Már Guðmundsson: Lifting capital controls in Iceland

Speech by Mr Már Guðmundsson, Governor of the Central Bank of Iceland, at a meeting organised by the European Economics and Financial Centre, London, 26 May 2016.

Central bank speech  | 
02 June 2016

Chairman, Ladies and Gentlemen,

Comprehensive capital controls were introduced in Iceland in November 2008, as part of the economic programme that the Icelandic authorities had developed with the IMF. The month before, around 90% of the banking system had collapsed when Iceland's three overblown cross-border banks failed. Earlier that year, Iceland had experienced a full-scale sudden stop currency crisis during which the currency plummeted by 26%. It depreciated by another 26% during the banking collapse and in the weeks afterwards. And if this were not enough, Iceland was at this point on its way into a recession, as the big, unsustainable macroeconomic imbalances - most clearly demonstrated in a double-digit current account deficit - were beginning to correct.

The economic programme with the IMF had three key goals: stabilisation of the exchange rate, fiscal sustainability, and reconstruction of the financial sector. Comprehensive capital controls were an important element in the programme, but the rationale behind them was to help stabilise the exchange rate after the currency had fallen more than 50% in 2008, and in addition, the stock of relatively liquid foreign-owned krónur - a legacy of carry trade and other capital inflows - amounted to around 40% of GDP and would have resulted in a further meltdown of the currency if it had begun to flow out through Iceland's shallow FX market.

After the split of the failed banks had been executed towards the end of 2010, it gradually became clear that it had not been a pure domestic/international split. The estates of the old banks held significant domestic assets, while only a small part of their claims were domestic. This posed a significant balance of payments problem for Iceland and, if left unmitigated at the point of resolution and distribution to creditors, could potentially have had a severely detrimental effect on the exchange rate and financial stability. The estates of the old banks were therefore brought under the capital controls in March 2012.

In late 2008, the combination of capital controls and exchange rate stability was in several ways a key to stabilising the economy. First, a very significant part of private sector debt was either directly linked to the exchange rate or indirectly linked to it through price indexation. The negative effect that further depreciation would have had on consumption and investment would therefore have been stronger than the stimulative effect on the traded goods sector. The latter effect comes with a lag in any case, and the real depreciation of about one-third, which had already taken place, was sufficient for that. Second, a large fiscal deficit had opened up and had to be financed domestically, as the sovereign was cut off from foreign markets. The capital controls facilitated this. Third, the capital controls gave monetary policy more scope to focus on the domestic economy, as interest rates would otherwise have had to be higher in order to defend the currency. Fourth, the capital controls provided the shelter needed to restructure domestic balance sheets, rebuild the financial sector, and develop a better framework for economic and prudential policies that would be suitable for a post-capital controls environment.

These stabilisation gains of the capital controls were greatest to begin with but have faded with time as balance sheets have been repaired and Iceland has recovered. The economic costs of the controls work in the other direction: they increase over time as lost business opportunities accumulate, as support to domestic asset markets begins to feed bubbles, and as risk diversification in domestic balance sheets decreases. In Iceland's case, this was partly mitigated through the liberalisation of capital inflows in October 2009 and legislative provisions that authorise the Central Bank to grant exemptions from the capital controls, provided that the balance of payments effects are not negative. But that does not change the big picture. Furthermore, the potential distortion stemming from the combination of free capital inflows and restricted capital outflows has materialised recently, as carry trade and other forms of capital inflows have increased following the announcement and subsequent execution of the resolution of the old banks in a manner that fully mitigated the negative balance of payments effect. This has created appreciation pressures that are not counteracted by outflows, as would have occurred under normal conditions. Finally, in the long run, comprehensive capital controls are inconsistent with Iceland's membership of the EEA and thus the EU internal market.

You might then ask: why have we not just lifted the capital controls? This is where the twist in the tale comes in. The problem is that our own success in preserving Icelandic asset prices and fostering the recovery of the economy has created the conditions for a stronger currency but has also, in some sense, preserved the obstacles to swift and orderly liberalisation of capital controls. If Parliament had not passed the emergency legislation in October 2008, thereby ensuring that there was a functioning banking system and that domestic deposits were secure, if capital controls had not been introduced in November 2008, if we had not been so successful in repairing domestic balance sheets, if the economic recovery had been weaker and the rate of return on domestic assets had been lower, then the domestic assets of the old banks would have been worth much less, as would the offshore krónur. In this scenario, these positions might have exited Iceland long since, at just a small fraction of the value they had just before the resolution of the old banks. Instead, the domestic assets of the old banks amounted to 25% of GDP and the stock of offshore krónur was estimated at around 15% of GDP. It was clear that free exportation of these positions without mitigating action would have resulted in a new currency crisis, with potentially serious effects on financial stability. That was both economically and politically unacceptable. As a matter of fact, the population had been promised solutions to these problems entailing no negative effect on the exchange rate.

