Kabula Irene Mulihano Bank of Tanzania Mentor: …mefmi.org/mefmifellows/wp-content/uploads/2016/10/The...1 Macroeconomic and Financial Management Institute of Eastern and Southern

1

Macroeconomic and Financial Management Institute of Eastern and Southern Africa

The Effects of Global Financial Crisis on Asset Allocation in East African

Central Banks

Kabula Irene Mulihano

Bank of Tanzania

Mentor: Nilakanta Venky Venkatesh

A Technical Paper submitted in Partial Fulfilment of the Award of MEFMI

Fellowship

March 2015

2

CONTENTS LIST OF TABLES AND FIGURES ................................................................................................. 4 LIST OF FIGURES ............................................................................................................................ 4 ABSTRACT ........................................................................................................................................ 5 ACKNOWLEDGEMENTS ............................................................................................................... 7 LIST OF ABBREVIATIONS ............................................................................................................ 8 CHAPTER ONE ............................................................................................................................... 10

1.0 INTRODUCTION .................................................................................................................... 10

1.1 Background ............................................................................................................................... 10

1.2 Problem Statement .................................................................................................................... 14

1.3 Objectives of the Research ....................................................................................................... 14

1.4 Research Questions ................................................................................................................... 15

1.5 Research Hypothesis ................................................................................................................. 15

1.6 Research Significance ............................................................................................................... 15

CHAPTER TWO .............................................................................................................................. 16 2.0 LITERATURE REVIEW .......................................................................................................... 16

2.1 Introduction .............................................................................................................................. 16

2.2 Empirical Review ..................................................................................................................... 16

2.3 Theoretical and conceptual framework .................................................................................... 18

CHAPTER THREE .......................................................................................................................... 21 3.0 RESEARCH METHODOLOGY ............................................................................................. 21

3.1 Research Design ....................................................................................................................... 21

3.2 Sampling plan and population .................................................................................................. 21

3.3 Data collection, sources and instruments ................................................................................. 22

3.4 Data measurement and Analysis ............................................................................................... 22

3.5 Reseacher reflexevity ............................................................................................................... 23

3.5.2 Validity and Reliability ......................................................................................................... 24

CHAPTER FOUR ............................................................................................................................ 26 4.0 EMPIRICAL SETTING ............................................................................................................ 26

4.1 Introduction .............................................................................................................................. 26

4.2 The East Africa Community ..................................................................................................... 26

CHAPTER FIVE .............................................................................................................................. 29 5.0 RESEARCH FINDINGS, ANALYSIS AND DISCUSSIONS .............................................. 29

5.1 Introduction .............................................................................................................................. 29

5.2. Sample description .................................................................................................................. 29

5.3 East Africa Central Banks asset allocation before the crisis .................................................... 29

3

5.4. Central Bank’s reaction to GFC .............................................................................................. 30

5.5 East Africa Central Banks investments after the crisis ............................................................. 48

CHAPTER SIX ................................................................................................................................. 55 CONCLUSION AND RECOMMENDATIONS ............................................................................ 55 REFERENCES ................................................................................................................................. 59

4

LIST OF TABLES AND FIGURES

Table 1: Sampled departments and bank officers selected.

Table 2: East Africa Community at a glance

Table 3: East Africa Central Banks investments before the crisis

Table 4: The liquidity of reserves and priorities of Central Banks Objectives

Table 5: Low interest rates and priorities of Central Banks objectives

Table 6: Level of reserves and priorities of Central Banks objectives

Table 7: Difficulties experienced in managing reserves during crisis era

Table 8: Actions taken by Central Banks to address crisis difficulties

Table 9: Triggers of the asset re allocation

Table 10: Assets that were reduced as a share of total reserves during crisis

Table 11: Reserves allocation and reserves management objectives

Table 12: CBs investment horizon for various tranches

Table 13: Low interest rates and the decision to increase fixed income securities

Table 14: East African Central Bank investments during crisis

Table 15: Adjustments in the currency allocation

Table 16: Allocation to Emerging Markets during the crisis

Table 17: Considerations for currency composition

Table 18: Other currencies considered for reserves investments during the crisis

Table 19: Frequency of investment policy review

Table 20: Central Banks benchmark revision

Table 21: Risk measurement

Table 22: Asset classes allowed for investment after the crisis

Table 23: Comparison of eligible Government bonds before and after crisis

Table 24: Comparison of eligible Supranational bonds before and after crisis

Table 25: Comparison of eligible Agency bonds before and after crisis

Table 26: Comparison of approved inflation protected bonds before and after crisis

Table 27: The likely review of strategic asset allocation

LIST OF FIGURES

Figure 1: The East African Map

Figure 2a: Eurozone Benchmark Curve Change

Figure 2b: US three months Treasury bill rates

Figure 3: EUR Volatility

5

ABSTRACT

The major problem that guided this study was to examine the effects of the global financial crisis

in East Africa Central Banks with a focus on how Central Banks reacted to the crisis in the

context of foreign reserves asset allocation. The global financial crisis of 2008, originated from

rapid defaults on subprime mortgage loans in the U.S. housing market and spread rapidly to

other sectors of the economy. It was witnessed since mid-2007 and had a significant impact on

the financial landscape that has led to changes in reserve management practices. Set in this

context, Central Banks have to choose an appropriate asset allocation of the foreign reserves in

agreement with policy objectives which are safety, liquidity and return. Asset allocation is a high

level decision which reflects the institution’s overall risk tolerance and investment objectives and

constraints over the planning horizon.

To focus on this important issue, the main aim of this research have been to investigate the

effects of the global financial crisis in East Africa Central Banks in terms of asset classes,

currency composition and maturity stuctures and explore ways in which CBs re-allocated their

reserves and evaluate the effects associated with such decisions. To achieve this, the general

research problem has been to identify the effects of the global financial crisis and its implication

to reserves management objectives.

Qualitative and quantitative approach have been deployed and findings are based on the data

derived from East African Community member countries. The entire sample was drawn from

five East African Central Banks and data was collected through the use of a

questionnaire/survey. The research is built upon a deductive approach; hence no new theory is

generated but rather conclusion have been drawn from the comparisons of collected data with

previous made researches.

The analysis led to the conclusion that, given the effects of the global financial crisis of 2008; all

East African Central Banks had choosen to reallocate their financial assets without changing

priorities of the key reserves management objectives of preserving capital, providing liquidity

and generating reasonable return. Furthermore, Central Banks re allocated financial assets by

investing a portion of their reserves in non traditional currencies as a way to diversify from

developed economies and to enhance return. In a way, this suggested that Central Banks risk

tolerance changed. When examined the asset allocation prior and after the crisis, the findings

from the research connotes that, before the crisis, Central Banks reserves were held in traditional

6

reserves currencies of USD, EUR and GBP, eligible assets were deposits and liquid government

bonds. However after the crisis, Central Banks diversified their currency composition to non

traditional currencies such as AUD, CAD and RMB, increased percentage allocation to

government bonds and supranationals and approved Inflation protected bonds and MBS/ABS as

other eligible assets in their portfolios.

In view of the analysis above, this paper recommends the following:

Improve risk management practices and procedures because many of the disastrous losses

would have been avoided if good risk management practices were in place.

Ensure robust portfolio and risk management infrastructure is in place: this is due to the

fact that reserves management has become more complex, resulting in Central Bank

reserves managers, needing to build capacity to up-skill staff and invest in sophisticated

IT and risk management systems which can handle the complex transactions CBs are

now engaging in.

