(b) Financial Market Integration

Financial markets are also more integrated within countries than they are between countries. While the reasons for the different degree of integration are not entirely clear, currency issues are an important factor. Given that low financial market integration can be quite costly for agents living in small currency zones, economists have identified enhanced financial market integration as one of the main benefits of a monetary union.

Economists have identified three consequences of financial market integration:

1. when finance markets are fully integrated, prices for similar assets in different regions are the same (except for trivial transactions costs);
2. when finance markets are fully integrated, agents in different regions have access to and use financial assets from different regions to save, borrow, invest, and insure; and
3. when both finance markets and goods markets are integrated, local saving and local investment decisions should be independent, and the capital account position should smoothly adjust to offset desired current account positions.

It should be stressed that both goods market and financial market integration are necessary for full economic integration, as any desired current account position entails the simultaneous exchange of financial assets. Consequently, if local people prefer to exchange financial instruments with other local people rather than with outsiders, goods market flows will be impeded.

(i)  Microeconomic Issues in Financial Market Integration.

Bayoumi (1997) provides a thorough review of the literature on financial market integration, splitting the topic into its microeconomic and macroeconomic aspects. He argues that international financial markets are now highly integrated in the sense that prices for the same wholesale instruments are similar around the world. There is little difference in the price of US bonds in Chicago, New York, London, or Tokyo, for instance, although this has not always been the case. [13]

While the price of identical financial assets is similar in different countries, agents in different countries purchase and use quite different assets. This contrasts the behaviour of agents in different locations in the same country, who appear to purchase and use similar assets. In part, this is because not all financial assets are available to people in different countries; in particular, poor agents may not be able to borrow in foreign currencies, as the local banking systems will not typically issue small value foreign currency loans.[14] [15] These asymmetries contribute to people having a pronounced home bias in the assets they hold. This is true in both debt-type instruments, which are overwhelmingly denominated in the home currency, and equity-type instruments.

A home bias in debt instruments means that people predominantly borrow and lend in their own currencies. This behaviour has the natural “insurance” advantage that variations in the value of the currency are offset with changes in the value of labour earnings and the cost of consumption, and consequently this home bias is likely to reflect risk aversion. This bias may have two disadvantages for people living in small currency zones. First, borrowers and lenders are limited to diversify over a smaller number of people, and thus are exposed to a greater amount of credit risk. Most New Zealanders currently lend to other New Zealanders, for instance, because there are few other people in the world who want to borrow New Zealand dollars. Secondly, since real interest rates differ across currency zones, people living in a small currency zone which is a capital importing region may face a high risk premium and thus high interest rates compared to the situation that they had a monetary union with a larger country.

The second issue warrants further comment. If people located outside the currency zone demand a currency risk premium in addition to a geographic risk premium to lend to people located inside the currency zone, the residents of the currency zone will be required to pay an additional premium to borrow if they are net debtors to the rest of the world (for a formal model investigating currency premiums see Carlson and Osler 1996). Because most currencies are national, there is little direct empirical evidence of differences in currency risk premiums and geographic risk premiums.[16] However, geographic risk premiums in common currency zones are typically very small, whereas risk premiums between currency zones are typically large, suggesting that most of the risk is currency risk.[17] Moreover, in small countries such as Austria or Netherlands which have fixed their currency to larger countries (Germany), interest rates in the smaller currency have converged to interest rates in the large country, often despite quite disparate economic indicators such as government deficit levels (Genberg 1990; Emerson et al 1992; Tatom and Proske 1994). If New Zealand borrowers are primarily facing a currency risk premium rather than a country risk premium, adopting another currency will lower the risk premium they pay. Given the high real interest rates New Zealanders have been required to pay over the last decade, this could be a substantial benefit to the economy.[18][19]

Because equity price volatility is greater than currency volatility, it seems less plausible to attribute a home bias in equities to currency risk than it does to attribute home bias in debt instruments. Nonetheless, home bias in equities is pronounced, and is a facet of the lack of integration between countries (French and Poterba 1991).

It is clear that the degree of home bias is falling over time, as people in most countries hold an increasingly large fraction of their portfolios in foreign instruments. Still, the degree to which most portfolios are home biased is enormous, and if this bias partly reflects currency concerns an enlargement of a nation’s currency zone will improve welfare by reducing the degree of home bias.

(ii) Macroeconomic Issues in Financial Market Integration.

The basis of the macroeconomic literature examining financial market integration is the Feldstein-Horioka result (Feldstein and Horioka, 1980). They found that domestic savings rates are typically highly correlated with domestic investment rates, suggesting that there is not a large degree of international financial market integration, for if there were savings decisions in one location would be independent of investment decisions in that location. This result was of considerable surprise, as most economists had thought that international capital markets were substantially integrated. While the degree to which national savings and national investment are correlated is declining over time, the result still largely stands: in aggregate, countries do not seem to use financial markets to fully separate savings and investment decisions. (See Bayoumi (1997), or Lewis (1995) for a discussion.) One implication is that countries do not use international financial markets as much as they could to diversify risk and smooth consumption shocks. Another implication is that capital account issues provide some constraints on the current account positions that countries can run.[20]

