The Federal Reserve and the Debt Crises
Thomas R. Saving
University Distinguished Professor of Economics Emeritus
Texas A&M University
All the countries in the developed world have been expanding sovereign debt at alarming rate ever since the 2008 financial crisis. At the same time, their central banks have engaged in unprecedented increases in assets by buying both own country debt and private assets.
Traditional monetary theory would have predicted, and many economists did so predict, that the result of these central bank asset expansions would have been raging inflation. What happened that made the combination of rapid sovereign debt expansion and central bank participation in that expansion not lead to inflation? Further, does the reason the first debt expansion did not lead to inflation suggest why the 2020 Pandemic debt expansion will not create an inflation crisis? To address the potential for a debt crisis to be controlled through appropriate central bank response, this work concentrates on the Federal Reserve and its actions during two onslaughts of United States federal debt. The first onslaught of debt was the period from 2008 through 2019. The second debt onslaught came with the deficits due to the Covid-19 2020 Pandemic.
Debt Crisis 1: The Great Recession
Until the beginning of fiscal year 2009 the first decade of this century was similar to other decades of the past half-century, at least in terms of Federal Reserve behavior. But then came the onset of the Great Recession and seemingly everything changed. First came an unprecedented increase in federal deficits, both in dollar size and as a share of GDP. Figure 1 shows the path of federal deficits as a share of GDP beginning in 2002. This series of deficits followed the also unprecedented series of surpluses of the late 1990s through 2001. Up to fiscal 2009 the annual deficit share of GDP was the same order of magnitude of the previous pre-surplus years. But then with the onset of the Great Recession, the fiscal 2009 deficit ballooned to almost 10% of GDP. Fiscal deficits then gradually declined both in nominal terms and as a share of GDP until beginning a gradual rise in fiscal 2016. Overall fiscal year deficits have averaged 5% of GDP for almost a decade, the longest stretch of deficits at this level in at least a century.
Figure 1. Federal Deficits as a % of GDP
Given the level of post-2008 federal deficits as a share of GDP it is not surprising that publicly held federal debt as a share of the nation’s income, GDP has also been steadily rising. Further, with the onset of a return to trillion-dollar federal deficits associated with the Covid-19 pandemic, federal deficits as a share of GDP are expected to rise faster than previously estimated by the Congressional Budget Office. Ten months into fiscal year 2020, July of 2020, the fiscal 2020 federal deficit was already $2.8 trillion, 13% of 2019 GDP, the largest since WWII.
Figure 2 shows the path of publicly held federal debt as a share of GDP for the 1992 to 2019 period and the pre-Pandemic CBO forecasts to 2030. The decline in debt share of GDP in the early part of the figure is a product of the mid-1990s to 2001 budget surpluses. The GDP debt share was relatively stable until the onset of the Great Recession, fiscal 2009. Then in the next decade the GDP debt share doubled and reached nearly 80% for fiscal 2019. The last CBO forecasts before the 2020 Pandemic CBO suggested that this trend of rising public debt share of GDP would continue into the long future reaching almost 100% by 2030. All that changed with the onset of the government response to the Covid-19 Pandemic so that by April of 2020, the CBO was forecasting that the publicly held debt share of GDP would be 100% for 2020 and would reach 110% of GDP in 2021.
Figure 2. Debt Held by the Public as a % of GDP 1992-2030
An important caveat to the information contained in Figure 2 is that it fails to recognize that the usual estimate of the level of publicly held federal debt considers the Federal Reserve as part of the public. That view is consistent with the Government Accounting Office that considers the Federal Reserve as not part of the government. But by law the earnings of the Federal Reserve must be remitted to the Treasury. In effect then, the Treasury, as the residual income recipient of the Federal Reserve, “owns” the Federal Reserve. It might be viewed as a non-voting common stock holder of Federal Reserve assets. Thus, the all Federal Reserve income earning assets, including Treasury securities and Mortgage Backed securities held by the Federal Reserve, are essentially held by the Treasury and not by the public.1 Thus, these Federal Reserve asset holdings reduce outstanding federal debt on a one-to-one basis.
Figure 3 shows the official publicly held federal debt share of GDP and the share adjusted by recognizing that Federal Reserve’s holdings of federal and equivalent debt are a full offset to publicly held federal debt. Two things are apparent in the figure. One, the federal debt held by the Federal Reserve prior to the 2008 economic meltdown was almost a constant share of GDP. Second, the tremendous expansion in the Federal Reserve holdings of both mortgage backed and Treasury securities increased the Federal Reserve’s share of the official publicly held federal debt.
- For a complete analysis of this issue see Thomas R. Saving, “Rethinking Federal Debt: What do We Really Owe?”,
Private Enterprise Research Center, Texas A&M University, PERC Study No. 1607, August 2016.