Over reliance on single credit rating models should be avoided as this led to similar risk

management strategies and decision. To supplement ratings from credit rating companies,

an in-house credit risk management system may need to be established. However, it

requires heavy investment in human, information technology and financial resources.

Diversification of assets to non-traditional countries and currencies with high credit

rating with caution that any increase in the range of allowable instruments and markets

that includes non-traditional investment areas should be done within the permissible risk

parameters and tolerance limits.

7

ACKNOWLEDGEMENTS

I would like to extend my sincere thanks to my mentor, Mr. Nilakanta Venky Venkantesh, Lead

Financial Officer, the World Bank Treasury, for his guidance in bringing up this technical paper.

I would like also to thank the MEFMI Secretariat for the funding of my customised training

program without which, this work would not have been produced. I appreciate Dr. Elias

Ngalande, Mr. Alphious Ncube, Ms. Michelle Mutinda and Mr. Simon Namagoa, Ms. Esther

Murahwa and the entire MEFMI family for believing in me and offering me such a distinguished

opportunity to be part of the fellowship program.

My sincere gratitude goes to my husband, Mr. Brilliant Cornel for his support and understanding

while busy with this work. My sons ( Speight & Dalton) and daughter (Bilhah) altogether

deserve my appreciations for they missed my care and love throughout the research session.

More importantly, I thank the almighty God for his glory and mercy on me till when this work is

produced on the table.

However, I remain responsible for errors and ommissions ( if any)

Kabula Irene Mulihano,

MEFMI Fellow,

Financial Sector Programme, Reserves Management

8

LIST OF ABBREVIATIONS

ABS Asset Backed Securities

AUD Australian Dollar

BDF Banque de France

BIS Bank for international settlements

BNR Bank Nationale du Rwanda

BOT Bank Of Tanzania

BoU Bank Of Uganda

BR Burundi

BRB Banque de La Republic du Burundi

BRICS Brazil, Russia, India, China and South Africa

CAD Canadian Dollar

CBK Central Bank of Kenya

CBs Central Banks

CDS Credit Default Swap

CHF Swiss Franc

COFER Currency composition of foreign exchange reserves

DKK Danish Kroner

EAC East African Community

EMCs Emerging Market Countries

EPFR Emerging portfolio Fund Research

EUR Euro

FED Federal Reserve Bank

GDP Gross Domestic Product

GFC Global Financial Crisis

IMF International Monetay Fund

JPY Japanese Yen

KE Kenya

MBS Mortgage Backed Securities

NZD New Zealand dollar

NOR Norwegian Kroner

OECD Organisation for Economic Co-operation and Development

9

RMB Chinese Remnbi

RW Rwanda

SAA Strategic Asset Allocation

SPSS Statistical package for Social scientist

TE Tracking Error

TZ Tanzania

U.S Unites States of America

UG Uganda

USD US dollar

VAR Value at Risk

10

CHAPTER ONE

1.0 INTRODUCTION

1.1 Background

In recent years, the growth in official foreign exchange reserves has led to renewed interest in the

way reserve management decisions are taken and their possible impact on financial markets

(Borio, Galati and Heath, 2008). This trend of increased reserve levels was particularly

noticeable in Emerging Market Countries (EMCs) especially in the BRICS as compared to

advanced economies (IMF, 2011). See also annex 1, chart 1a.

The magnitude and management of these vast resources can have a profound effect on markets

and Central Bank balance sheets Morahan and Mulder (2013). Reserve managers face important

decisions on their asset allocations, including currency composition and asset classes, to ensure

that the reserves meet the key goals of safety, liquidity and return. Asset allocation is a high level

decision on how to allocate reserves to the broad asset classes such as stocks, bonds and cash

(Idzorek, 2006). The so called strategic asset allocation which determines currency composition,

maturity structure and asset classes, has long term effects on Central Bank’s (CB) balance sheets

and it forms a back ground against which they respond to the crisis environment (IMF, 2013).

The critical lesson of the global financial crisis (GFC) and its aftermath is that asset allocation is

critically important (Shane, 2013).

The global financial crisis originated from rapid defaults on subprime mortgage loans in the U.S.

housing market and spread rapidly to other sectors of the economy (IMF 2009). Many leading

economists believe that it was the worst crisis since the great depression, and argue that many

factors catalyzed the emerging of the 2008 financial crisis; among which were credit market

failure, inefficient regulatory framework and under-estimation of risk by credit rating agencies

(see also Aryeetey and Ackah 2011). The collapse of Lehman Brothers in September 2008, sent

a wave of fear around the world financial markets. After the collapse, banks stopped lending to

each other and risk premium on inter bank borrowing rose sharply to 5 percent where as it was

close to zero (Warwich and Stoeckel, 2009). Chart 1b in Annex 1, depicts this rise in short-term

inter bank borrowing during September 2008 from as far back as July 2001. The speed and

transmission of the crisis was particularly aided by global financial linkages across national

boundaries and by intersectoral linkages within countries.

11

The effects of the global financial crisis on developing countries, East African Community

(EAC) countries included, were forecast to be comparable to those on developed countries

(Lunogelo, mbilinyi and Hangi, 2009). The East Africa Community is a regional inter-

governmental organization of the republics of Burundi, Kenya, Rwanda, Uganda and the United

Republic of Tanzania. Kenya, Tanzania and Uganda were the original three Partner States

whose established community came into force on 7th July 2000. The Republics of Rwanda and

Burundi became full Members of the Community with effect from 1 July 2007. The East Africa

community dreams of a prosperous, competitive, secure, stable and politically united East

Africa; and plans to widen and deepen Economic, Political, Social and Culture integration in

order to improve the quality of life of the people of East Africa through increased

competitiveness, value added production, trade and investments1.

Literature suggests that EAC grew faster than the rest of the continent both before and during

the the global financial crisis, the lack of natural resources notwithstanding (El Sayed and Hegazi

2013). In spite of the negative impact of the GFC, East Africa posted 5.8% real GDP growth in

2009 and has already recovered some of the lost growth momentum in 2010 and 2012. Rwanda,

Tanzania, and Uganda have led the regional economic expansion, however, Kenya, which grew

rapidly in 2006 and 2007, suffered a setback in 2008 due to the violence that broke out after the

elections at the end of 2007. Among the EAC’s members, only Burundi’s growth has been low

throughout the 2000s, reflecting in part the country’s fragility (Aryeetey and Ackah, 2011).

Although the global financial crisis created a ripple effect, that is having a devastating effect on

the economies of other countries in the world (Njoroge, 2009); the integrated nature of the global

economy means that East Africa region has not been spared as future growth prospects are

dependent, to an important degree, on global developments. The global financial crisis affected

East African countries through three primary channels. First, as growth in trading partners

slowed, EAC economies suffered from a decline in external demand for their good and services.

Second, growth in domestic demand was dampened due to reduced income and low prices of

commodities, Finaly, reduction in capital flow including foreign direct investments constrained

and dampened growth prospects( IMF 2009).

1 A paper of EAC development strategy on Deepening and Accelerating Integ ration

12

The financial crisis spilled over into the global financial system during the summer of 2007 and

its breadth increased following the collapse of Lehman Brothers as stated earlier. In response,

Central Banks around the world took actions including cuts in interest rates, the provision of

ample liquidity and other unconventional measures. Central Banks in their respective countries

acted as the lender and market maker of last resort (Nakaso, 2013). In pursuing this, Central

Banks used their foreign exchange reserves. This is in line with the IMF revised guidelines for

Foreign Exchange Reserve Management objectives that, Central Banks hold foreign reserves to

protect a country from external vulnerability by maintaining sufficient liquidity to absorb shocks

during financial crisis (IMF, 2013). Hence, the robustness of foreign reserves is important for

the stability of the domestic currency and confidence in the economy.

