“With a Federal debt which is so large a part of all debts, public and private, which permeates and dominates to some extent the whole securities market, and which has become the principal medium for adjusting portfolios of financial institutions, and the reserves of banks and others, we are not and won’t be wholly free to administer credit policy without regard to the Government security market, and without regard to Treasury financing requirements.  It won’t be enough to say to the Treasury, here is the credit policy we are going to follow, now you manage the debt.”

 Allan Sproul, President, Federal Reserve Bank of New York, January 27, 1953.

“The executive committee voted unanimously to direct the Federal Reserve Bank of New York to […] make such purchases, sales or exchanges [….] as may be necessary in light of current and prospective economic conditions […] with a view to exercising restraint upon inflationary developments, [or] correcting a disorderly situation in the Government securities market...”

Minutes of the Executive Committee of the FOMC, March 5, 1953., January 27, 1953.

Overview

Background

The Government Budget Constraint and the Fed’s Balance Sheet

DGDP(t)= \dfrac{(1+R(t))}{(1+growth(t)) } DGDP(t-1)+deficit(t)-remittances(t)

remittances(t)=R_Fed(t)∙FedAssets(t-1)-i∙(FedAssets(t-1)-currency(t-1))

Where FedAssets and currency are the ratios of Fed assets and currency to GDP, respectively, R_Fed(t) is the weighted average interest rate on Fed assets held at the end of period t – 1 , and i is the weighted average rate on the Fed’s interest-bearing liabilities.  For the purposes of this note, we take i to be the interest rate on reserve balances.   The first term in equation (2) is just the interest earned on Fed assets.  The second term captures the interest costs of the Fed’s liabilities.

Equations (1) and (2) provide some insight into the various ways in which Fed balance sheet actions might affect the trajectory of the DGDP ratio.^[8]  First, increasing the size of the Fed’s balance sheet relative to GDP, FedAssets , can boost the level of Fed remittances as long as the Fed enjoys a positive net interest margin.  In addition, Fed balance sheet actions may affect interest rates on government securities captured by R(t).

The Fed’s balance sheet decisions may indirectly affect the yields on longer-term debt by influencing the level of the so-called term premium.  This effect is thought to work through basic demand and supply channels.  If the Fed acquires a larger share of longer-term securities outstanding, all else equal, that will reduce the quantity of longer-term securities held by the non-Fed public and should drive up the price of those securities.^[9]  The rise in the price of long-term securities drives down the yield on longer-term securities and boosts the value of other financial assets. The rate that matters for the evolution of the DGDP ratio is the weighted average rate on all debt outstanding.  As a result, Fed decisions about the setting of short-term interest rates and any increases in term premiums affecting longer-term yields tend to feed through only gradually to the weighted average rate on Treasury debt outstanding.

Prior to 2007, the size of the Fed’s balance sheet was a little above 6% as a share of nominal GDP.  Moreover, the size of its balance sheet was constrained by the need to maintain reserves in a very narrow range consistent with the desired level of short-term interest rates.  In the current monetary policy implementation regime, the Fed’s balance sheet is much larger—currently about 22% of nominal GDP.  Under the ample reserves regime, the Fed has greater latitude to expand the size of its balance sheet while retaining control over the level of short-term interest rates than was the case pre-GFC.  As a result, Fed decisions about the size and composition of its balance sheet may have greater import for the fiscal position of the government than was the case pre-GFC.  Below we use a simple debt-to-GDP “calculator” based on equations (1) and (2) to illustrate the potential effects of alternative Fed balance sheet policies on the DGDP ratio over time.

