Revision of the Weights for Calculation of the
Nationalbank's Effective Krone Rate Index

Introduction

Danmarks Nationalbank regularly publishes an index of the development in the strength of the krone, called the effective krone-rate index. Changes in trading patterns make it necessary from time to time to recalculate the weights at which the individual exchange rates are weighted together when the index is calculated. This took place most recently in 1992 using a set of weights based on trade in manufactured goods in 1989¹⁾. As from June 2, 1998 the Nationalbank will publish an effective krone-rate index calculated using weights based on trade in manufactured goods in 1995.

The new index is linked to the previous index as of May 30, 1997, i.e. the krone-rate index has been recalculated for the last year. There are no changes in the published index prior to the date of linkage.

The new set of weights includes four new countries, South Korea, Poland, the Czech Republic and Hungary, bringing the number of countries up to 25. No other changes have been made to the basic method of calculating the index.

As a consequence of the revision, at the beginning of May 1998 the krone-rate index using the new weights was approximately 3/4 per cent higher than the previously published index, primarily due to the depreciation of the South Korean currency, the won, against the previous year.

What is an effective exchange rate?

In the currency market a large number of currencies are traded directly or indirectly against the Danish krone. Often the krone will strengthen against certain currencies while simultaneously weakening against others. By weighting the exchange rates together the effective krone rate provides an overall measure of the development in the international value of the krone as a consequence of fluctuations in the individual bilateral exchange rates. This requires a set of weights which shows the relative importance of each currency. How these weights are to be calculated is not determined beforehand since it depends on the problems to be studied.

The basis for the Nationalbank's effective krone rate is the international price competitiveness of Danish industry. It is assumed that the costs of Danish and foreign manufacturers, including wage costs, are to a great extent determined in national currencies. Since domestic manufacturing enterprises sell their products in competition with foreign enterprises the exchange rate is of central importance to the relative competitiveness of the Danish enterprise.

However, fluctuations in the effective krone rate can only be taken to indicate an equivalent change in Danish industry's competitiveness if costs in Denmark rise in step with those of our competitors. Otherwise adjustment must be made for the development in the relative wage costs in the same currency. This is the role of the real effective exchange rates. But it should also be emphasized that an index of effective exchange rates by no means reflects all aspects of competitiveness as a concept.

Effective exchange rates are overall measures which do not express the change in the competitive situation of an individual company. Relevant bilateral exchange rates are often more suited to this purpose.

In the following the new set of weights is first presented and then the principles behind the Nationalbank's choice of methodology are outlined. The more technical aspects of the calculation can be found in the Appendix.

The new krone-rate weights

In order to arrive at the final weights, as before a set of export weights (double-weighted export weights) is weighted together with a set of import weights (bilateral import weights), with greatest weighting given to the export side. The double weighting of the export weights reflects that competition with a given country not only concerns that country's domestic market, but also international markets. This is described further in the Appendix.

The set of weights based on trade in manufactured goods in 1995 is presented in Table 1. The general picture is of slightly more pronounced changes in the weights this time than on the last revision of the weights. During the intervening period all EU trade statistics have been re-organized in connection with the introduction of the EU single market in 1993. This has given rise to data quality problems.

The inclusion of additional countries will naturally entail an underlying trend for the weights for the individual countries to be reduced from the 89 set of weights. For some countries, however, the weights have decreased considerably more than can be attributed solely to the inclusion of the additional countries. This applies in particular to Sweden, the UK and the USA, but this is merely a continuation of a historical trend, cf. the Table. With regard to Sweden the entire decrease lies by and large on the export side, which quite certainly also reflects the economic downturn in Sweden at the beginning of the 1990s, and thereby low demand for foreign goods. The weighting of the Japanese yen has also diminished. This is due exclusively to a decrease in the import share, while the export weight is unchanged from the last revision of the weights.

The relatively large decrease in the weights for the four aforementioned countries is counteracted by a comparatively strong increase in the weights for Germany and the Netherlands. In the case of Germany this e.g. reflects the consequences of the German reunification. The increase in exports to

Germany has an extra strong impact on the overall krone-rate weighting, since in the 95 set of weights the export side is weighted more strongly on calculation of the final weighting than was the case in the 89 set of weights (60.3 per cent against 53.1 per cent). This reflects that exports are becoming increasingly more important to Danish industry.

In the new set of weights the overall weighting for the countries participating in the euro is just below 60 per cent. With the addition of countries which can be presumed to seek to maintain a generally stable exchange rate against the euro bloc the percentage reaches more than 70. This entails considerable cushioning of the future fluctuations in the effective krone rate. The possible inclusion of the four new countries, whose total weighting is 3.6 per cent, is a counteracting factor. In the event of very strong exchange-rate fluctuations this weighting is sufficiently high to have a marked impact on the overall index.

Comparing the development in the previous and new krone-rate indices the latter is approximately ¾ per cent higher at the beginning of May 1998, cf. Chart 1. This reflects among other things the depreciation of the South Korean won during the period, cf. also the breakdown in Table 2. Even though the South Korean currency is included with only a moderate weighting its very strong depreciation still has a marked impact on the overall index.

