Financial globalisation and challenges for prudential policies and macroeconomic management

Speech by Mr Már Guðmundsson, Deputy Head of the Monetary and Economic Department of the BIS, at a meeting of the Institut International d'Etudes Bancaires, Reykjavik, Iceland, 18 May 2007.


Financial globalisation is a process of ever tighter cross-border financial integration that will, at the limit, result in risk-adjusted real returns on assets with the same maturity and other characteristics being equal across countries. Although financial globalisation has made big strides in recent years, this state of affairs has not materialised. Financial globalisation has important benefits but also poses significant challenges for prudential and monetary policies. For prudential policies the fundamental problem lies in the mismatch between the international scope of banking institutions and the national scope of the frameworks for banking supervision and crisis management. For monetary policy, the problem is that the weakening of the interest rate channel of monetary transmission at the same time as the exchange rate might have a tendency to decouple from fundamentals in the short- to medium-term.

Full speech

Chairman, ladies and gentlemen,

It is a great honour for me to speak in front of such a distinguished audience of European bankers. My topic today is financial globalisation and the challenges it creates for prudential policies and macroeconomic management. In choosing this topic, I am partly influenced by my own personal experience of having been chief economist at the central bank of a very small open economy, that is here in Iceland, and then subsequently having a more global perspective at the Bank for International Settlements.

As the process of ongoing globalisation unfolds, it becomes progressively clearer that the world economy is more than the sum of individual countries. To that sum we have to add all the interdependencies, the markets that clear at the global level, international flows of goods, labour and capital, and many of the institutions that you represent, internationally active banks; and even this would not be a complete list. When analysing the world economy we therefore need first to look at the whole before going to the individual parts. We at the BIS have been trying to develop our analytical tools, our indicators and our data with this in mind.

When I moved from the Central Bank of Iceland to the BIS, I found it relatively easy to switch from the small open economy perspective to the global perspective. This may be partly due to the fact that my own country cannot have much relevance for global economic developments, although it was able to upset global markets for a few days in February last year. However, that episode, along with related experiences of other small mature open economies like New Zealand, woke the small open economist in me, but now within the global context. The key question is how ongoing financial globalisation is changing the environment within which domestic prudential and macroeconomic policies are framed and, consequently, how these policies might have to change? This is the motivation for my theme today.

The progress of financial globalisation

Let me, for good order, start by defining what I mean by financial globalisation. I take it to mean cross-border financial integration that is reasonably spread around the globe. Financial integration, in turn, is the process by which financial markets and institutions become more tightly interlinked and move closer towards a fully integrated financial market, where economic agents in different locations face a single set of rules, have equal access to financial assets and services and are treated equally. By implication, the law of one price would hold in such a fully integrated market, that is risk-adjusted real returns on assets with the same maturity and other characteristics would be equal.

The benefits of cross-border financial integration in terms of financial sector development, potential for risk-sharing and promotion of economic integration and growth are widely appreciated and so are the associated risks of surges and sudden stops in capital flows, contagion and vulnerability to financial crises. Almost everybody is trying to find the Holy Grail of reaping the benefits and mitigating the risks.

What is probably less well understood are the implications of the fact that financial globalisation is a process rather than a state of nature. This is maybe partly because macroeconomic textbooks jump from totally controlled capital movements to full global interest rate arbitrage from one page to the next. It is important to bear in mind that this process is only partly driven by government action. In addition, we have a market-driven process of financial innovation and evolving financial structures that will work to gradually increase cross-border financial integration.

Financial globalisation being a process has at least two implications. First, we would like to be able to measure where we are in the process. Second, different countries and regions will at any point in time be at a different stage.

Financial globalisation should manifest itself in stronger co-movement of risk-adjusted real asset returns across countries, a reduction in home bias in domestic portfolios, a higher level of gross cross-border capital flows and stocks of cross-border assets and liabilities, and an increase in cross-border banking and foreign direct investment in the financial industry. Theory also suggests an increase in the scope for international risk-sharing, which would be reflected in a lower correlation of domestic consumption and GDP.

