Why the capital account matters.

Pain, Nigel ; Westaway, Peter

WHY THE CAPITAL ACCOUNT MATTERS

Introduction

Most discussion of the balance of payments and its implications for exchange-rate prospects and economic policy falls into two distinct categories. Some authors focus on the current account alone while others argue that in a world of liberalised capital markets information from the volume of trade flows will simply be swamped by flows of highly mobile international capital. In this note we argue that both these viewpoints are too extreme; in aggregate both the current and capital accounts will matter.

Of course this is simply another way of stating the more sophisticated theoretical argument that the exchange rate must act as a market clearing price, moving continuously to reconcile the total demand and supply for sterling in the face of both expected and unexpected changes in current and future activity in both goods and financial markets. Despite the underlying complexity, this note argues that it is still possible to use the capital account to obtain a measure of exchange-rate pressure which provides more information than the current account alone. Moreover, the useful information contained in the composition of the capital account allows us to distinguish between long-term structural trends in the flow of capital and more short-term speculative movements.

We begin by outlining recent developments in the balance of payments and then relate them to the theoretical determination of the exchange rate in foreign exchange markets. We later deal with arguments against the use of the capital account in such analysis while the final section discusses some policy implications.

Recent developments in the balance of payments

In this analysis of the capital account we intend to focus upon foreign direct investment and portfolio investment, excluding that undertaken by the banking sector. These two categories are both the largest and most long-term of the various identified capital investments. The net sum of these two flows plus the current account will be referred to as the basic balance.

This does not rule out the possibility that other elements within the capital account are non-speculative; we will return to this issue below. For the moment we would simply note that while national definitions of the basic balance can vary, the concepts of long-term and short-term capital are widely used around the world by the IMF, the Bank of Japan and the Deutsche Bundesbank amongst others.

Recent developments in the UK balance of payments are summarised in table 1. Over 1979-83 the current account surplus was more than offset by net outflows of direct and portfolio investment, implying a small deficit in our basic balance measure. In part this may have been the result of adjustment effects induced by the abolition of exchange controls at the end of 1979. In the subsequent five years to 1988 the basic balance deficit increased somewhat as the current account moved in to deficit and net outflows of direct investment rose quite substantially. Net inflows within the banking sector increased in magnitude although there was actually a net increase in official reserves. [Tabular Data Omitted]

The data for 1989 (annualised using figures for the first three quarters) show a further deterioration in our basic balance measure to an annual deficit of 50.3[pounds] billion, some 9.8 per cent of GDP. In part this is due to a sharp turnaround in portfolio investment with net outflows rising to 25.2[pounds] billion. Marked changes are also apparent elsewhere within the capital account, particularly in the reserves and other investment. This latter development reflects an increase in net borrowing from overseas banks. We return to some of the factors behind these shifts and some potential caveats to our approach below.

It is instructive to place these changes in the basic balance in an international context. As chart 1 illustrates not only is this particular measure of changes in the composition of the balance of payments now much larger than anything in the recent past, it is also far in excess of those run over the 1980s by the world's major net importer, the United States, and the world's major net investor of long-term capital, Japan.

Theoretical implications

Why do we even worry about capital flows, long-term or otherwise, when we know that in total they are automatically equal and opposite to the current account? It is often suggested that any deficit on the current account has to be actively financed by attracting the requisite level of so-called `hot money' inflows in order to balance the books. The composition of the capital account has no role to play in this story and is of no concern to policymakers. Both these implications are seriously misleading.

Consider the purchase of an import by a UK resident. Payment involves a transfer of funds from the bank account of the UK resident to that of the foreigner. This transaction will be recorded as a banking sector capital inflow. Instantaneous balance prevails. Similarly, consider the purchase of a foreign security by a domestic institution. The purchase of the share is counted as a portfolio investment capital outflow. Correspondingly, the proceeds of the share sale will accrue to the bank account of an overseas resident. Again there is automatic balance. Neither of these transactions require consideration of hot money or long-term flows. Instead they comprise an underlying transaction (the import or the share purchase) and an accommodating banking sector flow. It is the sum of the underlying transactions which is usually identified as the basic balance.

The key point is that the ultimate implications for sterling of both the import and the share purchase are identical. In both cases there is a net transfer of funds from the bank accounts of domestic residents to those of overseas residents. If the desired proportion of foreign currency in the bank accounts of overseas residents is greater than that of domestic residents then, ex ante, there will be more net sterling deposits than investors want to hold. Attempts to switch into foreign currency will lead to banks entering the interbank market in order to attract the foreign currency funds required to satisfy the demands of foreign depositors. Interbank deposits are required as the banks will actively attempt to match the currency composition of their assets and liabilities.

The exchange rate must fall or the interest rate must rise to match demand and supply, with the size of the change depending upon the sensitivity of asset demands to the relative return on sterling compared with foreign currency, as measured by the difference in interest rates plus the expected fall in the exchange rate. In contrast to the usual explanation, the expected relative return must rise to persuade investors to hold on to sterling, rather than to attract the `hot money' inflow in the first place. The exchange rate is not moving in order to finance a given deficit but to determine how it is financed.