Figure 3. Adjusted Debt Held by the Public as a % of GDP
Official Publicly Held Debt % GDP
60 Adjusted Publicly Held Debt %
1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
It is significant that the difference between the debt percent of GDP curves becomes much larger after 2008. The scale of Federal Reserve holdings of securities as a % of GDP remained approximately constant from 1992 through 2007. Then the Great Recession’s almost unprecedented increases in the GDP share of federal deficits led to two changes in Federal Reserve behavior. One, the Federal Reserve’s securities share of GDP rose from its former equilibrium of about 7% to a peak of 25%. Two, the payment of interest on reserves made reserves essentially negative securities holdings. The total effect of these two events resulted in the significant post-2008 increase in the Federal Reserve net holdings of securities, as is evident in Figure 3.
Federal Reserve Response to the 2008 Crisis
To get a better feel for how dramatic the Federal Reserve’s post-2008 change in policy was, consider the period from October 2008 to December 2014, the period of massive federal deficits coupled with Federal Reserve unprecedented asset expansions. During that period PCE inflation averaged 1.56% and real GDP growth was 1.8%, both historic lows for any similar length period. Moreover, federal deficits for fiscal years 2009 through 2014 expressed both in terms of absolute dollars and as a share of GDP were post-WWII records.
Figure 4 shows the levels of fiscal year federal deficits and the corresponding changes in Federal Reserve assets, as reflected in changes in the monetary base. The monetary base changes presented in the figure are the result of the changes in the level of Federal Reserve
assets. In general, changes in Federal Reserve assets represent either running the money printing press in the case of increasing assets, or destroying money in the case of Federal Reserve asset reductions. Based on the data in the figure, it appears that for the fiscal years 2009 through fiscal 2015, Federal Reserve asset acquisitions financed just over 55% of the federal deficits.
Figure 4. Federal Deficits and Federal Reserve Monetization
Fiscal Year 2009 – Fiscal Year 2014
Federal Deficit Base Increase
Historically the level of monetary expansion, if measured by the increases in the monetary base reported in Figure 4, would have resulted in a return to the double-digit inflation of the late 1970s. We know that did not happen and the important question is why? But before the answer, a more detailed look at Federal Reserve behavior subsequent to the financial breakdown of September 2008 will prove useful.
Except for Federal Reserve responses to the financial crisis in the last four months of 2008, the increases in the monetary base during this period were the result of increases in Federal Reserve holdings of securities. But the securities held were not all federal debt. The Federal Reserve was buying assets, principally both Treasuries and Mortgage Backed Securities (MBSs). But only their Treasury purchases were directly reducing the level of federal debt held outside the government.
Figure 5 shows the aggregate level of Federal Reserve assets and its composition for the entire period from the recession’s beginning to the end of Federal Reserve asset expansion. An important question is: do Federal Reserve holdings of MBSs and private market assets contribute in any way to the level of Federal Reserve of financing of federal debt? To solve this puzzle requires analysis of what happens to all Federal Reserve revenue. By law, profits of the
Federal Reserve after all costs revert to the Treasury. Thus, the Treasury is the residual income recipient of Federal Reserve asset holdings and therefore in one sense at least the Treasury owns the Federal Reserve. As a result, all Federal Reserve earnings on Treasuries, MBSs and other private market holdings accrue to the Treasury and reduce the net servicing cost of the federal debt and as such are equivalent to purchases of federal debt.
Figure 5. Total Assets, Total Securities, Treasuries and MBSs
January 2008 to January 2015
|Bear Stearns Crisis||September 2008 Liquidity Crisis|
Figure 6 shows the special efforts the Federal Reserve took to alleviate the financial crisis that began in September of 2008. Since some of the facilities were in place at the beginning of 2008, the data in the figure begin at the end of December 2007. The effect of the September 2008 financial crisis is apparent in the figure as Federal Reserve loans spiked at that time. These loans were mutual fund and broker assistance. Then the commercial paper market all but disappeared as money market funds were pressured by customers for liquidity. The Federal Reserve, for all practical purposes, became the commercial paper market. The figure also shows the Term Auction program that provided money to financial institutions at rates of interest determined at auction. All of these special facilities were gone by the end of 2010 and all were very profitable for the Federal Reserve and the Treasury.
|Table 6. Early Great Recession Interventions|
|January 2008 to January 2011|
|Loans||Repurchase Agreements||Term Auction||Commercial Paper Funding Facility|
Federal Reserve’s Role in the Great Recession Debt Expansion
A common measure of the burden of the federal debt is the share that this debt is of the nation’s GDP. At the close of initial debt expansion, 2015, the CBO estimated that the publicly held federal debt was 73.6% of GDP. Presumably, the ratio of debt to a nation’s GDP is an indication of a nation’s ability to at least pay the cost of servicing that debt. Since taxpayers are on the hook for servicing the federal debt, this federal debt is on the same footing as personal debt. In that sense, the same logic used to value the net debt of any private citizen is also relevant to valuing federal debt.