A DGDP Calculator and Scenarios

The appendix describes a simple “DGDP calculator” that can be used to trace through some of the potential interactions among the fiscal outlook, Treasury debt management, and alternative Fed balance sheet policies.  The core elements of the calculator are equations (1) and (2) combined with a set of assumptions about the structure of federal debt outstanding, the Treasury’s approach to debt management over time, and the Federal Reserve’s balance sheet management along with a specification for the evolution of the term structure of interest rates.^[10]

In the analytical framework underlying the DGDP calculator, Treasury and Fed actions that remove longer-term securities from the private sector tend to reduce federal government financing costs over time through two channels.  The first channel is the supply and demand effect on term premiums discussed above.  In addition, Fed purchases of longer-term securities may drive down government financing costs if the yields on long-term debt are above the level of the interest rate on reserve balances.  In that case, larger holdings of longer-term securities in the Fed’s portfolio tend to increase Fed remittances and so reduce the government’s funding need.^[11]

A Baseline Scenario

In the baseline scenario, the calculator is calibrated to roughly match key characteristics of the longer-term projections for federal debt from the Congressional Budget Office and the Yale Budget Lab.  Key assumptions underlying the baseline scenario include: (i) the Treasury seeks to maintain the weighted average maturity (WAM) of total debt outstanding at 6 years; (ii) the Fed seeks to reduce the WAM of its Treasury securities portfolio from 9 years at present to 6 years over time; (iii) the Fed maintains the size of its balance sheet at 22% of nominal GDP over time (a level assumed to be consistent with ample reserve conditions); nominal GDP growth over the projection period runs at 3.5%; (iv) the interest rate on reserve balances is set at 3%; (v) each 1 percentage point increase in the ratio of the ten-year equivalent value of private holdings of Treasury debt relative to GDP pushes up term premiums for longer-term securities by about 2 basis points.^[12]

Under these assumptions, as shown in Figure 4, the DGDP ratio increases from about 1 at present to about 1.8 by 2055—about the same increase projected by the Yale Budget Lab.  As shown in Figure 5, as the WAM of the Fed’s portfolio declines over time to match the Treasury’s targeted WAM for all Treasury debt outstanding; the WAM of Non-Fed debt increases gradually to 6 years as well.  This upward trend in the WAM of non-Fed debt along with the expansion of non-Fed debt relative to GDP puts upward pressure on term premiums over time.  As a result, as shown in Figure 6, the yield curve gradually ratchets up over time.  As shown in Figure 7, the government’s net interest expense—gross interest expense less remittances from the Fed—as a share of GDP increases from about 3% in 2025 to more than 6% by the end of the projection period.  Figure 8 shows the evolution of the “gross interest accumulation factor”—the term

\dfrac{(1+R(t))}{((1+growth(t)) }

​ in equation (1) above.  This ratio is a key factor influencing the trajectory of federal debt over time.  When the level of this factor is below 1, the DGDP ratio tends to decline over time if the federal government can keep the primary deficit at zero.  When the factor is above 1, the government needs to run primary surpluses just to keep the DGDP ratio constant over time.  In the baseline projections, the gross interest accumulation factor moves above 1 in 2040.

Alternative Scenarios

This section considers six alternative scenarios embedding changes to some of the key parameters in the baseline scenario.  Three of the scenarios focus on possible adjustments to the Fed’s balance sheet.  The short Fed WAM scenario assumes that the Fed adjusts the desired long-run WAM of its asset portfolio from 6 years to 4 years.  The long Fed WAM scenario assumes the Fed adjusts the desired long-run WAM of its portfolio from 6 years to 8 years.  The large Fed scenario assumes that the Fed increases the long-run size of its balance sheet as a share of GDP from an assumed level of 22% in the baseline to 30%.

The remaining scenarios focus on broad demand and supply factors in the Treasury market and the economy.  The short Treasury WAM scenario assumes the Treasury adjusts its long-run target WAM for federal debt from 6 years to 4 years.  The increased Treasury demand scenario assumes that increases in the demand for Treasury securities over time offset upward pressure on the level of term premiums and longer-term interest rates associated with Treasury supply pressures in the baseline scenario.^[13]  The high growth scenario assumes that nominal GDP runs at 4% over the projection period (2% real GDP growth and 2% inflation) versus the 3.5% rate assumed in the baseline (1.5% real GDP growth and 2% inflation).