Furthermore, Central BankCentral Banks manage reserves by investing part of it in deposits and

fixed income securities of foreign banks (IMF, 2010). Put in this context, East African Central

Banks are not on their own as they also invests in such banks. This is supported by an IMF

survey which revealed that until the start of the global financial crisis in 2007 a large share of

CBs reserves were held as deposits in foreign banks and proved that when the crisis hit, many

Central Banks withdrew these investments (IMF, 2010). During 2008, CBs walked to safety by

reducing the weight allocated to commercial bank deposits and increased treasury and agency

bill holdings and when Lehman defaulted, they shifted from “walk to safety” to “flight to

quality” where they increased treasury holdings(Louis, 2010). In view of this, financial crisis

posed a great challenge to official foreign exchange reserve managers as it reminded them on

how to deal with emergencies (BIS, 2009 p.19).

The extent to which Central Banks have been susceptible to financial crisis varies depending on

the specific circumstance. Central Banks being public institutions performing a public policy

task in which credibility is crucial and the loss of reputation potentially devastating to their

effectiveness, are more constrained in their investment opportunities than their private sector

counterparties (Bakker and Ingmar, 2007). However, they have been looking at ways to expand

their universe of possible investment to earn higher returns with only minimal concession to

liquidity and safety. Set in this context, Central Banks have to choose an appropriate asset

allocation of the foreign reserves in agreement with policy objectives. Central Banks use

Strategic Asset Allocation process to distribute reserves into various asset classes to achieve

their long-term investment objectives. Strategic Asset Allocation reflects the institution’s overall

13

risk tolerance and investment objectives and constraints over the planning horizon (Cardon and

Coche, 2004).

The period of poor returns and high volatility that characterized the global financial crisis and

subsequent years demonstrated that asset allocation is critically important going forward (Shane,

2013). For Central Bank’s foreign reserve portfolios, the asset allocation process comprises

decision on the currency composition and within each currency, on the allocation to various

fixed income asset classes, mainly government bonds and other highly liquid, highly secure

instrument types (Cardon and Coche, 2004). These authors insist also that, the allocation will

vary over time due to changes in investment opportunities, the investment horizon and long term

macro economic risk factors like inflation and interest rates.

Accordingly, foreign exchange reserves of the East Africa region have been invested in assets of

developed markets which were affected negatively by the crisis. The post-crisis period was

characterized by several credit downgrades, currency volatilities, low yields and narrowing of

investment universe (Srimany, Gayen and Ranjeev, 2011; Davis, Roger and Patterson, 2011;

IMF, 2011; and World Gold Counsel, 2011). As a result Central Banks changed their asset

allocation in the search for diversification to generate return on their portfolios (Morahan and

Mulder, 2013). This is supported by Cardon and Coche (2004) who said that, “… a general

requirement might be that changes in the investment universe should be made in connection with

a review of strategic asset asset allocation”. The low level of yields on fixed income markets

which still represent the bulk of Central Bank’s portfolios led them to consider diversification

towards new currencies and asset classes as well as more actively managing their asset allocation

(Bernad, 2009).

Most of the studies on asset allocation in response to the effects of the global financial crisis

have largely focused on advanced economies, leaving out African economies’ perspective

particulary in East Africa (Were, 2012). Against this background, this paper identifies the effect

of the global financial crisis on asset allocation particulary in East African Central Banks.

Emphasis will be placed on exploring the ways in which Central Banks re-allocated their reserves

and evaluate the effects associated with such decisions.

14

1.2 Problem Statement

In recent years, issues related to the management of foreign exchange reserves have gained

importance and reserve management practices have developed rapidly. The financial crisis that

was witnessed since mid-2007 had a significant impact on the financial landscape that has led to

further changes in reserve management practices. Central Banks have been looking on how to

expand their universe of possible investment to earn higher returns with only minimal concession

to liquidity and safety. Set in this context, Central Banks have to choose an appropriate asset

allocation of the foreign reserves in agreement with policy objectives.

The approach has been varied from country to country but six years after the event, enough

knowledge has been gained to draw experience and good practice going forward. Empirical

studies regarding effects of the financial crisis on asset allocation have been carried out

(Morahan and Mulder 2013; IMF, 2011; Nichlas and Nicklas 2012; Srimany et al, 2011) and

focused on developed countries with little attention to developing economies particulary East

Africa region. This paper is going to focus on how East African Central Banks reacted to the

GFC in the context of foreign reserves asset allocation. The effects of the global financial crisis

on asset allocation in East African Central Banks in terms of asset classes, currency composition

and maturity stuctures is of concern.

1.3 Objectives of the Research

The objectives of this study are shaped around the three issues below and stem from the title of

the paper, that is, to identify the effects of the global financial crisis on CBs asset allocation.

From the foregoing, the following specific objectives are set:

1. Evaluate the extent to which priorities of Central BankCentral Bank’s objectives (safety,

liquidity and return) changed during the financial crisis.

2. Examine the asset re allocation after the financial crisis and assess whether it reflects the

change in risk torelance.

3. Examine the asset allocation prior and after the financial crisis;

15

1.4 Research Questions

In order to achieve the objectives stated above and provide answers to the general research

problem, the following research questions are addressed.

1. To what extent are the priorities of Central Banks reserves management objectives

changed during the financial crisis?

2. Is there any relationship between asset re allocation and the change in risk torelance?

3. Is there a difference between asset allocation prior and after the financial crisis?

1.5 Research Hypothesis

This research determines the extent to which GFC affects East Africa’s Central Banks asset

allocation and its associated implications to reserve management objectives. To achieve this,

three hypotheses will be tested.

i. The extent to which priorities of Central Bank’s objectives changed is related to the

financial crisis.

ii. Asset reallocation is related to change in risk tolerance.

iii. There is no difference between asset allocaction prior to and after the financial crisis

1.6 Research Significance

The study adds to new knowledge on how Central Banks in developing economies such as

Tanzania, Uganda, Rwanda, Burundi and Kenya dealt with asset allocation of their reserves

during the crisis. Based on this, Central Banks and other decision makers can use research

findings and policy recommendations to make an informed and effective investment decisions in

case of similar situations in future. The study also adds value to the creation of sufficient

information and knowledge required as a basis for further research in Central Banks in

developing economies. Similary, it is imperative that the various effects posed by the global

financial crisis on asset allocation especially in East Africa Central Banks be effectively

addressed if the financial sector is to remain competitive in line with policy objectives.

16

CHAPTER TWO 2.0 LITERATURE REVIEW

2.1 Introduction

This section presents an overall review of current knowledge on different theories and ideas in

relation to the effects of the global financial crisis (GFC) on asset allocation. It is primarily a

review drawing on existing research work on various topical areas relevant to asset allocation in

relation to GFC, with a particular focus on recent empirical studies at a national or comparative

level. The empirical review gives a description and analysis of what other scholars have written

in the same subject area. Thereafter, it is followed by a theoretical framework.