But in reporting the publicly held federal debt, the Federal Reserve is considered to be part of the public. But because the Treasury owns the Federal Reserve’s net income, debt held by the Federal Reserve is essentially owned by the government and not the public. As a result, the interest payments on that debt accrue to the Federal Reserve and are then turned back into the Treasury. Thus, in the most important sense, the federal debt owned by the Federal Reserve is not federal debt at all. But more than that the revenue from all Federal Reserve assets, Mortgage Backed Securities (MBSs) for example, also accrue to the Treasury so that these non-Treasury asset holdings are the equivalent of federal debt.
Figure 7 displays the level the net debt servicing cost of the federal debt and the level of Federal Reserve transfers to the Treasury for the fiscal years, 2009 through 2015. Importantly it shows that by the close of 2014 Federal Reserve transfers to the Treasury were paying for more
than 40% of total debt servicing costs. Since Federal Reserve holdings of MBSs create income for the Treasury these assets offset federal debt their purchase is equivalent to the Federal Reserve buying Treasuries.
Figure 7. Net Debt Servicing Cost and Federal Reserve Transfers to the Treasury Fiscal Year 2009 – Fiscal Year 2015
|Net Servicing Cost||Federal Reserve Transfers|
Returning to the post-2008 Deficits and Monetization depicted in Figure 3, it seemed that the Federal Reserve asset increases did indeed constitute monetization of a significant share of the massive federal debts of this period. But this simple view of the federal debt ignores the fact that the Federal Reserve increased its liabilities almost in lock-step with its increase in assets. The increase in Federal Reserve liabilities was the result of the introduction in October of 2008 of the payment of interest on bank reserves. Essentially, the obligation to pay interest on bank reserves made these reserves a short-term debt of the Federal Reserve. Since all earnings of the Federal Reserve offset the servicing cost of the federal debt, and the interest payments on reserves reduce these transfers by the full amount of the payments, makes bank reserves the equivalent of short-term federal debt. For all practical purposes, the Federal Reserve throughout the QEs was buying long-term federal debt and selling short-term federal debt.
Before interest payments on bank reserves, Federal Reserve actions that increased bank reserves led to an increase in the money supply by a multiple of the increase in reserves. When bank reserves earned nothing, banks moved any excess reserves into market investments, loans, Treasury securities or other investments. This activity increased the money supply by a multiple of the change in reserves. Then, these money supply increases affected the price level. But now these same bank reserve increases represent member bank income earning assets.
Figure 8 shows the level of bank holdings of excess reserves for the period from the beginning of interest payments on bank reserves through the first quarter of 2016. The figure also shows the difference between an important market rate of interest, the rate of interest on 1-year Treasuries, and the interest on excess reserves. This difference is a measure of the advantage or dis-advantage to the banks of holding excess reserves instead of market investments. At the onset of reserve interest payments, the rate of return for holding reserves matched the rate on 1-year Treasury Notes. Then for the period from January 2009 until September 2010 the rate of return on the 1-year Notes exceeded the return on reserves. From September 2010 the return to holding reserves exceeded the 1-year Note rate until August 2015. Then in response of rising 1-year Note rates the interest rate on reserves was raised to 37 basis points and then to 50 basis points. As is clearly shown in the figure the tremendous growth in bank excess reserves occurred when the return on these reserves exceeded the return on 1-year Treasuries.
Figure 8. 1yr Treasury – IOER Spread and Excess Reserves Jan. 2008 – Jan. 2016
In effect, the Federal Reserve sterilized their asset growth by making reserves a Federal Reserve short-term liability. This sterilization eliminated the normal effect of an increase in bank reserves on the money supply as member banks were paid to hold reserves rather than increase their holdings of economy assets, either loans or securities. Thus, any potential effect of these increased reserves on the money supply was mitigated or eliminated. In fact, excess reserves only contribute to money growth as they are used to create loans and the concomitant increase in bank demand deposits. That is excess reserves only become part of the monetary base when they cease to be excess, i.e., become required reserves.
The fact that bank excess reserves are the equivalent of short-term federal debt requires that the level of Federal Reserve financing of the post-2008 federal deficits shown in Figure 4 must be amended. Figure 9 amends Figure 4 by netting out any increase in Federal Reserve liabilities stemming from member banks excess reserve holdings. Essentially, it accounts for the inclusion of the increase in Federal Reserve liabilities, i.e., bank reserves, as an offset to Federal Reserve asset growth. The difference between traditional monetary base growth and the growth in excess reserves for each fiscal year is a measure of the effective monetary base growth. The effective monetary base growth is an estimate of the actual level of monetization of each fiscal year deficit. Using this measure of net Federal Reserve involvement in financing federal deficits for the entire seven fiscal year period the level of deficit monetization was just over 9%.
Figure 9. Federal Deficits and Federal Reserve Monetization: Effective Monetary Base Fiscal Years 2009 – 2015
|Federal Deficit||Base Increase|
The reported measure of publicly held federal debt includes Treasury debt held by the Federal Reserve and, as of the end of 2015, the debt was $13.11 trillion. However, the Federal Reserve asset holdings belong to the Treasury. If the Treasury debt and Federal Reserve holdings are consolidated, the publicly held federal is reduced by Federal Reserve holdings of $2.5 trillion of federal debt to $10.6 trillion. There is a further adjustment to the level of publicly held debt that is required due to the income of the Federal Reserve being the property of the Treasury. The other principle asset held by the Federal Reserve are the $1.7 trillion in Mortgage Backed Securities (MBSs) the income of which belongs to the Treasury. In effect the Federal Reserve holdings of MBSs are Treasury assets that offset Treasury liabilities. When all Federal Reserve assets Treasuries, MBSs and other assets are subtracted from reported publicly held debt the above $10.6 trillion becomes $8.9 trillion.