Figure 9 displays the effects of these alternative scenarios on the term structure of interest rates at the end of the projection period in 2055.  The dotted lines display the scenarios related to the Fed.  In the short Fed WAM scenario, the Fed holds lower amounts of longer-term securities and the non-Fed public holds more, putting some modest upward pressure on yields relative to the baseline (shown by the solid black line).  Conversely, the long Fed WAM scenario reduces the quantity of longer-term securities held by the non-Fed public, putting some downward pressure on term premiums and longer-term interest rates.  The large Fed scenario (orange dotted line) also reduces the quantity of longer-term securities held by the non-Fed public and pulls yields a little lower relative to the baseline.  The effect of the large Fed scenario is relatively muted because it has little effect on the WAM of debt held by the non-Fed public.  A large Fed scenario combined with a long Fed WAM assumption would have a more powerful effect on yields.

Turning to the other scenarios, the short Treasury WAM scenario (solid pink line) reduces the total supply of longer-term securities quite substantially over time.  As a result, relative to the baseline, the yield curve revises down appreciably. The increased Treasury demand scenario (solid green line) illustrates the behavior of the yield curve when an increase in Treasury demand offsets the effect on term premiums stemming from the rising quantity of Treasury securities held by the non-Fed public.  In this case, the Treasury yield curve is constant over time and yields are substantially lower than in all of the other scenarios that incorporate term premium effects.  The high growth scenario (solid blue line) also tends to pull yields lower because the trajectory of debt held by the non-Fed public relative to GDP does not increase as rapidly as under the baseline.

Figures 10 to 12 show the corresponding results across the various scenarios for other key variables of interest.  The long Fed WAM scenario has a noticeable effect in reducing federal financing costs and the DGDP ratio over time.  Conversely, the short Fed WAM scenario puts some upward pressure on those same variables.  The large Fed scenario has meaningful, albeit smaller, effects in reducing federal financing costs over time.

The short Treasury WAM scenario tends to generate sizable effects on all of the key variables because it dramatically reduces the WAM of federal debt held by the non-Fed public and therefore damps the upward pressure on longer-term interest rates from the term premium effect.  Absent the term premium effect, a short Treasury WAM strategy would still have some effect on the government’s financing costs by substituting low-cost short-term debt for relatively high-cost long-term debt.  That effect just wouldn’t be amplified by the additional impetus from a decline in the term premium.

The high growth scenario also points to significant benefits in reducing the DGDP ratio over time.  Effects like this are sometimes noted as supporting the possibility that sufficiently robust economic growth would allow the economy to “grow out” of federal debt burdens over time.  However, it’s important to note that this scenario assumes that higher long-run real GDP growth is not accompanied by an increase in interest rates.  As a result, the gross interest accumulation factor in equation (1) declines relative to the baseline.  In many models, an increase in long-run real GDP growth is reflected in a corresponding increase in the long-run real short rate.  In that case, there would be very little change in the gross interest accumulation factor and little change in the DGDP ratio relative to the baseline path

A Bottom Line and Some General Comments

A bottom line in the baseline and alternative simulations discussed above is that Treasury debt management decisions and Fed balance sheet decisions cannot fundamentally alter the trajectory of the DGDP ratio.  The baseline and all of the alternative scenarios point to a substantial increase in the DGDP ratio over time driven by persistent primary budget deficits.  However, the simulations do suggest that Fed balance sheet decisions, and especially Treasury debt management decisions, can have meaningful effects on the DGDP ratio and related statistics over time.  In general, policy steps that effectively replace high-cost debt with low-cost debt or that work to lower the cost of longer-term debt tend to generate a somewhat less steep increase in the DGDP ratio over time.

The discussion above points to the possible benefits of reducing the WAM of Treasury debt outstanding.  Indeed, in the framework underlying the DGDP calculator, fiscal costs over time would be minimized if the Treasury issued only very short-term debt, thereby avoiding the term premium costs associated with longer-term debt.  Of course, there are important considerations that would weigh against this approach.  Issuing large volumes of short-term debt can expose the Treasury to substantial variation in interest payments over time reflecting the volatility of short-term interest rates.  Large volumes of short-term debt could also leave the Treasury exposed to rollover risk in the event of an operational problem or other disruption that resulted in a failed auction of Treasury debt.  These types of factors are important in debt management and are not included in the DGDP calculator.  However, the DGDP calculator may at least provide a sense of the pecuniary benefits of reducing the WAM of debt outstanding that may, in turn, be weighed against these potential costs of a shorter WAM debt portfolio.