2.2 Empirical Review

More recently, Morahan and Mulder, (2013) conducted a survey on how Central Banks arrived

at their asset allocation and the way in which they executed their risk management frameworks

during the crisis era. The survey involved 156 reserve managing Central Banks from Americas,

Europe, Africa, Asia and Oceania continents and adopted a cross-sectional approach in which a

survey was sent to member countries for them to fill out. Out of 156 countries surveyed, 67

responded which is a response rates of about 43 percent. The main findings from the survey

indicate that, 70 percent of reserve managers changed their asset allocation. Specifically, about

50 percent of all CBs pulled back from their commercial bank deposits and 35 percent of all

respondents reduced their exposure to unguaranteed bonds. Respondents who were considering

adjusting the currency composition of their reserves were about 50 percent. Respondents

indicated a high interest in commodity currencies such as AUD and CAD. Middle and other

higher income countries (countries classified as having reserves over $3,975 million) were

considering investing in Renminbi (RMB).

This is supported by the IMF survey which highlights that, asset allocation for investors differs

markedly by country and the diversity in asset allocation across countries reflects in part

differing investment structures, but not differences in holdings by type of investor, which are

similar across countries (IMF,2011).

Likewise, during 2011, IMF staff researched on the fundamental drivers for asset allocation

decisions and determined whether their influence has been altered by the global financial crisis

and the subsequent low interest rate environment in advanced economies. The study employed

17

cross-sectional approach in which questionnaires were sent to unleveraged investors such as

individuals, public and private pension funds, insurance companies and sovereign wealth funds.

During this survey, questionnaires were sent to approximately 300 largest asset management

companies and 200 largest pension funds and plan sponsors in the world, and a total of 122 firms

participated (out of which 68 were asset management companies and 54 pension funds or plan

sponsors). In addition to the survey, the reseach used available public (OECD data) and private

data, the views of investors and other market participants. To determine the longer-term trends

in asset allocation OECD data and dataset compiled by Emerging Portfolio Fund Research

(EPFR) were used.

The research found that, the asset allocation of institutional investors differs markedly by

country where U.S. investors hold about equal shares of equities and bonds, while investors in

France hold a majority of assets in bonds and those in Germany hold almost one-third of their

assets in currency and deposits. The study also revealed that, the diversity in asset allocation

across countries reflects in part differing investment structures, but not differences in holdings by

type of investor. The empirical results and survey responses indicated that asset allocation

strategies of private and official institutional investors have changed since the onset of the global

financial crisis. Investors are more risk conscious regarding the risks associated with liquidity

and sovereign credit. Also, the structural trend of investing in emerging market assets has

accelerated following the crisis and with many investors taking advantage of the relatively better

economic performance of emerging countries.

Set in this context, this paper looks at how Central Banks arrived at their asset allocation in the

context of East Africa region particulary on whether and how CBs adjusted their currency

composition and asset classes.

In their working paper series at Reserve Bank of India, Srimany et al, (2011) researched on the

effects of financial crisis on optimal asset allocation of foreign reserves. The aim was to

understand the effect of the financial crisis on optimal asset allocation patterns for different

duration targets and the consequences thereon. They discussed some strategies for the Central

Banks to create optimal portfolios given their level of reserves under present low yield

scenarios.

18

In their dataset and methodology, they considered investment of foreign reserves in the US

government bond market and in search of better return, they included the government bond

markets of developing countries2 to understand the effect of such inclusion in the optimum asset

allocation pattern and consequences thereof. The data was collected for the period from January

2005 to December 2010 from Barclays Capital short Treasury indices and Datastream

Government bond indices. However, investment universe was restricted to a maximum maturity

of 10 years.

Their study attempted three optimizations namely, maximization of portfolio return,

minimization of portfolio risk and maximization of return per unit risk of the portfolio given

certain constraints. Based on USD-denominated government bond data, they studied the

characteristics of the optimal portfolio in the pre-crisis and post-crisis periods. Their findings

reveal the following: (i) the optimal asset allocation during post-crisis period suggests a

concentrated portfolio which is against the barbell structure for pre-crisis period, irrespective of

low, medium or high target duration; (ii) return and yield, in absolute term, are very low in the

post-crisis period; (iii) Central Banks can get higher pick-up in return in the post-crisis scenario

than in the pre-crisis scenario for same amount of increase in duration; (iv) in the post-crisis

period, the probability of negative return increases substantially compared to the pre-crisis

period.

Finally, the paper concluded that Central Banks with medium or long target duration can get

substantial yield and return pick up and also reduce the probability of negative returns by

including the emerging Asia within the investment universe without even extending the duration.

2.3 Theoretical and conceptual framework

Given the impact of The global financial crisis and subsequent European sovereign debt crisis,

recent years have witnessed a growing literature on strategic asset allocation for foreign

exchange reserves. Strategic asset allocation is the core of an investment plan. It seeks to

2 IMF defines Advanced economies as one with a high level of gross domestic product per capita, as well as a very significant

degree of industrialization; an Emerging market is a country that has some characteristics of a developed market, but does not

meet standards to be a developed market and a Developing country as a nation with a lower standard of living, underdeveloped

industrial base, and low Human Development Index relative to other countries.

19

maximize the benefits of diversification by arranging asset classes as efficiently as possible,

often through the use of mean variance-based optimization (Reilly and Ebsworth, 2013).

According to Idzorek (2006) SAA is a set of long term target allocations to applicable investable

asset classes with the highest likelihood of meeting long term investment goals. It is the choice

of equities,bonds or alternative assets that an investor wishes to hold for the long run, usually

from 10 to 50 years (Karl et al. 2011).

An important consequence of the chosen asset allocation is its impact on overall performance

and risk over time. Empirical study by Ibbotson and Kaplan (2000) indicates that asset allocation

decision explain about 90% of the variability of return over time. However, given the high

correlations and investment losses among many asset classes during the turbulent markets of

2008-2009, some investors have since questioned the effectiveness of strategic asset allocation as

a longterm investment approach (see also Reilly and Ebsworth, 2013).

Other authors also commented that, the global financial crisis in 2008 caused investors to

question what went wrong with many of their portfolios, which were believed to be diversified

(Lee, 2011). In addition to this, Pfleiderer and Marsh (2013) found that, a big question for most

investors was what tactical adjustment they should make to their portfolio holdings given the

large losses in wealth they had suffered and the extreme market conditions they faced.

In lieu of this, Central Banks restrict the eligible investment universe to highly liquid

government bonds and instruments issued by international institutions, government sponsored

institutions and supranationals and then derive the strategic asset allocation for the whole

portfolio through an optimization exercise, Cardon and Coche (2004). However, Fisher and Lie

(2004) provided an altenative SAA framework for reserves that considers various assets more

than what is suggested by Coche and Cardon such as non government bonds, equities and

currency and which still provide assurance of sufficient liquidity for trade and intervention

requirements. Fisher and Lie contended that the typical asset allocation process by CBs is

overconstrained, leading to portfolio inefficiency.

To overcome this concern, they came up with an altenative SAA framework where the

investment universe was broadened to include non government bonds such as MBS, ABS and

corporate bonds. While currency and country allocation in the typical SAA are treated

identically, in an altenative SAA frame work, currency and country allocation can differ. Fisher

20

and Lie (2004) relaxed also the duration and credit constraints. In their demonstrated

framework, the source of efficiency gains comes from exploiting the correlation between;

currency and bond returns, full maturity yield curves across countries and non government

securities.

Against the point by Coche and Cardon that Central Banks derive SAA for the whole portfolio, a

horizontal separation of reserves is created through creation of liquidity and investment tranches

to give focused attention on liquidity and return objectives separately (IMF, 2005 a & 2005 b).

