UK £ default is not only unlikely, it is de facto impossible

★ Document history

The first version of this document was written in late 2011, shortly before joining Société Générale. It was published under the SocGen name on 04 October 2011. Readers, public and private, sent some helpful comments, so a second version was published on 09 November 2011.

The © of the first version was mine, but the improvements in the second version belong to SocGen. SocGen has kindly allowed this third version to be based on the second version, though obviously responsibility for it lies with the author and not with that former employer.

I believe this to be an important thesis, worthy of much close attention from the rating agencies.

Usual disclaimer and copyright terms apply.

The first version of this essay was greatly improved by the legal assistance of Jacob Gifford Head, Barrister. This assistance is gratefully acknowleged.

★ Abstract: UK sovereign debt almost entirely in £, and its internal decision-making machinery could command the printing of £ in a crisis. Inflation possible, sure; devaluation of the curreny possible, sure; but default is de facto impossible.

UK £ default is not only unlikely, it is de facto impossible

★ Gilts may be issued: no ceiling

The legal authority to borrow stems from the National Loans Act 1968, which gives great discretion to His Majesty’s Treasury (HMT)†2. In particular:

So the law allows HMT to raise money, without legal limit, at any maturity from short to perpetual, in £, in other currencies, in SDR, and in precious metal, provided only that “the Treasury think fit”†3. Arguably it would even allow oil-denominated or oil-linked borrowing. There is no issuance ceiling, no debt ceiling, nor any other restriction that could cause a sudden stop.

★ Compelling the Bank of England to buy gilts

So HMT may issue, but what if investors don’t want to buy? More relevantly, what if investors don’t want to buy at a price at which HMT wants to sell? Can the Bank of England (BoE) be compelled to do so? Yes, by various means.

★ Maastricht rules

Until the UK actually leaves the EU, it has not left, which might be thought to introduce some international awkwardness. So the Treaty on The Functioning of The European Union (as the 1992 Maastricht treaty was renamed) continues to apply in the UK. From page 101 Article 123 of that treaty (“ex Article 101 TEC”):

This seems to prohibit the central bank purchasing government debt at auction. Seems to, but actually doesn’t.

Let us compare to the US. US Federal law is not only supreme, if necessary, it is enforced. For example, in 1957 Arkansas Governor Orval Faubus used his state’s National Guard to resist the implementation of the Supreme Court ruling in Brown v. Board of Education. So the central government, in the person of President Dwight D. Eisenhower, sent the Army, in the form of the 101st Airborne Division, to enforce Federal law. In very real practice, Federal law was enforceably supreme.

EU enforcement is far weaker. If a national politician insists on a course of action contrary to EU treaties, rules or judgements, well, that is how it is. A summit might be called. There might be a statement hinting at criticism of the recalcitrant. But there are no troops. There is no action. And indeed, even the statement might not happen: the sovereign states of which the EU is comprised have been even lighter on countries acting in accordance with a genuine and important national interest.

In other words, the treaty functions as a guideline broadly defining good behaviour. It is not enforceable. But the UK has a reputation for being bound by its word, and particularly in matters of national debt, might wish to maintain this reputation. Not a problem. Rearranging the phrases for legibility, “national central banks” are “prohibited” from “the purchase directly from” “central governments” “of debt instruments”. The purchase “directly”, eh? So the BoE bidding at auction is prohibited. But the BoE may still instruct a gilt-edged market maker (‘GEMM’) to bid on its behalf.

Indeed, this isn’t so far from practice during the BoE’s Quantitive Easing, when: the DMO sold; the GEMMs bid and bought; and shortly after the BoE bought from the GEMMs. If needed, the temporal gap could be shrunk.

★ The Bank of England Acts

This power is very general, and subjective. That is, if challenged, HMT would need to show that the direction was not “in relation to monetary policy”, and that it thought its order was in the public interest. But there is no requirement that it objectively be in the public interest—the thought suffices.

Strangely, it seems that neither Act specifies what is and what is not monetary policy.

It is arguable whether, in 1998, the purchase of gilts would have been construed as monetary policy. But events since then have expanded the practice of monetary policy: multiple central banks’ equivalents of the Monetary Policy Committee have bought debt, public and private, by various means. Whether or not it was, it has definitely become a tool of monetary policy.

