities of the Fed district banks, did not pay interest.
According to the 1913 Federal Reserve Act, should there be a need to fortify any Federal Reserve district bank’s resources because of operating losses, member banks were subject to call on the second half of the par value of their equity subscription. Moreover, the Act includes the little-known requirement that member banks contribute additional funds to cover district reserve bank operating losses up to an amount equal to the par value of their membership subscription. In other words, member banks were to be assessed for district bank annual losses in an amount up to twice the par value of their Federal Reserve district bank stock subscription. Note especially the use of the term “shall” and not “may” in the original 1913 Federal Reserve Act language:
“The shareholders of every Federal reserve bank shall be held individually responsible, equally and ratably, and not one for another, for all contracts, debts, and engagements of such bank to the extent of the amount subscriptions to such stock at the par value thereof in addition to the amount subscribed, whether such subscriptions have been paid up in whole or in part under the provisions of this Act.” (bold italics added)
Despite Congressional revisions to the Act over more than a century, the Federal Reserve Act still contains this exact passage—this provision of the law has never been changed.
Has the Federal Reserve System Ever had an Operating Loss?
In the early years after the Fed was organized, district reserve banks operated fairly independently. Member banks had a strong voice in appointing the officials who managed the operations of their district banks. District banks earned revenue primarily from discounting bills of exchange with a small amount of revenue from interest on government bond holdings. Gold and other eligible reserve assets did not generate revenue. Bills of exchange were discounted at a penalty rate by design, a feature not conducive to generating district bank revenues.
In 1915, the district reserve banks had combined negative earnings before dividends of $141,000. At a September 1915 meeting, the Board of Governors voted1 to approve assessing member bank stockholders to cover district bank operating losses. The district reserve banks, however, never made the assessment, reasonably fearing that an assessment would discourage state chartered banks from applying for system membership.
In the wake of the 1915 experience, district banks focused on generating revenue. Facing weak discount revenues, district banks bought tax anticipation notes and Federal government notes and bonds2 to generate interest income. According to Meltzer (p. 77):
“[Open market operations were combined] to avoid any effects of competitive purchases on market rates. Although effects on the market were recognized, purchases were made principally to increase the earning of reserve banks and were allocated to the individual banks in part based on their need for earnings. Reserve banks retained the right to purchase independently. Some claimed that New York did not buy enough so their earnings were held down.”
The pressure to generate revenues eased as district banks began doing a brisk business discounting the Liberty Bonds issued to finance World War I.
Modern Federal Reserve Board Policy Regarding Federal Reserve System Losses
Today, the Federal Reserve official position regarding gains and losses in the market value of it SOMA portfolio is,
[T]he fair value of the Federal Reserve’s portfolio as well as its earnings, gains, or losses do not affect the ability to carry out its responsibilities as the nation’s central bank, which is to conduct monetary policy to achieve its statutory goals of maximum employment and stable prices.
Regarding realized losses on its SOMA portfolio, the Fed’s official position is,
[I]n the unlikely scenario in which realized losses were sufficiently large enough to result in an overall net income loss for the Reserve Banks, the Federal Reserve would still meet its financial obligations to cover operating expenses. In that case, remittances to the Treasury would be suspended and a deferred asset would be recorded on the Federal Reserve’s balance sheet, representing a claim on future net earnings that the Reserve Banks would need to realize before remittances to the Treasury would resume.
Today, the Federal Reserve Board’s official position is that, should it face operating losses, it would not reduce its book capital surplus, but instead would just create the money needed to meet operating expenses and offset the newly printed money by creating an imaginary “deferred asset” (Section 11.96) on its balance sheet. Subsequently, sometime in the future when reserve banks start making positive operating earnings, after paying dividends, district reserve banks will apply any remaining income to reduce the deferred asset balance to zero before resuming their remittance payments to the U.S. Treasury.
By accounting for losses in this manner, the Federal Reserve’s reported capital and surplus account balances are not depleted by system operating losses. According to the Financial Accounting Manual for Federal Reserve Banks (p. 201), bank dividend payments will continue to be paid as long as a reserve bank has a positive surplus account. Under this policy and the rules that have been proposed to account for operating losses, it would appear that member banks dividends will be paid regardless of Fed operating losses.
Could Federal Reserve Losses Impact Monetary Policy?
The current Federal Reserve Board plan to manage losses is: (1) ignore any mark-to-market losses on its SOMA portfolio; (2) recognize realized losses on securities sales, if any; (3) monetize any operating losses and offset the liability on the Fed’s balance sheet by creating or increasing a deferred asset account. The Board has adopted this accounting policy notwithstanding an explicit Federal Reserve Act requirement that member banks be held liable for district reserve banks’ operating losses—a requirement still codified in law.
