Central banks and financing fiscal deficits

The Bank of England has finally decided to initiate the direct finance of the government budget. The UK thus becomes the first major economy, with surely more to follow, to announce deficit financing. The joint statement by the Treasury and the BoE noted,

As a temporary measure, this will provide a short-term source of additional liquidity to the government if needed to smooth its cashflows and support the orderly functioning of markets, through the period of disruption from Covid-19. The government will continue to use the markets as its primary source of financing, and its response to Covid-19 will be fully funded by additional borrowing through normal debt management operations. Any use of the Ways and Means (W&M) facility will be temporary and short-term. As well as temporarily smoothing government cash flows, the W&M facility supports market function by minimising the immediate impact of raising additional funding in gilt and sterling money markets. The W&M facility is the government’s pre-existing overdraft at the Bank. Any drawings will be repaid as soon as possible before the end of the year.

In other words, the W&M facility expansion provides an insurance cover for the government in its bond issuance efforts. In case the gilt markets are not able to absorb the debt offerings, the Treasury can borrow unlimited amounts in the short term from W&M without having to tap the gilt markets. This comes on top of the commitment by BoE to temporarily print £200 bn to pump into the government bond market to ensure there was sufficient demand for gilts. 

There has been mounting support within the country for  taking the plunge on deficit financing. An FT editorial board calling for printing money wrote,

The scale of today’s downturn means even the most direct monetary financing, such as “helicopter money”, or handing cash to the public, should remain an option. This will require co-ordination with democratically elected officials, who are responsible for the public finances. The debate should not be over whether monetary financing can happen — in QE, it already is — but over keeping the process under control via independent central banks.

But in order to prepare ground and allay fears that it would undertake "monetary financing" through a permanent expansion of the central bank balance sheet to fund the government, the Governor of the Bank of England Andrew Bailey had just three days back publicly rejected any such plans,

This type of reserve creation has been linked in other countries to runaway inflation. That is because it could undermine a central bank’s ability to control monetary conditions over the medium term. Using monetary financing would damage credibility on controlling inflation by eroding operational independence. It would also ultimately result in an unsustainable central bank balance sheet and is incompatible with the pursuit of an inflation target by an independent central bank.

Accordingly, the announcement has stressed the temporary nature of this funding program. 

The Covid 19 outbreak has necessitated unprecedented co-ordinated fiscal policy stimulus by governments across the developed world. The fiscal measures announced till end of March stood as below.

However, it is worth highlighting that only a small portion of the fiscal spending measures announced  by western governments so far are on-budget. The vast majority are deferrals (of tax and other social security dues etc) and liquidity provisions and guarantees. These are either revenues  deferred or contingent liabilities on the government.

In other words, the direct stimulatory effects of the measures announced in countries like UK may not be as large as it appears. Deferrals are only about buying time, and does not entail any revenue foregone by the government. And guarantees and liquidity are contingent on the extent of its drawl and default on the credit.

So how does deficit financing work?

It is important to understand the nature of transactions involved. Central banks control the level of benchmark interest rates by varying the quantity of their reserves, which are interest-bearing deposit accounts held at the central bank by various commercial banks and which are backed by the central bank's assets, namely Treasuries. The commercial banks, in turn, multiply these reserves through fractional reserve banking whereby banks lend a portion of the deposits they have on hand.

Andrew Bailey the Governor of the BoE points to two ways in which central banks typically create reserves as part of their regular operations to maintain monetary and financial stability,

The first type is when we undertake liquidity provision operations which are too short term to have an enduring influence on monetary conditions, but nonetheless have a short-term effect on the money supply. Examples of these include our recent provision of liquidity to the banking sector (the £200 bn facility) and purchase of commercial paper in the new Covid Corporate Financing Facility. The BoE also works with the Treasury to support the orderly functioning of the gilt and money markets. Short-term operations play an important role in stabilising market conditions and counteracting any immediate tightening of monetary conditions. These have only a very temporary effect on monetary conditions and are not primarily tools that can be used to achieve the inflation target in the medium term.

We also create reserves when we undertake operations that are also temporary but are designed to have an impact on monetary conditions in the medium term. Quantitative easing, where the BoE buys bonds, is one example. QE increases bond prices and therefore reduces yields, which in turn lowers borrowing costs and support spending. The crucial point is that the MPC remains in full control of how and when that expansion is ultimately unwound. The goal is to ensure that borrowing costs and spending are consistent with achieving the inflation target. If the recent expansion of bond buying appears to threaten that goal, the MPC can react.

