This section looks at how taxes influence the international movement of capital, and in particular considers how the tax reform programs implemented in several industrial countries in recent years may have affected international capital flows.

This section looks at how taxes influence the international movement of capital, and in particular considers how the tax reform programs implemented in several industrial countries in recent years may have affected international capital flows.

In general, taxes may affect capital flows in three basic ways. First, taxes affect aggregate savings and investment in a country and hence have an impact on the net flow of capital into or out of the country. Second, taxes influence portfolio composition—including the form in which individuals hold their wealth, the mix of assets held and liabilities incurred by financial intermediaries, and the manner in which firms finance their investment; these decisions have implications for the gross flows of capital among various countries. Finally, taxes affect the location of trading activity in particular financial assets, and hence gross flows, by providing an incentive for trading to take place in lower-tax jurisdictions.

Taxes and Net Capital Flows

Taxes that reduce aggregate savings or raise aggregate investment tend to raise a country’s net financing need and hence increase net capital inflows. There are several channels through which taxes may affect aggregate savings.53 First, an increase in tax collections, for a given level of government spending, results in an increase in government savings and will tend to raise national savings (although there may be some offsetting decline in private saving). Second, the taxation of interest income alters the terms on which individuals can exchange present for future consumption. In particular, an increase in income tax rates will reduce the degree to which present consumption may be exchanged for future consumption, and therefore tends to lower private savings. A third consideration is that if tax changes alter the distribution of after-tax income, total savings may be affected to the extent that individuals in higher-income groups typically save a larger percentage of their incomes. Fourth, the taxation of dividends provides an incentive for corporate savings, or retained earnings; however, such changes in corporate savings may be offset by changes in individual savings, to the extent that individual shareholders “see through the corporate veil.”54

The tax incentives for private saving have been affected by the tax reforms that have recently taken place in a number of countries (e.g., the United States, Japan, the Federal Republic of Germany, the United Kingdom, and Canada). These tax reforms have typically entailed a reduction in basic rates of income tax, a flattening of tax rate schedules, and a reduction in the number of rate brackets. In addition, there have been changes in the specific tax treatment of particular forms of savings. For example, in the United States, the conditions for individual retirement accounts were made more stringent; in Japan, tax-free savings accounts (maruyu) were abolished; in the United Kingdom, restrictions on the preferential tax treatment of owner-occupied housing were tightened; and in several countries (including the United States and Japan) the preferential treatment of capital gains was attenuated.

Taxes also affect aggregate investment in an economy. The corporation tax is particularly important in this regard. A tax on pure corporate profits would have no effect on a profit-maximizing firm’s activities, since the tax reduces the after-tax cost of and returns to capital in equal proportion. In practice, however, corporate taxes may encourage or discourage investment because taxes are not levied on pure profits. In particular, tax allowances for depreciation do not correspond to an economically relevant measure, both because of the difficulty of measurement and because of the deliberate use of accelerated depreciation schedules as an incentive for investment. In addition, nominal, rather than real, interest expenses are typically deductible from the taxable base, ignoring the concomitant effect of inflation on the value of a firm’s assets and liabilities. Moreover, in some countries investment tax credits may provide an incentive for investment. The effects of these factors interact with the level of the corporate tax rate itself.55

Many countries have recently undertaken reforms of the corporation tax. In general, basic rates have been reduced, but the recent tax reforms have typically increased the total amount of corporate tax levied, by broadening the tax base by a greater proportion than the reduction in rates. For example, recent tax reform efforts in the United States have reduced the implicit subsidy for investment by cutting basic rates of corporate tax while eliminating the investment tax credit and tightening depreciation allowances; similar changes to rates and depreciation allowances were also made in the United Kingdom in 1984–86 and in several other countries.

Taxes and Gross Capital Flows

Portfolio Composition

Individuals in each country may diversify their wealth over a variety of domestic and foreign assets, while financial intermediaries and other firms may diversify the sources of their borrowing as well as their portfolio of assets. Taxes affect both individuals’ and firms’ preferences between assets and liabilities, domestic and foreign. Changes in taxes will result in shifts in portfolio preferences, which will typically be reflected in changes in the relative rates of return on alternative financial assets and in gross capital flows as portfolios are adjusted.

If investors were taxed on their worldwide income at the rate applicable in their own country, then taxes would have little influence on what assets they hold (although taxes could induce changes in residency by investors seeking a “tax haven”). Analyzing the effect of taxes on capital movements consists of identifying the many important exceptions to this general principle.

