How likely is a sterling crisis or: is London really Reykjavik-on-Thames?

With the pound sterling dropping like a stone against most other currencies and credit default swap rates on long-term UK sovereign debt beginning to edge up, this is a good time to revisit a suggestion I made earlier on a number of occasions (e.g. here, here and here), that there is a non-trivial risk of the UK becoming the next Iceland.

The risk of a triple crisis – a banking crisis, a currency crisis and a sovereign debt default crisis – is always there for countries that are afflicted with the inconsistent quartet identified by Anne Sibert and myself in our work on Iceland: (1) a small country with (2) a large internationally exposed banking sector, (3) a currency that is not a global reserve currency and (4) limited fiscal capacity.

The argument is simple. First consider the case where the banking sector is fundamentally solvent, in the sense that its assets, if held to maturity, would cover its liabilities. Iceland’s banks were supposed to have been in that position, although I have seen no verifiable information on the quality of the three formerly internationally active banks. There is no such thing as a safe bank, even if the bank is sound. Without an explicit or implicit government guarantee, there is always the risk of a bank run (a withdrawal of deposits or a refusal to renew maturing credit and to roll over maturing debt) or a sudden market seizure or ‘strike’ in the markets for the bank’s assets bringing down a fundamentally sound bank.

To prevent a fundamentally sound bank succumbing to a deposit run or to asset market illiquidity, the central bank has to be able to act as lender of last resort, providing funding liquidity and as market maker of last resort, providing market liquidity to liquidity-constrained banks.

If the country has an internationally active banking and financial sector and if its foreign currency liabilities have a shorter maturity than its foreign currency assets, and especially if these foreign currency assets have become illiquid, the central bank has to be able to act as foreign currency lender of last resort and market maker of last resort if it is to be able to guarantee the survival of the banking sector when faced with a deposit run and/or illiquid markets for its assets.

The central bank of Iceland could be an effective lender of last resort in Icelandic krona, as it can print the stuff in unlimited quantities. It can be a lender of last resort and market maker of last resort in other currencies only to a limited extend – limited by the fact that the Icelandic krona is not a global reserve currency and by the fiscal spare capacity of the Icelandic sovereign.

If Iceland had been a member of the euro area, its central bank would have been part of the Eurosystem – the euro area central bank consisting of the ECB and the (currently 15) national central banks of the euro area member states. The euro is the junior of the two global reserve currencies. First is the US dollar, with around 64 percent of global official foreign exchange reserves held in US dollars. The euro’s share is around 27 percent. After the euro, there is nothing. Sterling’s share of 4.7 percent (at the overly flattering strong sterling exchange rate of late 2007) reflects its minor-league legacy reserve currency status. The Japanese yen and Swiss franc are completely irrelevant as global reserve currencies.

Clearly if a country has a major-league global reserve currency as its national currency, two consequences follow. First, it is likely to be able to borrow abroad using instruments denominated in its own currency rather than in that of the currency of the lender or some other global reserve currency – they are less affected by ‘original sin’ – in the currency-denomination-of-external-debt sense of the expression. Second, it will be possible for both private parties and for official parties like the central bank, to arrange access to foreign exchange (through swaps with other central banks, credit lines etc.) more easily and on better terms than are available to private parties, central banks and other official agents not blessed with a global reserve currency of their own.

As a member of the euro area, it would have been much easier and cheaper for Iceland to defend itself against speculative attacks on its banks – provided the banks and its government were indeed solvent and perceived to be so. With the krona, not only could solvent banks be brought down, even a solvent but illiquid (in foreign exchange) government could be brought down by a sufficiently large speculative attack on the banks, the currency and the public debt.

Of course, even with the euro, the banks could not have been saved by the Icelandic authorities if the banks were fundamentally unsound and if the government did not have the fiscal strength to recapitalise the banks. Under current circumstances, if the government injects capital into a bank to compensate for past and anticipated future losses, it may not achieve a risk-adjusted expected rate of return on this investment equal to its borrowing cost. The difference will have to be recouped through higher future primary surpluses, that is, higher future government budget surpluses excluding interest payments. If there is doubt in the markets about the ability or willingness of current and/or future governments to raise future taxes or cut future spending to generate the required increase in future primary surpluses, the default risk premium on the public debt will rise. We are seeing such increased default risk premia even for the most credit-worthy sovereigns, including the German government, the US government and the UK government. On Friday October 10, 2008, the spreads on 5 year sovereign CDS were 0.456% for the UK, 0.33% for the USA ad 0.265% for Germany, well above their post-war historical averages. On October 28, 2008, Bloomberg wrote:

“Credit-default swaps on [U.S.] Treasuries have risen nearly 40 percent since TARP was signed into law Oct. 3, and are now about the same as Mexican and Thai government debt before the credit markets began to seize up in June 2007.”