A final adjustment is required since the introduction of paying interest on bank reserves has made these reserves short-term liabilities of the Federal Reserve. Further the Federal Reserve engages in providing non-member financial institutions the ability to earn interest on some of their reserves by buying Federal Reserve reverse repos. Because all these liabilities reduce the level of Federal Reserve transfers to the Treasury, they are the equivalent of publicly held debt. Thus, it is appropriate to add the sum of these liabilities, together they equal to $2.8 trillion, to the $8.9 trillion adjusted publicly held debt. The adjusted level of the federal debt taking into account the adjusted Federal Reserve balance sheet is then $11.7 trillion about 89% of the 2015 measured $13.11 trillion and 63.3% of GDP.2
Is there an Inflation Mystery in the Federal Reserve Response to the First Debt Crisis?
Considering the virtual unprecedented rate of monetary base growth following the onset of the Great Recession, the usual economic projection would have been inflation to match. The natural question then is: what happened? In a simple world of constant currency to deposit and bank reserve ratios the rate of change in any of the popular definitions of money would equal the rate of change in the monetary base. It is just this growth in the monetary base that represents the Federal Reserve’s effect on the economy through its effect on the money supply, however defined. Then in a simple world, the 15.7% rate of growth in the monetary base of the last seven years should have, but did not, result in the significant inflation that many economists predicted. Why did this not happen?3
First, consider the traditional equation of exchange that relates the rate of change in money growth to the rate of change in prices adjusting for both velocity growth and real GDP growth. This equation can be expressed as, π = μ + ν – g. Essentially the rate of inflation, π must equal the rate of growth in the money supply, μ, plus the rate of growth in the velocity of money ν minus the rate of growth in real GDP, g. For this seven-year period, the rate of growth in the expanded definition of the money stock, M2, was 6.3%, much less than the 15.7% rate of growth in the monetary base. For this same period the GDP velocity of M2 fell at a rate of 2.9% and real GDP grew at 1.8%. The result using the simple equation of exchange is an inflation rate of 1.6%, not significantly different than the measured rate of inflation of the GDP deflator of 1.4%.
Second, what is the reason for the disparity between the 15.7% rate of monetary base growth and the much smaller 6.3% growth in M2? Here the answer lies in the change in the nature of the monetary base, in particular, the reserve component. Traditionally when the
- An alternate way of looking at this issue is to consider the burden of the public debt. That burden is the taxpayer responsibility of pay the interest cost of the debt. Using that criterion and the information contained in Figure 6 for fiscal year 2015, transfers from the Federal Reserve to the Treasury covered 44% of the net federal debt servicing cost. Adjusted for who has to pay the publicly held debt would be 44% of the official 2015 debt of $13.11 trillion, or $7.36 trillion.
- For a complete analysis of the mystery of the missing inflation see Thomas R. Saving, “The Federal Reserve, the Great Recession and the Lost Inflation”, Private Enterprise Research Center, Texas A&M University, PERC Study No.
1604, July 2016.
Federal Reserve increases its asset holdings the proceeds of these purchases became bank reserves and as such become part of the monetary base, often referred to as “high-powered money” because of the fact it can support a multiple of itself in the money supply. But now all reserves are not high-powered money. In fact, excess reserves are pure bank investments and cannot affect the money supply. As a result, for purposes of bank money creation, only required reserves matter. What must be done is to correct the usual measure of the monetary base by removing excess reserves. Then the relation between the money supply and this new adjusted monetary base should be unchanged from the before interest on reserves relation between the traditional monetary base and the money supply.
Figure 10 depicts the multiple of the monetary base that the M2 measure of the money supply was when total reserves and reserves adjusted for the investment component of bank reserves are included in the monetary base measure. The adjustment is a simple reduction in total reserves by the amount of bank investment in excess reserves. Since excess reserves before the payment of interest on reserves were for all practical purposes zero, the two multipliers were virtually identical until the payment of interest on reserves. However, once we get to October 2008, the traditional M2 base multiplier falls off the chart while the M2 adjusted base multiplier remains almost unchanged.4
Figure 10. M2 Monetary Base and M2 Adjusted Base Multipliers
Jan. 2000 – Jan. 2015
- Two periods of Federal Reserve support of market liquidity are easy to see in the Figure as sudden drops in the M2 base multiplier. One, January 2000, the continuation of the Y2K scare. Two, September 2001, the World Trade Center terrorist attack. In both these cases the Federal Reserve entered the market to ensure market liquidity by increasing the monetary base.