Thus a direction compelling the BoE to bid for gilts could not be made under s.4 of the 1946 Act; it would have to be made under s.19 of the 1998 Act, which requires that such a direction be “approved by resolution of each House of Parliament”.

This is one of the standard types of statutory instruments†4, specifying the ‘affirmative procedure’. According to the relevant Parliamentary Factsheet (HTML, PDF), “Approximately ten per cent of instruments subject to Parliamentary procedure are subject to the affirmative procedure”. Rejection is not unprecedented, but is rare: there are about 3½k SIs per year, yet “the last occasion that the House of Commons annulled an SI was in 1979 and the House of Lords voted to annul the Greater London Authority Elections Rules 2000 in February 2000.”

★ Each House of Parliament

The UK is not the USA. By design, the USA has a divided government. And if one part should be determined to destroy the other, there might be collateral damage. In contrast, a former Lord Chancellor described the UK’s system of government as an “elective dictatorship”.

There is an election. Members of Parliament (‘MP’, ‘MPs’) are elected, for example, at the 2015 election, for the constituency of Maidenhead in Berkshire, Theresa Mary May. The MPs of the House of Commons choose from amongst their number a Prime Minister (‘PM’). The PM is from the House of Commons, remains in the House of Commons, and is answerable to the House of Commons. If the government, headed by the Prime Minister, cannot win a no-confidence vote in the House of Commons, the government falls and there is another election. Fixed-term Parliaments Act 2011, s.2(3)(a) and s.2(4): Parliament is dissolved and there is a fresh election if “the House of Commons passes a motion” “That this House has no confidence in Her Majesty’s Government.”. So the PM can always command a majority in the House of Commons, because if the PM can’t, the PM ceases to be PM.

In contrast, the PM might not be able to command the confidence of the House of Lords, yet the approval of “each House of Parliament” is needed. With respect to primary legislation, the House of Commons can over-ride the House of Lords (the 1911 and 1949 Parliament Acts), but, as explained by the Companion to the Standing Orders of the House of Lords:

Even the Salisbury Doctrine provides only a little help, as it is customarily applicable only to a Government Bill mentioned in an election manifesto. Nonetheless it does make the unelected Lords unwilling to over-rule the Commons.

So by one means or another, even without the support of Parliament, a UK government can compel some large gilt purchases by the Bank of England. With that support it could rewrite the legislation as it deemed necessary.

★ Modes of failure

A British government has a lot of power over legislation, and over UK entities such as the Bank of England. For that reason, it is difficult to construct modes of failure. Described are one non-failure, and one implausible possible failure.

Even so, the worst that can happen is to complicate a direction to the BoE, made under “extreme economic circumstances”, compelling gilt purchases. In other words, the worst that can happen is to hinder the government doing something that it has never needed to do.

★ Summary: the BoE can be commanded to pay

Thus the government has emergency means of accessing unlimited £ funds. In extraordinary circumstances compelling central-bank buying of gilts might be awkward. Constructing circumstances in which it is impossible requires some implausible facts, central-bank folly, as well as extraordinary stretching of constitutional norms.

★ Civil emergency

During a temporary civil emergency (such as flood, terrorism, fire, power outage) a bank holiday could be proclaimed. From the Banking and Financial Dealings Act 1971:

Using bank holidays primarily as a means of delaying payment, without genuine reason, would be frowned upon by the rating agencies. And it wouldn’t be done, partly the previously discussed powers suffice. But if there were a civil emergency causing temporary operational problems, payment could be lawfully delayed without default.

★ Willingness to pay

Non-default entails both an ability to pay and a willingness to pay. The above has argued that there is ability with spare. What about willingness?

For over three centuries the UK (preceded by England & Wales) has been a good payer.†7 There were outstanding eight ‘perpetual gilts’, which had no set final maturity but were callable by the Treasury. In 2015 all were called. Of these, the oldest was 2½% Annuities, which had paid every quarterly coupon since issue in 1853, the proceeds having presumably been spent on the Crimean War.†8

Starting with that record of payment, what is the government’s optimal strategy? Such a long-standing reputation is surely worth a lot of defending. Further, it seems that the belief of UK politicians, of whatever political colour, is the same: that reputation is worth a lot of defending.

