Today, central governments may have to issue bonds to cover net spending, but the central bank of the currency in which the debt is denominated has long been able to control the price. With the advent of Real-Time Gross Settlement (RTGS) systems and Central Securities Depositories (CSD), the central bank has had to provide the liquidity needed to keep payment and securities settlement systems functioning. That has given it an increasingly direct influence over the cost of financing government issues.

Over the past few months, as demand for longer-term government debt has weakened and yields have risen, some commentary, particularly within MMT discussions, has suggested that the private market freely determines the price of government debt, including the sub-sovereign debt issued by Eurozone member states. Taken at face value, this implies that the private sector can set those yields at whatever level it chooses.

This post explains why that isn’t the case, why interest rates on government debt have a cap, regardless of debt or deficit size, and why that cap is tightest and easiest to engage at the short end.

As institutional arrangements stand, all government deficit spending, across all types of governments, is covered by issuing bonds. Only governments constrained in how many bonds they can issue end up being limited in their deficit spending. This includes the UK’s sub-sovereign assemblies, the state governments of various federations, and even the US federal government itself due to the infamous “debt ceiling”. In contrast, the UK central government has no explicit limitations.

What’s interesting about the Eurosystem is that many member states have no entrenched legislative restrictions on issuing government debt either. They have a treaty restriction on direct central bank financing, which has broadly the same effect as the Federal Reserve’s inability to lend directly to the US federal government. France, for example, is a sovereign state but a currency user within the Eurozone. Its government does not issue or individually control the currency it spends, and it finances its deficit through market borrowing, with parliamentary authorisation.

The national legislature sets the quantity and form of government borrowing, and the settlement system caps the price. That cap on interest rates on government bonds stems from the shift towards gross settlement systems. Banks used to pay each other during the day and settle only the net difference at the end of the day. That left each bank exposed to the other’s failure in the meantime. If a global financial crisis hit in the morning and a bank went bust that day, its counterparties could take a massive loss. The demand for certainty, and for money to move faster within the day, has led to systems with immediate full settlement. Payments become “gross”, in other words. The catch is that gross settlement requires substantial liquidity at the central bank.

The result is a Mexican standoff in which banks become “receipt-reactive” 1, leaving payments to the end of the day in the hope that another bank will pay them first so they can recycle the liquidity.

To address this, central banks introduce liquidity facilities and liquidity-saving mechanisms. Intraday credit and auto-collateralisation facilities, in particular, depend on the availability of “high quality collateral”, with government bonds forming a particularly important part of that collateral. Alongside the development of the wholesale repo market and the slow withdrawal of Quantitative Easing (QE), government bonds have become essentially liquid at no cost during the day and at low cost overnight.

In effect, regulated banks have become money changers at the temple, swapping one sort of money for another and charging a fee. It is no longer tenable to argue that bonds “sterilise” liquidity in quantity terms. In addition, many years of QE, involving substantial increases in central-bank balances and commercial-bank deposits without a corresponding or proportionate increase in consumer prices, make the argument that monetary expansion automatically produces inflation indefensible. In the recent Treasury bill consultation run by the UK Debt Management Office, no respondent mentioned that expanding bank deposits has had or would have any effect on inflation 2. Holding government paper is, of course, routine banking practice and has been for generations.

The net result is that the central bank is committed to providing ever-increasing liquidity support, particularly within the CSD 3. Even within the Eurosystem, the Correspondent Central Banking Model and Home Central Bank liquidity provision tightly bind the national central banks together 4. Removing or adjusting collateral limits then carries significant systemic and political risk.

This evolution and entanglement means the interest rate on government bonds is capped at a level that reflects the expected path of the base rate, the collateral conditions and the regulatory treatment of the intermediary’s balance sheet: the so-called “all-in” cost.5 If the yield on government bonds rises above that level, an eligible system participant can buy the bonds and finance the purchase through central bank facilities, using the bonds themselves as collateral. The difference between the bond yield and the cost of financing the position provides the intermediary’s profit. Other participants have an incentive to do the same, competing to acquire the bonds, bidding up their price and driving down their yield. In essence, the government can borrow from the central bank at a set rate, with the intermediary acting as a middleman and charging a fee.6 That fee is subject to competitive compression.

The cap is tightest on bonds with maturities under twelve months, which banks will hold directly. Longer bonds are also capped, since an intermediary can repo them at the central bank, but the range is much wider because the expected path of the base rate has a much greater dispersion, making the haircut much larger and the risk much greater.7

As the base rate rises, the discount to redemption value increases. Through the pull-to-redemption effect, short-term government bonds with low haircuts can accrete enough value through their remaining lifetime to cover the initial haircut, reducing the “all-in” cost.

Any government wishing to cap its interest costs should move to short-term paper 8 as it limits the dispersion effect of the expected path over the lifetime of the bond, engages the accretion effect and taps into the multiple bid effect (with both banks and their customers bidding for the bonds with effectively the same money). In the Eurosystem’s case, governments should use at least some intermediaries in other member states to increase the entanglement effect and bring wider EU Treaty rights and protections into play.

The private market cannot set government debt prices at will. Governments may have to issue bonds to cover net spending, but RTGS and CSD have forced central banks to supply liquidity using government bonds as collateral. That caps the yield at an “all-in” cost, because any system participant can step in as a middleman above it. The range is tightest with short-term government bonds and widens with maturity, which is why Warren Mosler’s three-month bills remain the way to keep interest costs down.

At least until sense prevails and we can get rid of government bonds altogether.


Chat about this and any other MMT topics on Discord with the growing New Wayland community . New members can click this invite link which will add the server to your Discord account