The Icelandic authorities have publicised three versions of their capital account liberalisation strategy: the first in 2009, the second in 2011, and the current one in June 2015. It was under the second strategy that the stock of offshore krónur was reduced through a series of auctions. That strategy, however, did not include a solution to the problem of the old banks. The split between the old and new banks was only finalised towards the end of 2010, and in any case, the old banks were largely exempted from the controls until March 2012.

The current strategy is based on a guiding principle and a set of criteria that any proposed solutions would have to fulfil. If I explain these to you now, you will better understand what we have done to date and what is forthcoming.

An overall guiding principle in the design of the liberalisation strategy was that it should entail comprehensive, cautious, and clearly communicated solutions, allowing the country to lift capital controls without undue risks to macroeconomic and financial stability.

The strategy is comprehensive in the sense of mitigating, in a sufficient manner, the risk that excessive capital outflows would overburden Iceland's small foreign exchange market and inflict systemic negative externalities on the general public. In particular, it involved reducing the size of capital outflows associated with the resolution of the failed banks' estates through composition agreements based on explicit stability conditions, with a Pigouvian tax as a necessary backstop mechanism.

The strategy is cautious in the sense that liberalisation is planned, timed, and sequenced so that the largest challenges are solved sequentially, in a manner that mitigates the risk of disorderly dynamics in the foreign exchange market. Furthermore, the general liberalisation of capital controls on domestic residents will take place once the largest obstacles have been successfully removed or neutralised and when certain economic criteria have been fulfilled, such as adequate FX reserves, market access for the sovereign and the banks, a well-balanced and growing economy, etc. The aim is to minimise the risk of capital flight.

The Icelandic authorities have emphasised public disclosure and transparency of relevant information about the liberalisation strategy. Last year, the solution to the failed banks' estates, as well as the economic analysis on which it was based, was presented in a comprehensive manner. A few days ago, a bill of legislation on the treatment of offshore krónur was passed by Parliament. An English translation of the bill was posted on the website of the Ministry of Finance and Economic Affairs, along with the Explanatory Notes and a Q&A (with links on the Central Bank website). Yesterday the Central Bank published terms of a foreign currency auction directed at offshore krónur, to be held on 16 June. This speech is, of course, part of the process of disclosure and explanation. Furthermore, in a week or two, the Central Bank will issue a special publication on the economic analysis of capital account liberalisation in Iceland.

One of the criteria was that solutions should solve Iceland's balance of payments problem in a manner that allowed Iceland to lift capital controls without taking undue risk with economic and financial stability; preserved stability in both the short and the long run; limited risk to Government finances, reserve adequacy, and credit ratings; avoided socialisation of private debt or risk; allowed the exchange rate to reflect economic fundamentals, unaffected by non-governmental legacy issues; honoured international obligations concerning equal treatment under the Icelandic Constitution; and gave preference to measures offering simplicity over complexity, minimal legal risk, and as much finality as possible.

The guiding principle and the criteria were used to design the solution for the estates of the old banks. The solution allowed the estates to choose between two options. Under the first option, they would receive exemptions from the capital controls in order to go into composition and make distributions to creditors while offering to fulfil stability criteria that fully mitigate the negative effect on the exchange rate and/or the FX reserves. According to the stability criteria, the estates would pay sufficient stability contributions (through the transfer of króna assets) to mitigate the negative effect, lengthen the maturity of their FX financing of domestic banks, and repay Government and Central Bank support and debt. Under the second option, a tax would have been levied on the estates that fully mitigated the negative balance of payments effect of the resolution. All of the estates decided to fulfil the stability criteria, and the solution has mostly been implemented.

The same principle and the same criteria will guide the Icelandic authorities in the next step, which is to mitigate the risks associated with the stock of offshore krónur before lifting capital controls on domestic residents in the near future. This will take two forms.