★ Other powers

The 1985 statement and these four paragraphs give HMT, acting via the DMO, considerable power over various matters relating to the debt of the United Kingdom. To the best of the author’s knowledge these powers have never been used mischievously, but they exist.

★ During an election campaign

Preventing default most definitely counts as “Essential business” which “must be carried on”. So the previous reasoning would not be different with Parliament dissolved and an election underway, nor during any post-election coalition negotiations.

As already quoted, s.19(5) of the Bank of England Act 1998: “no account shall be taken of any time during which Parliament is dissolved”, which gives additional flexibility during campaigning.

★ Scotland

The United Kingdom consists, unsurprisingly, of more than one Kingdom: that of England, and that of Scotland. The UK is England, Scotland, Wales, and Northern Ireland. Scotland has about 8.4% of the UK’s population (≈5.2 million out of ≈62.2 million total), and about 8% of UK GDP. Scotland does however have a lot of the empty space, being 32% of the land area. The Scottish National Party wants Scotland to be independent, and indeed there was a 2014 Scottish independence referendum which resulted in a decision to stay in the UK.

If Scotland were to leave, the UK’s debt and international assets, including the permanent seat on the UN Security Council, would stay with England and Wales and Northern Ireland. The loss of 8% of GDP would raise UK’s indebtedness ratio by a factor of about 1÷0.92, equivalent to increasing it from ≈89.2% of GDP to ≈97%. Such a change would be unwelcome, but would not alter any of the previous legal arguments. (Observe in the already-quoted ¶124 of the Information Memorandum that it “shall be construed in accordance with the laws of England”.†9)

But those who lost that referendum want a second go (but if they had won and Scotland were now independent, they wouldn’t now say that $50 oil is a significant change in circumstances requiring a re-vote). But if a second referendum reversed the verdict of the first, what would happen to the UK’s debt? HMT, 13 January 2014:

★ Non-£ default?

A non-£ default by the UK is, of course, not impossible, but nonetheless is very unlikely.

Not impossible? Though the UK parliament has ultimate authority over the Bank of England, it doesn’t over the US Federal Reserve, European Central Bank, or the Bank of Japan. So the UK cannot compel unlimited access to $ or € or ¥.

Very unlikely? Nonetheless, the author believes a non-£ default by the UK is very unlikely, and that that one of the primary reasons for this has been heralded too little: Since late 1992, the value of the £ against foreign currency has been free-floating. Not even a managed float: floating very free.

For the government and for a bank, owning £ assets funded in foreign money is risky. And since 1992†10, that risk has been very obvious. The variability of £/$, £/DEM, £/€, £/¥, £/CHF, etc, have been starkly visible, not ignorably theoretical. This is why the assets and liabilities of large entities, public and private, have been approximately FX matched. And if a private entity should collapse onto the government’s balance sheet, that combined balance sheet would remain approximately matched.

Observe what happened in practice. From July to November 2008, £/$ fell from above 2.00 to under 1.50, yet the stresses in the banks were not ascribed to this fall. Why? Because £/$ had long been a known unknown, so market participants were approximately hedged.

Indeed, it is easy to find examples of previous defaults by countries with currencies that were pegged, or nearly so. Such a story can start with a currency peg, foreign interest rates being lower. Local entities borrow in foreign currency, seemingly risklessly. Then the peg breaks; foreign liabilities become much larger when measured in local currency; much default follows. Much rarer have been defaults by countries with almost all the government debt in free-floating domestic currency. The UK has almost all of its government debt in the currency over which it is sovereign; most of the government debt is long-term and fixed-rate; and that currency is perhaps the freest-floating of all currencies.

★ Ratings

This essay explains the de facto impossibility of UK £ default. This essay does not model the decision process of a rating agency, nor does it replicate nor predict the future opinion of any rating agency.

Indeed, a rating agency might have a different opinion. Even though that different opinion would be wrong, it would still be the opinion of the rating agency.