In this context, Gavyn Davies writes about the types of deficit financing,

A fiscal stimulus can be financed in three main ways. The government can sell almost unlimited quantities of short-term Treasury bills. This is normally the first recourse in the case of an unexpected surge in the budget deficit. Slightly later, the Treasury may increase the sale of longer-term debt issues to the public. Alongside that, the central bank may also increase its purchases of government debt from the public, in effect “monetising” the deficit for as long as the central bank balance sheet increases.

In more general terms, deficit monetisation happens when the government issues debt and simultaneously the central bank puts that debt on to its balance sheet as an asset by purchasing it and replacing with credit (or cash). It does this not by printing money but issuing the credit to the reserves held by banks (and putting that on its own balance sheet as liabilities). The banks in turn lend out these excess reserves added.

This often interchangeably used with helicopter money. But there is a subtle difference. Ben Bernanke, the modern originator of the term, defines helicopter money this way,

A “helicopter drop” of money is an expansionary fiscal policy—an increase in public spending or a tax cut—financed by a permanent increase in the money stock. To get away from the fanciful imagery, for the rest of this post I will call such a policy a Money-Financed Fiscal Program, or MFFP.

To illustrate, imagine that the U.S. economy is operating well below potential and with below-target inflation, and monetary policy alone appears inadequate to address the problem. Assume that, in response, Congress approves a $100 billion one-time fiscal program, which consists of a $50 billion increase in public works spending and a $50 billion one-time tax rebate. In the first instance, this program raises the federal budget deficit by $100 billion. However, unlike standard fiscal programs, the increase in the deficit is not paid for by issuance of new government debt to the public. Instead, the Fed credits the Treasury with $100 billion in the Treasury’s “checking account” at the central bank, and those funds are used to pay for the new spending and the tax rebate. Alternatively and equivalently, the Treasury could issue $100 billion in debt, which the Fed agrees to purchase and hold indefinitely, rebating any interest received to the Treasury. In either case, the Fed must pledge that it will not reverse the effects of the MMFP on the money supply.

In other words, unlike with normal deficit monetisation, in case of helicopter money, the central bank purchases and holds the debt indefinitely. This money financed fiscal program, unlike a debt financed one, does not increase future tax burden. 

Or, as the FT writes, there is a subtle distinction between the QE being followed now and monetary financing, 

There is no clear distinction between quantitative easing and monetary financing. Central bankers say asset purchases under QE are temporary, meaning the newly-created money will one day be removed from the economy... Recent QE programmes, in fact, look increasingly likely to become permanent. Central bankers were unable to complete a much-discussed programme of “normalising” monetary policy between the financial crisis and today’s crash. They are not going to be able to do so any time soon. The scale of previous schemes means the Bank of Japan — which holds government bonds worth more than 100 per cent of Japanese national income — may never be able fully to unwind its purchases. The difference between QE and direct monetary financing is mostly one of presentation: whether asset purchases are deemed temporary or permanent. This matters: credibility and messaging are important features of central banking.

But like with all fiscal policy, there are no free lunches. The central banks, when it credits private banks with additional central bank reserves. But the central bank has to pay interest on these additional reserves. Gavyn Davies again,

Since central banks are wholly owned by their governments, their balance sheets should be consolidated fully into the public sector. These reserves are therefore equivalent to interest-bearing loans from the private to the public sector, a form of public debt almost identical to the issuance of Treasury bills by the government.

Narayana Kocherlakota explains this in terms of two important implications,

Pure monetary financing of budget deficits is more restricted than is assumed by the proponents of helicopter money. It can be done by printing extra banknotes, but the scale is severely limited by the private sector’s willingness to hold physical cash instead of bank deposits.

As David Mericle of Goldman Sachs explained in research for clients last week, almost anything that can be achieved by an expansion in the central bank’s balance sheet can also be achieved by the fiscal authorities. For example, the finance ministry could issue Treasury bills to fund an increase in the budget deficit, without involving the central bank. This would change the overall composition of public-sector liabilities just as if the central bank financed the deficit by increasing reserves.