A number of extensions can be made to the basic story. Firstly, it is possible that domestic residents will choose to make payment by running down any deposits they may hold with overseas banks. The size of any adjustment in the return to sterling in this case will depend upon the currency preferences of depositors with both domestic and overseas banks. Secondly the extent to which the price of sterling adjusts may be limited by changes in government holdings of foreign exchange reserves. However, since reserves are finite, the counterpart to any sustained imbalance must ultimately consist of net flows within the banking sector. We would expect that such inflows, particularly of the order of 9 per cent of GDP, would cause the premium offered on sterling to rise in order to reconcile demand and supply.

Of course this does not imply that the present structure of the balance of payments necessarily implies an imminent collapse in the exchange rate. Indeed if a string of deficits in the basic balance has been anticipated, the information should already be incorporated in the current value of the exchange rate.

Once asset demands are in equilibrium we might expect net speculative flows to tend to zero. This does not however imply that our chosen measure of the basic balance should return to zero, that is, net banking sector flows are zero; it is quite possible that investors wish to hold a certain proportion of their growing portfolios in the form of deposits with overseas banks, although the demand for sterling may be less than the demand for dollars or marks.

The main point is simply that marked changes in the composition of the balance of payments are likely to prove unsustainable. Both the current account and the capital account can provide an indication of present and future pressures on sterling, ultimately resulting in price changes that will help to reduce imbalances and restore stock equilibrium by improving competitiveness and changing portfolio preferences.

Criticisms of the capital account approach

Three main arguments have been advanced to suggest why the basic balance and other measures of imbalance derived from the capital account may not be an appropriate policy concern; the uncertainty generated by the existence of the sizeable sectoral balancing item, the increased liquidity of international investments and the importance of speculative elements in determining the reported capital flows.

The balancing item is the statistical discrepancy between the recorded current account and capital flows. While it is undoubtedly large (as can be seen from table 1), it does not negate the essential message from the capital account. This point is illustrated in chart 2 where the original basic balance measure is plotted alongside a revised measure obtained by adding the balancing item to the recorded basic balance. The balancing item may be able to `explain' away part of the current account deficit or part of net longer term capital outflows but it cannot account for both.

As for the liquidity argument, whilst individual investments are highly liquid, the share of the portfolio held within a particular region may be less so if international portfolio diversification is seen as a means of reducing overall risk. Sceptics may point to the events in the last quarter of 1987 when, following the stock market crash, some 7 billion [pounds] worth of overseas portfolio investments were repatriated, in part to provide a source of funds to take up several large rights issues at that time. This does not however negate the risk argument. In a world characterised by equilibrium in asset stocks, flows may at times be perverse as investors seek to restore the desired composition of their portfolios.

The liquidity argument can also be applied to direct investment. By direct investment we mean an investment which is made to acquire a lasting interest in an enterprise operating in an economy other than that of the investor and which gives the investor an effective voice in the management of the enterprise. Such investment includes mergers and acquisitions as well as capital investment in greenfield sites. To the extent that recorded investment flows are increasingly dominated by new cross-border mergers and acquisitions it is possible to argue that direct investments are becoming increasingly liquid as firms may more readily be sold to other investors. The commitment to overseas production may be less in those companies who have simply acquired existing assets.

We view this argument somewhat sceptically. Both joint ventures and the purchase of the ownership rights to existing overseas assets may be preferred to investment in greenfield sites in order to overcome the transactions costs arising from the need to find suitable sites, obtain planning permission, hire labour and establish a new product. Joint ventures in particular provide existing managerial expertise and a potential source of additional equity for capital constrained firms.

The third criticism, related to the liquidity argument, is that the so called `long term' flows merely reflect expectations about the short term relative returns from investing in alternate locations. This oversimplifies the aggregate investment decision which in the case of portfolio investment will depend on both structural (domestic and world wealth, exchange controls and investment costs) and speculative (expected relative return) factors. Further details can be found in Westaway and Pain (1989).

It is possible to obtain an estimate of the structural influences in the recorded levels of direct and portfolio investment by using the estimated expectational effects embodied in the econometric equations in the NIESR Domestic Econometric Model. For portfolio investment we subtracted the expectational effects and revaluations from the actual end quarter stock. Our estimate of the structural flow is the difference between the revised stock and the actual stock at the end of the previous quarter.

Speculative factors may also influence direct investment. Ultimately expectations must dominate given that flows reflect the expected profitability from investing in alternate locations. However such expectations are long-term in nature and unlikely to be influenced by quarterly fluctuations in, say, relative labour costs. Moreover adjustment costs for fixed investment are such that it may take many years for differentials in national rates-of-return to be eroded. This is not to say that short-term expectations are unimportant although price movements are more likely to affect the timing of any investment rather than its long-term level. For the purposes of this analysis we have assumed that the dynamic exchange-rate terms in our equations for inward and outward direct investment (see Pain(1989), NIESR(1989)) adequately capture the importance of short-term expectations. They were therefore subtracted from the recorded flows to obtain our estimate of structural investment flows.

