Separate minutes from the Executive Board meeting on 8 November 2001
Urban Bäckström, Chairman
§ 1. Assessment of the interventions on the foreign exchange market
It was noted that Anders Borg and Javiera Ragnartz would prepare draft minutes of § 1 on the agenda.
It was also noted that the Executive Board had decided at its meeting on 5 July that an assessment should be made of the interventions on the foreign exchange market that were decided at the meeting on 14 June.
The discussion began with a review of the academic research regarding interventions (see the annex to the minutes). This was followed by a report of the measures taken and of the development of the exchange rate, as well as other relevant data that had been released after the interventions. Finally, the Executive Board discussed the effects of the measures taken.
1.1 Account of the Riksbank's interventions in June 2001.
On Thursday, 14 June, the Executive Board decided to intervene on the foreign exchange market. This decision was made because the krona had depreciated considerably, without a similar depreciation taking place for other, comparable currencies, and the weak exchange rate was assessed as the largest individual risk that the inflation target might be exceeded (see press release no. 41 and the separate minutes of the Executive Board meeting on 14 June 2001, § 1 for a more detailed explanation). At this meeting the right to make decisions on interventions was delegated to Riksbank Governor Urban Bäckström in consultation with First Deputy Governor Lars Heikensten.
Figure 1. TCW and USD/EUR.
That same afternoon a speech was published that Urban Bäckström held in Stockholm, later in the evening. It was indicated in this speech that the recent weak development of the krona comprised an upside risk for inflation and that interventions could be necessary. The krona appreciated markedly after publication of the speech, while short interest rates rose.
Figure 2. SEK/USD and SEK/EUR on 14 June 2001
On Friday, 15 June, Lars Heikensten made a similar statement in connection with a speech in Malmö, and this also affected the krona. The first intervention was made at 11.10 a.m. when the Riksbank sold EUR 150 million for SEK. A press release was issued in connection with the first intervention (see press release no. 41). The krona appreciated markedly against both the dollar and the euro. A further intervention was made at 11.40 a.m., when the Riksbank sold USD 190 million for SEK, which cause the krona to appreciate further against both the euro and the dollar. Late in the afternoon, at 3.40 p.m., the Riksbank sold EUR 150 million for SEK, after which the krona showed an initial appreciation, although it later depreciated slightly once again.
Figure 3. SEK/USD and SEK/EUR on 15 June 2001.
Looking at the time from the afternoon of the 14th until the market closed on the 15th, the krona appreciated a total of 11 öre against the euro and 27 öre against the dollar, to 9.126 and 10.59 respectively. At the same time, the short interest rates rose, for instance, the 3-month rate rose by approximately 15 basis points. Earlier during the week the dollar had depreciated against both the euro and the krona as a result of rumours of interventions by the ECB and in connection with the US National Association of Manufacturers publishing an assessment stating that the dollar was overvalued. The krona tends to covary with the euro against the dollar, i.e. when the euro strengthens against the dollar, the krona also appreciates against the US currency. In the situation that prevailed, the Riksbank's intervention thus reinforced a positive trend.
Monday, 18 June, began with a continued appreciation of the krona against the euro and a further rise in the short interest rates. It is probable that an interview with Urban Bäckström on a Swedish radio news programme on the Saturday contributed to this. After the interview, the idea of an imminent interest rate hike was reinforced. This extensive communication was aimed at reinforcing the signalling effect and at avoiding the creation of uncertainty concerning the framework for general monetary policy.
The Riksbank intervened on Monday afternoon at 2.55 p.m. by selling EUR 170 million and at 3.15 p.m. by selling USD 240 million for SEK. The krona once again appreciated against both the euro and the dollar, rising to 9.05 and 10.50 respectively. Moreover, pricing on the options market indicated that the probability of an appreciation of the krona was now estimated to be greater than the probability of a depreciation. On Monday afternoon, the krona had appreciated by just over 2 per cent on the TCW index since the start of the interventions. The euro had remained largely unchanged against the dollar during the course of the day.
Figure 4. SEK/USD and SEK/EUR on 18 June 2001.
During the following days the krona depreciated against both the dollar and the euro. The fact that the euro depreciated against the dollar probably contributed to this. There could also have been some profit taking after the large appreciation.
On Monday, 25 June, at 3.15 p.m. the Riksbank sold EUR 100 million and at 3.48 p.m. it sold a further EUR 200 million for SEK. These interventions did not affect the krona to the same extent as the previous occasions. Over the following week, up until the Executive Board's monetary policy meeting on 5 July, the krona continued to weaken. The short interest rates rose marginally.