This was summarised by Gertjan Vlieghe, external member of the Bank of England monetary policy committee,

If the central bank pays interest on reserves, helicopter money is really just a fiscal expansion, financed by interest-bearing reserves. Interest is still payable. It is not that this would make it ineffective, it is just that it makes it little different from debt-financed fiscal expansion, other than unnecessarily making the central bank more involved in fiscal policy.

However, one way to overcome this problem is to raise the minimum required reserves themselves, thereby eliminating the need to pay interest on them.

This is a brief history of RBI's deficit monetisation practice,

Historically, India’s deficits were automatically monetised. This was done through issuance of 91-day non-marketable ad-hoc treasury bills to the RBI which in turn increased reserve money. Attempts were made since the late 1980s and early 1990s to gradually stop monetisation of the deficit. The government and the RBI, through two agreements, signed in 1994 and 1997 agreed to completely phase out funding through these ad-hoc treasury bills. To ensure fund flow to the government, a system of ways and means advances was introduced in 1997 that allowed the RBI to give short term advances to the government that were completely payable in three months. Despite the two agreements that were signed, the RBI still continued to subscribe to the primary issuances of public debt when the issuances were not fully subscribed to in the market.

However, with the enactment of the Fiscal Responsibility and Budget Management (FRBM) Act, 2003, this practice was stopped... The Act barred the RBI from participating in primary issuances of government securities from 1 April 2006. The FRBM Act provided an escape clause and said the RBI could subscribe to the primary issue of central government securities in case the government exceeds the fiscal deficit target on “grounds of national security, act of war, national calamity, collapse of agriculture severely affecting farm output and incomes, structural reforms in the economy with unanticipated fiscal implications, decline in real output growth of a quarter by at least three per cent points below its average of the previous four quarters”. The deviation allowed in the Act was 0.5 percentage points.

In India too, in the context of the fiscal deficit breach, there have been suggestions that it be financed not by issuing bonds, which would crowd-out private borrowers, but by RBI directly funding the government. Sample this from Jahangir Aziz,

Any large bond auction by the government, even if it is offset by the RBI through open market operations, is not likely to calm market nerves and bring down lending rates. What is needed is for the government to invoke the “escape” or the “natural disaster” clause in the fiscal responsibility act (FRBM) that allows the RBI to directly fund the budget deficit without having to go through market auctions.

In this case, as the BoE did, the RBI will have to clearly communicate that its balance sheet expansion is for a defined period and for specific amount. This is essential to generate market confidence that the monetisation is only a temporary measure given the extraordinary circumstances. However, the RBI or GoI, for a variety of structural reasons, do not have the same level of credibility and confidence of the markets (foreign investors in particular) as the BoE and UK Treasury enjoy. The impact on the rupee can therefore be uncertain. 

If the RBI finances the government directly, then it would purchase Treasuries from the government directly and issues credit to the government's account held at the RBI. Government's expenditures will be made through the commercial banks, in turn driving up their (banks') reserves with the RBI, probably not proportionately.

Since the reserves would now exceed the cash reserve ratio (CRR), the RBI would have to pay interest as notified by the reverse repo rate. However, this cost will have to be offset against the interest receipts by RBI for holding the treasuries. Since the yields on the G-Secs in India are typically higher than the reverse repo rate, the RBI would still continue to make a profit in the transaction. However, on the consolidated RBI plus government balance sheet, it would be a net outflow in terms of the reverse repo interest payouts.

See this on the US Federal Reserve's lending program.

Update 1 (17.04.2020)

Niranjan Rajadhyaksha argues in favour of deficit monetisation to the extent of 0.75% of GDP.

Subhash Chandra Garg, the former Secretary Department of Economic Affairs, assesses the fiscal deficit financing requirement to be about Rs 10 trillion, divided equally between additional expenditure requirement and revenue shortfalls. He argues in favour of deficit monetisation,

The savers cannot take this up. Asking banks and other financial institutions to provide financing would roil the credit markets. The real economy will get starved of credit, which is much more needed in these disruptive times. RBI has provided direct subscription to GOI bonds earlier. I have dealt with the issue of how amendment of FRBM law would allow the RBI to meet this extraordinary requirement. It is advisable that the RBI subscribes to these bonds directly. Subscribing indirectly i.e. RBI buying virtually equal amount of bonds from the secondary market to create space for these institutions to subscribe to new government bonds means the same thing effectively.

Primer on RBI's W&M Advances here.