As a final comment, all of the scenarios considered in this note assume that the economy remains on a steady state growth path with the Fed conducting an independent monetary policy aimed at maintaining inflation at 2 percent over time.  Of course, in the current environment, there is a great deal of interest in alternative scenarios in which the independence of the Fed is compromised and inflation moves higher.  The DGDP calculator is highly simplistic and analyzing a scenario like this would require a much more complete model of the economy and the yield curve.  That said, the high growth scenario hints at some of the issues involved.  In an adverse scenario involving loss of Fed independence, inflation would move higher and nominal GDP growth would then move higher as well assuming that real GDP growth was unaffected.  Higher inflation would also tend to boost longer-term yields, but the Fed might be able to offset some of that increase by adjusting the size and composition of the balance sheet.  As a matter of arithmetic, these types of effects could mechanically improve the DGDP ratio for a time.  However, in this type of very adverse scenario, the economy would likely not perform well and financial markets could be severely disrupted.  These effects would almost certainly further weaken the fiscal position of the government and greatly harm the economic well-being of all Americans.

Potential Longer Run Policy Issues

The discussion of the various scenarios above points to some ways in which the Fed could find itself indirectly influenced by government debt management decisions over time in pursuing its statutory mandate to promote maximum employment and stable prices.  All of the policy considerations discussed below are hypothetical.  However, as discussed more in the following section, some of these types of interactions between federal debt and the central bank have been present in other countries with heavy fiscal debt burdens.  And some of the policy considerations below are similar to those in prior periods of U.S. history—for example, the Fed’s role in wartime finance during World War II and in the immediate post-Accord period discussed in the introduction.^[14]

Potential Effects on Policy Implementation: One type of indirect effect on the Fed could operate through money markets and the Fed’s balance sheet.  In a scenario in which term premiums are increasing gradually over time, it would also likely be the case that demand for securities financing by dealers and others would put upward pressure on overnight interest rates and particularly the overnight repo rate.  The upward pressure on repo rates, in turn, could pull up other short-term rates, including the federal funds rate.  In this scenario, the Fed might conclude that the “ample” level of reserves had risen and that it needed to expand the size of its balance sheet.  In a sense, the ample level of reserves, broadly defined, would be higher but not because banks needed more reserves to operate efficiently but rather because the level of the Fed’s securities holdings needed to be higher to relieve financing pressures in securities markets.  As shown in the large Fed simulation, operating with a larger Fed balance sheet would help put downward pressure on longer-term interest rates and could reduce financing pressures in repo markets.

Potential Effects on Treasury Market Functioning: Yet another type of effect on the Fed could be connected with its role in liquidity provision.  In the simulations, the effect of increasing debt on yields plays out slowly and smoothly over time.  In the real world, those types of effects could occur in very unpredictable and large jumps accompanied by a “disorderly situation” in the Treasury market.  Indeed, strains in Treasury market functioning were intense in the early days of the pandemic, requiring extraordinary interventions by the Fed to support Treasury market functioning.  And the United Kingdom experienced a severe case of dysfunction in the gilt market in 2022 on market concerns about the federal budget outlook. In part to address the risks of Treasury market dysfunction, the Fed implemented two standing facilities—the Standing Repo Facility (SRF) and the FIMA repo facility—that provide financing to primary dealers and foreign official institutions, respectively, at a rate currently set at the top of the target range for the federal funds rate.^[15]  In the event of repo market pressures, primary dealers can turn to the SRF for financing to support their own securities inventories or to lend to other dealers and market participants.  Faced with the type of rate pressures in the baseline scenario described above, one could easily imagine that the Fed would find it necessary or expedient to expand access to these liquidity facilities or to create additional types of standing facilities over time to support smooth functioning in the Treasury market.^[16]  The Treasury may also take actions to support the liquidity of the market.  Indeed, the Treasury established a program of regular “liquidity support buyback operations” beginning in May of 2024 with the aim of relieving liquidity pressures in particular market segments.