There are two types of evidence indicating that geographic borders and separate currencies limit international financial market integration. First, personal savings and investment decisions are basically uncorrelated in separate regions within countries, in contrast to the high degree of correlation between countries, implying a high degree of capital market integration within countries. (See the summary in Bayoumi 1997 (chapter 3) for this literature; also Helliwell and McKitrick (1998) and Helliwell (1998) for a more recent analysis of the Canadian provincial data.) This evidence is supported by an analysis of cross-country savings and investment flows during the Gold Standard era prior to 1914. During this period exchange rates were fixed and capital was highly mobile, and the correlation between savings and investment in different countries was considerably smaller than it is now (Bayoumi 1990)[21]. This evidence is interesting as it suggests that the issue is not only one of international borders, but of exchange rate regimes as well.[22] The current relatively low degree of international capital mobility corresponds to the relative importance placed on international current account imbalances. Internal current account positions are not measured or cared about; and during the Gold Standard era countries maintained large current account deficits or surpluses for extended periods of time.[23]

The second body of evidence concerns the extent of risk sharing between countries and risk sharing within countries. In theory, people should be able to use capital markets to both borrow and lend in the face of income shocks, and to insure against income fluctuation. Either way, consumption should be less volatile than income. The international data shows that there is not much international risk sharing of this sort: local consumption tends to be highly correlated with local GDP, and little correlated with consumption in other countries. This appears to be a major anomaly, if we assume a world of risk averse agents, for there appear to be significant welfare gains from wider risk sharing (van Wincoop, 1994). There is growing evidence that risk sharing between countries is considerably smaller than risk sharing between regions within countries, suggesting that borders matter. (Bayoumi and Klein, 1997, analysing Canada; Asdrubali, Sorensen, and Yosha (1996) analysing the United States). This data also suggests that risk sharing of capital income is much more affected by the border than risk sharing of labour income, for the simple fact that little labour income risk tends to be shared using capital markets within countries or between countries.[24] This data is clearly consistent with the other evidence that savings is less correlated with investment within countries than it is between countries.

The extent to which the differences in intra-national capital flows and international capital flows are caused by border issues rather than currency zone issues is not clear. However, there are reasonable grounds for believing that currency zones are a major part of the story. Certainly the evidence of the high degree of capital market integration during the Gold Standard era supports this interpretation. Moreover, it seems reasonable to suppose that capital flows within countries are mobile because transfers within the banking system of a single currency zone are more or less automatic. Consequently, it seems likely that a country that chooses to have its own currency chooses to limit the extent to which it enjoys capital market integration and risk sharing - and that these costs are likely to be higher for small countries than for large countries because of the low degree of risk diversification available in a small country. [25]

(c) Economic Integration  and the Location of Economic Activity

In the last decade, a large literature has emerged analysing the effects of increased economic integration on the spatial distribution of employment and output. This literature has demonstrated that even if a country promotes greater international integration to reduce home bias, increase trade, and increase specialisation, the benefits of integration will not necessarily flow to residents remaining in the country. Rather, a region can decline following increased integration with another region either because of a migration of resources from one region to another, or because of the detrimental effect of increased competition on local firms.[26] For this reason, while policy makers have generally viewed measures taken to enhance integration as a “good thing”, some caveats should be noted.

Most of the literature examining this issue has been based upon models of industries with increasing returns to scale operating under conditions of imperfect competition. These models suggest that such industries will limit production to a few locations, because of economies of scale, and that these locations will be in regions with the best market access such as large cities. Wages in these regions will be high, and these regions will become net exporters of goods subject to increasing returns. When transaction costs fall, new and existing firms may choose a new location. There are two countervailing forces affecting where firms choose to locate. As transactions costs fall from high levels to intermediate levels, the benefits of increasing returns to scale dominate, leading to a shift to regions with better market access (the core). As they fall further, these costs cease to matter as much and the disadvantages of high wages in the centre offset the returns to scale and some firms relocate back to the periphery. Consequently, at some stages of a decline in transaction costs it is possible that the periphery region will lose high paying jobs to the core region.

The evidence on these theories is mixed. Krugman (1993) argues that the lower transaction costs and higher labour mobility in the USA compared to Europe is reflected in the greater regional concentration and specialisation of manufacturing industries. Other authors analysing the manufacturing and agriculture sectors have supported this observation, but it does not appear to be true for all industries. Labour market mobility appears to be a crucial factor, because if people are mobile they are more likely to migrate out of a region with low employment than companies and jobs are likely to migrate in, speeding up forces favouring large agglomerations.

The New Zealand economy is more integrated with the Australian economy than any other, and it is reasonable to consider cities such as Sydney or Melbourne as the local “core”, and smaller cities such as Brisbane, Auckland or New Plymouth as the periphery, particularly as international evidence suggests that large cities tend to be more productive than small cities (Ciccone and Hall, 1996).^[27] Decreases in transactions costs already cause a relocation of economic activity around New Zealand, and between New Zealand and the rest of the world, and there is no reason to believe that increasing economic integration further would be any different. Some New Zealand industries would move overseas, and some overseas industries would move to New Zealand. In general, policy makers in New Zealand have proceeded as if falling transaction costs and lower barriers to trade will be good for New Zealand residents, not bad.[28] [29] In the absence of research demonstrating that closer integration will be detrimental to New Zealand, the assumption that closer integration is beneficial to New Zealand has been maintained in this paper.