This is in line with Borio et al (2008) who asserts that there has been a growing tendency to

separate the reserve portfolio into tranches with different objectives. The growing tendency was

witnessed during a survey when Borio and others found that around two thirds of the Central

Banks responding to the survey had established two (or more) separate tranches.

In his article, Bernad (2009) described the challenges facing institutional investors and came out

with observations of new trends in the investor’s asset allocation behavior. According to the

paper, the financial crisis underlined the difficulty of predicting risk and returns and provoked

investor interest in new approaches to asset allocation such as pure risk based strategies and

allocation by risk factors. Also, the study found that low level of yields on fixed income markets

which are the bulk of Central Banks’ portfolios increasingly leads them to consider

diversification towards new currencies and asset classes, as well as more actively managing their

asset allocation.

The study recommended to Central Banks, currencies that benefit from long term appreciation

potential, RMB in particular as it is not over valued against other major currencies and it is

expected to benefit from increased internationalization. However, Liquidty and market depth for

RMB was cited as a challenge. Commodity linked currencies such as AUD and CAD and safe

haven currencies like CHF were also cited.

21

CHAPTER THREE

3.0 RESEARCH METHODOLOGY

According to Zikmund (2000), a research design is a master plan specifying methods and

procedures for collecting and analyzing required information. It is “…An action plan for getting

from here to there, where here is defined as the initial set of questions to be answered, and there

is some set of conclusions (answers) about these questions” (Yin, 1994). This chapter has five

main sections: research design, sampling plan and population, data sources and collection

techniques, data measurement and analysis instruments, and researcher reflexivity.

3.1 Research Design

This research involved a combination of a descriptive design and an exploratory survey method

through literature search. The descriptive study adopted a cross sectional approach in which all

sampled respondents were studied at one period of time (Churchill, 1996). As supported by

Bryman and Bell (2011, p. 55), researcher was interested to examine the patterns of association

on many cases at one point in time and collect a body of quantitative data in order to examine

one or several variables.The research also involved a pilot study to pre-test the questionnaire and

due to geographical proximity, a convenience sampling method was used to select BOT for the

pilot survey.

Similary, the research used an exploratory design through experience survey in which, a semi-

structured questionnaire was addressed to staff responsible for financial markets function. The

great strength of the survey as the primary data collecting approach is its versatility (Bloomberg,

Cooper, and Schindeler 2011, p.207).

3.2 Sampling plan and population

The targeted population for this research were Central Banks based in five countries in East

Africa region namely Tanzania, Uganda, Kenya, Rwanda and Burundi. From each Central Bank,

the Directorate of Financial Markets was used as a sampling frame to obtain the relevant

department which deals with foreign exchange reserves management, and in each department an

experienced officer was selected to fill in a questionnaire. This is in line with Kothari (1990)

who says, the optimum sample is one that fulfills the requirements of efficiency,

22

representativeness, reliability and flexibility. Table 1 below displays the sampled department for

each country in the East African region.

Table 1: Sampled departments and bank officers selected.

Country Central Bank Sampled department Position of the bank

officer(s) selected

Burundi BRB Foreign Banking Operations Manager

Kenya CBK Financial Markets Senior Officer

Rwanda BNR Financial Markets Manager

Tanzania BOT Foreign Markets Senior Financial

Analyst

Uganda BoU Financial Markets Portfolio Manager

Source: Field findings, August 2014

3.3 Data collection, sources and instruments

Data collection methods reflect the design of data collection instruments, and how data will be

collected and administered. In this case, the study combined both primary and secondary data. In

order to obtain primary information, a structured questionnaire attached in Appendix 2 was

deigned with a few open-ended questions ( unstructured questions) to allow managers/ officers in

the sampled departments to express their thoughts about the subject in question.

The questionnaires were designed to cover all the research concepts as they are described in the

research problem, objecives and questions. Additionally, the questionnaire was constructed

based on the literature reviewed as well as the knowledge from previous studies. Thus, the

questionnaire that was used in collecting data from EAC Central Banks is attached at the end of

this paper.

3.4 Data measurement and Analysis

According to Zikmund (2000), the choice of the statistical technique for data analysis depends on

the type of question to be answered (central tendency, or distribution of a variable), the number

of variables (single or multiple), and the scale of measurement (interval / ratio scaled –

parametric statistics or ordinal/nominal scaled - nonparametric statistics). Both descriptive and

quantitative tools of analysis were applied to the responses and an analysis of the relationship

23

between the global financial crisis and asset allocation through Chi- square and cross tabulation

tests were performed using Statistical Package for Social Scientists (SPSS 21).

However, since the data set is very small (only 5 Central Banks were surveyed), the statistical

analysis may not provide conclusive or meaningful results in all cases. Therefore, the

conclusions drawn (to address the questions and hypotheses listed in the previous section) are

not solely based on statistical analysis. A qualitative analysis of the survey responses, in addition

to the statistical analysis, has been used to form conclusions

3.5 Reseacher reflexevity

3.5.1 Ethical issues

Ethics is the study of the “right behavior” and addresses the question of how to conduct research

in a moral and responsible way (Blumberg, Cooper & Schindler, 2011). Thus ethics not only

addresses the question of how to use methodology in a proper way to conduct sound research,

but also addresses the question of how the available methodology may be used in the right way

(Sounders, Lewis and Thornhill, 2003). Other scholars ( for example, Marczyk, DeMatteo and

Festinger 2005) insist that, various ethical issues need to be taken into consideration when

conducting business research particulary if human subjects are involved. Sekaran, (2003) is of

the opinion that, ethical issues arise in each stage of the research process from problem

identification to the dissemination of reseach results. Besides, a number of ethical codes have

been developed to provide guidance when doing research. These research ethics include

protecting human participants (safety), such as respecting the respondents, doing no harm to the

respondents and selecting the respondents scientifically (Blumberg et al, 2011).

Other ethical considerations relate to informed consent, protection of anonymity, fabrication of

data, neglecting the limitation of the research, speculative interpretation of the results and

confidentiality of the information provided by respondents. In this context, the researcher

believes that the study poses a number of ethical issues that need to be dealt with during the

whole process of research. The first ethical issue relates to informed consent, which means that

prospective research participants must be fully informed about the procedures and risks involved

in the research and must give their consent to participate. Thus participants were informed in

advance of the overall purpose of the study and effects of their participation and this was

communicated in a written form and verbally through telephone.

24

Furthermore, with regard to confidentiality, the research is set to guarantee the participants that

identified information will not be made available to anyone who has no link with the study.

Importantly, the speculative interpretation of the results alongside fabrication of data will be

avoided and controlled to the maximum possible and this means all limitations as well as sources

accessed during the study are fully communicated and acknowledged.

3.5.2 Validity and Reliability

Many forms of validity are mentioned in the research literature, and the number grows as we

expand the concern for more scientific measurement (Blumberg et al. 2011). Once the data is

collected, it is necessary to determine the degree to which it is valid and reliable. Validity can be

defined as… “the extent to which a test measures what we actually wish to measure” (Blumberg

et al. 2011). Reliability, on the other hand, is the degree of consistency between multiple

measurements of a variable. In other words, are the variables or a set of variables consistent with

what they are intended to measure? Reliability differs from validity in that the former does not

relate to what should be measured, but instead to how it is measured (Blumberg et al. 2011).