Neutrality and Dueling WAMs: The simulations above note that the Treasury can lower the cost of debt finance over time by reducing the WAM of its issuance and the overall debt portfolio.  A parallel move by the Fed to shorten the WAM of its securities holdings has the opposite effect on interest rates.  As a result, if the Treasury opts to reduce the WAM of its issuance in an effort to reduce its financing costs and the Fed follows suit by reducing the WAM of its securities holdings to remain “neutral” with respect to the Treasury’s new policy, that would have the effect of offsetting a portion of the interest rate reductions that the Treasury was seeking to achieve.  Of course, a similar type of interaction may stem from Fed actions to lower long-term yields through large-scale asset purchase programs.  Some of the intended effects of such LSAP programs can be offset if the Treasury seeks to lengthen the WAM of its issuance at the same time.

More General Policy Interaction Effects: The simulations highlight the potential for federal debt growth to put upward pressure on term premiums and longer-term interest rates.  The simulations implicitly assume that short-term interest rates are the key factor driving aggregate demand in the economy.  As a result, in the simulations, the Fed does not need to respond to changes in term premiums and long-term rates in pursuing its mandate to promote maximum employment and stable prices.  However, many economic models posit that aggregate demand depends importantly on the level of longer-term interest rates.  All else equal, in this type of model, the increase in term premiums would tend to put downward pressure on spending and inflation over time.  As a result, the Fed would need to adjust the stance of policy in response to rising term premiums resulting from the increased supply of Treasury debt.  Echoing the theme from the opening quote from President Sproul, monetary policy and federal debt management in this case become deeply intertwined. Lowering the path of short-term interest rates would be one possible way the Fed could adjust to rising term premiums.  In effect, the long-run equilibrium level of the federal funds rate, r*, would have moved lower.  A lower value for r* makes the Fed’s job more difficult because it tends to bring the zero lower bound on interest rates into play more frequently in the event of adverse shocks to the economy.  Another type of policy response could involve the Fed adjusting the size and composition of its balance sheet to blunt the rise in term premiums and the corresponding decline in r*.  This could be accomplished by lengthening the WAM of the Fed’s securities holdings and/or increasing the size of its balance sheet. Other types of policy interactions are possible as well.  For example, as discussed above, if the Treasury sought to shorten the maturity of debt outstanding to reduce debt service costs, that would tend to put downward pressure on term premiums over time.  But the associated fall in longer-term interest rates could then boost aggregate demand and inflation pressures, leading to some form of countervailing interest rate or balance sheet response from the Fed.  In this case, the Fed’s actions in pursuit of maximum employment and stable prices could offset some of the savings on debt service costs that the Treasury was seeking to achieve.

Recent Case of Central Bank Long-Term Interest Rate Targeting – Bank of Japan’s Yield Curve Control Policies 2016-2024

Economic policy in Japan presents an interesting (and extreme) case of the nexus between monetary policy, federal finances and debt management.  While the Bank of Japan (BOJ) yield curve control regime is an extreme form of central bank intervention in sovereign debt markets, the Japanese experience highlights some of the issues that may arise in softer forms of central bank government debt involvement of the type discussed in this note.

In 2016, the BOJ broke from modern central banking norms when they began targeting longer-term interest rates. Specifically, under the yield curve control policy (YCC), the BOJ set a 0% target for 10-year Japanese government bond (JGB) yields with a permitted fluctuation range around the target. YCC policy was an extension of the BOJ’s quantitative and qualitative monetary easing (QQE) policies, which had been in effect since 2013. While JGB yields were already hovering around 0% when YCC was introduced, the effects of the policies were evident in financial markets: 10-year JGB yields remained remarkably stable for years. Yield volatility declined (and remained much lower than yield volatility in other developed economies).

In the case of YCC and the BOJ’s QQE policies more generally, the goal was to overcome deflation and to achieve the BOJ’s new higher inflation target of 2%, not to minimize financial costs for the government. As former BOJ Governor Haruhiko Kuroda described it:

Recently, the cooperation between monetary and fiscal authorities increasingly has been drawing attention globally, so let me explain the basic thinking on it. If the government implements fiscal policy when a central bank pursues monetary easing to achieve its inflation target, economic stimulus effects will become more powerful through their synergy effects. Such a combination of policies is called a policy mix, and this is a standard approach of macroeconomic policy…. In economic theory, a policy mix is assumed to be conducted in an appropriate manner through monetary and fiscal policies based on their own particular policy objectives. In recent discussions, however, some seem to confuse a policy mix with debt monetization by central banks.

