a study of the factors influencing exchange rates in a...

40
Copyright UCT A study of the factors influencing exchange rates in a small, open economy – the case of the Zambian economy A Research Report Presented to The Graduate School of Business University of Cape Town in partial fulfillment of the requirements for the Masters of Business Administration Degree By Bedah Salasini December 2006 Supervisor: Dr Barry Standish

Upload: nguyentu

Post on 10-May-2019

218 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

A study of the factors influencing exchange rates in a small, open economy – the case of

the Zambian economy

A Research Report

Presented to

The Graduate School of Business

University of Cape Town

in partial fulfillment

of the requirements for the

Masters of Business Administration Degree

By

Bedah Salasini

December 2006

Supervisor: Dr Barry Standish

Page 2: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

2

ABSTRACT

World over, stable and competitive exchange rates continued to be a major

macroeconomic target of most countries over the past decade. But to maintain the

exchange rate at stable and competitive levels requires extensive knowledge on what

factors influence the exchange rate to fluctuate. This paper studied factors influencing the

foreign exchange rate in Zambia, a small open developing economy from 1999 to 2004

whose currency, the Kwacha, generally depreciated over this period against the United

States Dollar. By using the balance of payments approach, the monetary approach as well

as the portfolio balance approach to exchange rate determination, the study found that

appreciation in the Kwacha is positively associated with rising yield on the 91 day treasury

bill, the rising copper prices, improvement in the overall balance of payments, a reduction

in the international reserve change, reduced budget deficits, reduced broad money supply

and lastly reduced total gross national product. Surprisingly, increased financial account

balance and debt relief were found to negatively impact on the kwacha appreciation. This

absurd result, most likely, was caused by the fact that US dollars earned from these

sources were not supplied into the market but must have ended up to build strategic

international reserves which the government can draw upon in times of calamities like

droughts, unprecedented fuel shortages. However, if these US dollars were supplied into

the market, then they only helped the kwacha not to depreciate to the full extent it would

have if the Zambian exchange rate regime were a free floating system. However, the

Zambian exchange rate regime is a dirty floating system.

Page 3: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

3

DECLARATION

Plagiarism is forbidden, and so I declare that this is my own work and all sources of

information have been cited and referenced.

Signed

Page 4: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

4

ACKNOWLEDGEMENT

I am greatly indebted to all those people who have assisted me in the various aspects of

this study: Dr Barry Standish for providing supervision and reviewing my work at

different stages of the study. My wife, Charity and our two girls Yongo and Tusankine for

their patience, encouragement and understanding. Joseph Simumba for his invaluable help

in collecting the data and his insight into monetary economic theory.

Page 5: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

5

TABLE OF CONTENTS

Abstract…………………………………………………………….……….2

Declaration………………………………………………………….………3

Acknowledgements……………………………………………….………...4

1. Introduction………………………………………………….…………..6

2. Overview of Zambia’s economy………………………………………..8

2.1 The first period 1964-1974…………………………………………….8

2.2 The second period 1975-1990………………………………….………9

2.3 The third period, post 1991…………………………………….......…11

3. The nominal exchange rate regime 1964-2006……..............................11

4. Literature review on exchange rate determination…………….….....17

4.1 Traditional or Trade approach…………………………..…….…….18

4.2 Modern approach and theory…………………………….…….….…20

4.2.1 Monetary approach………………………….…………………..….20

4.2.2 The portfolio balance approach…………………….…………..….22

4.3 Empirical evidence: An overview……………………….………...….23

5. Methodology and definition of variables……………………….…......25

6. Description of the data………………………………………….…..…..27

7. Data analysis and discussion of findings.……………………..…….....29

8. Conclusion and policy recommendations……………………..…...….34

Bibliography

Appendices

Page 6: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

6

1. INTRODUCTION

Zambia, like other small but open developing economies, has struggled to maintain a

stable and competitive exchange rate. A stable and competitive exchange rate continued to

be a major macroeconomic target over the past decade in the national economic plans. The

need for a stable and competitive exchange rate emanates from government realisation that

big swings in the exchange rate do impose real costs on business planning. Anticipated

revenues, production costs, household expenditure, foreign direct investment and the

budget balance of the government can all be adversely affected as the exchange rate

fluctuates.

Every time the currency fluctuates, regardless of direction, one entity in the economy

usually gains while another loses. In the case of domestic currency appreciation, producers

of export commodities lose on revenue whilst consumers of imports including importers

record a rise in their real income. In the case of domestic currency depreciation the

importers as well as the consumers now experience a decline in their real incomes while

the exporters gain extra revenue.

The phenomenon of gains and loses as exchange rate fluctuates does not spare foreign

investors or the government. On the part of government, volatility in the exchange rate

may lead to unforeseen deficit in its budget. Evidently, the recent appreciation in the

Kwacha undoubtedly forced the finance minister to seek parliament’s approval of a

substantial supplementary expenditure bill.

The tendency of one group benefiting and another losing out simultaneously as the

exchange rate fluctuates usually creates conflicts in a nation. To avoid conflict arising

from exchange rate volatility, it is imperative that central banks endeavour to promote a

stable and competitive exchange rate.

Managing a stable and competitive exchange rate is not easy. It requires policy makers and

regulators to effectively understand factors which affect the observed volatility in the

exchange rate. Studying the factors responsible for exchange rates volatility has for long

preoccupied much theoretical and empirical work of international finance. Most

researchers have suggested a wide range of probable factors that determine volatility in the

exchange rate. These factors are discussed and studied at length in the literature review.

Page 7: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

7

Recent works on exchange rate determination have shown that traditional models of

exchange rate behaviour provide less significant results when applied to an emerging or

small open developing economy. Such studies tend to conclude that some conditions

exclusive to these nations are treated differently in the earlier models. As such there is

increasing desire amongst economists to study the determinants of exchange rates in small

open developing economies.

This paper analysed factors that influence the Kwacha to fluctuate against US dollar over

the period 1999 to 2004. By using annual data while being guided by the balance of

payments, monetary and portfolio balance approach to exchange rate determination, the

study found that appreciation in the Kwacha is positively associated with rising yield on

the 91 day treasury bill, the rising copper prices, improvement in the overall balance of

payments, a reduction in the international reserve change, reduced budget deficits, reduced

broad money supply and lastly reduced total gross national product. Conversely, increased

financial account balance and debt relief were found to negatively impact on the kwacha

appreciation. This result is absurd and contradicts the prediction of the balance of

payments approach. Nonetheless it adds to the evidence that exchange rate determination

in the small open developing economies does not show systematic results which have been

observed in the developed economies.

From the beginning, it should be emphasised that the results of this study are indicative

rather than conclusive. The Pearson product moment correlation method used is not a

robust method. A robust treatment of the data required that the exchange rate should have

been regressed on its determinants. Although, this could have been done using the

ordinary least squares method, this method was deemed inappropriate because variables in

this study were found to be highly correlated, as shown in table 3, such that the errors in

the ordinary least square method would have been serially correlated thereby invalidating

all the results. The other available econometric methods which would have overcome this

problem are beyond the ability of the Author. Alternatively even if the author had the

ability to handle the other econometric models, the data available would have not been

sufficient to guarantee good results. Although monthly data on 91 day treasury bill yield

rates and exchange rates was readily available, monthly data on the balance of payments

proved a challenge.

Page 8: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

8

2. OVERVIEW OF ZAMBIA’S ECONOMY

Zambia is a landlocked country in southern Africa. It covers a land area of 752614 square

kilometres with an estimated population of 10million. The country gained its

independence in 1964 from the British Rule with most production units bring under being

under foreign ownership. The major economic activity was mining with copper production

and exports dominating the sector and the economy. Since independence copper has

remained as the major export commodity although government has in the recent past

supported diversification to production of Non Traditional exports. The importance of

copper can be seen in table 1 below

Table1: Importance of Copper Mining to the Zambian Economy

1970

1975 1980 1985 1990 1996

Mining and Quarrying as % of GDP 36 14 16 16 7.4 5.9

Mineral Tax as % of Government Revenue 58 13 5 8 0.1 2.3

Copper Exports as % of Total Exports 95 91 85 83 84 52

Mining Employment as % of Total

Employment

17 17 17 16 15 10

Source: IMF: International Financial Statistics, 1997; Mkenda, 2001

Due to challenges faced in the data collection process, this information was only available

up to 1996

To comprehensively understand how Zambia’s economy has evolved, three distinct

periods ought to be noted (Ng’ambi, 2004). These are 1964 to 1974, 1975 to 1990 and

1990 to date.

