After the lows following September 11, the Nasdaq had surged by about 30 percent at year-end, while the more broadly based Standard and Poor 500 index was up by half as much (see Diagram 46). The tendency on stock markets in the euro zone, Sweden, and Japan was similar. The eurozone Dow Jones, EUROSTOXX, rose by some 15 percent, the Swedish SAX by more than 20 percent, and the Japanese Nikkei by almost 10 percent.
Between January 1 and early March this year, however, both the Standard and Poor 500 and the Nasdaq have receded by some 3 percent and 10 percent, respectively, even though the US economy now appears to have reached its low point. The decline may be due in part to uncertainty about principles of business valuation after the Enron bankruptcy. However, the profit potential of US businesses is considered high, to judge from the P/E ratios for the Standard and Poor 500 index. By historical comparison, these ratios remain high, despite the weak tendency so far this year.
Economic growth also seems to be picking up in the euro zone. Nevertheless, the euro-zone Dow Jones, EUROS-TOXX, showed a slight declining tendency at the outset of the year. The decrease is partly attributable to the ICT sector and to continued uncertainty about the profitability of 3G investments in telephony. Moreover, European countries have probably also been affected by the increased uncertainty about principles of business valuation after the Enron bankruptcy.
The development of the Swedish SAX index so far this year has largely resembled that of the broader-based US indices. From January 1 to early March this year, the SAX declined by about 4 percent. The international economic slowdown has particularly affected ICT companies. These previously accounted for a large share of the SAX, one reason being the high valuation of Ericsson. With the pattern of stock prices over the past year, however, the share of ICT-related companies in the index has decreased in favour of the more "traditional" manufacturing companies. Even after declining in 2000 and 2001, however, the SAX index was still nearly 20 percent higher in early March this year than at the outset of 1999.
Tighter Monetary Policy Next Year
From early November last year, the tendency in long-term bond rates has been rising after a declining trend since last summer. Part of the reason for the rise in bond rates is probably that there is now less uncertainty about what would happen to the world economy after September 11. Even before the terrorist attacks, weakness in equities was putting demand pressure on the bond market. When the stock market appeared to be recovering, bond rates turned upward. Swedish bond rates have closely followed the tendency in the euro zone since the start of the year. The differential between the Swedish 10-year bond rate and the corresponding rate in the euro zone therefore remains at about 0.5 percentage points (see Diagram 47).
In the USA, the differential between the return on high-risk corporate bonds and the return on government securities has been diminishing somewhat for a while. However, since flaws and elements of uncertainty in the principles for business valuation have been exposed in connection with the Enron bankruptcy, the differentials have begun to widen again. Despite the small corporate-bond market in Sweden, Swedish companies may also have been affected; for example, the differential between the interest rate on an Ericsson corporate bond and that of a 10-year government bond increased in February.
On the basis of the yield curve, the repo rate is expected to be about 5 percent at the end of 2002 and about 5.5 percent at the end of 2003 (see Diagram 48). By this measure, expectations of Swedish monetary policy and the monetary policy in the euro zone differ for the years covered by the forecast. It is estimated that the Swedish official interest rate will exceed the rate in the euro zone by an average of almost 1 percentage point. In the longer run, however, the expected differential in the official interest rate is virtually zero (see Diagram 49). However, since the yield curve incorporates various risk premiums, repo-rate expectations based on the yield curve are often higher that the expectations presented in various surveys. The expectations for the repo rate in the latest Prospera survey, dated February 20 of this year, are also below the estimates based on the yield curve. According to the survey, the repo rate should remain unchanged for the next three months and should be 4.5 percent in two yea rs.
The same survey also showed that money-market expectations of the long-term inflation rate were unchanged since the previous measurement, but that they still are marginally higher than the 2-percent inflation target of the Riksbank. As in the preceding survey, employee organizations and purchasing managers in retail trade and manufacturing indicated expectations of a high inflation rate in five years.
Another indicator of the inflation rate expected in the long term is the differential between the yield on a nominal-rate bond and the corresponding real-rate instrument. This differential reflects the average expected inflation rate until the maturity of the bonds, and despite the high inflation rate since last April, it is still consistent with the target inflation rate of the Riksbank.
Since winter last year, monetary conditions have become somewhat less expansionary. The exchange rate has appreciated more than expected, and real interest rates appear to be somewhat higher, particularly shorter-term rates. The real-yield curve has shifted upward since winter last year. The slope of the curve has also changed, since the short end of the curve has increased more than the long end (see Diagram 50).