Many of these manifestations lend themselves to some kind of measurement. Using the BIS international banking statistics and other sources, we see that the trend of financial globalisation is unmistakable. Let me mention a few stylised facts:

First, countries: from 1995 to 2005 there was a doubling of gross external positions of a representative sample of 29 countries, measured by the sum of foreign assets and liabilities as a percentage of GDP.

Second, banks: international claims of banks located in mature economies have increased fivefold as a percentage of GDP since 1980 to reach 50% last year.

Third, asset returns: international co-movements of asset returns have increased significantly, as demonstrated in numerous studies. Furthermore, as predicted by theory, these co-movements are stronger further along the maturity spectrum. However, it is possible that they are, at least partly, due to other factors than financial globalisation, such as common shocks.

But it is a mixed bag. Although home bias has fallen, it remains substantial, even among countries that have operated open capital accounts for decades. Consumption remains more correlated with domestic output than predicted by theory. The global integration of financial markets has therefore so far provided less insurance against idiosyncratic shocks than expected. This could be because capital flows have effectively been more constrained than appears on paper, and financial integration has thus been less advanced. Alternatively, capital flows may be inherently volatile due to information problems and herding, thus becoming a source of shocks as much as smoothing.

Internationalisation of banking and prudential policies

Let me now turn to the internationalisation of banking and the challenges it creates for prudential policies. This is a big subject and my exposé will of necessity be sketchy. However, you might be interested to know that this will be treated in greater depth in the forthcoming Annual Report of the BIS.

As you are better aware than most others, internationalisation of banking is not a uniform process. There are cycles in the overall pace and uneven patterns of development in terms of whether banks engage in cross-border banking from their home base or have a local presence in host countries through branches or subsidiaries.

In recent years it seems that direct presence in host countries has become more prevalent. However, cross-border lending still dominates in the euro area and emerging Europe, while local presence is more usual in the United States and Latin America.

The pace of development has been diverse across countries and regions. According to the IMF's latest Global Financial Stability Report, the share of foreign controlled bank assets in total bank assets increased globally by 8 percentage points between 1995 and 2005, to reach 23%. However, this share grew much more strongly in eastern Europe and Latin America, and it is currently significantly higher in these two regions than in other parts of the world.

A similarly strong regional variation emerges when seen from the standpoint of home countries. Cross-border activities, measured by the unweighted average shares of assets, revenues and employees, accounted for 45% of the total activity of 50 major European banks, but only 23% of the activity of 20 such banks in North America, and 14% of 20 major banks in Asia and the Pacific.

The drivers of the internationalisation of banking are partly the same as for globalisation in general, a reduction in transaction costs, progress in communication, economies of scale and scope, and liberalisation of trade and capital flows. Globalisation, in turn, gives a strong impetus to further internationalisation of banking. Thus following important domestic clients in international ventures is often an important motive for bank's cross-border activities. However, there are also drivers more specific to the financial industry and the same applies to the domestic regulatory and competitive environment in which banks operate. Advances in computing and communication, in measurement and management of financial risk and financial innovation have increased the returns from expanded scope and scale, thus facilitating international expansion. Limits to domestic growth and a reduction of oligopolistic rents as competition increased then helped to provide incentives for expansion.

In many ways, Iceland fits this bill. Membership in the European Economic Area in 1994 provided access to EU markets and a level playing field. Privatisation of the banking system in the late 1990s and early this decade created the incentives and the dynamism. Then, the country's banks embarked on an international expansion that was partly designed to follow Icelandic firms in their own outreach. This has now created a situation where three-quarters of the total lending of the largest commercial bank groups is to non-residents. This episode raises an interesting question that has to my knowledge not been given a satisfactory answer: what determines which countries become predominantly home countries and which become predominantly host countries? Why is Iceland a home country but New Zealand and the Baltic countries, to name a few examples, host countries? I leave the question with you, but I suspect that accidents of history might have as much to do with it as current assessment of economic efficiency.