2.1 The first period: 1964 to 1974

The first period 1964 to about 1975 reflect a period of high copper prices, rapid economic

expansion, growth in public sector and a corresponding rise in government expenditure.

During this period the nation was one of the prosperous in Africa. In 1967/8 the nation

embarked on the mulungushi reforms, a set of policies which aimed at empowering the

indigenous people in the running of the economy. This led to nationalisation of many

Page 9: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

9

manufacturing and financial services firms. However the revenues obtained from copper

exports still helped the government to continue with its ambitious programme under this

strong command economic structure.

In 1973, the nation was hit by an external shock when OPEC increased oil prices by over

400 percent. (Ng’ambi, 2004:2) writes that following the oil shock, the country’s GDP fell

by 5 percent in 1973.Gross international reserves decreased from US $700 million to

about US $ 200 million by 1974, threatening the country’s ability to import; the current

account balance went into deficit and inflation started to rise steadily. All this was

happening at a time when copper prices at the world market had started to decline

significantly. As can be seen from the table, copper being the largest export commodity

accounting for about 95 percent of total exports with mineral tax averaging between 58

and 13 percent of government revenue, undoubtedly government revenues contracted

tremendously.

Coupled with a huge programme of infrastructure development and human capital

formation, given that the nation had only 168 graduates at independence, with most

existing infrastructure only sufficient for a minority group, the governments budget

balance by 1974 was negative 4.5 percent as a proportion of GDP. As a result government

had to borrow externally to finance its expenditure such that the external debt as a

proportion of GDP stood at 37.3 percent in 1974.

2.2 The second period; 1975 to 1990

During this period, mineral earnings continued to decline as copper prices continued to

sour on the global market. The performance of the economy also continued to deteriorate

resulting into one of the worse depression the country has ever recorded in its economic

history. Although the copper and oil external shocks were key factors in the country’s

economic malaise, the failure by government to respond with favourable macroeconomic

adjustment policy perpetuated the crises. Table 2 below shows selected macroeconomic

indicators 1973 to 1990.

Page 10: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

10

Table 2: Selected Macroeconomic Indicators; 1975-1990

Variable

Year

ToT1 Reserves2 Current3

Account

Budget4

Balance

GDP5

Growth

Copper6

Prices

1975 126.3 142.0 -726.1 -340.8 -2.4 56.10

1976 139.6 92.7 -132.8 -231.3 4.3 63.64

1977 119.8 66.3 -232.3 -190.3 -4.8 59.41

1978 114.0 51.1 -321.1 -208.8 0.6 61.92

1979 135.9 80.0 4.7 -139.9 -3.0 89.49

1980 125.5 78.2 -544.6 -295.0 3.0 99.12

1981 100.1 56.2 -766.6 -210.2 6.2 79.05

1982 88.9 58.2 -592.6 -276.5 -2.8 67.21

1983 97.8 54.5 -310.0 114.6 -2.0 72.23

1984 Na 54.2 -162.7 -120.1 -0.4 62.66

1985 Na 200.1 -404.1 -232.3 1.6 64.29

1986 Na 70.3 -372.1 -388.1 0.6 62.13

1987 Na 108.8 -256.7 -232.3 -0.2 80.79

1988 Na 134.0 -324.7 -205.2 6.7 117.93

1989 Na 116.2 -292.0 -71.7 -1.1 129.15

1990 Na 193.1 -489.8 -43.8 -0.4 120.72

Notes: 1= terms of trade (1987=100); 2=total reserves minus gold (current US$ million); 3=current

account balance before official transfers (current US$ million); 4=Government budget balance (current

K’mn); 5=annual growth of real GDP (%); 6=London Metal Exchange Prices (US cents/pound)

Source: As tabulated in Ng’ambi (2004) from World Tables and International Financial

Statistics, 1992

Given the structural imbalance and the huge external debt, that the nation in 1983

implemented the first structural reform programme. The government restricted wages, new

public sector employment and the budget deficit with a view to reduce aggregate demand

given the high and rising inflation rate. However in 1987, after the food riots on the

copperbelt and Lusaka provinces, the IMF structural reforms were replaced by the New

Economic Recovery Programme following the broad criticism against them. With a partial

command system at its core, NERP managed to reduce the budget deficit and inflation, at

Page 11: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

11

the cost of serious commodity shortages, but it too collapsed after the 1991 elections when

the UNIP government lost the elections.

2.3 Third period: post 1991

When the MMD won the 1991 elections, it introduced the economic recovery programme

under the auspice of the IMF and World Bank. Between 1992 and 1995, most parastatal

companies were privatised. A cash budget approach was adopted in 1993 and values

Added Tax as well as user fee arrangement for social services were introduced. The

central Government was downsized through worker retrenchment. Prices and exchange

rates were decontrolled. In this period, GDP growth was volatile especially in the

agriculture sector where the private sector usually failed to absorb output in bumper

harvest years as government had withdrawn from agriculture marketing activities. Another

significant decline was observed on LME mineral prices which had fallen from 94 US

cents/pound in 1990 to 79 in 1994 before bouncing to 100 in 1995 and dropping again to

78 in 1996&97 and 56 in 1998. Data sources used in this study did provide information

subsequent to 1998.

In 2000, Zambia was classified as a highly indebted poor country (HIPC). By virtue of this

status creditors pledged to cancel the Zambia’s debts stock once it met certain bench

marks. The most notable condition was the requirement that government must reduce its

deficit to below 10 percent of gross domestic product. Therefore beginning 2001,

government tightened fiscal and monetary policies with the view to reduce the rising

aggregate demand which had caused inflation to rise while contributing to the depreciation

of the Kwacha. Although the nation failed to meet the conditions at due date, it managed

to qualify later on after the qualification period was extended to 2004. Huge debt

cancellations and restructuring in the post HIPC period have been reported to have a

positive impact on the appreciation of the Kwacha which recently characterised the

economy.

3. THE NOMINAL EXCHANGE RATE REGIMES: 1964 – 2006

Since independence, Zambia’s nominal exchange rate has undergone significant policy

changes. From 1964 to 1982 the country used a fixed exchange rate (to the US$) regime

which was abandoned in favour of the crawling peg which as well just lasted for a two

Page 12: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

12

year period. In 1985 the country introduced the floating exchange rate regime but it too

was abandoned in 1987 when the nation resorted back to the fixed exchange rate again.

Stringent government controls and decrees among them prosecution of erring player and

possible exclusion from trading in the market enforced the fixed exchange rate system in

the nation.

The major thrust for abandoning the fixed exchange rate regime earlier was caused by the

bullish behaviour the US dollar exhibited against major currencies in the early 1970s when

the International Gold Standard system was pronounced since its re-emergence in 1946

(Encyclopaedia of American History). In 1971 the International Gold Standard Exchange

Rate Regime collapsed after the US treasury department under the administration of

president Richard Nixon announced that it would no longer redeem US dollars for gold in

foreign exchange transactions following the continued deterioration in its balance of

payment position and dwindling gold reserves after past attempts to devalue the dollar

failed (Encyclopaedia Britannica; Encyclopaedia of American History).

In Zambia the bullish US dollar led to an overvalued and misaligned Kwacha which in

addition to the external souring of copper and oil prices deteriorated the current account

balance. However the replacing of the US dollar peg with the Special Drawing Rights later

saw a series of devaluations in the Kwacha.