Actual inflation was surprisingly high this winter. The rise in domestically generated inflation, as well as unexpectedly large increases in wages and salaries, suggest that resource utilisation in 2001 was somewhat higher than previously assumed.
At present, resource utilisation is decreasing, dampening inflationary pressure this year and next year. UND1X inflation is therefore expected to subside and to be consistent with the Central Bank's target in the next one to two years. The repo rate can therefore be maintained at 3.75 percent for a while longer, in the NIER's judgment. Next year, as the economy improves, economic growth is expected to exceed the rate that is sustainable in the long term. Inflationary pressure will gradually build up to the point where it might jeopardize the target rate in 2004--2005. Since monetary policy has a delayed effect, it should be appropriate to raise the repo rate in 2003 (see Diagrams 51 and 52).
Long-term interest rates will likely show a marginal increase in 2002-2003, given the forecast of only a modest rise in short-term rates. In the forecast, however, allowance has been made for the substantial differential between Swedish tong-term rates and those in the EMU, and for the likelihood that this differential will continue for a portion of the forecast period.
Swedish Krona Increasingly Strong
Since reaching record-low levels after the terrorist attacks in the USA, the Swedish krona has successively strengthened. In terms of the TCW index, the krona has appreciated by about 5 percent since September (see Diagram 53). Part of the reason may be that pressure on the krona from temporary factors has been steadily diminishing. The previous massive underlying net outflows of capital subsided toward the end of 2001 (see Diagram 54), thus helping to relieve the pressure on the krona. (3) Probably the continued strengthening early in 2002 is also due in part to growing expectations that Sweden will join the European Monetary Union.
The development of the krona will probably be affected both by a decision on the timing of a referendum on entering the EMU, and by the outcome of such a referendum. If Sweden votes "yes" to the single currency, the krona will probably be tied to the ERM2 shortly thereafter. Even before this, the krona will probably adjust to the level regarded by the market as the future conversion rate against the euro.
In accordance with the assumption of no change in economic policy, the exchange-rate forecast does not consider the consequences that a referendum and a possible linking of the krona to the ERM2 may entail for the development of the krona.
In addition to effects of a temporary nature, the exchange-rate forecast takes account of the real effective (TCW-weighted) exchange rate and its underlying fundamental determinants. The weakening of the krona in recent years can be explained to some extent by changes in these fundamental factors. Above all, deteriorating terms of trade in relation to Sweden's principal trading partners have contributed to a weakening of the krona in recent years. However, even in light of these circumstances, and despite its appreciation in recent months, the krona is still regarded as undervalued in real effective terms. This is particularly true in relation to TCW-currencies other than the euro (see Diagram 55). The analysis below describes the NIER's calculation of the real effective equilibrium exchange rate for the krona.
A number of factors -- both temporary and more fundamental -- are expected to contribute to an appreciation of the krona in the future. Among other things, the improvement in recent months in the underlying flows of capital is expected to continue. Moreover, the dollar is likely to weaken against the euro during the forecast period, thus tending to strengthen the krona in terms of the TCW index. The fact that the present surplus on current account is being used to decrease Sweden's foreign debt is also an indication that the krona will strengthen over time.
It is therefore believed that the krona will adjust in the direction of its underlying equilibrium exchange rate. In terms of the TCW index, the krona should appreciate to about 128 by the end of 2003. The krona should strengthen slightly against the euro during the forecast period, but much more against the dollar. It is estimated that at the end of 2003 a euro will cost SEK 8.90, while a dollar will cost somewhat less than SEK 9.40.
Table 6 Interest Rates and Exchange Rates 1999 2000 2001 2002 2003 Official interest rates (1) Repo rate, Central Bank of Sweden 3.25 4.0 3.75 3.75 4.25 ECB refi rate 3.0 4.75 3.25 3.25 4.0 Federal funds target rate 5.5 6.5 1.75 3.25 4.75 Long-term interest rates (2) Sweden 5.6 4.9 5.2 5.5 5.7 Euro zone (4) 5.2 4.9 4.8 5.0 5.5 USA 6.3 5.2 5.0 5.3 5.7 Exchange rates (3) SEK/USD 8.48 9.66 10.56 9.99 9.37 SEK/EURO 8.59 8.66 9.44 9.03 8.90 TCW-index 124.4 128.0 138.6 132.0 128.7 (1)At the end of each year (2)Rate on 10-year government bonds, at the end of each year. (3)In December of each year. (4)Yield on a benchmark bond. Sources: NIER and the Riksbank.