The challenges for prudential authorities created by the internationalisation of banking are many. First, internationally active banks are more complex, as you know better than most, requiring more sophisticated resources from supervisory authorities. Second, in the absence of global or regional supervisors, internationalisation of banking requires information, views and local knowledge to be shared between supervisors in different jurisdictions, in particular between home and host authorities. Third, crisis management and resolution will necessarily involve several jurisdictions, with all the complexities and tensions over burden-sharing that this implies. Fourth, emergency liquidity assistance will be complicated or even impossible for central banks to deliver when internationally active banks face liquidity problems in currencies other than that of their home country. Iceland is a case in point.

The fundamental problem is of course that there is a mismatch between the international scope of banking institutions and the national scope of the frameworks for banking supervision and crisis management. The home-host principle is designed to meet these challenges, at least partly. However, it is currently being strained in several respects. Let me mention three examples:

First, banks might be systemically important in the host country but not in the home country, leading to differences in perspectives on the importance of supervision or at the time of stress. An example of that are many countries in central and eastern Europe.

Second, internationally active banks might be engaging in activities in host countries that are risky in terms of the macrofinancial stability in those countries, but where the risks are not sufficiently large to threaten the stability of the institutions concerned. An example is the foreign currency lending of some major European banks to unhedged borrowers in central and eastern Europe.

Finally, shocks to the operations of institutions in host countries might have systemic effects on the banking system in the home country, if the host country operations are sufficiently large. An extreme example of that is the Icelandic banking system. However, the other side of that coin is that the banking system in Iceland is more resilient to domestic macroeconomic shocks as its relative exposure to the domestic economy is much less than it used to be.

The setting-up of supra-national (that is global or regional) supervisors requires political will and initiative that is hard to muster. Although there has been some discussion about setting up such a supervisor for internationally active banks at the EU level, it is in the best of cases some time off. However, that does not mean that progress is not being made. On the contrary, policymakers have been dealing with these issues through various other means. Let me mention few examples. First, international standard setting has aimed at co-ordination and harmonisation of rules and supervisory practices. Second, Basel II has set up a common regulatory infrastructure which further reinforces cross-border harmonisation and co-ordination among supervisors. Third, at the level of the European Union there are significant efforts taking place, including joint crisis management exercises. Fourth, as you know very well, the most important supervisors for major global banks get together to discuss supervisory strategy. For example, I could mention the close collaboration of supervisors in the Nordic countries on the supervision of Nordea.

Monetary policy

My final topic is how financial globalisation is affecting monetary policy, especially among small and medium-sized countries that operate open capital accounts and flexible exchange rates.

Although sometimes forgotten when discussing the current plight of countries like New Zealand or Iceland, we know from theory that full financial globalisation will result in real returns on financial assets with similar maturity and risk being equalised across countries. For the small open economy that is unable to affect global financial conditions, this means that monetary policy will not be able to influence domestic real interest rates. Its ability to affect domestic demand through the interest rate channel would then disappear. That still leaves the exchange rate channel, which is sufficient for monetary policy to hit any inflation target in the medium to long run and potentially retain some countercyclical force in the short run, provided of course that monetary authorities do not try to fix the exchange rate.

These results are of course not new. Nobel Laureate Bob Mundell demonstrated in a series of articles in the early 1960s that for the small open economy with free and frictionless capital movements, monetary policy working only through the exchange rate would be a powerful stabilisation tool when the exchange rate floats but totally ineffective when it is fixed. The reverse would hold for fiscal policy.

We are still some way from this state of affairs, although a few small open mature economies might be getting closer. However, in order to know where we are heading, even if we might never completely get there, it might be interesting to speculate on what would happen to monetary policy if globalisation, both real and financial, were to run its full course.

In a similar fashion to financial globalisation, I define real globalisation as the cross-border integration of markets for goods, services and factors of production. In the extreme case when real side globalisation has run its full course, all goods would be traded, there would be no domestic non-traded goods sector. Furthermore, there would be instant factor mobility, implying that real factor returns are equalised across borders and the domestic output gap becomes irrelevant and meaningless. In fact, there would be no specific national resource constraint, except land.

