The first layer is withholding at source on dividends. Most countries tax dividends paid to foreign investors, at rates that vary by country and reach substantial levels in common cases. The tax is deducted before the dividend arrives, so investors see a net figure. Return calculations using published dividend yields therefore overstate actual income systematically. The second layer is treaty relief. Where a treaty exists between two countries, the withholding rate is usually reduced, and obtaining that relief requires documentation and an application process. The complexity and practicality of that process differ by country, and in some cases the administrative cost of reclaiming exceeds the amount involved, so it is commonly abandoned.
The third layer is taxation in the country of residence. Foreign income generally remains reportable at home, and whether tax already paid at source can be credited, and to what extent, depends on local rules. Well-designed systems avoid double taxation, and complete elimination is less common in practice than the principle suggests. The domicile of a fund changes the whole structure. The same underlying holdings, held through funds registered in different jurisdictions, can face different withholding rates, because the fund's own tax status determines which treaties apply to it. Two funds tracking the same index can therefore show observable differences in long-run return for this reason alone. Accumulating and distributing share classes present another distinction.
Accumulating funds reinvest dividends without distributing them. Under some tax systems this defers taxation in the country of residence, and under others it makes no difference at all or produces unfavourable treatment. This depends heavily on where the investor is resident. Capital gains treatment varies more widely still. Some countries do not tax non-residents on gains, others do under specific conditions, and some markets set thresholds based on holding period or ownership percentage. These rules are easy to overlook in cross-border investing, since they generally do not appear anywhere in a trading interface. Estate and gift rules are a further layer receiving even less attention.
Some countries levy estate tax on assets held locally by non-residents, with exemption thresholds that can sit far below those available to residents. For investors holding large foreign positions, the potential effect of this can exceed all other tax costs combined. What these items share is that they can be established in advance and seldom enter return estimates. Several choices that appear minor turn out to matter. Which jurisdiction's fund to use, whether to take accumulating or distributing shares, and whether to hold directly or through receipts all produce considerable differences when compounded. One general caution applies.
Rules in this area change, and treaty terms are renegotiated periodically, so a structure that is efficient today may not remain so. Anything requiring highly specific arrangements carries a risk that the arrangement itself is altered. Simpler structures tend to be more durable, even where they are marginally less efficient at any given moment. One practical starting point exists for anybody wanting to quantify this. Comparing the published gross and net return series for a fund shows the aggregate effect of withholding on that vehicle. The difference between the two is stated rather than estimated. It captures only one of the layers described above, and it establishes the size of the largest one.