By bailing out the banks, and other bits of the financial system, the authorities reduce bank default risk but by increasing sovereign default risk. As long as there is sufficient fiscal spare capacity (the technical, economic and political prerequisites are met for raising future taxes and/or cutting future public spending by a sufficient amount to service the additional public debt and maintain long-run government solvency).

Iceland’s government did not have the fiscal resources to bail out its banks. All three internationally active banks were put into receivership. The domestic bits then were bought by the government out of the receivership. The Icelandic krona collapsed and is no longer internationally convertible: exchange rate restrictions have been imposed. It is an open issue whether Iceland will default on some of its sovereign debt obligations as well.

How and to what degree is this relevant to the UK? Iceland is a tiny country (about 300,000 people – the size of the city of Coventry). The UK has a population of over 61 million. Nevertheless, the UK is a small open economy for economic purposes: it is a price taker in the markets for its imports and exports and in global financial markets. Its share of world GDP in 2007 was 3.3% (at PPP exchange rates – somewhat higher at market exchange rates). Its currency is no longer a serious world reserve currency.

The UK banking sector’s balance sheet is about half the size of the Icelandic banking sector as a share of annual GDP: just under 450% at the end of 2007 as compared to Iceland’s almost 900%. Switzerland, another vulnerable country (small, no currency with global reserve currency status , large banking sector relative to GDP and limited central government fiscal capacity) has a banking sector balance sheet of just over 650% of annual GDP. With UK annual GDP around £1.5 trillion, that gives us a banking sector balance sheet of well over £ 6 trillion.

The first Chart below shows the size of the balance of the UK banking sector. This includes the Bank of England. If we exclude the Bank of England, the latest observation on the balance sheet of the banking sector and a percentage of annual GDP would still be around 420 percent. The deleveraging of the banking sector, visible at the very end of the sample period, has much further to go. The Chart also shows that foreign currency assets and liabilities of the banking sector are very evenly matched – the two lines are almost indistinguishable. Both now are just below 250% of GDP. I don’t have any data on the degree of mismatch by individual currency. Just the aggregate foreign currency exposure is shown.

While there is no net foreign exchange exposure of the banking system in the UK, banks are banks. The foreign currency liabilities of the banking system are therefore likely to have shorter maturities than the foreign currency assets. The foreign currency assets are also likely to be less liquid than the liabilities. I don’t have information on the maturity and liquidity composition of foreign currency assets and liabilities to confirm or refute this presumption. Let me just say that Iceland’s banks were brought down despite an aggregate match between foreign currency assets and foreign liabilities.

Chart 1


Source: Office for National Statistics

Not only are the UK banks rather large relative to the size of the economy, the gross external assets and liabilities of the British economy are also hefty – about the same size relative to UK GDP as the total assets of the banking sector (there is no deep reason for this coincidence). Chart 2 below shows the gross external asset and liability position and the net foreign investment position of the UK. While not in the Iceland league (Iceland had gross foreign assets and liabilities of around 800 percent of annual GDP at the end of 2007) the UK, with gross foreign assets and liabilities of well over 400 percent of annual GDP does look like a highly leveraged entity – like an investment bank or a hedge fund. By contrast, gross external assets and liabilities of the US straddle 100 percent of annual GDP.

Chart 2


Source: Office for National Statistics

Foreign currency illiquidity risk for the UK banks and authorities

Assume for the sake of argument that the UK’s banks are sound. Most of them obviously are not, which is why so many of them have had capital injected into them by the government, and why all of them benefit from explicit government guarantees on new bank debt issuance and implicit government guarantees that the government will come to their assistance should they be at risk of insolvency. With foreign currency assets of longer maturity and less liquid than foreign liabilities, the banks and the country would still be vulnerable to a foreign currency run on the banks (a refusal to renew foreign currency credit) or a seizing up of the markets in which the banks’ foreign currency assets are traded. The Bank of England’s foreign currency reserves are puny and the government’s foreign currency reserves are small – around US$43 billion, pocket change, really.

No doubt the Bank of England would be able to arrange swaps, credit lines or overdraft facilities with the systemically important central banks – the Fed, the ECB and the Bank of Japan. Given sound banks and sound fiscal fundamentals, it should be possible for the UK to defend the banking sector against runs or market strikes. There would, however, be a cost involved – the cost faced by any issuer of a currency that is not a global reserve currency and who therefore either has to insure ex-ante against the possibility of running short of global reserve currencies, or risk getting clobbered on the terms of an emergency currency swap or similar arrangement cobbled together when the enemies are already scaling the ramparts.

This cost of insuring against foreign currency illiquidity risk will make the City of London less competitive as a global financial center than rivals based in global reserve currency jurisdictions. It provides another strong argument for the UK adopting the euro and for the Bank of England becoming part of the Eurosystem as soon as the other EU member states will let it.