One important exception is related to tax evasion. Investors who wish to evade tax typically seek to hold assets that can be held anonymously and free of withholding tax. This has been an important element in the development of the Eurobond market: Eurobonds have always been bearer—that is, unregistered—bonds and are not subject to direct supervision and tax withholding by the national authorities. The incentive to evade taxes depends on income tax rates, especially the top marginal rates to which wealthy taxpayers are subject; thus recent tax reforms that have reduced the maximum marginal tax rates have tended to reduce the flow of funds into the Eurobond market.

The withholding of tax from investment income has recently been a controversial issue. The impact of withholding tax on capital flows is well illustrated by the recent withholding tax changes in the Federal Republic of Germany. The announcement, in October 1987, that a 10 percent withholding tax on interest income would be introduced as of January 1, 1989, led many German investors to purchase deutsche mark Eurobonds and other offshore assets (including assets denominated in other currencies), and many German financial firms, including at least one major bank, set up offshore investment funds in Luxembourg to facilitate such capital outflows. As a result, domestic bond yields rose abruptly, increasing the cost of deutsche mark borrowing for German residents, including the Government, over the cost for nonresidents.56 Accordingly, many German borrowers withdrew plans for bond issues on the local capital market, preferring to raise funds through their offshore subsidiaries. Despite these offsetting inflows, the overall result was substantial downward pressure on the deutsche mark notwithstanding Germany’s substantial current account surplus.

The considerable economic impact of the withholding tax led to its abolition, which was announced in April 1989 to be effective July 1, 1989. However, despite the policy reversal, there is a widespread perception that the situation has permanently changed, since investors have now learned to use the offshore markets and the institutions have been established to allow them to do so. Thus, it is not expected that all of the “capital flight” will now be repatriated.

A similar episode occurred in connection with the announcement by the U.S. authorities in June 1987 that beginning the following January they would impose a 30 percent withholding tax on interest payable on any bond issued in the Netherlands Antilles since 1984. This announcement was followed by a plunge in prices of the affected bonds, and some observers of the Eurobond market claimed that there was a detrimental effect on the market for zero-coupon bonds that continued long after the announced tax withholding rule was rescinded.57

The impact of tax withholding on capital movements depends on whether tax is withheld from all assets and all holders equally. In some countries, such as Canada and Australia, taxes are withheld only from interest paid to nonresidents; in others, such as France, tax is withheld from all interest regardless of whether or not the recipient is a resident, but tax is refunded to nonresidents. The first arrangement discourages capital inflows, while the second encourages capital outflows.

In light of the strong potential impact of withholding tax on capital flows, there is now concern to harmonize withholding tax policy in the European Community in the context of the attempt to create a single Community-wide market by the end of 1992. Without such harmonization, there are fears that, as barriers to capital flows are removed, investors’ funds will flow from member countries that withhold tax (such as France and Belgium) to countries that do not (such as Britain and Luxembourg). The European Commission suggested that all countries withhold tax on interest income at the uniform rate of 15 percent, but this proposal met with resistance from some member countries that were concerned that it would reduce the attractiveness of financial assets everywhere in the EC and that, as a result, funds would seek a withholding tax-free environment elsewhere in the world, thus jeopardizing the status of Europe’s international financial centers.58 The proposal to harmonize withholding taxes has therefore now been abandoned; in its place is a plan for tighter cooperation among the national fiscal authorities of the EC member countries with a view to discouraging tax evasion.

In addition to tax withholding, there are several other tax provisions that favor one form of financial asset over another, and have had an important effect on capital flows. The tax treatment of capital gains played an important role in the development of the Eurobond market, especially in the early 1980s. Zero-coupon Eurobonds, which provided income exclusively in the form of capital gains, were particularly popular with Japanese investors, who faced high tax rates on interest income but no tax on capital gains. Tax reforms such as those in the United States and Japan, which have reduced the discrepancy between taxation of income and capital gains, as well as earlier changes that refined the definition of capital gains in the case of zero-coupon bonds, have reduced the tax advantage of zero-coupon bonds and therefore reduced the level of activity in these bonds.

A feature of the tax system that has tended to discourage international capital flows is the preferential tax treatment of owner-occupied housing in many countries, including the United States and the United Kingdom.59 This creates a strong incentive for households to save in residential real estate as opposed to holding other assets, including foreign assets. The special treatment of pension funds and life insurance companies also may affect portfolio composition, and thus international capital flows, to the extent that these institutions invest their funds differently than their customers would as individuals. This may be the case, as the nature and size of these institutions may lead them to have different investment criteria and to face different regulatory constraints. For example, in Japan, both life insurance companies and pension funds are restricted in the proportion of their portfolios that may consist of foreign assets. The progressive liberalization of these restrictions has been followed by an increase in insurance companies’ foreign assets.