★ Elsewhere

Germany, France, Italy, Greece, and the other members of the eurozone do not have the sovereign ability to print the numéraire in which their debt is denominated. That alone gives them all—yes, including Germany—a materially higher risk of default than has the UK. Obviously some have income exceeding expenditure (Wilkins Micawber: “result happiness”); others don’t (“result misery”).

The USA in theory has these sovereign powers, but its internal decision-making process means that it can’t necessarily use them. Sometimes the Executive wants to act as if the UK’s National Loans Act 1968 applied there, while the Legislative—in demanding a silly price to raise the debt ceiling—seems hell-bent on economic suicide. Perhaps if enough members of Congress could see themselves as a future President then a USA version of NLA68 could pass.

Denmark: the DKK is the only member of version II of the Exchange Rate Mechanism. The Danmarks Nationalbank keeps €/DKK within ±2¼% of EUR 1 = DKK 7.46. Currently DKK policy interest rates are below those of the EUR, so one might expect that Danish players currently prefer to borrow in DKK. But over the lifetime of a long bond that might change, in which case Danish players would prefer to borrow in EUR. And if the peg were then to break, those euro liabilities could land in the lap of the Danish government. So this thesis about de facto impossibility does not apply to Denmark—even though, for other reasons, Denmark is currently an excellent credit.

Japan: perhaps somebody expert could say whether Japan’s distribution of decision-making powers more closely resembles that of the UK or that of the USA.

Emerging-market countries often have large dollar borrowings, or state-owned enterprises with same, or a currency that doesn’t float freely. In theory an EM could have a deep local-currency debt market, a floating currency and hedging vehicles for it, and an effective government capable of taking honest competent decisions. But that doesn’t seem very “emerging”.

★ Conclusion

Her Majesty’s Treasury have all the authority they need to issue debt, within or outside the UK, without limit as to quantity, in such manner as they think best.

The Treasury have the power to compel the Bank of England to purchase gilts. If Parliament were to disrupt this, the matter would be settled by an election.

Debt markets look ahead, and because such emergency means of funding can be seen, that should reduce the need for them to be used. Nonetheless, they exist.

So the UK will pay its £ debts: sovereign default in £ is de facto impossible.

But that does not say anything about the purchasing power of the pounds in which the UK will pay. Also, it does not say anything about the relative merits of alternative investments, such as compounded short-term deposits. These matters are not riskless, but are not default.

There are other countries that are very unlikely to default. But a sovereign default in £ by the United Kingdom is de facto impossible.

★ Footnotes

(Many of the footnotes in the earlier versions of this essay were just links; as this is published in HTML those links have become active links in the text.)

†1 There have been several versions of the Memorandum: 24 Mar 2016; 23 Nov 2015; 22 Aug 2013; 15 Aug 2011; 22 May 2009; 20 Aug 2007; and 21 Dec 2004. They are not identical, but the differences have not been material—at least, not yet. Some of these changes have been accompanied by a ‘Notice of Amendment’, that of 24 June 2011, but only that one, saying: “For the avoidance of doubt, the above amendment shall apply to all Stocks currently in issue, whether issued under the Information Memorandum, under previous versions of the Information Memorandum, or otherwise.”

†2 His Majesty’s Treasury is legally plural as there are there are at least six Lords of the Treasury who serve concurrently, jointly serving as a commission for the Lord High Treasurer. The First Lord of the Treasury is the Prime Minister; the Second Lord of the Treasury is the Chancellor of the Exchequer; the Junior Lords typically being other MPs. Hence “the Treasury may … as they think fit”.

†3 This phrasing was not new, first appearing in the National Loans Act 1939, s.1 (which was repealed by the 1968 Act): “Any money required … may be raised in such manner as the Treasury think fit. … For the purpose of raising money under this section, the Treasury may create and issue such securities as they think fit.” (HTML, PDF).

This 1939 Act was introduced to the House of Commons on 7 December 1939 by Sir John Simon, the then Chancellor of the Exchequer, who eloquently justified the need for these broad powers—even today, much of that speech rings true (start column 94), and the author commends it to US policy makers. That speech made reference to various WW1 financing acts—alas not on www.legislation.gov.uk as they were wholly repealed before that website’s base date of February 1991—which imposed a £ limit to the borrowing. Perhaps those limits might have been an inspiration, or partial inspiration, for the US debt ceiling in the Second Liberty Loan Act of October 1917.

