A Study of the Real Interest Rate Differential Mode and Nominal Inter-Bank Lending Rate Differential as a determinant of the Swiss-Euro exchange rate

Transcription

1 A Study of the Real Interest Rate Differential Mode and Nominal Inter-Bank Lending Rate Differential as a determinant of the Swiss-Euro exchange rate

2 TABLE OF CONTENTS Introduction... 3 The Real Interest Rate Differential Model of Exchange Rate Determination... 3 Theory... 4 Real Interest Differential Model... 4 Nominal Short-term Inter-bank lending Interest rate Differential... 5 Results of Testing... 6 Real Interest rate Differential (Long-run)... 6 Short-term Nominal Inter-bank lending Rates... 9 Overall Analysis... Conclusion... 2 Bibliography

3 CHF per Euro INTRODUCTION The Swiss franc Euro exchange rate is currently an interesting exchange rate with headlines such as Switzerland abandons floating exchange rate in dramatic 'currency war' twist. The Swiss National Bank (SNB) operated a floating currency exchange policy until in September 20 it deem it appropriate to drop that policy in substitute for a fixed exchange rate so as the Swiss franc did not drop below.20 CHF/Euro CHF/Euro THE REAL INTEREST RATE DIFFERENTIAL MODEL OF EXCHANGE RATE DETERMINATION With the movement of capital becoming ever more uncomplicated in recent decades due to the development of technology and lifting of international barriers interest rate differentials have become more important determinant of exchange rates. Well known models and studies underscore the role of interest rates in determining real exchange rates. These include the 3

4 International Fisher Effect, The Sticky Price Model (Dornbusch & Mussa) and optimising models (Grilli & Roubini)which highlight the effect of liquidity impulses on real interest rates and consequently exchange rates. 2 Basic economics tells us that if the interest rate (return) of an asset such as Government bonds goes up, the currency needed to buy that asset should appreciate relative to others with the increase in demand, ceteris paribus. Empirical models have found that the trend in real long term yield spreads which accounts for the influences of inflation is better at explaining exchange rate movements than the trend in nominal short term yield spread which do not adjust for the effects of inflation. Real yields will evidently show the mainstay structural shifts in savings and investments and the activities of Central Banks. The long term aspect is better because long term rates tend to be sustained, whereas short term rates can be transient and affected by speculative activities. The relationship between interest rates and exchange rates under the Real Interest rate Differential model is built on two main blocks being Purchase Power Parity (PPP) and Uncovered Interest Rate Parity. Hoffmann & MacDonald through their paper on PPP and Real Interest Rate Differential have found that there is little evidence to support PPP, however there is a valid stationary relationship between interest rates and exchange rates, in particular for small open economies such as the Swiss economy. THEORY REAL INTEREST DIFFERENTIAL MODEL q s = q s + (r Swiss r German ) The RID model is useful in the study of this particular exchange rate due to its large focus on capital movement which Switzerland has been subject to over the past 2months. It stipulates that the spot exchange rate of two currencies is determined based on the interest rate differential between those two currency areas. As interest rates rise in a country, the increased return will attract capital from international investors and there will be an inflow of capital. The model looks at the difference between the interest rates of the two assets, the Real Interest rate Difference (RID). If for example the RID was to fall. This could be caused for 2 Institute for empirical research in economics, University of Zurich Hoffmann & MacDonald 4

5 two reasons either the Swiss interest rate has gone up on the 0 year bond or Germany s interest rate on their 0 year bond has gone down. The net result is the interest rate on the Swiss bond is now relatively speaking after increasing relative to the German 0 year bond. In turn there is an increased demand for the Swiss franc by international investors due to the increasing interest rate of Switzerland. Thus in theory the RID model of exchange rate determination predicts a positive relationship between the real exchange rate and real interest rate differential. NOMINAL SHORT-TERM INTER-BANK LENDING INTEREST RATE DIFFERENTIAL This examination is to see whether the short-term lending rates show any correlation in the exchange rate determination. In theory the inter-bank rate which includes inflation should reflect economic and market conditions and market sentiment which are inherently tied to exchange rate determinations 5