Different methods are available for assessing validity and reliability. In this study several

methods were used to enhance the reliability and validity of the data. For example; as pointed in

section 3.1 above, this study will involve a pilot study to pre-test the questionnaire, as such it

reflects testing of validity and reliability of each question in capturing the information. Carrying

out a pilot survey helps to provide some conceptual clarification to the research design and can

detect problems in the design of a questionnaire (Zikmund, 2000). In this regard, a convenience

sampling method was used to select units for the pilot survey on the basis of geographic

proximity and ease of personal contact. The pilot survey was conducted through both personal

and telephone interviews with selected officers of EAC Central Banks. This is in line with

Blumberg et al. (2011:213); who suggests that an interviewer can answer questions about a

survey, probe for answers, use follow up questions and gather information by observation.

Commonly though, for almost all quantitative research, internal validity is usually weak

(Sounders et al. 2003).This is in line with the discussion made in the previous section in which, it

is argued that a cross-sectional approach was adopted, where in, it does not have the ability to

manipulate the independent variables. This lowers the credibility of the research and negatively

affects one of the preoccupations of quantitative research which is causality.

25

Although the research is meant only to investigate Central Banks currently in the five countries

in East Africa, one could argue that they are favored before other Central Banks in the region.

But, this research is only supposed to be representative for the chosen geographical areas hence

consider the external validity to be solid. Nevertheless, when data is collected, it is important to

determine the degree to which it is reliable and valid. Hence, the use of the quantitative

techniques makes replication perhaps easier to execute since it does not involve any situations of

interpretations by its subjects so that consistent results was taken into account as a measurement

of reliability.

3.5.3 Limitation of the study

Since the study is set to explore the extent to which the global financial crisis affected asset

allocation in East Africa CBs, only selected GFC indicators and asset allocation variables were

the focus of the investigation. Set in this context, financial market departments from each CB in

the region were selected. Notably however, time is always not sufficient and is one of the

challenges that have arisen during the execution of this study.

26

CHAPTER FOUR

4.0 EMPIRICAL SETTING

4.1 Introduction

This chapter is designed to provide a brief background information about the East African

Community (EAC), the region in which the study was conducted. The chapter highlights on a

general overview of the geographical and economic features including reflection on the level of

foreign exchange reserves held in each member country. Therefore, the chapter provides a brief

overview of the five countries involved in this study.

4.2 The East Africa Community

The East African Community is the regional inter-governmental organization grouping the

Republics of Burundi, Kenya, Rwanda, Tanzania, and Uganda with its headquarters in Arusha,

Tanzania. The Treaty for Establishment of the East African Community was signed on 30

November 1999 and came into force on 7 July 2000 following its ratification by the original

three partner states – Kenya, Tanzania and Uganda. The Republics of Rwanda and Burundi

acceded to the EAC Treaty on 18 June 2007 and became full Members of the Community with

effect from 1 July 2007.

4.2.1 Aims and Objectives

The EAC aims at widening and deepening co-operation among the Partner States in political,

economic and social fields for their mutual benefit. The target for the Monetary Union was

2012, leading ultimately to a Political Federation of the East African States. The protocol for the

EA Monetary Union was finally signed on 30th November 2013 by the five heads of state.

4.2.3 Region and People

The Community has a combined population of 125 million, land area of 1.85 million sq

kilometers and a combined GDP of $ 44 billion. The people of East Africa comprise many ethnic

groups, including the Bantus, Nilotes, Hamites and Cushites. The dominant religions are

Christianity and Islam. Other religions include Hinduism and traditional African religions.

27

Kiswahili, which is widely spoken, is a Bantu language with significant word borrowings from

Arabic, Persian and European languages. Many local languages are also spoken.

Table 2: East Africa Community at a glance

Member States Republique du Burundi; Republic of Rwanda;

Republic of Kenya; United Republic of

Tanzania; Republic of Uganda.

Heads of States: Burundi: President Pierre Nkurunziza

Rwanda: President Paul Kagame

Kenya: President Uhuru Kenyatta

Tanzania: President Jakaya Mrisho Kikwete

Uganda: President Yoweri Kaguta Museveni

Secretary General of the EAC Dr. Richard Sezibera

Main institutions of the EAC: Summit; Council of Ministers; East African

Court of Justice; East African Legislative

Assembly; Secretariat.

Surface area: 1.85 million sq. km

Population: 125 million

GDP: US$ 44 billion

Currency: Burundi: Francs

Kenya: Kenya Shilling (KSh)

Rwanda: Francs

Tanzania: Tanzania Shilling (TSh)

Uganda: Uganda Shilling (USh)

Principal religions: Christian and Islam

Climate: Climatic conditions Vary from tropical to

temperate, depending on

elevation. Two rainfall seasons: the long

rains, from late March to early

May, and the short rains, from late October to

28

early December.

Capital cities Burundi: Bujumbura

Rwanda: Kigali

Kenya: Nairobi

Tanzania: Dar es Salaam

Uganda: Kampala

Source: EAC information guide for Investor,2013

Official name East African Community (EAC)

Figure 1:The East African Map

Key:

The city where Central Bank is located

29

CHAPTER FIVE

5.0 RESEARCH FINDINGS, ANALYSIS AND DISCUSSIONS

5.1 Introduction

This chapter presents the key issues for the effects of GFC on asset allocation decision and how

East African CBs reacted in terms of asset classes, currency composition and maturity structure.

As part of this, data has been gathered, analysed and interpreted according to the methodologies

discussed in chapter three. Moreover, findings from the survey are linked with theoretical

background; in which different kinds of analysis have been applied so as to get a good overview

of the findings. The hypotheses that are linked with the main research aspects are also analysed.

Other relevant issues that have been investigated during the study are also reported.

5.2. Sample description

The survey was carried out in five Central Banks based in East African countries, namely:

Burundi, Kenya, Rwanda, Tanzania and Uganda. The Questionnaire was mailed to each Central

Bank in the Directorate of Financial Markets, particulary the department which deals with

foreign exchange reserves management and the response rate was 100 percent; and as a result, all

the questionnaires were usable in the analysis.

Consequently, the experienced officers who deal with strategic asset allocation were interviewed.

The respondents surveyed were classified into two general categories of positions held namely:

Managers (60%) and Senior Financial Analysts (40%). This supports the view that officers with

experience in the department dealing with FX reserves were in a position to provide the relevant

informantion with regard to the subject matter (see also table1above).

5.3 East Africa Central Banks asset allocation before the crisis

Findings indicate that, before the global financial crisis of 2008, about 60% of East Africa

Central Banks had highly invested in government bonds, supranationals and short term deposits

each as depicted in table 3 below. However, the study revealed that, BOT had also large holdings

in agencies while BoU had less holdings in the same. Moreover, none of the EAC Central

Banks had investments in other asset classes like inflation and corporate protected bonds, MBS

or ABS, Equities and/or Commodities. This suggest that Central Banks were not ready to

30

assume more risk given their level of reserves, staff capabilities and existeing infrastructure and

this was highlighted by respondents during interviews.

Table 3: East Africa Central Banks investments before the crisis

Choice Total % response BR KE RW TZ UG

Government bonds 60

Supranationals 60

Deposits 60

Agencies (more) 20

Agencies (less) 20

Inflation protected bonds 0

Corporate protected bonds 0

MBS / ABS or Asian (excl. Japan) bonds 0

Equities and Commodities 0


5.4. Central Bank’s reaction to GFC

To examine the Central Banks’ reaction to GFC, difficulties encountered alongside decisions on

how assets were allocated are identified. Consequently, to determine the way Central Banks re-

allocated their reserves during the crisis, these difficulties and decisions are descriptively

analysed. Statistical as well as descriptive tools of analysis were used to determine the

relationship between effects of GFC and asset allocation.