In July 1983, the fixed exchange rate regime was abandoned and replaced by the crawling

peg after the 20% devaluation of 1983 (Aron and Elbadawi, 1992; Ng’ambi, 2004).

Nonetheless the crawling peg, unlike the fixed peg, was the trade weighted average basket

of currencies for the country’s major trading partners. Under this regime, the Kwacha was

allowed to fluctuate within a narrow band to allow for adjustment unlike in the earlier

cases when no adjustment was made except for occasional devaluations (Mkenda, 2001).

In October 1985, the foreign exchange auction system was adopted in favour of the

crawling peg. Aron and Elbadawi (1992) reports that dissatisfaction with the downward

rate of adjustment and the inefficiencies of accompanying annual exchange allocation led

to the adoption of the foreign exchange auction system. Following the introduction of the

foreign exchange auction system, the Kwacha per US dollar exchange rate on 3rd October

1985, was quoted at K2.2 depreciating (127.7 percent) to K5.01 at the close of the first

week and 8.3 on 11th October 1986 a year later which marked the 53rd weekly auction In

Page 13: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

13

the week 60 auction, conducted on November 29, the Kwacha recorded a sharp decline

closing at 15.25 representing a depreciation of 86 percent since the introduction of the

system.

On the back of these figures, the rules, regulations and operations of the auction system

had changed notably. The notable changes included the change in documentation

accompanying bids, from the 41st auction, the introduction of the Dutch auction on 2nd

August, 1986 and the three-fold increase in the amount of foreign exchange by the Bank

of Zambia (BoZ).

Looking at the trends in macroeconomic indicators presented in table 2, which reveals a

weak state of the economy at this time, it is not surprising that the Kwacha depreciated

even sharply in most time periods. The depreciation of the domestic currency after

inception of the auction system shows that the Kwacha was simply trying to discover its

true value (equilibrium value) given its demand and its scarcity in the country.

Unfortunately in towards January 1987, the foreign exchange auction system began to

collapse. The system was blamed for having caused the depreciation of the Kwacha and

that it had contributed to the already rising inflation which emanated from rising food

prices after the subsidies on maize meal were abolished. Furthermore the system was

accused perpetuating poverty by worsening the distribution of income in favour of the

minority rich at the expense of the majority poor and lastly the system was mismanaged.

The Bank of Zambia in most cases failed to honour its commitments which were in most

cases above its stock of foreign reserves, such that other market agents lost confidence on

the credibility of the auction system. As the unfilled promises rose, fears on sustainability

of the system also rose obviously resulting into adverse market reactions amongst the

participants whose effects on the stability and value of the Kwacha must have been

disastrous.

The panic must have been much stronger towards January 24, 1987 (68th week of auction)

when the system was temporarily suspended after being an issue of concern in the

December 5th, 1986 Copperbelt and Lusaka riots which were sparked by increased food

prices after maize meal subsidies were abolished as part of government austerity measure.

Page 14: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

14

On 28th March 1987 the suspension on the system was lifted and the official trading rate

was set at K9.0. Beginning this date, six more auctions were conducted under a two-tier

auction arrangement. The first arrangement, the official window was established with a

fixed official rate of K9.0 and it was restricted to debt service, importation of essential

commodities and receipts of loans and grants, whilst the auction rate was allowed to

fluctuate between the official rate and the K15 upper ceiling only.

The upper limit was abandoned after four weeks such that in the fifth auction, of the six

auctions conducted after the auction resumed, held on the 24th of April, saw the value of

the kwacha fall to its lowest at K21.02/US$. Therefore on 1st may, after the sixth auction

which served to allocate foreign exchange at the K15 rate was done, President Kenneth

Kaunda abolished the auction system when he announced his governments rebuff on the

IMF supported programmes in favour of the New Economic Recovery Programme dubbed

growth from own resources.

From 5th may, the government re-introduced the fixed exchange. The kwacha value was

fixed to the dollar first with the initial value at K8 and later after the dollar was replaced

by the basket of currencies of the country’s major trading partners. However the

government established a Foreign Exchange Management Committee (FEMAC) to guide

all foreign exchange allocations and issuing of import licences. FEMAC operation had a

number of restrictions on the activities agents could engage in and requests for foreign

exchange were filed through commercial banks. Under the FEMAC, exporters of non

traditional commodities were allowed to retain fifty percent of their earnings with free

trade of title to this retained earnings occurring at a premium on the official rate but

required approval from FEMAC. Imports were only permitted if considered essential and

not produced locally or as long as they won’t out-compete local production.

In February, 1990 the FEMAC introduced a dual currency pricing and allocation

mechanism as the two concurrent devaluations of the Kwacha in June and December1989

to K16 and K24 per US dollar respectively. Under the first window, the initial official rate

of K27.8 fixed for Bank of Zambia to sell, under existing FEMAC guidelines of imports,

foreign exchange for imports while simultaneously using it to all the foreign currency

earned by ZCCM. Opposed to the official window, the open general licence (OGL)

window attracted a market rate of K40 per US dollar.

Page 15: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

15

After the 1991 elections, the foreign exchange market was liberalised with most controls

abolished. This caused the exchange rate to depreciate rapidly. During this liberalisation

period, the retention rate of export proceeds was increased from the previous 50 percent to

100 percent with the government adopting a negative list of imports under the open

general licence window, opposed to the more detailed one under FEMAC. In 1992, the

bureau de change system was authorised while the exchange rate system was unified with

the official rate now determined by the weighted average bureau de change rate. In

December the following year, the foreign exchange auction system was re-established at

Bank of Zambia. By 1994 the currency was fully convertible after most foreign exchange

controls on the current account had been abolished following the suspension of the

Exchange Control Act of 1965. Further attempts to strengthen the foreign exchange

market in this year saw commercial banks being allowed to establish foreign currency

accounts which could even be held on behalf of the non bank public. Increased market

ability to handle foreign exchange made the authorities to abolish the OGL window. After

the abolishment of the OGL window, currency auctions increased from three times a week

to daily although ZCCM continued to be the major supplier of foreign currency dealing as

the non traditional sector was still small.

In 1996, ZCCM was allowed to retain 100% of its earnings when authorities resolved that

the supply of currency in the market emanating outside BoZ official intervention, was

imperative in creating a vibrant market. At this stage, Caroline (2004) reports that at this

stage the BoZ official buying and selling rate were de-linked from the weighted average

dealing rate and linked to commercial banks daily weighted average rates.

After most mines under ZCCM were privatized, monopolistic pricing of foreign currency

in the market emerged after it became evident that new mine owners began dealing with

only a selected faction of commercial banks which consisted of big banks, as they are

usually perceived to be more secured and can finance required mine financial capital. This

led to huge depreciation of the Kwacha the coupled with rising dollarisation in economy, a

situation where business houses illegally price goods and services in US dollars

demanding payment at a predetermined rate (Ng’ambi, 2004). In addition, with low

intermediation process in the market, the foreign currency holding at commercial banks

increased. Figure 1 below shows the exchange rate from 1995 to 2004.

Page 16: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

Figure 1

QUARTELY FLACTUATION IN THE NOMINAL ZMK/US$ SPOT EXCHANGE RATE: 1995-2005

0

1000

2000

3000

4000

5000

6000

PERIOD (QUARTER ENDING)

ZMK

/US$

Figure 1 show that the kwacha generally depreciated although towards the end of 2003 the

rate of depreciation became low with the currency gaining value from quarter ending

March 2005. Observation on the behaviour of the rate after 2002, show that tight fiscal

conditions under the Highly Indebted Poor Country initiative (HIPC) and subsequent

introduction of the broad based Interbank Foreign Exchange system (IFEM) were key in

stabilising the exchange rate.