†4 Many Acts have similar requirements, even if not identically worded. For example, in the Banking Act 2009 in s.2(5) and in s.75(7)-(8); and in the modifications to other Acts done by the Financial Services Act 2012 in s.4(1)(9N) and in s.6(1)(3H).

†5 The Parliament Acts are rather slow, but an alternative has precedent. In order to pass the 1911 Parliament Act the Prime Minister, H. H. Asquith, threatened to appoint new Lords until the government had a majority of the Lords. That could be done, or at least re-threatened.

†6 One might think that this problem could be avoided by delaying the laying of the Statutory Instrument before Parliament until after the auction. However, s.4(1) of the Statutory Instruments Act 1946 requires that the SI “shall be so laid before the instrument comes into operation”, the allowed exceptions probably being insufficient for this purpose.

†7 Charles II, king from the restoration in 1660 until his death in 1685, was a notoriously unreliable debtor. In the words of a contemporary, “A merry monarch, scandalous and poor.” Other comment of the time concurred, but too bawdily for reproduction here. So the Bank of England was founded in 1694 to borrow money on behalf of the government, with a greater perception of creditworthiness. Since then the E&W/UK government has been a good payer.

†8 The author owns, in bearer form, £100 nominal of 2¾% Annuities (first issued 1884). I haven’t been assiduous about collecting the quarterly coupons (“£0·68” each as payments truncate fractions of a penny), nor have I redeemed the principal. Should I?

†9 Standard legal wording has shifted over time from “the law(s) of England” to “the law(s) of England and Wales”. In legislation from 1189 to about 1950 it was always the former; from then to about 1980 it became about half of each; since when “… and Wales” has predominated. So the DMO’s usage of just England is a little dated but not unusual. Yet this old-fashioned wording was introduced in the 2007 version of the Information Memorandum (the 2004 version not specifiying a governing law).

†10 GBP ignominiously left the Exchange Rate Mechanism on 16 September 1992, though, arguably but irrelevantly, the FX risk was obvious before then. The exchange-rate policy has been a very clean float since 25 February 1993, the UK intervening only in concert with others to affect the value of a non-£ currency, and even that only twice and in small size. On 22 September 2000 the BoE “purchased €85 million against sterling” (¶1 of announcement). And on 18 March 2011, “The Bank sold ¥12bn for £93m. This was the first coordinated G7 intervention since 2000” (statement, and more in ¶47 on p9 of EEA accounts July 2011).

★ Afterwords

• In December 2017 the DMO’s website was re-arranged, and links have been updated.

I agree. Nonetheless, the ultimate cause of this “co-operation between the institutions” comes from one side having, in a crisis, clear parliamentary authority over the other. Even though the rules make that difficult to abuse.

• August 2023, during the afternoon of Thursday 8 September 2022, upon the The Accession of King Charles III, Her Majesty’s Treasury became His Majesty’s Treasury. Quotations have retained their original pronoun.

• August 2023, HMT had rotted links, which have been fixed. A link to www.webarchive.org.uk returned “Available in Legal Deposit Library Reading Rooms only”, so replaced with a link to web.archive.org. Also, a minor punctuation error and a minor typo were fixed.

UK £ default is not only unlikely, it is de facto impossible

★ Document history

UK £ default is not only unlikely, it is de facto impossible

★ Gilts may be issued: no ceiling

★ Compelling the Bank of England to buy gilts

★ Maastricht rules

★ The Bank of England Acts

★ Each House of Parliament

★ Modes of failure

★ Summary: the BoE can be commanded to pay

★ Civil emergency

★ Willingness to pay

★ Other powers

6. Government Statement

7. Jurisdiction

8. Governing Law

9. Revision

★ During an election campaign

Preface

General Principles

★ Scotland

★ Non-£ default?

★ Ratings

★ Elsewhere

★ Conclusion

★ Footnotes

★ Afterwords