Location of Market Activity

Taxes affect the costs of trading securities themselves, and therefore influence the location of activities in securities markets. For instance, stamp duties and turnover taxes are applied to securities’ transfers in many countries—including the Federal Republic of Germany, the Netherlands, and Switzerland—and have discouraged market development particularly at the short end. These taxes provide sufficient disincentives for large transactions to be sourced in London and explain in part the slow development of the money and short-term securities markets in those countries.60

The scope for conducting business offshore is increased by the possibility of assembling “synthetic securities”—combinations of securities and derivative products that achieve a given combination of risk and expected return in a more tax-efficient way. For instance, a common practice recently has been to assemble a “synthetic Bund,” by combining the Bund futures traded on LIFFE with other securities in order to duplicate the risk-return characteristics of the Bund while avoiding German taxes.61 The use of futures markets to assemble such synthetic securities increases the potential for funds to flow, typically in both directions, in response to tax differences.

As turnover and other taxes have often shifted trading to offshore markets, conversely, tax concessions have often been of considerable importance in the establishment of new financial markets. For example, the establishment of the MATIF was facilitated by changes in the French tax laws in May 1987. Tax incentives were important in the development of futures trading in Singapore; for example, the success of the newly introduced fuel oil contract was facilitated by the decision, in February 1989, to extend to this contract the 10 percent tax concession applied to other contracts traded on SIMEX.

To the extent that taxes affect the location of market activity, this is inevitably reflected in gross—although not necessarily in net—capital flows. For instance, if a Japanese investor purchases a Japanese company’s shares issued on the London market, this results in a two-way flow of funds between Japan and the United Kingdom. The location of trading need not correspond to the identity of either the issuer of the securities or the investor; taxes may therefore, in principle, have effects on capital movements that do not correspond to any effects either on savings and investment or on the portfolio structures chosen by households and firms.


Many of these issues are surveyed in Roger S. Smith, “Factors Affecting Saving, Policy Tools, and Tax Reform: A Review,” IMF Working Paper No. WP/89/47 (unpublished, Washington: International Monetary Fund, 1989); in A. Lans Bovenberg, “Tax Policy and National Savings in the United States: A Survey,” IMF Working Paper No. WP/88/110 (unpublished, Washington: International Monetary Fund, 1989); and in A. Lans Bovenberg and others, “Tax Incentives and International Capital Flows: The Case of the United States and Japan,” IMF Working Paper No. WP/89/5 (unpublished, Washington: International Monetary Fund, 1989).


In the sense that individuals may perceive the actions that the firms whose shares they own are taking on their behalf, and adjust their actions accordingly. For example, a shareholder in a firm that retains earnings may reduce his own personal savings to achieve a desired overall level of savings. See Franco Modigliani and Merton Miller, “The Cost of Capital, Corporation Finance and the Theory of Investment,” American Economic Review (Nashville), Vol. 48, No. 3 (June 1958), pp. 261–97.


For an analysis of these effects, see Robert Hall and Dale Jorgenson, “Tax Policy and Investment Behavior,” American Economic Review (Nashville), Vol. 57, No. 3 (June 1967), pp. 391—414.


For example, yields on World Bank deutsche mark bonds moved below those on German Government bonds, rather than trading at the normal 30-40 basis point premium.


See Peter Gallant, The Eurobond Market (Cambridge: Woodland-Faulkner, 1988).


See Fiscal Affairs Department, Tax Harmonization in the European Community (Washington: International Monetary Fund, forthcoming).


See A. Lans Bovenberg, “Tax Policy and National Savings in the United States: A Survey,” IMF Working Paper No. WP/88/110 (unpublished, Washington: International Monetary Fund, 1988).


For example, it is estimated that 30–50 percent of all trading in Dutch state bonds is done in London.


As a simple example, one could hold a 250,000 deutsche mark Eurodeposit maturing in one year and purchase one Bund futures contract with a face value of 250,000 deutsche mark, for delivery in one year. If the annual percentage yield on the Eurodeposit were the same as the coupon on the Bund specified in the futures contract (e.g., 6 percent), this would be identical to purchasing a Bund now (with one year longer maturity than the one delivered). In practice, synthetic Bunds may be much more complex, also including securities in other currencies hedged with currency futures.