The IFEM system was launched on the 23 July 2003 after BoZ abolished the earlier

measures government had instituted to curb speculation in the market and stabilise the

exchange rate in January 2001. In 2001 the government directed all foreign firms to retain

at least 75 percent of their foreign currency earnings within Zambia and only repatriate the

25 percent. In addition BoZ the ruled that suppliers trading at least US$100, 000 per week

should transact through its dealing window. Nonetheless the measure failed to curb

further depreciation of the Kwacha as evidenced by the 20.0 percentage point depreciation

of the kwacha barely six months after the measures were enacted.

16

Page 17: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

17

With the rationale to foster a transparent and efficient exchange market, the IFEM was

established to enhance availability of liquidity in the market and improve the flow of

information among participants. The major notable difference in the IFEM is that the

market agents are categorically, based on some criteria, classified either as primary dealers

or market makers.

Primary dealers, mainly commercial banks, do quote two- way prices all the time without

disclosing which side of the market they wish to deal thereby transmitting symmetric

information to the market which in the process generates competition. They are ever

present in the market and affirmatively deal on the prices they quote. The Bank of Zambia

deals based on the lowest bid in the market and highest ask such which makes its role

passive in the market without undermining its critical role as the regulator. Therefore the

exchange rate regime in Zambia remains a managed float, opposed to the extreme fixed

and free floating exchange rate regimes.

4. LITERATURE REVIEW ON EXCHANGE RATE DETERMINATION

Developed theory on exchange rate determination stands in two distinct categories, at

least; Traditional and Modern exchange rate theory (Susmel, 1998). The notable difference

is that most part of modern theory view exchange rates from a purely financial

phenomenon. Modern theory, first developed in the late 1960s, does explain much on the

short run volatility of exchange rates and their tendency to overshoot their long run

equilibrium level, as observed over the past two and half decades in the Dornbursh (1976)

exchange rate overshooting model . However it must be born in mind that prediction of

theory may differ depending on whether exchange rate regime under consideration is

fixed, free floating or managed float. Nonetheless space in most models allows for

modification of prediction to reflect ruling exchange rate regime.

This section presents the theoretical models of exchange rate determination although much

focus is on modern thought which is more useful in understanding exchange rate

behaviour today. Nonetheless pertinent issues from the traditional models are emphasised

and noted especially in the long run.

Page 18: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

4.1 TRADITIONAL OR TRADE APPROACH

The traditional exchange rate determination model is based on the trade flows and the

speed of adjustment depends on how responsive imports and exports are to exchange rate

movements hence traditional model can also be called the trade or elasticity approach

(Salvatore, 2004). This approach treats international private flows as passive responses to

compensate for temporary trade imbalances. Changes in the exchange rate according to

traditional models occur as a result of policies aimed at restoring the BOP equilibrium.

Beginning from a Balance of Payment (BOP) equilibrium position we know that;

0=++++= SDORFAKACABOP 1

Where CA is the current account, KA is the capital account, FA is the Financial, OR the

official Reserve account and SD the usual statistical discrepancy.

Under the flexible exchange rate regime, as long as equation one does not balance

inevitably the exchange rate will move to correct the imbalance. The traditional model

therefore seek to show how exchange rates will move as the nation tries to correct for a

trade imbalance. As mentioned earlier, the model implores the trade flow approach. Let’s

consider a simplified trade balance expression;

( )tfd ryyfMXTB ,,=−= 2

For showing the domestic income, foreign income and the real exchange rate

(ratio of domestic to foreign prices)

dy fy tr

The implication of equation two is that we expect the nominal exchange rate to change as

long as any underlying factor in the TB equation varies. A priori, a rise in domestic prices

and income will create a rise in imports over exports and vice-versa. In effect the

exchange should depreciate, or in its absence, measures to devalue must be instituted to

ensure a balanced position is restored. But to fully understand this model we need to

estimate trade elasticities (price and income). In his International economics text,

Salvatore (2004: 560) while citing empirical studies by Harberger (1957), Houthaker and

Magee(1969), Goldstein and Khan(1985), Marquez (1990) to mention a few, reports that

real world trade elasticities are likely to be much smaller in the short run than in the long

18

Page 19: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

run. Therefore much of the prediction from the elasticity approach may not hold in the

short run period. This problem is aggravated by the famous J-curve argument, which

postulates that when the domestic currency values depreciates, the trade balance can

actually deteriorate before improving overtime due to the tendency of domestic currency

price of imports to rise faster than export prices after a depreciation or devaluation.

Another way to look at exchange rate behaviour in a traditional way is to use the

absorption method. Introduced by Alexander in 1952, the absorption approach begins with

the income equilibrium condition;

( )MXTGICY −+−++= 3

Where is consumption, C I is investment, G is government spending andT national

taxes. By definition absorption, A = GIC ++

The intuition behind the absorption approach is total domestic spending does not

determine domestic output but spending on domestic goods determines domestic output.

The proportion of domestic spending not spent on domestic output is spent on imports

instead. Therefore the trade balance in here in defined as the difference between output

and domestic absorption (Susmel, 1998).

.

Rearranging equation 3 and introducing the after-tax private savings we obtain;

({ TGISMXTB − )}+−=−= 4

The implication of this model is that it highlights important factors from the actions of

government, independent of the central banks intervention, which can be manipulated to

correct for a trade imbalance whose movements inevitably affect the exchange rate

Traditional approach has over the years failed to explain why some nations with trade

balance deficits have experienced appreciating currencies just like why some trade surplus

nations have seen depreciating currencies. The explanation seem to be the fact that

international private capital flows are much larger than trade flows today such that

exchange rates reflect mostly financial rather than trade flows especially in the short run.

This in fact is the premise upon which modern theory is founded (Salvatore, 2004: 501)

19

Page 20: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

4.2 MODERN APPROACH AND THEORY

Rather than considering trade flows, modern theory views exchange rates and the balance

of payments as a financial phenomenon. At the core of this approach, is the Purchasing

Power Parity (PPP) theory, as introduced by the Swedish economist Gustav Cassel. In its

absolute form, the PPP postulates that the equilibrium spot exchange rate between two

currencies is equal to the price ratios in the two nations. However given the many

weaknesses of the absolute PPP the much flexible and much useful version is the relative

PPP. According to the relative PPP theory the change in the spot exchange rate over a

period of time should be proportional to the relative change in the prices levels of the two

nations over the same period.

4.2.1 THE MONETARY APPROACH

To show how monetary approach is developed to determine exchange rates, the starting

point is to model the money market. From the famous Irving Fischer (1911) Quantity

theory of money, as commonly modified to exclude double counting, if we let and

be the nominal amounts of money demanded in the domestic and foreign nation

respectively, then;

dM

*dM

kPYM d = and 5 **** YPkM d =

Where is the desired ratio of nominal money balances to nominal national income, k P

is the price level and Y is the real output is both nations respectively.

If we assume equilibrium in the money market, such that = , where is money

supply, substituting for is equation 5 as well as dividing by and then

solving for the domestic-foreign price ratio, we obtain;

dM sM sM

sM dM sM *sM

kYMYkM

PP

S

S*

**

* = 6

20

Page 21: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

By definition from the PPP theory, *PP is the spot nominal exchange rate quoted as the

number of domestic currency units per unit of foreign currency as long as the PPP holds

continuously.

Equation 6 can be rewritten to reflect growth rates in the underlying variables by assuming

that money velocity does not significantly change in the short run as;

( ) ( )TTTSTSTt YYMMS −+−=+**

, , 7

Equation 7 shows that movements (rise or decline) in the nominal spot exchange over a

given period is equivalent to the sum of relative growth in money supply, a variable

mostly influenced by central banks, and relative growth in the two nations income over the

same period. Therefore fluctuation of the currency can also be explained by the relative

action of central banks involved and the relative growth in incomes between the two

nations.

The monetary model is usually extended to incorporate the effects of future expectations

on exchange rate movement through the interest rate differential, which is the difference

between the home and foreign nation yield rates on bonds (Salvatore, 2004:521).

Exchange rate expectations in this model are viewed from the uncovered interest arbitrage

theory since monetarists perceive foreign and domestic bonds as perfect substitutes hence

additional risk of a foreign bond is ignored. According to the uncovered interest arbitrage

theory, when capital is perfectly mobile between nations, interest rates will always be

equal between them in the long run.