This might seem far-fetched and in some sense it is. Yet it demonstrates the direction we are heading in. Moreover, we already see several signs of these phenomena emerging. You only need to think about how the inflow of labour has played a significant role in relieving labour market pressures in countries like the United Kingdom and Iceland in recent years. Switzerland is another long-standing example.

We now add full financial globalisation to the picture. Then the real risk-adjusted yield curve is, through speedy arbitrage, completely determined by the global curve and unaffected by domestic monetary policy, even in the short run. Monetary policy would then lose all its countercyclical force. It is anyhow not needed as there is no domestic output gap that needs to be stabilised. However, monetary policy would, through the exchange rate channel, be able to deliver any inflation target that the authorities want, by creating deviations of the domestic nominal policy rate from the global rate. Monetary policy has, however, no real effect. It can only determine the inflation rate, which is also neutral in its effect on the real economy.

We are still far from this extreme case. Furthermore, a plausible argument could be made that financial globalisation might progress more rapidly than real globalisation. We would then have a situation where the countercyclical force of monetary policy would still be useful but the interest rate channel would be significantly weakened, or even fully blocked. To what degree that would constitute a problem would depend partly on how well the exchange rate channel operates.

Yet this is where the concerns rise. Evidence seems to suggest that foreign exchange markets exhibit excess volatility and that exchange rates diverge from fundamentals for lengthy periods. The existence of carry trade can in some sense be taken as evidence of this, as it involves a bet that interest rate differentials are not fully compensated by exchange rate movements, that the so-called uncovered interest rate parity does not hold. According to the theory of uncovered interest rate parity, low-yielding currencies should be expected to appreciate and high-yielding currencies should be expected to depreciate. However, what we observe over lengthy periods is the reverse, followed by sharp corrections. In this regard, one could reflect on the yen versus the New Zealand dollar, or the situation here in Iceland.

What can small and medium-sized countries do when faced with a weakened interest rate channel and a misbehaving exchange rate? It seems to me that they have basically three options.

First, they can decide to live with it. After all, they will eventually be able to deliver their inflation target. The road will be bumpy in terms of exchange rate volatility and potential misalignments but it is not clear how strong the detrimental effects on the traded goods sector really are, partly because a sophisticated financial sector can provide hedging instruments.

Second, they can try to sharpen and realign existing instruments in order to reduce the burden on monetary policy and exchange rate adjustment with the aim of reducing adverse effects on the traded goods sector. This would involve measures like shifting the policy mix in the direction of fiscal policy, recalibrating prudential instruments with a view to minimising procyclicality, reviewing the tax and incentive structures of asset markets, particularly housing, and, maybe, even an occasional foreign exchange intervention. This seems to me to be the road that New Zealand, the pioneer of inflation targeting, is currently embarking on. However, it remains to be seen how much mileage can be got out of measures of this type; there are, for instance, well known problems with using fiscal policy for short-run stabilisation purposes, especially if there is already a significant fiscal surplus.

Finally, they could radically change the framework by entering a monetary union. Each country faces different options and its particular pros and cons in such regard. However, it is clear that, as both real and financial globalisation progress, the relative attractiveness of entering a monetary union increases, everything else being equal. The reason is of course that we are heading in a direction where for small economies that are unable to influence global interest rates; countercyclical monetary policy will be both impossible and unnecessary.

Concluding remarks

Chairman, ladies and gentlemen, let me conclude. Financial globalisation has big benefits and we hope that it will continue to make progress. Yet, as with most good things in life, it carries risks, and creates challenges for prudential and monetary authorities. There is no blame intended when I say that the public sector response is lagging. These are difficult problems and political structures are still to a significant degree nationally based. However, prudential and monetary regimes will probably have to adjust. There are several possible scenarios and it will be interesting to see how this unfolds. Could it be that after several decades we will have significantly fewer currencies in the world and fewer financial regulators as well, at least for internationally active banks?

Thank you very much.

Már Gudmundsson is the Deputy Head of the Monetary and Economic Department of the Bank for International Settlements. The views expressed are the author's and not necessarily those of the BIS.