The issue of maintaining a positive Federal Reserve system capital cushion, once a necessity to maintain public confidence in convertibility under the international gold standard—and still a requirement in the current version of the Federal Reserve Act—is no longer an issue of practical importance. Federal Reserve notes and member bank reserve bank balances have not been convertible into gold in the US for more than 90 years and there has been no required gold backing for Federal Reserve notes for more than 50 years. The pure fiat currency the Federal Reserve issues today has no commodity backing and there is no longer any constraint on the amount the Fed can issue. Given the Fed’s stated intention to monetize operating losses and back any newly created currency with an imaginary “deferred asset”, the Federal Reserve Board has demonstrated it no longer has a concern in maintaining the value of loss true absorbing assets backing the Federal Reserve system’s capital and surplus accounts.
The Federal Reserve Board’s proposed treatment of system operating losses is wildly inconsistent with the treatment prescribed by the Federal Reserve Act. In all likelihood, operating losses, should they occur, will in large part be a consequence of the interest payments made to member banks for reserve balances held at Fed district banks. It is impossible to imagine that the authors of the Federal Reserve Act would have approved of allowing the Fed to create an imaginary “deferred asset” as a mechanism to hide the fact that the Fed is depleting its cushion of loss-absorbing assets while paying banks interest on their reserve balances when the Act itself makes member banks liable for Federal Reserve district bank operating losses. Under the international gold standard, before and after the founding of the Federal Reserve system, banks earned nothing on their gold reserves, so today’s arrangements where the Fed pays interest on bank reserves would have never been considered at the time the Fed was founded.
If the Federal Reserve were to comply with the language in the Federal Reserve Act and exercise its right to assess member bank resources to cover operating losses, monetary policy could be significantly impacted in a number of ways. As short-term interest rates rise and the interest expense needed to fund reverse repurchase agreements and member bank reserve balances consumes more of the interest earnings on its SOMA portfolio, the Fed’s willingness to shrink its balance sheet by liquidating SOMA assets at a loss could become constrained to member bank assessments to cover Fed operating losses. Just as it did in 1915, the issue of operating losses would focus the attention of the district bank presidents who vote on Federal Open Market Committee monetary policies.
The prospect of passing on the Federal Reserve system’s operating losses to member banks could also create pressure to attenuate these losses by lowering the interest rate paid to member banks. Should this occur, it would directly impact Fed operations by constraining the short-term interest rate increases the Fed uses to constrain inflationary pressures.
Under its post-crisis monetary operating policies, as the Fed raises rates, banks will earn larger interest payments on the reserve balances held at the Fed district banks while continuing to accrue dividends on their Federal Reserve district bank shares. Meanwhile, the Fed’s actions, though necessary, will impose higher interest rates on the public at large, losses in the value of the public’s bonds and stocks in their savings and retirement accounts, reduced growth, and likely cause a significant increase in unemployment before the Fed successfully arrests inflationary pressures. If Fed policies lead to operating losses, and the Fed follows its plan to monetarize these losses, the losses will only contribute to the inflationary pressures the Fed seeks to control. Should the public understand the implications of these policies, the Fed could well face a contentious “communication problem.”
For only the second time in its history, the Federal Reserve system is facing the prospect of losses, only this time the losses are massive. The Fed already has huge market-value losses on its SOMA portfolio that it chooses not to recognize in its formal financial accounting statements. Any financial institution other than the Fed faced with market-value losses greater than 13 times its capital would have already lost public confidence and probably be in receivership. And soon the Federal Reserve will face large operating losses, losses which it must recognize on its financial statements.
While the Federal Reserve Act explicitly requires that Federal Reserve member banks be assessed to cover operating losses, the Federal Reserve Board’s stated plan is to monetize these losses and still report a positive capital and surplus position through the use of “creative accounting” entries not seen since the 1980s savings and loan crisis. Those that recall that historical period know that relying on “regulatory accounting standards” to create phantom capital cushions did not turn out well. In the Fed’s case, failure is not an issue because the Fed can literally print as much money as needed to pay its expenses and member bank dividends. Monetizing operating losses will however enrich the Fed member banks that are supposed to be bearing the loss, while the public at large will face higher interest rates, higher unemployment, reduced growth, and the inflationary consequences of the new money printed to cover Fed losses. The Fed seems to be hoping that nobody notices.