Plan B: Dollarization?
Statements from Scotland’s First Minister last week suggest that a Currency Plan B is beginning to emerge. It appears that the Scottish Government is committed to a sterling currency union regardless of the UK Government's view. The fall-back option then appears to be dollarization using sterling as the medium of exchange. This blog considers some the possible consequences of dollarization.
Currency unions are simply defined as two or more countries using the same currency. This definition includes several possible currency regimes. At one extreme this includes sharing a central bank (which means possible access to tax-payer funds) in a formal monetary union as laid-out in the While Paper (and ruled-out by the UK Government). At the other extreme a country can ‘dollarize’, which simply means using the currency issued by another country as legal tender and without any formal agreement. This appears to be the emerging Plan B of the Scottish Government.
Countries which are dollarized broadly fit into three categories. First, tiny city-states such as Andorra, Monaco and the Vatican use the euro with a special dispensation from the EU. Second, states in transition, such as Montenegro and Kosovo use the euro without any EU agreement. Third, countries which had legacies of economic or political instability, the most famous examples being Panama, Ecuador and El Salvador. If Scotland were to dollarize, it would be by far the biggest and wealthiest country to do so. The Scottish economy is eight times bigger than Panama's.
It is, of course, true that sterling is an internationally traded currency and so an independent Scotland, or any other country for that matter, could use sterling as its medium of exchange. However, this is not the same thing as saying that Scottish notes and coins are internationally accepted; clearly they are not.
So the first question is what would happen to Scottish notes and coins? At present they are issued by three Scottish clearing banks under license from the Bank of England and are fully backed by mega sterling notes called giants and titans. The most likely outcome is that the mega notes (or the equivalent value in Bank of England notes) are handed over to the Scottish government which then has 100% asset backing for its new state liability of notes and coins. Scottish and Bank of England notes would continue to circulate alongside each other and be freely convertible at the new central bank of Scotland. To keep the dual currencies at the same value, the issue of each new Scottish note would need to be backed by an increase in Bank of England notes held by the Scottish central bank as foreign reserves.
One way to see dollarization is as a formal monetary union but without any fiscal or banking union. Independence means ending many of the risk sharing aspects of a fiscal union, such as a single centrally funded social welfare system. Armstrong and Ebell (2014) have argued that because of the amount of debt that an independent Scotland would inherit, there would be limited space for counter-cyclical fiscal policy. In another blog David Bell argues that North Sea oil tax revenues alone do not get Scotland out of this bind. The lack of a banking union means an independent Scotland would have to provide its own regulation, deposit insurance and lender of last resort facilities for its financial system.
For financial regulation, much depends on whether an independent Scotland could be fast-tracked into Europe. If it could, then it is possible for non-eurozone countries to join the single supervisory mechanism. A Scottish regulator could then work with the European Central Bank to supervise an independent Scotland's largest banks. Whether this makes sense while using sterling is doubtful because this would open up the possibility of arbitrage between the two sets of regulators. The White Paper proposes a shared single authority or at least a fully harmonised regulatory system with the UK. Andrew Bailey, Head of the Prudential Regulatory Authority of the Bank of England, noted that so far only Greenland and the Faroe Island successfully share a foreign financial regulator.
Deposit insurance would need to be provided ultimately by the Scottish government. Based on a population share of the stock of UK household deposits, and assuming governments in reality insure all household deposits, the insurance cover would be for £95.2bn. This is for all the assumed deposits in Scotland. The critical question is the probability of the policy paying out and what would be the total cost. This in turn depends on the volatility of the real economy together with the soundness of the financial system. The lack of fiscal capacity for significant counter-cyclical policy is likely to mean the economy is more vulnerable to significant economic shocks. This puts even greater importance on the soundness of the financial system. Households are smart enough to figure out when a state can, and cannot, cover the insurance.
This brings us to the third and critical aspect - lender of last resort. The nature of banking is to transform short term deposits into longer term loans (such as mortgages and business loans). This inevitably involves a risk, however distant, that the bank might be caught short of liquidity and hence the need arises for a lender of last resort. The lender of last resort must have deep pockets to provide enough liquidity - and the deepest pockets are when a central bank can create its own money. If no lender of last resort exists, and a bank does not have access to liquidity, it may have to sell off its loans cheaply to raise funds. If depositors even doubt a bank’s ability to repay all of its deposits, then a run might start which, without a lender of last resort, could fast turn into a system wide crisis. Put simply, given how banks currently operate, the absence of a lender of last resort makes the probability of paying out deposit insurance much greater.
Countries which dollarize do not generally act as lender of last resort as they cannot issue their own currency unless to the extent it is backed by additional foreign reserves. Panama does not even have a central bank and Ecuador and El Salvador have central banks but without responsibilities for providing liquidity to commercial banks. They simply facilitate the transfer of one bank’s reserves to another to ensure a smoothly functioning payments system. No dollarized country provides lender of last resort support across its banking system.