(3.) The underlying flows of capital are defined here as the total, but with the opposite sign, of the current account, direct investment, and portfolio investment in shares. Thus, a positive figure corresponds to an underlying net outflow of capital. An underlying net outflow must be offset by a net inflow of capital via trade in interest-bearing securities.
RELATED ARTICLE: The Development of the Repo Rate in the NIER's Forecasts
The development of the repo rate in the forecast is not necessarily the one considered most probable in light of the measures taken by the Riksbank. Discrepanices between forecasts and the monetary policy actually pursued arise in cases where the Central Bank and the NIER differ in their analysis concerning inflation.
The NIER's assessment of the repo rate is based on the principles for monetary policy announced by the newly appointed directors of the Central Bank in the spring of 1999. Therefore, the path of the repo rate shown in the forecast is based primarily on the NIER's assessment of UND1X inflation in one to two years.
Calculations of the Real Effective Equilibrium Exchange Rate (4)
The NIER's forecasts for the future development of the Swedish krona are based on the so-called real effective exchange rate - that is, the relationship between Swedish prices and a weighted average of prices in other countries in terms of a common currency. The forecasts are based among other things on the relationship between the real effective exchange rate and its equilibrium value, and on the normal rate of adjustment toward this equilibrium value. The equilibrium value is calculated on the basis of the long-term relationship between the real effective exchange rate and its fundamental determinants. (5) Thus, the equilibrium value is not necessarily constant, but may vary over time if the fundamental variables change. The short-term equilibrium rate is calculated from the values of the fundamental variables currently prevailing. The long-term equilibrium rate is based on the long-term equilibrium values of the fundamental variables, which can be calculated with the model.
The real effective exchange rate is defined here as the ratio between Swedish consumer prices and consumer prices in 14 OECD countries, weighted according to their TCW weights, all consumer prices being expressed in a common currency (see Diagram 56). (6) An increasing index thus stands for a weakening of the real effective exchange rate. Three fundamental variables are included in the model:
* Terms-of-trade, measured as the ratio between the unit value of Swedish exports and the unit value of ratio for other countries (TCW-weighted, 14 countries). Between the end of 1995 and the end of 2000, Sweden's relative terms of trade deteriorated by approximately 10 percent (see Diagram 57). One reason for this development was the weak tendency of prices in the telecom sector during the period. Declining terms-of-trade tend to lower the real effective equilibrium exchange rate in the model. The weak tendency since the mid-1990s should therefore have contributed to a depreciation of the real equilibrium exchange rate during the period.
* The relative price of internationally traded goods and services (so-called tradables), and nontraded goods and services (so-called non-tradable), measured as the ratio between the Wholesale Price Index and the Consumer Price Index, in relation to the corresponding relative price in other countries and the Consumer Price Index, in relation to the (TCW-weighted, 14 countries). As intuitively described, this variable functions in the model as follows: If the development of productivity in the tradables sector in Sweden is better than in other countries, wages throughout the Swedish economy will tend to rise more rapidly. If the development of productivity in the non-tradables sector is the same in Sweden as in other countries, the relative price of Swedish tradables to non-tradables in relation to the corresponding relative price in other countries will decrease. Such a tendency will contribute to higher inflation in Sweden than in other countries, thus leading to an appreciation of the real exchange rate. fro m the mid-1990s to 2000, this relative price gradually increased by almost 1.0 percent in total (see Diagram 58). This development, which. is believed to have contributed to a depreciation of the real equilibrium exchange rate as measured with consumer prices, is probably explainable in part by deregulation in the nontradales. The deregulation has helped to increase productivity and to squeeze profit margins, thus limiting price increases in this sector. The dampening effect on consumer-price inflation has been particularly noticeable, and has thus contributed to a rising relative price between tradables and non-tradables in Sweden.
* Foreign debt in proportion to GDP. A lower proportion of foreign debt to GDP should help to strengthen the real equilibrium exchange rate. As is shown in Diagram 59, foreign debt increased substantially in relation to GDP in the early 1990s. Thereafter, the proportion of foreign debt has decreased somewhat. Since the change between the mid-1990s and 2000 is modest, it is probably of little significance for the development of the equilibrium exchange, rate. In time, however, this factor may become more important as surpluses on current account gradually reduce Swedish foreign indebtedness.