The structural flow estimates permit calculation of the underlying level of our basic balance measure. This is shown in chart 3, alongside the original measure from chart 1. Although the uncertainty surrounding the calculation of the underlying measure should not be underestimated, it seems clear that the recent changes in the capital account are due to identifiable structural trends rather than short term speculative factors. However, the chart also shows that speculative effects can be important and have grown in magnitude over the sample period, peaking in the middle of 1987 prior to the downturn in the stock market. For 1988 and the first three quarters of 1989 we estimate that such effects account for some 25 per cent of recorded net portfolio outflows.

There are a number of identifiable factors lying behind the upsurge in longer-term capital flows in the 1980s. Firstly, the abolition of exchange controls at the end of 1979 had significant effects on the level of both direct and portfolio investment outflows. Financial deregulation has been accompanied by increasing international diversification of investment portfolios. The adjustment process was by no means instantaneous due to the time required to acquire information on suitable investments and expand overseas dealing operations. Deregulation was also instrumental in improving the attractiveness of the UK as an investment location in the mid 1980s with the abolition of fixed brokers commissions at the time of the `Big Bang' reforms.

Secondly, there has been a rapid increase in the size of total portfolios over the 1980s. For example, the total assets held by domestic insurance and pension funds grew at an annual compound rate of 18.2 per cent between the end of 1979 and 1988. Institutions are continuing to acquire some 20-25 billion [pounds] of contractually related savings from the personal sector. At the present time this is combined with an absence of suitable domestic investments with the public sector maintaining its buyback of gifts, firms issuing little new equity and takeovers limiting the size of the equity base.

Foreign direct investment has been aided by the growth of world markets and the rise in real corporate profits which has provided much of the requisite investment finance. Expanding overseas has also become an essential part of minimising currency risks as it enables companies to obtain a closer match of their cost and income currency mixes.

Policy implications

A number of important implications for policy arise out of our analysis of the capital account. Most obviously it suggests that a blinkered focus on the current account alone may be misguided. Marked changes in the structure of portfolio preferences are of equal importance in understanding exchange-rate fundamentals. Correspondingly policy at both micro and macro levels might be widened to incorporate measures designed to encourage longer-term inward investment. Such policies would be desirable both for their beneficial supply-side consequences and also for their macroeconomic implications arising from their role in the determination of the exchange rate.

A large number of microeconomic policy measures fall into this category; investment incentives, tax structure and competition policies may all be important determinants of the scale and chosen locality of international investment. Recent work by the OECD (OECD 1989) suggests that international investment incentives can have a significant impact on the intra-regional regional location of investment within a wider market such as the EEC. The structure of business taxation also plays an important role in stimulating investment and it will be interesting to see whether the run-up to 1992 produces a round of successive tax reductions within Europe as national governments compete to attract multinational companies. Competition policy is another obvious source of uncertainty to the potential investor, particularly if applied inconsistently or if so-called strategic industries are protected. One of the main reasons for the negative inflow of foreign portfolio investment in the first quarter of 1989 was the force sale of part of stake held in BP by the Kuwait Investment Office.

Apart from explicit supply-side measures, the government can also affect the capital account through its own actions as an important player in financial markets; in particular through its role as a provider of long-term securities. In the recent past the policy of fully funding the PSDR has curtailed the supply of fixed interest sterling securities. Contrary to the government's hopes the gap has only partly been filled by private sector issues. With pension funds becoming increasingly mature the demand for fixed interest or index linked gifts to match known financing requirements has grown. A return to a policy of over-funding may ameliorate the consequences of these developments which have resulted in an increased outflow of portfolio investment into overseas government securities and corporate debentures.

So far we have mentioned policies which will have downstream macroeconomic implications. However it is equally important to emphasize that standard macroeconomic policy measures can have consequences for the capital account and hence influence the overall policy stance. For example it is possible to argue that any policy-induced improvement in the current account will help to stem the tide of outward portfolio investment, at least to the extent that it is driven by expectations of capital gains from a future depreciation in the exchange rate (see Davies 1989). However it is also possible that deflationary measures may worsen the structural capital imbalance by lowering expected profitability and making the UK a less attractive place to invest.

While the government could rightly claim to have undertaken a number of measures to attract overseas investment such as the reductions in corporation tax and the changes to stamp duty at the time of the `Big Bang', there is little sign that the capital account figures highly in the overall macroeconomic policy framework. This note suggests that a focus on both the current and capital accounts is required in order to fully assess any financing problems arising from the balance of payments. So long as imbalances of the present magnitude remain, it is difficult to see how domestic interest rates can be significantly reduced without the risk of a substantial fall in the exchange rate.

REFERENCES

Davies G (1989) `Key Aspects of the Balance of Payments Problem', Appendix 9,`The 1989 Autumn Statement', First report from the Treasury and Civil Service Select Committee, Session 1989-90 NIESR (1989), `National Institute Model 11', December 1989-90 OECD (1989) `Investment incentives and disincentives: effects on international direct investment" (OECD Paris) Pain, N (1989), `International direct investment flows and the UK economy', National Institute Discussion Paper No. 158 Westaway P and N Pain (1989) `Towards a structural model of the UK exchange rate' National Institute Discussion Paper No. 165