Figure 5. SEK/USD and SEK/EUR on 25 June 2001.
On Thursday, 5 July, the Executive Board decided to raise the repo rate by 0.25 percentage points (see press release no. 47 for the motivation behind this). According to the pricing on the money market, this decision had been expected. Following the decision, the krona appreciated, above all against the dollar, but also a little against the euro. The strengthening against the dollar was primarily the result of a general weakening of the dollar.
During the course of the summer there were increasing signs of a more marked economic slowdown, both internationally and in Sweden. The Executive Board decided at its monetary policy meeting in August to retain the repo rate unchanged. A more subdued growth in demand and lower international export prices indicated that inflation would be in line with the Riksbank's target (for a more detailed explanation, see press release no. 50).
In mid-August, when it was possible to detect signs of a more marked slowdown in economic activity, the stock markets fell and the krona began to depreciate once again, after a relatively stable period (Figure 6). The depreciation of the krona accelerated considerably after the terrorist attack on 11 September. However, the pattern now was completely different from that during the spring. It was more similar to the developments in 1998, when the krona, like certain other currencies, weakened substantially in connection with unease on the international financial markets. The short interest rates began to fall, primarily because of portfolio adjustments from equities to bonds, but also because monetary policy expectations changed in line with the altered picture of economic activity.
Figure 6. TCW and OMX.
To summarise, the Riksbank sold euro and US dollars to a total value of SEK 12 billion.
The monetary policy signalling in connection with the interventions appears to have been understood by the market, as short interest rates rose and pricing on the money market indicated that a repo rate hike was expected as early as the monetary policy meeting in July (Figure 7). Following the repo rate raise in July, the krona began to appreciate, the long-term interest rate differential against Germany declined and inflation expectations, which had previously increased, were dampened slightly (Figure 8). The long-term interest rate differential against Germany, which had declined somewhat over the summer, once again rose and was thus at the end of August remained largely unchanged from the start of the period (Figure 9).
Soon after the interest rate was raised, signs began to appear of a more profound economic slowdown than most analysts had expected. The situation was accentuated after the terrorist attacks in the USA. With considerable unrest on the international financial markets, the krona depreciated, as did certain other currencies. The Riksbank lowered its repo rate by 0.50 percentage points on 17 September (for the motivation behind this, see press release no. 57). Despite the fact that the krona had developed less strongly than expected, it was estimated that the effects on inflation would be counteracted by, for instance, falling international export prices and subdued domestic demand.
Figure 7. Expected repo rate in three months time and the TCW index.
Figure 8. Expected inflation in two years' time according to pricing on the money market.
Figure 9. Long-term interest rate differential against Germany and the TCW index.
1.2 The Executive Board's discussion
The Executive Board's discussion began with an introduction by First Deputy Governor Lars Heikensten. This partly concerned a discussion held in the monetary policy group.
The interventions in June were a feature of the bank's monetary policy. The background to the interventions was a situation with resource utilisation that was still relatively high, despite the slowdown in economic activity, surprisingly rapid price rises during the spring and a severe depreciation of the krona. All in all, there was a risk that this would lead to an inflation rate above the Riksbank's target. There were also some indications that inflation expectations were rising. It was considered important, in order to break this trend, to send a signal that the Riksbank takes its inflation target extremely seriously. Given the krona's important role in the developments of import prices - and thereby inflationary trends - and its close connection to the formation of expectations, the Executive Board considered it appropriate to intervene on the foreign exchange market.
The preparatory work on which the decision was based included the review of academic research attached to this assessment. Previous experiences by the Riksbank and other central banks, as well as academic studies of interventions were all considered when making the decision.
The assessment of the interventions cannot be based solely on information regarding their scope and their direct effects on the exchange rate, but should also take into account the effects on other important variables, such as pricing on the interest rate market and inflation expectations. This is because the interventions formed part of a larger monetary policy context that ultimately aimed to safeguard price stability. It is also important to evaluate whether the purpose of the interventions was communicated clearly, as well as how the interventions were perceived by external analysts. This can be partly achieved by studying various market commentaries in connection with the interventions in June, and also later in the summer.