5.4.1. General difficulties encountered by Central Banks

In this study the general difficulties encountered by CBs are defined and divided into the

following: Level and liquidity of reserves, increased needs for certain currencies, low interest

rates and concerns for specific asset classes. Several authors including Pringle and Carver

(2006); Mulder and Morahan (2013) indicate that Central Banks experienced these difficulities.

Each of the above difficulties were analyzed and the following hypothesis was articulated:

31

Hypothesis 1

The extent to which priorities of Central Bank’s objectives changed is related to the financial

crisis.

o The liquidity of reserves

From table 7, it is notable that each Central Bank experienced at least some and or multiple

difficulties over the period, which indicates the severity of the crisis. The findings revealed that,

80% of EAC Central Bank experienced difficulties with liquidity of reserves assets. Given the

crisis, experience has confirmed that many assets became increasingly illiquid (Minsky, 1986).

Based on the Chi-Square test statistic on the relationship between the liquidity problems and the

the priorities of the Central Banks’ reserves management objectives, the finding indicates that

there is significant relationship at 5% significance level (p-value = 0.171) as shown inTable 4 .

Although the relationship is substantiated, the EAC Central Banks did not change their

priorities of reserves management objectives. Instead, they withdrew their money from foreign

commercial banks, possibly because of concerns about the bank’s financial condition or because

they were worried about the bank runs. This is an indication that EAC Central Banks were

struggling to protect their capital from loss. Accordingly, Mulder and Morahan (2013) also

supports the idea that 50 percent of the 156 Central Banks surveyed across the IMF member

countries pulled back their commercial banks depoists.

Table 4: The liquidity of reserves and priorities of Central Banks Objectives

Value Df Asymp. Sig. (2-sided)

Pearson Chi-Square 1.875a 1 .171

Continuity Correctionb .052 1 .819

Likelihood Ratio 2.231 1 .135

Fisher's Exact Test

Linear-by-Linear Association 1.500 1 .221

o Low interest rate

The findings in table 7 revealed that, 60% of the respondents were affected by low interest rate

environment. The quantitative findings reveal that , only BRB and BOU were not affected by the

low rates. However, when reserves managers from the two Central Banks interviwed, they

pointed out that low rates affected their scope of generating returns and hence lower income

from foreign investments were realized. Notably, the low interest rate environment was a crucial

32

concern of the difficulties encountered because it significantly dwindled the earnings of EAC

Central Bank’s over the period. For instance, since the start of the crisis in 2008, yields on highly

rated Eurozone Sovereign paper and the US treasurt bill rates have declined significantly as

depicted in Figure 2a and 2b below:

Figure 2a: Eurozone Benchmark Curve Change

Figure 2b: US three months Treasury bill rates

Source: Bloomberg and Resource Center of the U.S Department of the Treasury

As depicted in table 5, when cross tabulated with priorities of reserves management objectives,

the Chi-Square test statistic at the 5% significance level (p-value=0.361) reveals significant

relationship between low interest rate challenge during the financial crisis and the change in the

priorities of reserves management objectives. However, low interest rates environment observed

33

during the crisis did not led to the change in the priorities of Central Banks’ reserves

management objectives, instead, the phenomenon lead to portfolio diversication where EAC

Central Banks concentrated their investments in products with a view to preserve capital. For

instance, BOT invested small part of its reserves in instrument issued by the Government of

China and Australia. On the other hand, the result from an interview with the Bank of Uganda

officials spells out that the Bank searched new counterparties with good credit ratings in other

regions apart from the G73 with a view that it would help to address the shrinking investment

room. The respondents added also that the BOU reserves, whether internally or externally

managed, were invested in instruments, which have remained largely unaffected by the crisis.

Like wise, CBK diversified their reserves and invested only on short term deposits, US-issued

Government securities, and in high quality private supra-national securities. Consequently the

Bank of Burundi decided to concentrate only with deposits with Central Banks and the B.I.S

while their neighbours, BNR authorised changes of the benchmark for one of their external fund

managers from 0 – 3 to 1 – 3 Year US Treasuries and authorized another fund manager to

include in its benchmark the Australian dollar. Both respondents disclosed that, despite the

changes, the Banks succeeded in preserving the capital and providing liquidity, while earning

low return than the previous period.

Table 5: Low interest rates and priorities of Central Banks objectives


Pearson Chi-Square .833a 1 .361

Continuity Correctionb .000 1 1.000

Likelihood Ratio 1.185 1 .276

Fisher's Exact Test

Linear-by-Linear Association .667 1 .414

N of Valid Cases 5

Source: Field findings, January 2015

3 G7 are the major advanced economies as reported by the International Monetary Fund: Canada, France, Germany, Italy, Japan,

the United Kingdom, and the United States

34

o The level of reserves

It is also worth noting that, the level of reserves was among the variables identified as a

challenge during the crisis. Findings in table 7 disclose that BRB and BoU were among the

Central Banks which cited the level of reserves as being of concern compared to the rest.

Possible explanations for this could be either an indication of the insufficient level of reserves to

cover the need for greater liquidity buffers and/or there was sufficient reserves which faced the

narrowed investment universe. Further examination of Chi Square statiscal tests at the 5%

significance level (p-value=0.709) indicates that, the level of reserves during crisis relates to

change in priorities of CBs reserves management objectives. However, these results do not

conform with responses from an interviewed respondents in which the study concludes that

there were no changes in priorities of reserves management objectives (See also table 6 below).

For example, respondents disclosed that reserves of the BOU fluctuated from October 2008

through mid 2011 largely due to interventions in the interbank foreign exchange market

performed to curb currency volatilities occasioned by the massive capital outflows observed

during the period. Nonetheless, this study found out that, the Bank embarked on daily purchases

of foreign exchange from the interbank market for reserve build up. The survey with BRB

officials, also disclosed that they have had low reserves and were worried in case of the need for

liquidity buffer and the Bank was prepared to set up policies for export promotion in order to

increase the level of foreign exchange reserves and improve business environment to attract

foreign investors. This observation confirms that BRB had insufficient level of reserves to cover

the need for greater liquidity buffers and yet did not consider change in the priorities of reserves

management objectives.

Table 6: Level of reseves and priorities of Central Banks objectives

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square .139a 1 .709

Continuity Correctionb .000 1 1.000

Likelihood Ratio .138 1 .710

Fisher's Exact Test

Linear-by-Linear Association .111 1 .739

N of Valid Cases 5

35

Table 7: Difficulties experienced in managing reserves during crisis era

Choice

Total %

response BRB CBK BNR BOT BoU

Level of reserves 40

Liquidity of reserves 80

Increased needs for certain Currencies’ as

part of the currency composition 20

Concerns related to specific asset classes 20

Low interest rates 60

Did not experience any difficulties 0


5.4.2 Central Banks decisions in addressing crisis difficulties

As indicated in table 8 below, the Central Banks decisions in responding to crisis were identified

as immediate change in asset allocation, strengthening risk management system, use of reserves

to provide liquidity and seeking other sources of liquidity. After classifying decisions in the

aforementioned categories, the results indicate that, more than half (80%) of the respondents

changed their asset allocation whereas strenghthening risk management system and seeking other

sources of liquidity decision categories both accounted for 20% each. This is to say that,

difficulties experienced by East African CBs led them to re think effective management of their

reserves.