In addition to these fundamental variables, the real interest rate differential in relation to other countries is included (see Diagram 60) in the model. Since this differential is intended to take account of cyclical effects in the real effective exchange rate, it is not considered in the long-term relationship.
The estimates have been based on quarterly data from the period 1982-2000. A number of different specifications of the model have been analysed to test its robustness. For example, so-called dummy variables have been included to capture the effects of devaluations and of the transition to floating exchange rate at the end of 1992. On the whole, the various specifications of the model yield very similar results, thus suggesting that these are quite robust.
The fundamental variables are all statistically significant, with the expected sign in the long-term relation ship. Estimates indicate that a one-percent increase - i.e. an improvement - in the terms of trade in relation to other countries will strengthen the real effective equilibrium exchange rate by approximately 3/.4 percent, all else being equal. The corresponding effect of a one-percent inorease in the ratio between the prices of trade ables to non-tradables has, as expected, the opposite sign and is roughly half as great.
Diagram 61 presents the short-term and long-term real effective equilibrium exchange rates from a representative specification of the model, together with the real effective exchange rate. It is apparent from the diagram that the long-term real effective equilibrium ex-change rate very closely follows its short-term counterpart. The breaks in the time series are due to the inclusion of dummy variables to eliminate the effects of the policy shift to a floating exchange rate - with exceptionally rapid weakening of the krona that ensued - in November 1992. The results suggest that the real effective exchange rate was roughly at parity with its equilibrium value until the end of the 1980s. With the rapid strengthening of the real effective exchange rate in the early 1990s, however, the krona grew considerably stronger than was justified by its fundamental determining factors. The weakening of the krona when the policy of a fixed exchange rate was aabandoned brought the real effective exchange rate back near its e quilibrium value. As the krona continued to weakening until mid-1995, it became undervalued, a situation that was corrected by the rapid appreciation of the krona to-ward the end of 1995.
From 1996 on, both the real effective exchange rate and its equilibrium value gradually weakened. The decline in the equilibrium exchange rate was due primarily to deteriorating terms of trade, but also to the rising relative price of tradables to non-tradables during the period.
In the final quarter of 2000, the real effective ex change rate was about 4 percent weaker than the shortterm equilibrium exchange rate calculated from the model, and somewhat more so in relation to the long term equilibrium exchange rate. The undervaluation may approximately be interpreted in terms of the socalled TCW-index, which is an index of the nominal effective exchange rate. In the fourth quarter of 2000, the TCW index was slightly more than 128. The undervaluation of some 4 percent implies that the short-term 'nominal effective equilibrium exchange rate, as expressed in a TCW index, was around 123 during the same quarter.
It should be noted, however, that calculations of the equilibrium exchange rate are generally associated with a considerable degree of uncertainty. It may therefore be interesting to compare the above results with other calculations that have been made previously. The above results indicate that the krona was at equilibrium in the fourth quarter of 1995 (see Diagram 61). At this time, the TOW index was roughly equal to 119, which then corresponded to the nominal equilibrium exchange rate. This is consistent with the findings presented by Alexius and Lindberg (1996) (7). In this study, which was based on quarterly data through the fourth quarter of 1995, it was concluded that the nominal equilibrium exchange rate of the krona at that time was located in the interval between 110 and 119 in terms of the TOW index.
(4.) For a more thorough description of the calculations of the real effective equilibrium exchange rate of the krona, see Kristian Nilsson, Do Fundamentals Explain the Behaviour of the Real Effective Exchange Rate?, Konjunkturinstitutet, Working Paper No. 78, March 2002.
(5.) The Swedish Economy -- March 2000 provides a description of the initial verson of the model for the equilibrium exchange rate. Here the new, re-estimated version of the model is presented.
(6.) The 14 OECD countries are Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, the Netherlands, Norway, Spain, the UK and the USA.
(7.) Alexius, A. and H. Lindberg, "Kronans reala jamviktskurs (The Real Equilibrium Exchange Rate of the Krona)", Penning och valutapolitik (Monetary and Foreign Exchange Policy), 1996:1, pp. 19-34, Central Bank of Sweden.
|Printer friendly Cite/link Email Feedback|
|Publication:||The Swedish Economy|
|Article Type:||Statistical Data Included|
|Date:||Mar 1, 2002|
|Previous Article:||International economy.|
|Next Article:||Demand and output in the Swedish economy.|