2.1 Scope of the study: key definitions
In this study, a cross-border trade is defined as a trade between counterparties located in different countries. A cross-border settlement is defined as a securities settlement that takes place in a country other than the country in which one or both counterparties are located. Under these definitions every cross-border trade results in a cross-border settlement, but cross-border settlements can also result from domestic trades.
In the simplest and most common case, a counterparty located outside the country in which the security is issued trades with a counterparty located in the country of issue, and the trade is settled in the country of issue. However, it may be noted that the country of issue of a security is not necessarily the country in which the security is most actively traded or the country in which the majority of trades are settled. In the case of Euro-bonds, issuers are located in a variety of countries, the most active trading is usually in London, and settlement usually occurs in Belgium (through Euroclear) or in Luxembourg (through Cedel). In the last few years, much the same pattern has emerged for certain European government securities much of the trading reportedly takes place in London rather than in the country of issue, and trades are often settled on the books of Cedel or Euroclear or via a link between one of these systems and a settlement system in the country of issue. For trades settled over these links, settlement should probably be considered to take place in more than one country. In any event, however, such settlements clearly should fall within the definition of the term cross-border settlement and, therefore, within the scope of this study.
It is worth noting that under the above definitions trading and settlement of a security outside the country of issue would not be considered cross-border trading or cross-border settlement if the counterparties are located in the same country and settlement takes place in that country. For example, secondary trades and related settlements of depository receipts would not be considered cross-border trades or cross-border settlements. In such cases, however, a bank or CSD in the home country must hold the securities in custody for a bank or CSD in the country in which the trading and settlement take place, so that a cross-border custody relationship is involved.
2.2 Expansion of cross-border trading and cross-border settlements
Comprehensive data on cross-border trading or cross-border settlements simply do not exist. The best data on cross-border trading are compiled from national balance-of-payments statistics. As shown in Table 1, these data indicate that in the G-7 countries the growth of cross-border trading of bonds and equities has far exceeded the growth of GDP over the last two decades and especially during the last few years. It should be noted that the quality of these data varies. The primary purpose of balance-of-payments statistics is to measure changes in net holdings of foreign securities by domestic residents and of domestic securities by foreign residents rather than gross volumes of cross-border trading. Indeed, the United Kingdom, which is probably the most important centre for cross-border trading, discontinued collection of gross transactions data after 1991. Nevertheless, that cross-border trading has grown extremely rapidly in recent years cannot be seriously disputed.
Table 1 Cross-border trading of bonds and equities1 (as a percentage of GDP)
| Country | 1970 | 1975 | 1980 | 1985 | 1990 | 1993 |
| United States | 2.8 | 4.2 | 9.3 | 36.4 | 92.1 | 134.9 |
| Japan | n.a. | 1.5 | 7.0 | 62.8 | 120.7 | 78.7 |
| Germany | 3.3 | 5.1 | 7.5 | 33.9 | 61.1 | 169.6 |
| France | n.a. | n.a. | 8.42 | 21.4 | 53.6 | 196.0 |
| Italy | n.a. | 0.9 | 1.1 | 4.0 | 26.6 | 274.6 |
| United Kingdom | n.a. | n.a. | n.a. | 366.1 | 689.0 | 1,015.83 |
| Canada | 5.7 | 3.3 | 9.6 | 26.7 | 64.1 | 152.7 |
Note: n.a. = not available. 1 Gross purchases and sales of securities between residents and nonresidents. 2 1982. 3 1991. The series has since been discontinued. Source: National balance-of-payments statistics. As will be discussed in greater detail, cross-border settlements can be effected through multiple channels and usually involve multiple intermediaries. In part for this reason, data collected by local CSDs generally do not permit cross-border settlements to be isolated. By far the best data on cross-border settlements are collected by the ICSDs (Cedel and Euroclear). However, these data are far from comprehensive. Because they have not been able to establish the requisite links to the local settlement systems, the ICSDs do not currently settle trades in US or Japanese government bonds or UK gilts, all of which are heavily traded and settled cross-border. Furthermore, although the ICSDs settle trades in a wide range of internationally traded equities, their share of total cross-border settlements of such instruments is negligible. It should also be noted that the ICSD data include both settlements of outright purchases and sales of securities and settlements of collateral transactions (repos and securities loans). Although precise figures are not available, the ICSDs estimate that repos are the fastest growing component of settlement activity and have already reached around 30% of total activity. Annex 2 describes repos and other collateral transactions and presents available information on the relative significance of various types of collateral transactions in the G10 countries. Even with these limitations, the ICSD data are quite informative. They clearly document that cross-border settlements have been growing very rapidly and have reached levels that are quite large relative to total settlement volumes in most national settlement systems. Chart 1 shows that average daily settlement volumes at the ICSDs reached the equivalent of US$ 112 billion in 1994 and have been growing over the last few years at a compound annual rate of around 50%. The left-hand panel shows that Euroclear has accounted for about three-quarters of the total during the last two years, compared with about a two-thirds share in earlier years. More dramatic changes have occurred in the currency composition of settlement volumes. As shown in the right-hand panel, in 1987 settlements of US dollar-denominated securities accounted for about 55% of the total volume, but that share has been shrinking steadily and fell to less than 20% in 1993 and 1994.