If, instead, yield rates are higher in one nation, investors of the second nation will find it

profitable to invest in the first nation. The implication is that increased demand for the first

nations bonds will raise their prices (a reduction in yield), whilst a reduction in demand for

the second nation’s bonds will reduce their price (a rise in their yield). The arbitrage

process will continue until interest parity is achieved. The buying of foreign bonds where

yield rates are high will cause the currency of the foreign nation to appreciate relative to

the domestic currency hence contributing to the reduction on expected returns. Therefore

the percentage difference in interest rates between nations indicates the percentage change

expected in the currency in the current period.

21

Page 22: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

22

4.2.2 THE PORTFOLIO BALANCE APPROACH

The portfolio balance approach is a more realistic application of the monetary approach

discussed earlier (Salvatore, 2004). This model treats exchange rates like any other

speculative price. It begins by arguing that money is just one asset in which individuals

and firms hold their wealth. The other assets are domestic and foreign bonds. Unlike the

monetary model, it assumes that foreign and domestic bonds are imperfect substitutes such

that holding a foreign bond comes with additional risk. Portfolio balancing depends on the

returns of each asset.

Holding domestic money is risk free but its cost is the forgone returns if that money was

invested into a bond. A domestic bond has a default risk but in addition to default risk, a

foreign bond face other risks like ban on capital flight, freeze on foreign capital as a result

of political and civil strife to mention a few. However foreign bonds are desirable because

they help diversify country specific risk that is losses on domestic returns may be

compensated by gains in foreign yields.

At any given time, an individual or a firm will hold a portfolio which maximizes his utility

given his preferences, wealth, yields on both domestic and foreign bonds, future value of

foreign currency, relative inflation rate, level of risk aversion and his income.

Equilibrium in assets is attained when its supply is equal to demand. However a change in

any of the underlying factors of any asset will cause the equilibrium level to change. This

is where the exchange rate is now determined. Let us consider an instance where yield

rates on the foreign bond declines. Investors will balance their portfolios by reallocation

their wealth from foreign bonds to domestic bonds keeping money balances constant.

In the foreign exchange markets, investors will be selling the foreign currency and buying

the domestic currency such that the domestic currency will appreciate relative to the

foreign currency. No wonder the model argues that exchange rates are determined in the

process of individuals and firms to balance there portfolios through asset reallocation

hence it is also called the Asset Model.

The interest differential is now equal to the expected appreciation less the risk premium

since foreign bonds are now considered to have additional risks over domestic ones.

Page 23: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

23

In summary determining exchange rate is rather complex. Various factors act quite

simultaneously and isolating them can sometimes be difficult. However it is generally

agreed that exchange rates today depends so much about the expected future movements

of the rate given all available information. However unevenness in information amongst

market agents has surely created unanticipated movements in the exchange rate making

them more random and difficult to predict with high confidence.

4.3 EMPRICAL EVIDENCE: AN OVERVIEW

Since 1976 after Frenkel published his influential paper which tested the validity of the

monetary model in explaining the German exchange rate during its hyperinflation period

of the 1920’s and which subsequently provided strong evidence favouring the model,

various studies have been conducted overtime to test the exchange rate determination

models although exchange rate volatility still remains a thorny issue to all economies

today.

Studies on exchange rate determination have evolved to estimate or fit models not only to

past data, also called the in-sample estimation, but also to predict the future exchange rate

values also known as out-of-sample estimation.

In their controversial paper, published in the journal of international economics, Meese

and Rogoff (1983) argued that monetary and traditional models failed to forecast exchange

rates well. However researchers like Mark (1994) argued that Meese and Rogoff

mispecified the models during estimation, mainly because they did not consider the effects

of change in economic policy and foreign exchange market arrangements which mostly

differ between countries. Furthermore baskets used to compute the overall price indices as

well as the existence of transaction costs across nations have affected the PPP as well as

exchange rates too. Actually Dornbusch(1987) and Levich(1985) provided evidence that

the PPP had collapsed in the latter part of the 1970s and much of 1980s.

However ten years after the Meese and Rogoff (1983) paper was published, McDonald

and Taylor (1993) examined the Deutsche-Mark to US dollar exchange rate by employing

the monetary model and ruling out price-stickiness effects between 1976 and 1990. They

found evidence that their exists a long run equilibrium between monetary variables and the

exchange rate after it co-integrated with relative money supply, relative income and

relative interest rates using the Johansen(1988) multivariate co-integration method.

Page 24: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

24

In 1999, when China had just started recording all time high economic growth rates, Wei

Weixian (1999), compared the absorption, the elasticity and the monetary models when he

empirically studied the foreign trade balance in China. His results showed that the trade

balance failed to co-integrate with the exchange rate hence he argued in favour of the

elasticity approach. Furthermore his study reported that a J-curve existed in China and that

devaluations have had significant impact on the trade balance.

With increased variations in the findings from empirical studies, rather than essentially

accepting or rejecting the importance of one variable over the other, emerging new

evidence from research on exchange rate determination, for much of the period beginning

1990, has now concentrated much on methodological issues like the econometric

specification and estimation of the models. For example McDonald et al (2003) reviewed

several non-stationary panel methods which have become popular in the nominal

exchange rate studies today. They included the Kao and Chen (1995) Ordinary Least

Squares (OLS) method, Chiang (2000) dynamic OLS method and Pedroni (2001) fully

modified OLS. Since exchange rates are driven by factors between two nations, at least,

panel data methods do capture not only time series but also cross-sectional properties of

the data which is inevitably lost from ‘pure’ time series models like the vector error

correction and the Johansen method for example.

Nonetheless McDonald et al (2003) provided evidence that augmented monetary model

provided a good description of the exchange rate in a panel of six transition countries

namely the Czech Republic, Hungary, Poland ,Romania, Slovakia and Slovenia.

At the country level unfortunately empirical work on the nominal exchange rate is scanty

and if any, unpublished. Much of the published empirical work on exchange rates in

Zambia is on the Real exchange rates. Empirical evidence on the Real exchange rate in

Zambia can be obtained from the works of Kalinda (1996), Mungule (2000) and Simatele,

(2004).

At the regional level, the work done by Zita and Gupta (2007), modelling and forecasting

the Metical- Rand exchange rate, was reviewed. This paper was insightful because it was

based on the Dornbursch (1976) sticky-price monetary model. Findings from this study

showed that the Metical-Rand exchange rate is best explained by changes in GDP and

Inflation rather than the interest rate differential which was found to have a negligible role.

Page 25: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

25

Results in this study were obtained by modelling 48 quarterly data points covering the

period 1994 first quarter to 2005 last quarter using the Bayesian Vector Error Correction

Model (BVEC) with a follow-up Johansen method. However during forecasting (out-of-

sample estimation) the vector autoregressive method was introduced and its outcomes

were compared to those obtained under the BVEC.

5. METHODOLOGY AND DEFINITON OF VARIABLES

From the reviewed pieces of research work as presented in the literature review, studying

factors which determine the nominal exchange rate can be seen to have become more

sophisticated such that difficulty in improving the existing methodology and estimation is

inevitable. Exchange rate modelling demands good statistical strength in both time series

and cross sectional data estimation on the part of the Researcher. However from a strict

business manager perspective, the value addition for such level of statistical demand can

be questioned although for an economist or policy maker in the central bank value

addition is obvious.

As such it’s not the preoccupation of this paper to either improve on the existing

econometric modelling methods or employ the existing sophisticated multivariate

estimation methods. Instead this paper just aimed to explore, using simple ratio analysis,

the one-to-one relationship that exists between carefully selected variables as guided by

theory presented in the literature review as well as the easy of data availability. Therefore

the major question addressed in this paper is does the value of the Kwacha appreciate or

depreciate with a given change in a given underlying variable. If yes, is the direction of

change consistent with the prediction of economic theory? As such the question of how

much amount of appreciation or depreciation did occur was beyond the scope of this

study.