In simplistic terms what happened to the inflation threat of the tremendous Federal Reserve post 2008 asset expansion was a remarkable reduction in the velocity of money coupled with a reduction in the effect of the monetary base on the supply of money.
Understanding what happened brings two perplexing issues into focus. First, did everything we thought we knew about how the monetary system worked become wrong at the beginning of the fourth quarter of 2008? Second, if not, then why did an unprecedented expansion of Federal Reserve assets not have any real effect on the economy?
The answer to both these questions is apparent once we understand the effect of the introduction of interest payments on bank reserves. These interest payments on bank reserves created investment opportunities for banks that did not involve the economy. Simultaneously, interest payments on reserves created liabilities on the Federal Reserve balance sheet.
Prior to these interest payments, the largest official Federal Reserve liability was currency. But currency could not be considered a liability in any real sense. To see this point, consider what you get when you bring currency to the Federal Reserve. You just get replacement currency. The Federal Reserve gets real resources when it issues currency, just as you would if you issued a personal bond, i.e., borrowed money from a bank. The difference is that you would be required to pay back the loan with interest, that is give up something real, while the Federal Reserve never has to pay back anything.5
In this new world of interest payments on reserves, the asset purchases of the Federal Reserve have suddenly been financed, at least partially, by issuing debt. Moreover, this new debt has real consequences because all Federal Reserve earnings accrue to the Treasury. Thus, the interest payments on reserves reduce Federal Reserve transfers to the Treasury. These reduced transfers increase the cost of servicing the federal debt and ultimately increase the federal debt burden. In effect, with interest payments on reserves bank excess reserves are equivalent to federal debt.
The Years Between the Two Debt Crises
The Federal Reserve’s involvement in the first debt crisis ended at the close of 2014, when it closed down its asset expansion program and essentially ended its role in reducing the level of the relevant measure of federal debt. From that point until the beginning of the second debt crisis, the onset of the Covid-19 pandemic and the economic shut-down, the Federal Reserve holdings of securities remained relatively stable.
Before interest on reserves a long period of stable Federal Reserve assets would have meant little or no monetary growth. Now monetary growth can happen with asset expansion or liability contraction. During this period monetary growth matched economic growth as Federal Reserve net assets rose because their liabilities, excess reserves, fell. These net assets belong to the Treasury. As such, increases in Federal Reserve net assets are direct reductions in a correct
- This issue was resolved as long ago as the 1960s. See Thomas R. Saving and Boris P. Pesek, Money, Wealth and Economic Theory, 1966, Macmillan.
measure of publicly held federal debt.
From the 1990s until the 2008 financial crisis and the accompanying huge federal deficits, the Federal Reserve held on average about 12% of the outstanding federal debt. Subsequent to 2008 the Federal Reserve share of total federal debt has averaged over 25%, more than double the pre-2008 share. Figure 11 shows the net federal debt servicing cost and the Federal Reserve transfers to the Treasury, as well as the % share that these transfers were of any given year’s debt servicing cost. The transfers peaked in 2014 at $99 billion, the year the Federal Reserve ceased its asset acquisitions. In that year Federal Reserve transfers to the Treasury covered over 43% of total net debt servicing costs. Then, as market interest rates rose the Federal Reserve had to raise the interest rate on reserves to control the net monetary base. This increase in Federal Reserve payments on their debt with little change in the earnings on their portfolio reduced Federal Reserve transfers to the Treasury. By 2019 the Federal Reserve transfers to the Treasury covered less than 15% of debt servicing cost.
Figure 11. Net Federal Debt Servicing Cost and Federal Reserve Distributions to the Treasury
|Fed Transfers||Net Interest|
There were three distinct periods of Federal Reserve policy between the two debt crises. The first such period, from the beginning of 2015 until November 2017, Federal Reserve assets were constant but as liabilities in the form of excess reserves fell, net assets rose. Then in November of 2017, the Federal Reserve initiated a program to reduce the assets held in its balance sheet. This asset reduction period was from November of 2017 until the close of that program in August of 2019. The third period was the resumption of securities purchases in October of 2019 until March of 2020, the onset of the second debt crisis.
All three of these periods saw federal deficits although in the first, federal deficits as a share
of GDP were declining. Even though in the first period the Federal Reserve was not adding to its securities holdings, excess reserves were falling fast enough to allow Federal Reserve holdings of net federal debt to rise. Thus, during this period, net publicly held federal debt as a share of GDP actually fell. In fact, as Figure 3 shows during this period net public debt as a share of GDP remained almost constant while the unadjusted publicly held share continued to rise. The same can be said for both of the remaining periods.