BoJ Governor Haruhiko Kuroda, December 2019

Still, when YCC policies were implemented, many voiced concerns about whether fiscal management issues would impact YCC exit strategies. The concerns were largely for naught – when Japan’s inflationary situation changed radically coming out of the COVID shock (as was the case across most of the developed world), exceptionally easy monetary policies were no longer justified and the BOJ dropped its YCC policies.

That’s of course not to say Japan’s exit from YCC policies was perfectly smooth sailing. Throughout the end of 2022 and into 2023, investors were betting against the BOJ’s commitment to the yield target. The BOJ was forced to rapidly expand its balance sheet to defend the YCC yield target band. This had not been the case through the YCC period. The BOJ’s credible commitment to the target yield band had effectively de-linked the pace of balance sheet expansion from the policy to maintain a target long-term rate, that is until the credibility came into question.

Partially as a result of the rapid expansion of the monetary base (increasing rate differentials also played a role, as rates rose across developed economies, with Japan as the key exception, during this period), the yen also weakened significantly throughout 2022. This yen weakness was politically unpalatable (particularly as it was accompanied by a rise in global commodity prices, creating significant pain for Japanese importers) and fanning the flames of political criticism against YCC policies.

Between the yen weakness and related political criticism, the undesirable rapid balance sheet expansion, and mounting evidence that the inflation target had credibly been achieved, the BOJ dropped the yield curve control target altogether in 2024 (though it could be argued the policy was dropped in 2023 when the target range was shifted to a reference range).  After YCC was dropped, the BOJ’s balance sheet expansion effectively halted and has now run in reverse with the BOJ winding down its holdings of JGBs.

Of course, this discussion is not intended to imply that the Fed is inexorably moving on a path toward YCC at some point.  But the Japanese experience highlights some of the issues that can arise whenever the central bank becomes more actively involved in managing longer-term interest rates, even when that is part of achieving the central bank’s mandate.  In particular, market speculation about the central bank’s balance sheet policies can become a key factor influencing pricing of longer-term securities.

The BOJ and the Fed are far from the only central banks facing questions about whether actions taken in the name of effective monetary policy (in line with monetary policy mandates) strayed into the territory of fiscal issues and debt management. For example, the European Central Bank’s Transmission Protection Instrument (TPI), introduced in 2022, allows the bank to buy an unlimited number of bonds from a euro zone country facing disorderly and unwarranted tightening in financial conditions, provided the eurozone country is in compliance with the European Union’s fiscal rules.  There’s been speculation this year on whether the ECB would use TPI to intervene in the French bond market.  France, which is under an excessive deficit procedure by the European Commission, is facing rising spreads as the government has struggled to come to an agreement for how to cut public spending and bring the deficit back in line with European Union fiscal rules. Critics of TPI claim that the program artificially depresses yields, distorting incentives away from fiscal consolidation.

Conclusion

This note focused on the federal budget outlook and how it may interact with Fed policies over time in seeking to promote maximum employment and price stability and support overall financial stability.  Assuming that federal debt remains on a steep upward trajectory, Fed policies and procedures may be influenced by the need to support smooth Treasury market functioning and to mitigate any undesired tightening of financial conditions associated with rising debt burdens.  Such actions could include changes in the composition and size of the Fed’s balance sheet as well as the expansion or addition of new liquidity facilities to address periods of Treasury market distress.  Of course, one could hope for favorable developments that might obviate the need for these types of actions; for example, improvements in the outlook for primary budget deficits or an increase in Treasury demand that offsets some of the market effects of the increase in Treasury supply.  Absent such favorable developments, however, many of the debates faced by the FOMC in 1953 seem likely to be a recurring theme over coming decades.  No doubt the voice of Chairman Martin will be present in those discussions emphasizing the need to preserve and defend the independence of the Fed by steering clear of policy actions that interact with Treasury debt management to the maximum extent possible.  But the voice of President Sproul will be there as well noting the singular importance of the Treasury market and the inescapable connections between monetary policy and federal finances.