First, the Central Bank of Iceland has offered to use part of its own FX reserves to buy offshore krónur in a single-price auction that will take place on 16 June. This will be the last auction of this type before the authorities begin to lift capital controls on domestic residents, which is expected to commence in the autumn. In the terms of the auction, which were announced yesterday, you can see that the exchange rate in the auction will vary inversely with the amount of offshore krónur traded in the auction. It will start at 210 krónur per euro for any amount and could rise as high as 190 krónur per euro if the volume exceeds 175 billion krónur. The total eligible stock of offshore krónur is 319 billion krónur. Outside the auction, anyone can sell offshore krónur to the Central Bank for 220 krónur per euro at any time until 1 November 2016.

Participation in this auction, as in the 21 similar auctions that preceded it, will be entirely voluntary. We are aware that some holders of offshore krónur may wish to remain in Iceland and await future developments. The authorities will revisit the situation concerning any offshore króna positions that remain after this auction, once sufficient progress has been made in liberalising capital controls on residents. We cannot predict how smoothly that liberalisation process will proceed; therefore, we cannot state exactly when the authorities will again be able to devote FX resources to mopping up any residual offshore krónur.

Second, last Sunday Parliament approved a bill of legislation whose principal objective is to segregate offshore krónur in a secure manner, so that it will be possible to lift capital controls on domestic residents and re-establish cross-border transactions with krónur, irrespective of the outcome of the auction. This involves some technical changes to the bank accounts involved. The environment will be largely the same, however. Offshore króna holders will be allowed to invest in the same instruments as they may invest in now, with the addition of special certificates of deposits issued by the Central Bank of Iceland. And they may continue to hold the instruments they currently hold, including long-term Government bonds and equities. And this is very important: there is no forced or distressed debt exchange involved, and all Government debt will be paid in full and on time, as it has been hitherto.

There is one change, though: the banks holding the accounts of offshore króna holders will be required to buy certificates of deposit from the Central Bank for that same amount. These instruments will initially bear an interest rate of 0.5%, which will be reviewed annually. This mapping into the Central Bank is part of the mechanism that is intended to keep the remaining offshore krónur segregated in the face of domestic liberalisation. But why 0.5%, you might ask? You see this if you look at the Central Bank balance sheet and take into account that we are accumulating reserves. Basically, this rate reflects the rate at which the Bank is, at the margin, investing these funds, and no one can expect us to incur a significant cost of carry because of these positions. And this rate should not be confused with our key Central Bank rate, which is 5.75% on seven-day deposits. This rate is as high as it is for monetary policy reasons; it is intended to affect the spending and saving decisions of domestic residents and has nothing to do with these positions.

I know that there are more questions, and Sturla Pálsson, Director of Market Operations and Treasury at the Central Bank of Iceland, and I are willing to answer them during the Q&A session. But let me take two before I conclude.

First, why 220-190 krónur per euro? Why not the onshore rate? First of all, this is a different asset class from onshore krónur and has been trading at very different exchange rates. Second, even if the authorities should, for some obscure reason, allow these krónur to go through the onshore market at this point, they would never command the current onshore rate. The position is simply too large, and that applies even if we look only at the current liquid part. It would simply push the market in front of it. Third, these numbers are within the range where offshore krónur have been trading in auctions and in the market. Fourth, the most important number is 190.

This is found partly by looking at what will happen to our FX reserves if we buy the entire position at this rate. It will take us below the metrics that both we and the IMF think we should prudently meet when we start lifting capital controls on domestic residents. We are willing to stretch that far, however, in order to facilitate an amicable solution, as we are confident that we can rebuild reserves to the required level in the next few months.

Second, why are the authorities not offering longer-term bonds as a part of the solution? That was the initial intention when work on auction design began. At that time, the idea was to have a multi-product auction offering foreign currency, long-term euro-denominated bonds, and long-term króna-denominated bonds. The deliberations revealed, however, that such a structure would have involved significant risks for Iceland. First, a multi-product auction carries more operational risk than a straightforward FX auction. Second, there was a risk that offering these bonds with special conditions might be construed as distressed debt exchange. Third, these bonds could be seen as legacy bonds and might have had detrimental effects on the Government bond market in both krónur and euros. Fourth, the Government does not need to borrow these funds, as it has a financial surplus and Government debt as a share of GDP is on a declining trend. Fifth, as capital controls were lifted on domestic residents, it would become possible for holders of the long-term króna bonds to exit through the FX market much earlier than they would otherwise, through hedging contracts with domestic residents. In principle, this could have been avoided through prudential and other measures, but it would have been messy. It is for these reasons that when Iceland - by accumulating significant reserves in recent months - became able to offer much more FX in the auction than was previously envisaged, it was decided to opt for a straightforward FX auction. In this regard, remember what I said earlier about our preference for finality.

Let me stop here, listen to the discussant, and then take questions.