At this juncture, it is also worth mentioning the Hong Kong Monetary Authority, which operates a currency board (similar to dollarization) and provides limited lender of last resort facilities to its banks. It can do this because it has fiscal reserves (net of debt) of 30% of GDP and $300bn of foreign exchange reserves. Any intervention would be limited because of the need to maintain enough reserves to back the Hong Kong dollar notes and coins (or monetary base – about half of the reserves). By contrast an independent Scotland is likely to have a fiscal debt at 80% of GDP ratio and around $9bn of reserves (see Armstrong and Ebell (2014).
So how could Scottish financial institutions survive without a lender of last resort facility? One option is that the large financial institutions de-camp to the rest of the UK and Scotland’s financial system becomes dominated by non-Scottish institutions. This is essentially the Jersey model. Foreign banks could operate in Scotland on a branch basis, so that deposit insurance would be provided by the overseas government (like Icesave, until the deposit insurance was suddenly withdrawn by the Icelandic government). However, it is unclear why, for example, the UK authorities (and hence UK tax payers) would effectively provide free deposit insurance and liquidity to what would become an off-shore sterling market. Alternatively, UK institutions could operate in an independent Scotland on a subsidiary (separate legal entity) basis. They would have to fund their activities in Scotland as separate entities. The cost of Scottish government borrowing is likely to be the floor on private borrowing costs which, without a lender of last resort, are likely to be significantly higher than estimated in Armstrong and Ebell (2013).
Another option is that Scottish institutions self-insure, accumulating sufficient buffers to make up for the lack of a deep-pocketed lender of last resort. In order to self-insure, Scottish financial institutions would have to become far more resilient and have far more prudent capital structures to minimise any shortfall of liquidity beyond that which could be comfortably funded in the interbank markets. It is noteworthy that banks in Panama have around 17% capital ratios and 25% liquidity ratios presumably because they simply have to manage their own risks. Subsidiaries of foreign banks would likewise require far higher buffers than in their home markets. While there is merit in this proposal, is it realistic that Scotland could require this degree of self-sufficiency from its banks, when the rest of the world (particularly the UK) is hell-bent on giving vast subsidies to banks?
The final plausible option is cross border liquidity insurance can be provided from the rest of the UK, presumably on a commercial basis. A new forthcoming paper by Armstrong and McCarthy examines whether this is economically feasible.
Unless cross border liquidity insurance is possible, dollarization would in all likelihood transform Scotland’s financial system. Many institutions would change their domicile. Apart from the serious impact on employment, income and tax on Scotland’s second largest on-shore economic sector, it would mean a dramatic reduction in the export of financial services to the rest of the UK. There are no formal current account figures for Scotland, but most economists believe an independent Scotland would have a rough current account balance (including a geographic share of oil). If this is correct, the loss of significant financial services could imply a substantial deficit. This will make it harder to generate the balance of payments surpluses necessary to accumulate foreign reserves.
Introducing the toughest of all currency regimes may reduce currency risk (the risk that a Scottish currency might fall) but this may come at the expense of greater country risk (the risk that Scotland could not fund itself). On top of the risks from operating with no macroeconomic policy levers (no interest rate or currency adjustment policy and only limited fiscal space), the cost of providing deposit insurance together with a lack of a lender of last facility is likely increase country risk further. It is worth noting that Panama, often held up as an example, had between no fewer than 17 IMF programmes between 1973 and 2000, second only to Pakistan. It is hard to believe that it is a coincidence that the two countries who have benefited most from IMF largesse also happen to be strategically immensely important.
 Scottish Government White Paper, pp111-117.
 Dollarization is the common term irrespective of the currency being used.
 Non-eurozone members of the EU can join the emerging European Banking Union. But as currently envisaged this really means European supervision as the ECB can only act as lender of last resort for banks in euro markets and deposit insurance looks to be staying with national governments.
 Oral evidence to the Scottish Affairs Committee, 25th February, 2014.
 At the end of 2013 the total of UK household deposits at banks and building societies was £1,120bn. An ideal way to fund deposit insurance is through a bank levy and for a limited amount of the deposits. However, in a systemic crisis it was found that a government back stop was necessary.
 It is important to note that the Sterling Monetary Framework (which, inter alia, includes the conditions for liquidity support) covers all financial institutions as well as banks.
 El Salvador has recently created a modest liquidity fund which draws on the reserves of other banks to assist a single bank with liquidity. Drawing on the reserves of banks makes this an inappropriate lender of last resort for the banking system.
 The main difference between a currency board and dollarization is that a domestic currency is the legal tender which generates seigniorage revenue for the state.
 See IMF Article IV, SAR Hong Kong, 2013.
 Jersey’s banking sector is thirty times its GDP but there are no Jersey banks.
 It may not be economically efficient for banks to self-insure for all eventualities.