In order to study the factors which underlie the movement of the exchange rate, this study

used annual data from 1999 to 2004. Nine independent variables were selected based on

their relative theoretical significance as guided by literature with the nominal Kwacha per

US dollar spot rate as the dependent variable.

The independent variables chosen were as follows;

1. Yield on the 91 day maturity treasury bill, in percentages.

2. London metal exchange copper prices, in US dollars per pound.

Page 26: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

26

3. Current account balance(C/Acc bal), in million US dollars.

4. The financial (F/Acc bal) balance, in million US dollars.

5. The overall balance of payments figure, in millions US dollars.

6. Net change in international reserves, in millions US dollars

7. Debt relief, in millions US dollars

8. Broad money (M2) in billion Kwacha

9. Total gross domestic product in billions of Kwacha

The nominal Kwacha US dollar spot exchange rate was used because its data was readily

available. Broad money used was M2, which is the sum of currency in circulation, demand

and checkable deposits held by non bank public and time and savings deposits that are

readily issued for money.

The study used annual data because, although monthly data on exchange rate and the

91day treasury bill were available, monthly data on the variables in the balance of

payments proved a challenge. The data used in this study was sourced from the

international monetary fund statistical appendices for Zambia, which is a data pool

reflecting the estimated figures by Fund, Bank of Zambia, Central statistical office and

Ministry of Finance and National planning

The 91 day treasury bill was included to represent the domestic instrument which investors

find desirable to hold with no risk. By implication this means the investors are assumed to

be risk averse. It was also taken to represent the interest rate differential which is cardinal

in indicating which country investors will invest in, as portfolio balance approach

suggests. A further plausible assumption made was that the interest rate differential

remained in favour of Zambia since interest rate on the similar treasury bill in the United

States hardly exceeded was five percent. Therefore we expect a rise in the yield rate to be

associated with appreciation of the Kwacha since Americans will find it more profitable to

invest in the Zambia government treasury bills.

To discover the influence the explanatory variables have on the exchange rate, the study

computed the Pearson product moment correlation coefficient using excel spreadsheet.

The Pearson product moment correlation coefficient shows the nature of a liner

relationship which exists between the explanatory and dependent through the coefficient

sign while the strength of the linear relationship observed is explained through the

absolute value of the coefficient. The Pearson correlation coefficient ranges between

negative one and positive with values close to negative one showing a high inverse linear

Page 27: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

relationship while zero indicates the absence of the linear relationship. Approaching

positive indicates a high direct linear relationship between variables.

However this method only indicates dependence or variable association and does in no

way explain causality among variables. Therefore all variables in this study are only

explained in relation to their association with the exchange rate. There exists a possibility

that exchange rate can be the one determining some of these variables and not strictly they

determining the exchange rate. In short we can know what is associated with the exchange

rate and not what causes the exchange rate. Causality is exchange rate can be like the

chicken egg spiral paradox. The next section describes the study data visually.

6. DESCRIPTION OF THE DATA

Figure 2 below shows the behaviour of the data over the sample period when plotted on a

line chart.

Figure 2

27

ANNUAL MOVEMENTS IN SELECTED EXCHANGE RATE DETERMINANTS:1995-2004

-2000

-1000

0

1000

2000

3000

4000

5000

6000

7000

1 2 3 4 5 6 7 8 9 10

PERIOD

91 days t/bill copper price exchange rate Bop balance Reserve changeC/A balance F/A balance Debt relief Govt Budget/bal Broad money

Page 28: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

28

From figure 2 it can be seen that over the sample period except after period 9 which

represents the year 2003, the exchange rate remained above the broad money although it’s

worth noting that units of measurements between them differed. Both broad money and

the exchange rate remained positive throughout the sample period together with the debt

relief and financial (F/A) account balance factor although movements of the later are seen

to be closer to the horizontal axis when compared to the former. The current account

(C/A) balance, the overall BOP balance and the government budget balance remained

below zero throughout the sample period whereas change in international reserves

although predominantly negative after period five at least showed some positive values in

period 1, 3 and 4. Interpretation of international reserves, unlike other factors, a negative

coefficient implies the nation increased it net holdings of reserves such that bigger

negative values translates into increased ability of the nation to defend its currency from

undesired fluctuations.

Over the sample period the price of copper fell significantly from the higher US$1.19 per

pound in period 1 to a low US$0.7 per pound price in period 5 and 9 before closing at

US$1.2 per pound in period 9.

Although a strong visual tool, figure 2 does not explain any meaningful relationship

between the independent variables and the dependent variable. Meaningful results on the

nature of the relationship between the dependent variable and the independent variable

was obtained by computing the Pearson product moment correlation coefficient using

excel spreadsheets. Results from the Pearson product moment correlation method are

presented in the next section.

7. DATA ANALYSIS AND DISCUSSION OF FINDINGS

Results from excel spreadsheets on the nature and strength of the relationships between the

independent and dependent variable are presented in Table 3 below.

Table 3; Findings of the Study

DETERMINANT MEAN STDEV1 PEARSON R2

DEP(EXCH) COUNT3

Yield on 91days t/bill (%) 33.9 14.6 -0.50 10.0

Copper price (US$/pound) 0.9 0.2 -0.18 10.0

Page 29: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

29

Exchange rate (ZMK/US$) 2818.6 1497.5 Nil 10.0

BOP balance (Million

US$)

-293.6 117.7 -0.49 10.0

Reserve change (Million

US$)

-44.0 135.5 -0.65 10.0

C/Acc balance (Million

US$)

-578.2 107.7 -0.72 10.0

F/Acc balance (Million

US$)

302.5 129.8 0.20 10.0

Debt relief (Million US$) 281.4 145.8 0.61 10.0

Govt Budget/bal (Billion

K’)

-589.7 427.1 -0.79 10.0

Broad money (Billion K’) 2384.8 1785.8 0.96 10.0

Total GDP (Billion K’) 11144.0 7679.7 0.95 10.0

Notes: 1= standard deviation, 2= Pearson correlation coefficients with exchange rate as

dependent variable, 3= total data points used in the analysis

Results from table 3 are indicative rather than a conclusive on the influence explanatory

variables exert on the movement of the exchange rate. The yield on the 91 days to

maturity Treasury bill carried a correct sign as expected from theory even after making the

assumption that its representative of the interest rate differential. As noted earlier the

qualifying argument that yields on the similar debt instrument in the United States over the

sample had remained low was feasible. The negative sign on the Pearson coefficient

shows that when the yield on the Treasury note is declining the exchange rate values are

rising implying local currency depreciation.

This relationship is explained by 50 percent variation in the yield rate. Therefore we can

argue that as the yield rate falls, investors substitute holding of the US dollar for 91 day

Treasury bill. Hence increased demand for the US dollar against a rise in the supply of the

Kwacha arising from their discarding the 91 day Treasury bills leads to the depreciation of

the Kwacha. Over the sample period the average yield rate was 33.9 percent with a

standard deviation of 14.6 against the average exchange rate of 2818.6 Kwacha per US

dollar with a standard deviation of 1497.5. Data points included in this analysis had a total

count of 10.

Page 30: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

30

Similarly rising copper prices can be seen to have led to declining exchange rate values

implying that as copper prices rose on the London metal exchange, the exchange rate

tended to appreciate. However the strength of this relationship looks to be quite low

considering the coefficient value of 0.18. Nonetheless the sign on the coefficient value is

correct as theory predicts. Therefore we can argue that although rising copper prices lead

to Kwacha appreciation we expect stronger relationship from the total copper revenue

because copper prices can be increasing while production lags behind.

The overall balance of the BOP, as defined earlier in the literature review, carried a correct

negative sign with a 49 percent variation in exchange rate movement being explained by

variation in the overall BOP balance. This is undoubted given the fact that every time

inflows rise more than outflows in a nation’s transaction with the rest of the world, the

balance of payments improves hence the appreciation of the domestic currency. Although

the overall balance remained negative throughout the sample period, periods of its

improvement did coincide with reductions in the exchange rate values implying that the

Kwacha appreciated. The opposite should hold when the overall balance deteriorates.