Figure 12 shows the path of Federal Reserve net assets, securities holdings plus other investments that contribute to asset earnings and liabilities that use up asset earnings in the forms of interest earning reserves and reverse repurchase agreements.6 The figure begins at the end of the 2008 asset expansion period, January of 2015 and ends at the beginning of the Covid-19 expansion, February 26, 2020.
|Figure 12. Federal Reserve Net Assets|
|January 1, 2015 – February 26, 2020|
|3,500||Net Assets||Asset Reduction Period|
There are three distinct periods represented in the figure. The first period is the time of relatively constant Federal Reserve assets from January 2015 until the onset of asset reductions, November 2017. During that constant gross asset period reserves fell consistently as the difference between market interest rates and the IOER rose making it profitable for banks to invest their reserve holdings. Then in the second period the Federal Reserve actually reduced its assets. During this period the monetary policy job of the Federal Reserve was to ensure that the
- The non-securities assets include repurchase agreements and loans to member banks, and other facilities designed to protect parts of the nation’s financial sector. The latter were for all practical purposes zero by 2015 and did not return until the onset of the 2020 Covid-19 crisis.
difference between market interest rates and the IOER was sufficient to incentivize the banks to reduce reserves faster than the Federal Reserve was reducing assets. That worked until March of 2019, when bank reserve holdings and Federal Reserve net assets stabilized. The failure of net assets to rise and then to actually fall forced the end of asset reductions in August 2019. The third period is from November of 2019 when the Federal Reserve began a new asset accumulation policy. This policy was successful in increasing assets more than the increase in reserves so that net assets rose.
As Figure 12 clearly shows the Federal Reserve does not have to add to its securities holdings to increase its net assets. In fact, except for the brief period in 2019 when the Federal Reserve failed to decrease the IOER to keep market interest rates above the rate paid on reserves, Federal Reserve net assets rose. The decision in 2008 to begin paying interest on bank reserves, particularly excess reserves (the IOER), on the path of Federal Reserve net assets has changed the role of the banking system in the determination of the nation’s money supply.
Paying interest on bank reserves essentially turned these reserves into investment opportunities for banks. The effective monetary base is the measured monetary base less the excess reserves held by banks. Now the effect of an increase in the monetary base on the money supply is reduced by the share of that base increase that is absorbed by bank reserves. As a result, banks are active, rather than passive, players in Federal Reserve monetary policy actions.
The failure of the Federal Reserve to maintain the market advantage over holding reserves is the reason for the end of the trend of declining bank reserve holdings. Figure 13 shows the difference between the yield on 1-year Treasuries and the interest paid on bank reserves for the period of Federal Reserve asset reductions. During the period where reserves were falling faster than Federal Reserve asset reductions the spread was heavily in favor of the market. But by March of 2019 the spread between 1-year Treasuries and the IOER was zero and then negative. The failure of the Federal Reserve to adjust the IOER to offset what was happening in the market resulted in its net assets failing to rise and even falling slightly leading to the cessation of the asset reduction program in August of 2019. Then the advantage of holding reserves did not change as the Federal Reserve continued its failure to reduce the IOER as fast as market interest rates were falling. As a result of falling net assets in October of 2019 the Federal Reserve began to expand assets. This asset expansion was enough to offset the advantage of reserves and once again allow net Federal Reserve assets to rise.
Figure 13. 1yr Treasury – IOER Spread and Excess Reserves October 2017 – August 2019
The fundamental tool of monetary policy is control of the monetary base and the nation’s money supply. It is clear that the banking system responds to the level of the IOER relative to market interest rates. As result we are now in a world where the banking system shares, almost on an equal footing with the Federal Reserve, the control of the money supply.
The Federal Reserve can conduct an expansionary monetary policy by entering the market for securities, but also by reducing the interest rate it pays on bank reserves. As a result, this interest rate, which is wholly determined by the Federal Reserve, is now an important part of monetary policy. But by the same token, market interest rates now play a role independent of the Federal Reserve in monetary policy. An increase in market interest rates, with a fixed Federal Reserve-determined IOER, will increase the effective monetary base and then increase the nation’s money supply.
But these changes do not mean that the Federal Reserve has lost its power to control the money supply. But to maintain its ability to control monetary policy, it must be constantly cognizant of financial market interest rates. Thus, the Federal Reserve is more subject to the vagaries of financial markets than at any time in its more than 100-year history.
Debt Crisis 2: The Covid-19 Pandemic
While there is some evidence that the onset of COVID-19 might have been even earlier than January 2020, its impact on the world economy was certainly no sooner than January 2020. From Figure 12 it is clear that the Federal Reserve asset expansion program that began in October 2019
was necessary to allow Federal Reserve net assets to begin to rise, as is required if the net monetary base is to increase. Further, as Figure 13 demonstrates the Federal Reserve was unwilling to set the IOER enough below market interest rates to incentivize banks to move excess reserves into the market. Thus, by October 2019 the required monetary growth necessitated a significant expansion of Federal Reserve assets.
In the first debt crisis that began in fiscal 2009 with a record $1.4 trillion federal deficit, the Federal Reserve essentially supplied resources to finance $800 billion of that deficit, just over 56%. But the decision to pay banks to hold reserves resulted in the majority of that $800 billion being offset by Federal Reserve liabilities that for all practical purposes were federal debt. So that in the final analysis the Federal Reserve financed only just over 5% of the monumental fiscal 2009 deficit.