The reason the costly handicap of a minor-league currency does not appear to have harmed the UK in the past is the same as the reason why I have not made the argument in the past. Before the current financial crisis, no-one could conceive of a world in which a financial crisis would start in the global financial heartland – Wall Street and the City of London – rather than in some developing country or emerging market, would paralyze most systemically important wholesale financial markets and lead to the government nationalising much of the north Atlantic region’s banking and wider financial system and underwriting or guaranteeing the rest. Well, most of the world now knows that this is the way things can be. If it retains sterling, the City of London will put itself at a competitive disadvantage (for those who remember then-Chancellor Brown’s Five Tests for euro area membership, this means that the fourth of these tests now has been met also).

Sovereign default risk for the UK

Even if the UK had the euro as its currency, its banks would still have been at risk if they were unsound (their assets, even if held to maturity, would not cover their financial obligations). In this case, bank insolvency would result unless the British authorities were both able and willing to bail them out. I assume in what follows that the government is willing to bail out the banks. The evidence thus far supports this.

Northern Rock and (rump) Bradford and Bingley were nationalised. The SLS allows all banks to swap illiquid asset-backed securities for Treasury Bills. For reasons that cannot be understood by ordinary mortals, the Treasury Bills lent/swapped by the SLS don’t count as public debt (something to do with Treasury bills with less than one year remaining maturity not being part of the public debt for some accounting and accountability purposes – don’t ask). The Bank of England is accepting a wider range of private securities as collateral at the discount window and in repos. The state has a 60 percent ownership stake in RBS and roughly 40 percent ownership stakes in HBOS and Lloyds-TSB. The government has made up to £25o billion available to guarantee new issuance of bank debt. The state stands behind the formal £50,000 deposit guarantee for bank retail deposits.

The key question is, can the government meet all these fiscal commitments, whether firm or flaccid, unconditional or contingent and explicit or implicit ? Does it have the resources, now and in the future, to issue the additional debt required to meet the growing volume of up-front obligations it has taken on?

To be solvent, the face value of the government’s net financial obligations has to be no larger than the present discounted value of current and future primary government surpluses (government surpluses excluding net interest and other investment income). The government argues that its net debt position is strong, with a net debt to annual GDP ratio still just below forty percent. That statistic is a prime example of lies, damned lies and government statistics.

The 40 percent excludes such old sins as the debt incurred through the PFI (private finance initiatives). This will be brought into the total soon. It also does not yet include the net debt of Northern Rock and Bradford and Bingley. It also excludes the debt of RBS, where the government owns a majority stake and the debt of Lloyds-TSB and HBOS, where the government has a controlling minority stake. Under normal accounting practices, the debt of all three banks will have to be counted as public debt in the future.

Three large UK banks, HSBC, Barclays and Abbey (Santander) have not yet taken the King’s shilling – they are attempting to meet the capital raising targets they agreed with the government from sources other than the government. All three banks are, however, heavily exposed to emerging markets (Santander mainly in Latin America, HSBC in Asia, the Americas, Europe, the Middle East, and Africa and Barclays in Europe, Africa and Asia). This has been a source of strength until recently, compared to their competitors who were mainly exposed to the USA and Western Europe. However, with all emerging markets now severely affected by the financial crisis (both directly and through trade links with Western Europe and the USA), what was a source of strength is become a further source of weakness. The likelihood that some or all of the banks that have not yet received capital injections from the government will do so in the not too distant future is rising steadily.

It is not at all far-fetched to hold the view that the British government has effectively guaranteed the balance sheets of the entire UK banking sector. Let’s value this conservatively at 400 percent of annual GDP, some £ 6 trillion. The value of this guarantee depends on the likelihood it will be called upon, and on the amount of money the government would have to come up with if the guarantee is called. Both numbers are highly uncertain and any guestimate is bound to be subjective. The expected payments under the guarantee are, in my view, hardly likely to be less than £300 bn (on top of any money already paid out), some 20 percent of annual GDP. It could be much higher. With a recession of unknown depth and duration looming, there is a material risk that the government would have to come up with a multiple of the £300 bn just mentioned.

Of course, the value of the assets acquired by the government as shareholder has to be set against the explicit and implicit liabilities it has taken on. I would like to see a valuation of the equity stakes of the government that does not benefit from the recent scandalous relaxation of fair-value accounting and reporting that was forced upon the IASB. I don’t believe any valuation that relies on managerial discretion. With the regulatory constraints likely to be imposed on banks in the future, and the lower returns associated with banking-as-a-public utility, the government may well be getting rather poor financial returns on its investment in the banks. While that does not mean the government should not have made the capital injections – the systemic externalities associated with the failure of large banks don’t show up in the share price – it does mean that the immediate fiscal burden of the capital injections is likely to be only partially offset by future dividends and (re-)privatisation receipts.