The inclusion of countries with strongly depreciating currencies in an effective exchange-rate index is not without its drawbacks, since usually such countries also have a high price- and wage-increase rate. Even in the short term this can make the value of the nominal effective exchange rate a problematic indicator of the development in competitiveness since the index will tend systematically to show relatively poorer development in competitiveness than is actually the case.

In order to make adjustment for the deviating development in the relative prices and wages in the same currency real effective krone rates with the four new countries have been calculated, including the two real krone-rate indices which are published regularly by the Nationalbank with respectively consumer prices and hourly wages in manufacturing as deflator²⁾.

The real effective krone rate with prices as deflator shows the same overall development as the nominal index, cf. Chart 2, but the rate of price increases in South Korea has doubled during the last six months and the price-increase rate in the three eastern-European countries included in the krone-rate index is also rather high (10-20 per cent).

It must be concluded overall that the revision of the krone-rate index does not significantly change the picture the index gives of the development in the competitiveness of Danish industry in recent years, not even with further retrocative projection of the date of linkage.

The basis for calculation of the Nationalbank's effective krone rate

When effective exchange rates are calculated a number of choices must be made - both operational and concerning issues of principle. The most

important are the delineation of which countries to include in the index, the choice of base year, and the goods composition. Moreover, the date of linkage with the previous index must be determined.

Since the effective krone-rate index is designed to measure the development in industry's competitiveness, the weights in the index should reflect the breakdown by country of trade in manufactured goods. In practice, the trading pattern does not change significantly from year to year, but considerable shifts may take place over a period, calling for revision of the weights. Ultimately the weights could be updated every year, although this might make interpreting the index more difficult, since a given change might now be due to both the exchange rate and/or changes in the weights. The choice of 1995 as the base year for the Nationalbank's index was made exclusively on pragmatic grounds, since it is the last year covered by the OECD trade statistics at the required level of detail.

Compared to previously, in recent years the importance of regions outside the group of industrialized countries to trade in manufactured goods has also grown steadily³⁾. However, as before, the delineation of countries has been limited to the OECD countries. Since the last revision of the weights the OECD has been enlarged to include South Korea, Poland, the Czech Republic and Hungary, and these countries are thus included in the calculation of weights. Mexico and Turkey are also OECD members, but are excluded since they are both characterized by strongly depreciating currencies and high inflation. Countries outside the OECD group are not included on operational grounds, i.e. problems in obtaining data with a sufficient level of detail and of a sufficiently high quality, and also for reasons of principle. In this regard two factors are involved. If countries with high inflation and continuously depreciating currencies are included a nominal exchange-rate index will be distorted, as previously described, and there is more focus on the real exchange-rate index, which in practice is more difficult to calculate and will only be available after a certain delay. Secondly, it is problematic to include countries with an industrial structure which differs significantly from Denmark's, i.e. actual low-wage countries. This aspect is discussed in more detail in the Appendix. In reality, there is no clear distinction between which countries to include and which not.

As stated, the basis for the calculation of weights is trade in manufactured goods. Denmark's substantial agricultural exports are thus not included. The reason is not a lack of statistics, but rather the extensive subsidy schemes and quantitative restrictions in this area which make the impact of exchange-rate fluctuations more difficult to isolate, since prices are not fully market-determined. The growing international trade of services is not included either. However, this is due to a lack of data. Manufactured exports constitute approximately 70 per cent of total Danish goods exports and 50 per cent of total exports. The proportions on the import side are a little higher.

In relation to some of Denmark's trading partners it is vital whether all trade in goods is included, or only trade in manufactured goods. This applies for example to trade with a country such as Russia. Measured in terms of total trade of goods this country is one of Denmark's largest trading partners in eastern Europe and the previous Soviet area. However, if only trade in manufactured goods is included, exports to Russia constitute only 1/3 of manufactured exports to e.g. Poland. The difference is even greater on the import side.

The inclusion of South Korea in the set of weights is the basis for the re-calculation of the index for the most recent year, in contrast to the procedure last time the weights were changed. In autumn 1997 the Korean currency depreciated strongly in line with a number of other Southeast-Asian currencies. It is the assessment that in South Korea's case at any rate this depreciation far exceeded what could be attributed to the real economic situation. Inclusion of the won only after the depreciation took place will make the krone-rate index biased when the currency begins to re-appreciate. If the weaker level of the won proves to be longstanding this will not present any problem in relation to the krone-rate index. However, in recent months the won has already recovered some of its former strength.

The single currency, the euro, will be introduced by a number of European countries as of January 1, 1999. This raises the question of the weighting given to the euro area in the krone-rate index. One possibility is to add together the new weights for the participating countries, another to calculate a new set of weights covering the euro area as one. This would eliminate the inter-regional trade flows in the euro area. The two methods do not give quite the same set of weights, which is a small disadvantage of the double-weighting methodology. In the first instance the Nationalbank intends to use the first-mentioned procedure in view of the wish to continue to use the price and wage statistics from the individual countries in the euro area to calculate real effective krone rates.

"to be continued"

Footnotes