Even before the onset of the pandemic the Federal Reserve initiated in October 2019, the first month of fiscal 2020, an asset acquisition program. Then with pandemic economic restrictions multiplying and having a real bite, the rate of asset acquisition virtually exploded. Figure 14 shows this asset expansion program from its October 2019 inception through November 11, 2020. Up to mid-March the program was a steady asset purchase of less than $40 billion a weekly and usually less than $20 billion weekly. Then, when the pandemic shutdown really took hold the Federal Reserve from mid-March 2020 to the end of April 2020 added another almost $1.6 trillion in securities to its portfolio. In all of fiscal 2020, the Federal Reserve has purchased the equivalent of $2.85 trillion of the fiscal 2020 federal deficit, just over 90% of the $3.13 trillion fiscal 2020 federal deficit.
|Figure 14. Weekly and Cumulative Federal Reserve Securities Held Outright Purchases|
|October 2, 2019 to November 11, 2020|
|$3,500||Cumulative Purchases||Weekly Purchases||$450|
In addition to the tremendous Federal Reserve securities asset expansion, the pandemic economic shutdown has brought a return to the Federal Reserve operating in private markets and aiding financial institutions. The private market aid for the pandemic includes all of the aids for the 2008 crisis plus several more specific to the characteristics of this broader crisis. Each of these so-called facilities have the characteristics of Federal Reserve investment in the economy and just as in the 2008 crisis will be revenue generating. The total list is: Commercial Paper
(PDCF), Primary Market Corporate Credit Facility (PMCCF), Secondary Market Corporate Credit Facility (SMCCF), Term Asset-Backed Securities Loan Facility (TALF), FIMA Repo Facility (FIMA), Main Street Lending Program (MSLP). The Treasury has an equity position in the both of the Corporate Credit Facilities and the TALF. The TALF will enable the issuance of asset-backed securities (ABS) backed by student loans, auto loans, credit card loans, loans guaranteed by the Small Business Administration (SBA), and certain other assets.
Figure 15 is the Covic-19 version of Figure 6. In it many of the items are the same but as the above paragraph shows there are at least four additional facilities. These recognize that the government mandated economic shutdown for Covid-19 affected markets much more than a simple recession, which absent the 2008 financial crisis, the 2008 Federal Reserve actions were adequate to handle. The additional facilities for this last crisis include the two corporate facilities, the municipal facility, the TALF, the FIMA, and the MSLP. The figure also shows total Federal Reserve loans and their composition among Money Market Fund Liquidity, Primary Dealer Credit and Paycheck Protection.
Figure 15. Covid-19 Special Facilities
March 4, 2020 to November 25, 2020
|$70||Commercial Paper Funding||Corporate Credit||Municipal||Primary Dealer Credit|
|Money Market Fund Liquidity||Paycheck Protection||MainStreet||TALF|
Table 16 shows the massive Federal Reserve response to the economic shutdown initiated in March of 2020. It is essentially the extension of Figure 12 for the pandemic period but begins on January 1, 2020 to show the magnitude of the tremendous Federal Reserve response. It is clear from the figure that in spite of much discussion of the facilities shown in Figure 15, the Federal Reserve’s major response was in the securities market. The Federal Reserve in the first three months after March 1, increased its holdings of securities by 55% from the February 26 level of $3.848 trillion to a June 3, level of $5.972 trillion. Then by November 11, 2020 securities holdings were $6.555 trillion up 70% from the last pre-Covid-19 reporting date.
|Figure 16. Federal Reserve Net Assets|
|January 1, 2020 – November 25, 2020|
|4,000||Other Income Earning Assets|
In marked contrast to this Federal Reserve Covid-19 securities response its 2008 response in the first five months was a 5% increase in securities holdings. It should come as no surprise then that the money supply responses to the two crises were remarkably different. Figure 17 shows the growth of two measures of the money stock, M1 and M2, relative to the beginning of both the Great Recession and Covid-19 crises. The base for the Great Recession is August 27, 2008, the last Wednesday reporting date before the September 2008 financial crisis. The base for the Covid-19 crisis is February 26, 2020 the last reporting date before the onset of the economic shutdown in March of 2020. The contrast between the two money supply responses is easily apparent. Not surprisingly considering the tremendous difference in the securities responses for the two crises, the money growth in the Covid-19 response is markedly greater than in the Great Recession response. The Covid-19 M1 growth has been more than 30% in just 20 weeks while the Great Recession M1 growth was just over 15% while the Covid-19 M2 response was almost
20% and the Great Recession response was just over 5%.
|Figure 17. M1, M2 Growth – Great Recession Versus Covid-19|
Once again, just as in the post 2008 budget crisis, the fear is that the tremendous expansion of the Federal Reserve assets will lead to double digit inflation. But also, just as in the 2008 budget crisis, the Federal Reserve did much more than expand its securities holdings. The 2008 non-securities expansion all went away within 16 months as shown in Figure 6. Once the economy opens up fully the special facilities associated with this crisis will disappear. Also, just as it did in the 2008 crisis the Federal Reserve will profit from its ability to supply liquidity when others could not. But even after the special facilities have all disappeared, what will remain is the tremendous increase in Federal Reserve securities holdings. The challenge will be to take actions that will prevent these assets from producing increases in the money supply that will lead to double-digit inflation. The Federal Reserve succeeded in doing just that in the 2008 crisis. The critical question is, can it do it again?