In addition to the debt that has been and will be issued to finance asset purchases by the government, there are the future debt issuance associated with the large cyclical and structural government deficits that will be a feature of the coming recession. If GDP falls peak-to-trough by, say 3.5 percent and recovers only slowly, we could have a seven percent of GDP or higher government deficit for 2009 and 2010. Together with the explicit or implicit fiscal commitments made to safeguard the British banking system, the numbers are likely to spook the markets.

With the true net public debt to GDP ratio probably already well above 100 percent of GDP and rising, and with massive public sector deficits, partly cyclical and partly structural, about to materialise, the markets will question the fiscal-financial sustainability of the government’s programme with increasing vehemence. The CDS spreads on UK public debt will start rising. The notion that, except for currency, there may not be a safe sterling-denominated asset may come as a shock. But the same is true in the US. In 2009, the US government will have to sell (gross) at least $ 2 trillion worth of government debt (the sum of the Federal deficit plus asset purchases plus refinancing of maturing debt). The largest such figure ever in the past was $550 billion. In the US too, the markets will have to learn to do without a US dollar financial instrument that is free of default risk.

The fiscal dire straits the UK government are in limits their capacity to engage in a discretionary fiscal stimulus to boost domestic demand. For it to be meaningful, a debt or money-financed stimulus of at least one percent of GDP and more likely two percent of GDP is called for. But if the market takes fright and believes that the government will not raise future taxes or cut future public spending by the amounts required to safeguard government solvency despite greater current borrowing, it will add higher default risk premia to the longer-dated UK sovereign debt instruments.

Such mistrust in the temporary nature of a fiscal stimulus would not be irrational. After its first term in office, the government have thrown fiscal restraint to the wind and have engaged in a steady increase in public spending as a share of GDP which has been only partly matched by an increase in the tax burden as a share of GDP. Rising debt and deficits and a fondness for fiscal and accounting gimmicks designed to hide the increase in the debt burden have undermined public confidence in the fiscal rectitude of the government. With enough mistrust, the interest rates will rise by enough to crowd out completely the stimulus to private demand provided by the tax cut or public spending increase. Lack of confidence in the government’s fiscal sustainability would also undermine confidence in sterling. In the worst case, we could see a run on the banks, on the public debt and on sterling all at the same time. This is not the most likely outcome yet, in my view. But it is a distinct possibility.

Could the government monetise the deficits instead (i.e. sell gilts to the bank of England)? The Bank would only be willing to buy such debt (either directly or indirectly in the secondary markets) if it was consistent with its interpretation of its price stability mandate. The Bank appears to believe that short rates may have to go down quite a bit further if it is not to undershoot the inflation target by the end of next year. It may also view the monetisation of gilt issuance as consistent with its mandate.

If there is a conflict between the Bank of England and the government, the government could invoke the Treasury’s Reserve Powers. This is a clause in the Bank of England Act that allows the government to take back the power to set rates from the Bank of England, under exceptional and emergency conditions. It has never been invoked.

If the deficits get monetised, there will not be the upward pressure on real interest rates that would result from debt financing. But the markets may fear the long-term inflationary consequences of the monetary financing, especially if it were to be done by the government after invoking the Treasury’s Reserve Powers. So long nominal rates would be likely to rise if monetisation of the government’s deficits were chosen. Monetisation of deficits would also weaken sterling further.

All may still end up well (cyclically adjusted well, that is). But the piling of fiscal commitment on fiscal commitment by the government is not a risk-free option. The British government has limited fiscal spare capacity. Among the larger European countries, the UK government’s exposure, formal or implicit, to its banking sector is by far the highest. Switzerland, Denmark and Sweden are in a similar pickle, with the banking sector solvency gap threatening to become larger than the fiscal spare capacity of the state.

The British government should go easy on the discretionary fiscal stimulus it applies, lest it risk a triple bank, sterling and public debt crisis. Better to first let the Bank of England use the 300 basis points worth of Bank Rate cuts that it still can play with. Even better to combine rate cuts with measures to directly target the disfunctionalities in the interbank market, such as government guarantees for (cross-border) interbank lending.

The UK shares with the United States of America the predicament that unfavourable fiscal circumstances make the wisdom of a significant fiscal stimulus questionable. In the US as in the UK the twin deficits (government and current account) severely constrain the government’s fiscal elbow room. Both countries need all the help they can get from fiscal stimuli abroad, in China, in Germany and in the Gulf. Beggars can’t be choosers.

Maverecon: Willem Buiter

Willem Buiter's blog ran until December 2009. This blog is no longer active but it remains open as an archive.

Professor of European Political Economy, London School of Economics and Political Science; former chief economist of the EBRD, former external member of the MPC; adviser to international organisations, governments, central banks and private financial institutions.

Willem Buiter's website