Chart 1
International central securities depositories
(average daily settlement volumes, in billions of US dollars*)
* Includes internal deliveries within each system, deliveries received by each system from the other (via the bridge) and deliveries to and from each system via links to domestic systems.
Sources: Cedel and Euroclear. Table 2 provides additional information on the types of instruments settled by the ICSDs. The ICSDs were created about twenty-five years ago to settle trades in Euro-bonds. However, as can be seen, settlements of Euro-bonds now account for a small fraction of their overall settlement activity. The bulk of the activity is now in other fixed-income bonds, predominantly European government securities. Deutsche Mark issues are by far the largest single category, but significant volumes of securities denominated in Dutch guilders, French francs, Danish kroner and Spanish pesetas are also settled. US dollar-denominated securities are the second most actively settled, but the instruments involved are largely Euro-bonds and money market instruments. As noted earlier, the ICSDs currently do not have links to the local CSD (Fedwire) that would permit them to settle trades in US government securities. Table 2 ICSD turnover by currency1 (daily averages in US$ bn equivalent, 1994)
| Currency | All | Fixed-income bonds | Other |
| instruments | Euro-bonds | Other | instruments2 |
| 1. Deutsche Mark | 44.8 | 0.6 | 43.0 | 1.2 |
| 2. US dollar | 22.1 | 8.0 | 0.8 | 13.3 |
| 3. French franc | 7.8 | 1.3 | 4.8 | 1.7 |
| 4. Dutch guilder | 7.0 | 0.2 | 6.7 | 0.1 |
| 5. ECU | 7.0 | 1.9 | 2.6 | 2.5 |
| 6. Danish krone | 5.1 | . 1 | 5.1 | . 1 |
| 7. Spanish peseta | 3.1 | . 1 | 2.8 | 0.2 |
| 8. Japanese yen | 4.2 | 2.6 | 0.2 | 1.4 |
| 9. Pound sterling | 2.3 | 1.4 | . 1 | 0.9 |
| 10. Italian lira | 2.2 | 0.7 | 1.0 | 0.5 |
| Memorandum item: All currencies | 111.8 | 18.1 | 70.5 | 23.2 |
1 Less than $50 million. 2 Lines 1 to 10 exclude $1.3 billion in turnover in warrants and equities at Euroclear, because the currency breakdown is unavailable. 3 Short and medium-term notes, floating rate notes, CDs and convertible bonds. Sources: Cedel and Euroclear. Table 3 shows that the volumes settled by the ICSDs are exceeded by only a few national settlement systems, and volumes at most of the national systems have not been growing as rapidly as at the ICSDs. Furthermore, in at least two cases (Germany and the Netherlands) the volumes of government bonds and other domestic securities settled in the ICSDs are comparable to the total volumes of securities settled in the national settlement systems (DKV and SCC respectively). Taken together, these data underscore the importance of cross-border trading and cross-border settlements in certain government bond markets and the importance of the ICSDs as a channel for effecting such cross-border settlements.
|