6 Monday, December 5, 2008 Thursday, January 5, 2009 Sunday, February 5, 2009 Sunday, March 5, 2009 Wednesday, April 5, 2009 Friday, May 5, 2009 Monday, June 5, 2009 Wednesday, July 5, 2009 Saturday, August 5, 2009 Tuesday, September 5, 2009 Thursday, October 5, 2009 Sunday, November 5, 2009 Tuesday, December 5, 2009 Friday, January 5, 200 Monday, February 5, 200 Monday, March 5, 200 Thursday, April 5, 200 Saturday, May 5, 200 Tuesday, June 5, 200 Thursday, July 5, 200 Sunday, August 5, 200 Wednesday, September 5, Friday, October 5, 200 Monday, November 5, 200 Wednesday, December 5, Saturday, January 5, 20 Tuesday, February 5, 20 Tuesday, March 5, 20 Friday, April 5, 20 Sunday, May 5, 20 Wednesday, June 5, 20 Friday, July 5, 20 Monday, August 5, 20 Thursday, September 5, 20 Saturday, October 5, 20 Tuesday, November 5, 20 Thursday, December 5, 20 RESULTS OF TESTING REAL INTEREST RATE DIFFERENTIAL (LONG-RUN) Real Interest Rate Diferential Swiss - German (%) CHF/Euro Fig. 6

7 Looking at the seismic moves of the exchange rate and real interest rate differential the results are interesting. We can indicate some increase in the exchange rate relative to an increase in the real interest rate differential, denoting the positive correlation which the RID theory supports. However this is not an immediately correlated. The time needed for the markets to react to interest rate movement of capital can be seen. We can see with the red connectors in the weeks following an interest rate augmentation there is an appreciation of the Swiss franc Euro exchange rate. This represents the inflow of funds into the Swiss monetary zone to take advantage of increasing returns. Au contraire, the purple arrows in the middle of 2009 show the depreciation of the exchange rate as a result of (but not solely) the corresponding interest rate declines representing the positive relationship in the form capital outflows. Despite recognizing a number of correlation points along the time series between December 2008 and December 20 there does not appear to be consistency within the correlations. The black box in the Real Interest Differential graph the period of July 2009 to August 20 highlights a decrease in the RID. This if the theory was to hold true this movement should be recognized with a depreciation of the exchange rate in the months of September and/or October, but this is not the case. In fact we see the currency appreciate. The size of the movement in exchange rate relative to the increase in interest rate shows the vigor of the correlation between the variables. A % increase in the RID would on average cause a corresponding % increase in the CHF/Euro exchange rate over the past 3 years. 3 When we study the reverse movement when the RID decreases there is less of a reaction in the exchange rate based on a movement of the same magnitude when increasing with a % decrease in the FX rate based on a % drop in the RID on average over the 3 years. 4 Thus the variables are positively correlated but illustrate a stronger correlation when the RID is increasing, this may be due to inflationary effects on the exchange rate in the long run. However the CPI Index for Switzerland does appear to be reasonably stable thus other factors would have to be considered (see fig.2.) 3 Workings in bibliography 4 Workings in bibliography 7

8 Fig. 2 No material consistency with the magnitude of changes in the exchange rate exists relative to the interest rate movement. Point on fig. shows a 750 basis point move and there is relatively small increase in the exchange rate compared with some of the other positive correlation movements. Also the two green connectors circa March 200 and September 200 should large not inconsiderable movements in the in the RID with results of minimal movement in the exchange rate. This could be due to a number of external factors outside the scope of the model. Market sentiment and activities play a huge part in shaping investors decisions on capital movements. Also liquidation timing and contract limitations on funds may restrict the movement of capital into Swiss francs and take advantage of the RID change. The green circa arrow the last quarter of 200 connects what appears to be a depreciation of the currency following the decrease in the RID. However based on my knowledge of the financial events of the last number years I believe this is a false representation by the RID theory. (Discussed later) 8

9 Monday, December 5, 2008 Thursday, January 5, 2009 Sunday, February 5, 2009 Sunday, March 5, 2009 Wednesday, April 5, 2009 Friday, May 5, 2009 Monday, June 5, 2009 Wednesday, July 5, 2009 Saturday, August 5, 2009 Tuesday, September 5, 2009 Thursday, October 5, 2009 Sunday, November 5, 2009 Tuesday, December 5, 2009 Friday, January 5, 200 Monday, February 5, 200 Monday, March 5, 200 Thursday, April 5, 200 Saturday, May 5, 200 Tuesday, June 5, 200 Thursday, July 5, 200 Sunday, August 5, 200 Wednesday, September 5, Friday, October 5, 200 Monday, November 5, 200 Wednesday, December 5, Saturday, January 5, 20 Tuesday, February 5, 20 Tuesday, March 5, 20 Friday, April 5, 20 Sunday, May 5, 20 Wednesday, June 5, 20 Friday, July 5, 20 Monday, August 5, 20 Thursday, September 5, 20 Saturday, October 5, 20 Tuesday, November 5, 20 Thursday, December 5, 20 SHORT-TERM NOMINAL INTER-BANK LENDING RATES Nominal Interest Rate Differential with 2 month inter-bank lending rates Swiss - German (%) CHF/Euro Fig 2. 9