Another interesting variable in the model is the change in net international reserves.

Countries do hold international reserves as a strategic buff stock which they can draw

upon either to defend their currency from undesired movements or which they can utilize

in the event of a serious shock like unanticipated rise in oil price or import of food in the

event of a drought. A negative value indicates that international reserves increased over

the period under consideration whereas a positive figure shows that international reserves

declined. The average net change in international reserves in the sample period was US$-

44.0 million indicating that on average, Zambia’s reserves rose by US$44 million with a

standard deviation of 135.5.

Its relationship with nominal spot exchange rate under the Pearson product moment

coefficient shows that rising net reserves are associated with rising exchange rate values or

Kwacha depreciation. At first thought this result looks absurd. But systematic thought to it

reveals that if the nation, through the central bank has to build strategic international

reserves in US dollars, this can be achieved by the central bank (Bank of Zambia)

withholding some dollars thereby inducing a relative shortage of the dollar in the market

hence the appreciation of the US dollar and depreciation of the Kwacha.

Page 31: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

31

Furthermore, the sustained negative overall BOP balance, which is a sum of above the line

elements in the BOP accounting, means that the below the line element, which is the

change in reserves, must always be positive if payments have to balance (sum to zero).

Logically to obtain a zero sum as required from the accounting identity, the change in

international reserves should have been positive implying that the central bank was dis-

saving official reserves to fund the deficit in the above the line items. But evidence from

this study showed that instead the change in reserve account increased by an average

US$44 million. Hence this explains only way equilibrium would have been reached was

through the depreciation in the Kwacha.

The Pearson correlation coefficient sign on the current account (C/Acc) balance shows

that the exchange rate appreciated when the current account balance improved. However

improvement in the current account is known to be highly associated to domestic currency

depreciation. This is because depreciating exchange rate improves the external

competitiveness of exports whereas imports became expensive. But this is true only when

the dependant variable is the current account. Therefore the result provided above can be

viewed to reflect the scenario where the Kwacha appreciates through increased net inflows

in the current account. With the correlation coefficient of 0.72, certainly the observed

relationship is robust. Descriptive statistics on the current account balance over the study

period show the mean value of –US$578.2 million with a standard deviation of 107.7.

The financial (F/Acc) account balance over the study period had an average value of

US$302.5 million with the standard deviation of 129.8. The correlation coefficient

however carried a positive sign implying that increased net financial inflows were

associated with rising exchange rate values (Kwacha depreciation). Although the financial

account was positive throughout the sample period as expected from the balance of

payments accounting identity given a sustained current account deficit, the positive sign

on the correlation coefficient is absurd and demands an explanation when compared with

established theory on the balance of payments.

The positive sign can be interpreted to mean that the nation either dis-saved with the rest

of the world and or borrowed externally to finance the deficit in the current account. As

such the dollar earning from this source most likely was not supplied to the market or if it

did, it must have only helped the Kwacha not to undergo a huge depreciation than what

was observed. This is possible because Zambia uses a dirty float exchange rate regime

which entails that the central bank does intervene to smoothen out undesirable currency

Page 32: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

32

movements using the official foreign reserve account. Therefore we can argue that,

although positive net financial inflows over the sample period were recorded, certainly

they either must have only helped reduced excessive Kwacha depreciation or were taken

into the strategic international reserve account which recorded an average net change of

US$44 million. Governments through central banks build strategic reserves which they

can draw upon in time calamities like a drought to import food, fuel in times of

unprecedented shortages and/or defend there currency from excessive depreciation.

Like the financial account balance, increased relief on debt, instead of being positively

related to appreciation in the Kwacha as expected, was rather found to be positively

related to Kwacha depreciation. This too looks absurd since we expect increased debt

relief, through an improvement in the supply of US dollar to the local economy, should

influence the Kwacha to appreciate. However the similar argument as in the financial

account holds that inflows from this source may have been inadequate to influence the

appreciation in the local economy although they must have had reduced depreciation

margins of the Kwacha. Similarly it can be argued that increased financial account net

inflows must have been channeled to increase net international reserves which over the

sample period rose quite significantly recording an average increase of US$ 44 million.

Therefore this study concluded that the positive correlation sign meant that savings from

debt relief were used to mitigate would be excessive depreciation in the Kwacha had the

exchange rate been a free floating system. Such would be excessive variations in the

domestic currency value is the most common reason advanced as to why governments

must always regulate and intervene in exchange rate markets

Like established in the literature review, the budget balance of the government plays a

critical role in the determination of the exchange rate of any given economy. Regardless of

its ever being below zero in the sample period, it was found to be negatively related to the

exchange rate implying that reduction in the deficit coincided with appreciation of the

Kwacha. This supports the evidence that reduction in government budget deficit entails

reduced government consumption on foreign goods and services at least thereby leading to

decline in the relative demand of the foreign currency against supply thereby causing the

foreign currency to depreciate whereas the domestic currency appreciates. As can be seen

from the correlation coefficient, variation in this variable explained about 79 percent

variation in exchange rate movement showing that the relationship was strong.

Page 33: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

33

Alternatively the influence of the reduction in government budget deficit can be seen from

equation 4. In equation 4, it can be seen that a reduction in G-T, for a given level of S and

I, the right hand side of the equation will approach S-I when G-T approaches zero. By

implication the result can be interpreted as an improvement in the current account which

as shown in the model was found to be related to appreciation of the Kwacha.

Of further importance is the role of broad money growth on the exchange rate. Increased

broad money in the economy means that economic agents will increase their volume of

transactions since the medium of exchange has risen. This is consistent with the findings

of this study. The positive linear relationship between broad money growth and exchange

rate depreciation, with 95 percent variation in the exchange rate being explained by the

changes in broad money can be explained in two ways as by applying the motives why

people demand money. Firstly from the transaction money demand motive, increased

broad money meant most Zambians increased their buying of goods and services mainly

from abroad such that more kwacha was supplied in the foreign exchange market against a

simultaneous rise in the demand for US dollar which was used to finance imports. The

second argument from the speculative motive for holding money, implies that when broad

money increases, the marginal utility of holding Kwacha balances by economic agents

reduces such that the rebalance there portfolio through converting the additional Kwacha

into US dollars either for speculative motive or further investment in the foreign countries.

In fact the evolution of exchange rate movements in Zambia as depicted in figure 1 seems

favorable to speculators. From the Pearson coefficient, as well as figure 2, it can be seen

that continuous growth in broad money is related to rising domestic currency figures or

depreciation of the Kwacha with about 95 percent variation in the dependent variable

being explained by the changing broad money.

Finally in the variable list is the influence total gross domestic product (GDP), which was

taken at market prices, exert on the exchange rate. With a good correlation coefficient at

95 percent, the rise in gross domestic product was related to rising values of the Kwacha.

This depreciation is expected from theory since rising incomes amongst individuals will

normally lead to higher consumption of imported goods especially that Zambia’s

manufacturing industry is still in infancy stage. The buying of the US dollar using the

Kwacha leads to more kwacha than the US dollar hence the depreciation of the kwacha.

Lastly in this section it should be emphasised that although the major aim was to study the

influence of some selected notable theoretical factors on the exchange rate, results above

Page 34: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

34

are not conclusive because amongst the variables themselves there exists reverse

dependence and causation may not strictly exist. However valuable information necessary

for policy can be sourced and need be studied further. The section below highlights a

number of policy implications from the findings of the study together with the conclusion

of the study.

8. CONCLUSION AND POLICY RECOMMENDATIONS

The study aimed at examining which factors influence the exchange rate and establish how

the exchange rate is affected and in which direction the Kwacha goes. To achieve its

objective, the study began by reviewing briefly selected changes in the macroeconomic

environment of the country since independence. It showed that since independence in

1964, the exchange rate regime has significantly evolved. Beginning with episodes of

fixed exchange rate systems, the country now uses the dirty floating system. The study

also explained that deterioration in the economy begun with the declining copper prices

and the oil price shock of the early 1970s.