The burden of the debt is essentially the cost to taxpayers of paying the servicing cost of the debt. The debt becomes a crisis when servicing that debt puts such a burden on taxpayers that will prevent the economy from achieving its full potential. From the perspective of the Federal Reserve the question is how much can the Federal Reserve absorb of the Covid-19 debt increases without inflation?
The April 28, 2020 CBO projection for fiscal 2020 forecasted a federal budget deficit of $3.7 trillion, more than double the previous record of $1.4 trillion achieved in fiscal 2009. The actual federal budget deficit for fiscal 2020 was $3.132 trillion, more than double the previous record fiscal 2009 deficit of $1.413 trillion. The question is how much of the fiscal 2020 deficit was financed by the Federal Reserve. In fiscal 2020 the Federal Reserve added $2.847 trillion to its
securities holdings, 91% of the entire fiscal 2020 deficit. In addition, the Federal Reserve added another $200 billion in earning assets in the form of loans and facilities holdings. In total then, in fiscal 2020 the Federal Reserve has added assets covering just over 97% of the fiscal 2020 deficit.
On the surface it would seem then that the Federal Reserve has essentially financed the entire fiscal 2020 deficit. If so, then the dire forecasts of the CBO that publicly held debt would exceed 100% of GDP in fiscal 2020, if corrected for the fact that Federal Reserve revenue is owned by the Treasury, will not happen. But while the Federal Reserve assets have ballooned, its liabilities have as well. As a result, while Federal Reserve net assets for fiscal 2020 have increased only $1.95 trillion. Even corrected by the increase in Federal Reserve liabilities publicly held debt will have increased by the difference in official fiscal 2020 deficit of $3.132 trillion and the Federal Reserve financed $1.95 trillion, or $1.12 trillion.
Assuming that fiscal 2020 GDP will be unchanged from 2019 the unadjusted GDP share of publicly held debt is 93%. But adjusting for Federal Reserve net debt holdings the close of fiscal 2020 publicly held debt share of GDP is 75.8%.
All Federal Reserve revenues after costs and payment to bank owners must be transferred to the Treasury. As a result, in the usual meaning of the **** “own”, the Treasury owns the Federal Reserve. It is the residual income recipient of Federal Reserve revenues. Importantly, the Treasury ownership status does not give Treasury the right to vote on Federal Reserve policy in the way that corporate owners have that right.
Whether or not the Treasury and Federal Reserve are connected in a political way, the Treasury is the residual income recipient of Federal Reserve revenues. Therefore, both Federal Reserve assets and income affect the taxpayer burden of federal debt. The federal debt can and should be adjusted for Federal Reserve’s connection to the Treasury when evaluating the burden of that debt on the taxpayers. Such an adjustment can be done using a balance sheet or an income statement approach.
Using the income statement approach the level of debt is determined by the net of Federal Reserve transfers debt service cost. In 2014 the transfers from the Federal Reserve were $99 billion and accounted for 43% of the net debt service cost. But these transfers have been falling and by 2019 were only $55 billion and covered only 14.8% of debt service cost. If the official publicly held debt at the close of 2019 of $16.6 trillion is reduced by the share of the servicing cost that is provided by the Federal Reserve the adjusted publically held debt for 2019 is $14.1 trillion.
An alternate approach to this calculation is to offset the federal debt by the assets held at the Federal Reserve. This approach has merit since it is a balance sheet approach given that the flows from Federal Reserve net assets accrue to the Treasury. Essentially, the net assets of the
Federal Reserve now stand at $3.5 trillion; the difference between primary asset holdings of $6.5 trillion and $3.0 trillion liability composed primarily of bank reserves on which the Federal Reserve must pay interest. This measure of an adjusted 2019 federal debt is $13.1 trillion. But given the fiscal 2020 deficit of $3.132 trillion the 2019 official publicly held federal debt of $16.6 trillion adjusted for the fiscal 2020 deficit is $19.7 trillion. Then finally adjusting the close of fiscal 2020 publicly held debt of $19.7 trillion for the close of end of fiscal 2020 Federal Reserve net assets of $3.76 trillion yields an adjusted the fiscal 2020 publicly held federal debt of $15.9 trillion.
How burdensome will the publicly held debt will be depends on the response of the economy post the pandemic. If we assume that 2020 GDP will match 2019 GDP then the adjusted for the Federal Reserve publicly held debt will be 74.5% of GDP. In contrast the 2019 adjusted for Federal Reserve holdings was 60.4% of GDP. At the time of the deadline for this chapter, much was still unknown. However, the essential point made here is that whatever the final result the burden of the federal debt on the taxpayers of the nation is affected by the fact that the Federal Reserve will play a significant role.