10 We examine the 2month inter-bank lending rates of Switzerland and Germany. These being the Swiss Libor and the Euribor respectively. These are nominal rates which include inflation to represent the effects of the economies on the exchange rate in study. With this interest differential paradigm as with the RID model previously we expect to see a positive correlation between the nominal inter-bank interest rate differential and the exchange rate. The two red arrows shows a positive correlation with increasing currency demands on the Swiss franc following the rise in interest rate differential on the inter-bank lending rates from the Swiss perspective between April 2009 and May 200. Similar as to the RID we encounter the time lag in investors moving funds in response to the swell in interest rates. The purple indicator highlights the positive correlation in the opposite direction with the nominal interest rate differential decreasing with regard to the interest rate differential. As with the long term RID study there is not a pure correlation. The short term correlation is proving to be considerably weaker to the long-term, which was expected on comments in the Theory section above. This is in conjunction with past studies of interest rate differential and exchange rate determination. The green connectors point out deviations from the expected result. The st green connector shows a negative correlation. With the decreasing interest rate differential we would expect there to be some depreciation in the currency. Moreover in August of 200 the 2 nd green connector points out the spike in the Swiss franc- Euro exchange rate following a decrease in the nominal inter-bank interest rate differential. If the theory was to hold true we should have also seen an increase of the interest rate differential in the months preceding August (reason discussed later). 0

11 OVERALL ANALYSIS It s evident the interest differential model has had some decree when it is determining exchange rates. This is more so evident in the long run with real interest rates as opposed to the short term nominal rates as expected. The models (both the RID and short-term nominal interest rate differential) have not however been robust in towards the end of 20. While it could be easy to compel ourselves to see positive correlations (which don t exist/exist for alternative reasons) in the period studied, the reality of economic events in the same period must be accounted for when critiquing the models. Toward the latter end of 20 we see a spike in the Swiss-franc Euro exchange rate. This was due to the ongoing Euro debt crisis inducing panic in international investor due to speculation on the collapse of the Euro single currency. Investors in turn wanted a safe haven for their funds and the Swiss-franc was a popular choice by many. The Swiss-franc operating as a floating currency appreciated as demand for the Swiss coinage soared. This over-valuation of the Swiss-franc posed a threat to the Swiss economy particularly tourism and exporters. Consequently in September of 20 the highest point of the Swiss-franc appreciation the Swiss National Bank (SNB) announced it would peg its currency against the euro at.2chf/euro, willing to enforce it by buying foreign currency at unlimited quantities. This is why we see the sharp rise in demand and following drop of the Swiss-franc Euro exchange rate and not necessarily due to the interest rate movements. The future of this exchange rate is somewhat more predictable due to the announcement by the SNB. This in itself is interesting as the Swiss-franc becomes a useful vehicle for the carry trade for investors investing in euro areas. Investors know the Swiss-franc should not appreciate further against the Euro based on the SNB announcement and auxiliary depreciation will make repaying their borrowings less expensive in terms of Euro. Moreover actions the SNB take to implement this new policy may require some form of Quantitative Easing. These actions could result in a further manipulation of interest rates if the SNB move to sterilize the inflows. Inflation could also be adversely affected putting further pressure on the RID model as a tool in exchange rate determination. How the SNB manage the exchange rate with the Euro and the inflationary affect are sure to have implications on the Swiss economy and exchange rate going forward.

12 CONCLUSION To conclude there appears to be some merit to the RID model to exchange rate determination and to a lesser extent the nominal interest rate differential on the inter-bank 2 month rate. Nevertheless the models have not held up in 20. They in essence rely too heavily on assumptions such as economic growth and stable economic activity to be used as a significant exchange rate determination utensil in his period where market sentiment and volatility played a more prevailing role. Despite this with the correct acceptance of the short comings of the RID model it can be used to good effect in the determination of exchange rates. Word Count excluding references and bibliography: =209 2

13 BIBLIOGRAPHY Department of economics University College Dublin: Interest Rate Linkages in the Exchange Rate Mechanism Rodney Thom. Foreign Exchange Markets Richard J. Sweeney (Book) The economist online RTE News online ADVFN Financial News online Prof Dr. Simon Evenett Can the Swiss national Bank tame the Strong Swiss Franc Global Financial Journal: Structural Breaks in the Real Exchange Rate and Real Interest Rate Relationship Institute for Empirical Research in Economics, University of Zurich: Real Exchange Rates & Real Interest Rate Differentials: a Present Value interpretation. Deutsche Bank Bundesbank: Monthly Report July 2005 Department of Economics National Chung Cheng University Taiwan: Working Papers Calculation on reference 4: Calculation on reference 5: % change of up tick movements of RID over time period/375 % change from corresponding changes in the exchange rate/375 % change of down tick movements of RID over time period/382 % change from corresponding changes in the exchange rate/382 3