Various attempts made by the earlier governments to save the economy, including

currency devaluations and exchange controls, failed. Therefore by 1991, with the help of

the International Monetary Fund and the World Bank, the newly formed government

liberalised most aspects of the economy such as exchange rate market, decontrol of prices

and many others. This brought new developments in the exchange rate policy with the role

of the private sector becoming more pivotal in the determination of the exchange rate.

The study showed that most aspects of exchange rate theory do significantly explain

exchange rate movement in Zambia except for the financial account balance and the debt

relief. However it should be understood that, Zambia, as a small and open economy is

prone to adverse external and internal shocks such that it can at times behave contrary to

what established theory predicts.

Furthermore the study emphasised that its results were indicative rather than conclusive.

This arose from the fact that the correlation method used was not a very robust method.

Finally the study through its findings, the study recommends that the Zambian government

should endeavour to pursue policies that will help the manufacturing sector grow. It was

observed that, although rising incomes lead to increased consumer expenditure, a bigger

Page 35: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

35

proportion of increased income is spent on imports which have an implication on the

stability of the local currency value. Although it’s true that Zambia cannot produce

everything by itself as comparative advantage theory predicts, surely the country can

produce competitively some of the consumer goods that are currently imported given the

right deliberate government institutional policy.

Lastly the Bank of Zambia should continue to regulate the market although its actions

must be credible and transparent. The credibility of the Central bank cannot be

overemphasised. As figure 1 showed, the generally depreciation in the kwacha was

perpetuated by the reduced confidence the public had for the central bank. As a result

unfounded fears amongst market participants generated speculative behaviour such that

banks started to withhold the US dollar for no apparent reason (Ng’ambi, 2004).

Page 36: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

36

BIBLIOGRAPHY

Ng’ambi, C. (2004), “The Changing Landscape of the Foreign Exchange Market in Zambia Over the Last Forty Years and the Challenges for the Future,” A paper Presented at the Bank of Zambia 40th Anniversary Commemoration Conference: Lusaka, Zambia. Aran, J. and Elbadawi, I. (1992), “Parrallel Markets, the Foreign Exchange Rate and Exchange rate Unification in Zambia,” Working Paper, WPs 909, World Bank. Bank of Zambia, Annual Reports, Various, Lusaka, Zambia. Bigsten, A. and Kayizzi – Mugerwa, S. (2000), “The Political Economy of Policy Failure in Zambia,” Working Papers in Economics, No. 23, Department of Economics, Gothenburg University. Babula, A and Oticer – Robe, I. (2002), “The Evolution of Exchange rate regime since 1990 Evidence from De facto Policies,” Internal Monetary Fund Work Paper, WP/02/155. Mkenda, B. (2001), “Long run and Short run Determinants of Real Exchange Rate in Zambia,” Working Papers in Economics, No. 40, Department of Economics, Gothenburg University. Hall, Brand Taylor, J. (1991), Macroeconomics, New York: W.W. Norton and Company. Meese, R. and Rogoff, K. (1983), “Empirical Exchange Rate Models of the Seventies: Do they fit out of sample?” Journal of International Economics, No. 14, PP. 3-24. Susmel, R. (1998), “Structural Models of Exchange Rate Determination,” URL:http://www.bauer.uh.edu/rsusmel/7386/in4.pdf, pp1-15 (accessed on the 21/11/2006). Ministry of Finance and National Planning, (2006), Macroeconomic Indicators, Modeling and Database Unit, Lusaka Zambia. International Monetary Fund. (2004), “Zambia: Selected Issues and Statistical Appendix,” IMF Country Report, No 04/160 :Washington DC. International Monetary Fund. (2006), “Zambia: Selected Issues and Statistical Appendix,” IMF Country Report, No. 06/118:Washington DC. Allsopp, C., Joshi, V., and Mistry, P. (1989), “Zambia: Exchange Rate Policy,” Studies in Macroeconomic Management, SIDA: Sweeden

Cresp-Cuaresam, J., Fidrmuc, J., and MacDonald, R. (2003), “The monetary approach to

exchange rates; Panel data evidence for selected CEECs.” Focus on transition No.2 of

2003

Page 37: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

37

Weixian, W. (1999) “An empirical study of the foreign trade balance in China” Applied

economic letters, vol.6, issue 8

MacDonald, R. and Taylor, M.P. (1993) “The monetary approach to the exchange rate”

IMF staff papers No.40

Zita, S. and Gupta, R. (2007) “Modelling and forecasting the Metical-Rand exchange rate”

A University of Pretoria working paper 2007-02

Dornbusch, R. (1976) “The theory of flexible exchange rate regimes and macro-economic

policy” Scandinavian journal of economics Vol.78 No.2

Frenkel, J. (1976) “A monetary approach to the exchange rate: Doctrine aspects and

empirical evidence” Scandinavian journal of economics Vol.78 No.2

Encyclopaedia of American History, “International Gold standard”

www.answers.com/topic/gold-standard (accessed on 20/04/2007)

Encyclopaedia Britannica, “International Gold standard” www.answers.com/topic/gold-

standard (accessed on 20/04/2007)

Salvatore, D. (2004) International Economics, Danvers: John Wiley & Sons inc

Appendix I. Data the study used year 91 days

t/bill copper price

exchange rate

Bop balance

Reserve change

C/A balance

F/A balance

Debt relief

Govt Budget/bal

Broad money

Total GDP

1995 41.5 1.19 873.28 -252 40 -463 43 37 -104 541 3,005 1996 60 0.86 1207.48 -103 -31 -525 296 310 -212 727 3,950 1997 20.3 1 1314.58 -195 49 -471 412 159 -212 901 5,140 1998 33.4 0.72 1861.83 -453 246 -579 285 122 -485 1105 6,028 1999 36.2 0.7 2,387.60 -156 -35 -429 452 443 -298 1428 7,478 2000 34.1 0.82 3,110.80 -420 -155 -622 202 217 -708 2486 10,072 2001 50.5 0.77 3,610.95 -292 -124 -758 466 436 -1056 2754 13,133 2002 32.5 0.7 4,306.80 -414 -225 -652 238 437 -1031 3620 16,260 2003 13.8 0.78 4,734.00 -319 -161 -700 380 389 -1349 4468 20,3772004 16.5 1.2 4778.7 -332 -44 -583 251 264 -442 5818 25,997

Note: All units of measurement as defined in the report

Page 38: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

Appendix II

91 day Treasury bill Yield rate

0

10

20

30

40

50

60

70

1 2 3 4 5 6 7 8 9 10

Period (1995-2004)

Perc

ent

91 days t/bill

38

Page 39: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

Appendix III

London Metal Exchange Copper prices:1995-2004

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1 2 3 4 5 6 7 8 9 10

Period

US$

per

Pou

nd

Appendix IV

Annual movements in study variables obatined from the BOP:1995-2004

-1000

-800

-600

-400

-200

0

200

400

600

1 2 3 4 5 6 7 8 9 10

Period

US$

mill

ion(

s)

Bop balance Reserve change C/A balance F/A balance Debt relief

39

Page 40: A study of the factors influencing exchange rates in a ...gsblibrary.uct.ac.za/ResearchReports/2006/Salasini.pdf · A study of the factors influencing exchange rates in a small, open

Copyright UCT

Appendix V

Study variables that were taken in Kwacha

-2000

-1000

0

1000

2000

3000

4000

5000

6000

7000

1 2 3 4 5 6 7 8 9 10

Period(1995-2004)

Bill

ions

of K

wac

ha

Govt Budget/bal Broad money

Appendix VI

Total Gross Domestic Product: 1995-2004

0

5000

10000

15000

20000

25000

30000

1 2 3 4 5 6 7 8 9 10

Period

Billi

ons

of K

wac

ha

40