It is in general difficult to evaluate economic policy. For instance, it is not possible to determine how the exchange rate would have developed if the Riksbank had refrained from intervention on the foreign exchange market. Nor is it possible, in practice, to adjust for all other factors, such as new macroeconomic data, political manoeuvring, other monetary policy decisions made at a later date, etc., which have also changed during the period under assessment. The signalling effect of implementing currency interventions normally also influences expectations of monetary policy, which should be taken into consideration, both when deciding on interventions and when assessing them. One should not, therefore, expect to be able to draw totally unequivocal conclusions.
An assessment is made difficult by the reconsideration of monetary policy that took place in the early autumn, in the light of the weaker economic activity abroad and the lower international export prices. Following the events of 11 September in particular, the money market came to price interest rate reductions and several central banks adjusted their steering interest rates.
The dramatic depreciation of the krona ceased after the interventions in June and the repo rate hike in July. The krona appreciated slightly until mid-August, when new data and developments on the financial markets indicated that the economic slowdown in Sweden and abroad would be more pronounced. Inflation expectations, when measured according to Statistics Sweden's survey of household purchasing plans, also fell slightly. The indications of weaker economic activity probably contributed to the latter.
The reception of the interventions on the financial markets was largely negative. This is despite the fact that the measures had been signalled in speeches and that they were implemented with greater transparency both with regard to communication of motives and the design of the operations than is normal in international practice, or has been for the Riksbank previously. Although the Riksbank motivated the interventions as part of its monetary policy, which was ultimately aimed at safeguarding price stability in line with the inflation target one to two years ahead, some uncertainty nevertheless arose on the financial markets.
The Riksbank's credibility was not negatively affected by the interventions on the foreign exchange market. The concept of credibility should in this context mainly refer to the inflation target and whether the measures taken contributed to achieving this or not. Following the interventions there were no clear signs of rising inflation, higher inflation expectations or increased long-term interest rate differentials in relation to other countries. Nor has the depreciation of the krona accelerated, rather the reverse. This indicates that the credibility of the inflation target has certainly not been affected negatively. At the same time, it can be concluded that many analysts, both market participants and academic research economists, have been critical of the measures taken and that this criticism has remained for longer than it did, for instance, after the interventions in 1998.
One way of reducing the risk of interventions contributing to increased uncertainty regarding the general framework for monetary policy is to further increase the openness and clarity as to the Riksbank's attitude to this instrument. The Riksbank has earlier announced that there is reason to clarify how interventions should be handled in the monetary policy process. In this context there may be good reason to further describe the process for drafting, deciding and communicating interventions on the foreign exchange market, as well as how the interventions are reported in the minutes. The Riksbank's earlier experiences and those of other central banks show that the role of interventions in the framework for monetary policy in an inflation target regime can be complicated. There may be reason to return to this issue, too, in the Riksbank's day-to-day work. There is also reason to aim for a greater understanding of the microstructure on the foreign exchange market. Some work has already been done in this field.
Lars Heikensten summarised the evaluation of the interventions in three conclusions:
1. It is possible that the Riksbank may also intervene on the foreign exchange market in the future. Normally, interventions are made as a part of monetary policy and are aimed at safeguarding the inflation target. However, it is not possible to rule out the possibility that other situations could arise where interventions may be motivated. It should always be possible to provide clear motives for such decisions.
2. Interventions are an instrument that will be used with great caution, as long as the floating exchange rate regime prevails. Normally, monetary policy is implemented through adjustments to the steering interest rates. Situations where interventions are necessary are rarely obvious from a decision-making perspective. It is therefore possible to have differing opinions within the Executive Board when making decisions on interventions.
3. The process for drafting and making decisions on interventions should be as open and clear as possible. This should be the starting point for proposals put to the Executive Board on procedures and communication, etc. regarding interventions.
Lars Heikensten also proposed that the separate minutes of the Executive Board meetings on 14 June 2001, §1 and 5 July 2001 no. 9:3 should be published as this could no longer be considered to comprise a risk of counteracting the measures. As publication of the intervention mandate delegated at the meeting on 14 June could make future interventions more difficult and counteract their purpose, this particular piece of information should not be published. It was proposed that publication of these earlier minutes should take place in connection with the publication of these minutes on 30 November 2001.
Discussion on the interventions
The discussion began with all Board members expressing support for the main elements of Lars Heikensten's report. All of the members agreed with the three conclusions Heikensten drew from the evaluation and the proposal for a decision.
One Board member asserted that it was natural for the Executive Board to decide now to publish the minutes taken during the decision on the interventions on 14 June, as well as those taken of the discussions regarding the interventions held at the meeting on 5 July. This member considered that publication would no longer risk disturbing the aim of the interventions made. Other members expressed agreement with this reasoning.
Another member pointed out that today's discussion could be a brief one, as the differences of opinion expressed at earlier meetings would thereby become public knowledge.
Several members emphasised, however, that there was good reason to refrain from publishing the framework established by the Executive Board when delegating the implementation of the decision on interventions. The motive was that to make this framework public could complicate the implementation of any future interventions.
A couple of other Board members felt that the theoretical and empirical research into interventions should be interpreted with caution. Some studies are based on strong assumptions, which do not always apply. More recent research, using improved methods, provides a more complex picture. One member pointed out that the interventions are a part of a larger economic policy context and thus difficult to evaluate separately. Interventions play a role in a central bank's range of instruments. The member went on to point out that there could be situations, for instance when there is considerable unrest on the market and trade is thin, where interventions are motivated not merely as a direct element of monetary policy aimed at maintaining price stability. In certain situations there could be motivation for interventions during a temporary failure to function of the markets. The other member pointed out that it was not possible to rule out problems caused by herd behaviour on a small foreign exchange market with a concentration in trading, and that such circumstances could motivate interventions by the Riksbank.
One Board member agreed that under certain circumstances, such as considerable unrest on the market and little trading, it could be reasonable for a central bank to intervene on the foreign exchange market. However, these situations seldom arise and this was not the case in June. This member considered that there were more effective means of communicating the direction of monetary policy than implementing interventions. In addition, the member felt that the market reactions indicated that the signals were not sufficiently clear until after the above-mentioned radio interview with Urban Bäckström. Following the interventions, there was a period of lack of clarity regarding monetary policy, according to this Board member. Nor did this member feel that the effects of the interventions were as intended.
One Board member pointed out that even if a decision on interventions is normally part of monetary policy, it is not self-evident that a decision on an intervention should always be followed by an interest rate adjustment in the same direction. Further information could be received that changed the situation.
This Board member shared the view that the reinforced slowdown in economic activity was an important factor behind the more subdued expectations of the rate of price increase during the summer. The member added that the Riksbank should therefore show caution in claiming that it was the interventions and interest rate hike that subdued inflation expectations.
The other members expressed objections to this reasoning. One Board member asserted that after the May Inflation Report there arose a picture among external analysts that the Riksbank was relatively unconcerned by the price rises taking place in the spring. Following publication of the report inflation expectations rose, said this member. Another Board member pointed out that the interest rate hike in combination with the economic slowdown appears to have had a clear effect on inflation expectations.
A third member shared the view that the Riksbank's measures had contributed to subduing inflation expectations and pointed out that the Riksbank cannot refrain from taking necessary measures merely because these are not expected by the markets, or because they could be assumed to arouse criticism from outside.
Another Board member agreed that there had been some concern in Swedish society that the Riksbank did not take the price rises in the spring too seriously. This was not clearly visible in surveys, etc. but could be read in the pricing on certain markets, and heard in many of the various contacts with market participants and companies. Nor can one rule out the possibility that this concern had an effect on the krona at the beginning of June. At the same time, the member pointed out that the Executive Board's decision was based on a total assessment of inflation prospects one to two years' ahead. This assessment balanced together many different factors, such as resource utilisation, the severe depreciation in the krona and inflation expectations.
This Board member also agreed with those who claimed that it would probably have been possible to achieve almost the same effect simply by signalling that an adjustment in the interest rate was imminent. The fact that this would be done on 5 July was otherwise for this Board member one of the bases for making the decision on interventions. The member also felt that interventions as a means of trying to influence the exchange rate should not be over-dramatised. The krona was tangibly undervalued and some restraint via the exchange rate appeared desirable at that time.
The Chairman summarised the discussion and concluded that all of the Board members had declared their support for the three conclusions of the evaluation reported. In addition, there was a proposal for a decision
to publish § 1 of the separate minutes of the Executive Board meeting on 14 June 2001 and no. 9:3 of the meeting on 5 July 2001, and that the information on the scope of the intervention mandate delegated on 14 June should not yet be made public.
The Executive Board decided in favour of this proposal.
1) This section is based on the presentation made by Javiera Ragnartz, Monetary Policy Department, at the Executive Board meeting.
2) This group consists of officials within the Riksbank and is headed by one of the deputy governors.