Table 8: Actions taken by Central Banks to address crisis difficulties

Choice

Total %

response BR KE RW TZ UG

Immediate changes to asset allocation 80

Strengthening risk management system 20

Use of reserves to provide liquidity or for

market intervention 0

Seeking other sources of liquidity 20

36


Furthermore, what triggered the change in SAA was grouped in six classes as shown in table 9.

The results show that credit risk aversion (shown by rating downgrades and flight to quality) was

the most highly ranked category that triggered a change in asset re allocation. This could be a

result of the observed increased currency volatility and rating downgrades. This is supported by

the view that, the episodes of high volatilities were associated with the global financial crisis,

Eurozone sovereign credit crisis and recent slowdown in the global economy and as shown in

Figure 3 the market has witnessed volatility spikes of above 15% during late 2008 and early

2009 compared to 10% observed prior to the crisis, implying the risk of holding Euro has

increased (Bloomberg, November 2014).

Though the study indicates that credit risk was a major concern for all EAC Central Banks, table

8 show that only one CB responded to strengthen to its risk management framework. When

interviewed the rest of the Central Banks with respect to credit risk management; the BoU

disclosed that, it focused on relatively default free instruments, set conservative minimum

counterparty credit quality standards and aimed at efficient diversification. The BOU also was

searching for new counterparties with good credit ratings in other regions apart from the G7

because the existing counterparties had been affected by the rating downgrades. Like wise,

BNRs’ minimum acceptable threshold (minimum rating for sovereign holdings which was at

least AA, with at least 50% with AAA rating) was no longer giving the possibility and

flexibility of investing in securities issued by sovereign entities which were downgraded, instead,

37

BNR’s Board revised the minimum sovereign credit rating from AA to A-, opening the

possibility of tapping into the lower pool of AA to A- rated, without taking risk.

While Bank of Burundi responded to have limited the number of counterparties to those who are

well rated by rating agencies, the CBK was preparing to enhance Risk Management function so

as to avert any financial crisis occurring in their econmy, however the challenge was the fact

that critical functions in the CBK have their own Risk Management Units, and Risk Management

operations are normally not harmonized.

Table 9: Triggers of the asset re allocation

Choice

Total %


Increased volatility 60

Rating downgrades 80

Government bonds that were downgraded 40

Change in reserves management objectives

priorities 40

Reputational risk 40

Low interest rates 40

Risk reduction/flight to quality in reserves 100

Change in risk appetite 40


38

Figure 3: EUR Volatility

Source: Bloomberg, November 2014

In addition, results in table 10 revealed that, during asset re allocation, all Central Banks reduced

deposits in foreign commercial banks and held more safe assets such as treasuries. The greater

allocation of reserves were directed to investment grade assets. Suprisingly BoU responded that

it did not change its SAA but reduced its deposits held in commercial banks. On the other hand,

BOT reduced the bonds that were downgraded and those without government guarantee for the

sake of protecting foreign exchange reserves values.

Table 10: Assets that were reduced as a share of total reserves during crisis

Choice

Total %

response BRB CBK BNR BOT BoU

Deposits with commercial banks 100

Bonds without government guarantee 20

Agency, ABS, MBS, Corporate bonds 0

Government bonds that were downgraded 20

Non-Core European government bonds 0

Equity 0


Apart from what triggered asset re allocation and other factors considered in changing SAA,

Central Banks surveyed (with an exception of BoU) had chosen to re distribute their reserves to

39

investment grade including holding more safe assets4. This indicates that the East Africa Central

Banks were concerned with credit risk and wanted to reduce risk of capital loss by allocating

reserves to quality and safe assets. Table 11 below summarises what policy actions were

considered appropriate during the financial crisis.

Table 11: Reserves allocation and reserves management objectives


Greater allocation of reserves to investment

grade assets 40

Holding more safe assets 60

Change in reserves management objectives

priorities 0

Source: Field findings, august 2014

Although Central Banks changed their asset allocation given the crisis difficulties, findings

indicate that all respondents (100%) did not change priorities of the reserves management

objectives. Therefore the sample findings has established that, during the financial crisis, the

priorities of reserves management objectives remained the same i.e capital preservation ,

provision of liquidity and generation of reasonable returns.

5.4.3 Yields search and priorities of Central Banks objectives

Although respondents were searching for yields and diversification, priorities of investment

objectives did not change as highlighted ealier. All investments were performed to make sure

that capital is preserved, liquidity requirements are met and finally the minimum return is earned.

When analysed in terms of investment horizon, findings in table 12 display that, investment

horizon for working capital tranche remained less than a year for all Central Banks. However,

the investment horizon for liquidity tranche varies. Whereas the investment horizon for BRB is

the same for the two tranches due to low level of reserves holdings, the investment horizon for

Uganda is on the other side longer than other Central Banks and these are mandated to external

fund managers. Infact, none of the Central Banks are investing longer than 5 years. This also

goes in hand with IMF (2005 a&b) that, a horizontal separation of reseves is created through the

4 Safe assets are those issued by Sovereign Governments

40

creation of liquidity and investment tranches to give focused attention on liquidity and return

objectives.

To summarize, statistical tests supports the hypothesis that there is relationship between financial

crisis and change in the priorities of reserves management objectives, however findings from the

sample results does not support this conclusion.

Table 12: Central Banks investment horizon for various tranches

CB

Working capital Liquidity Investment

Less than 1 year 1 to 3 years 2 to 5 years

BRB

CBK

BNR

BOT

BoU


5.4.4 Characteristics of financial assets and Central Banks investments

Hypothesis 2

Asset re allocaction is related with change in risk tolerance

The relationship between the global financial crisis of 2008 and types of asset classes that the

Central Banks considered to be eligible was determined in this study. The findings depicted in

table 16 below confirm that a great proportion (80%) of East Africa Central Banks allocated

their financial assets with issues and issuers that are rated between the range of A to AAA while

increasing fixed income investments or holding more of US treasuries (60%) and agencies

(60%) . In this regard, none of the respondents allocated part of their reserves to Gold or

Equities. This phenomenon suggests that selection of eligible asset classes is an important part of

the asset allocation process. In other words, the choice of eligible asset classes was influenced

by investment objectives, risk-return considerations, reputational risk issues and staff

capabilities. To analyse the Central Banks risk torelance, relationship between the effects of

rating downgrades and decision taken to invest in AAA was tested. The observed relationship

41

was significant at 5% level (χ2=10.00; 8 d.f. or p=0.265). Furthermore, when tested, the

relationship between low interest rates and the decision to increase fixed income securities, the

relationship was also significant at 5% level (χ2=2.222; 2 d.f. or p=0.329). Therefore asset re

allocaction is related to the change in risk tolerance. Table 13 summarises this Chi-Square Tests

results. Besides, Cardon and Coche, (2004) also assert that, Strategic Asset Allocation reflects

the institution’s overall risk tolerance and investment objectives and constraints over the

planning horizon.

Table 13: Low interest rates and the decision to increase fixed income securities


Pearson Chi-Square 2.222a 2 .329Likelihood Ratio 2.911 2 .233Linear-by-Linear Association

1.500 1 .221


Moreover, some reasons were established with respect to why the Bank of Uganda invested more

in assets rated as BBB. These include: the BoU’s part of reserves were and are managed by

External fund managers who were given relaxed credit risk limits given their investment

expertise. Actually, this observation is supported by the view that, External fund managers

remained in the investment grade limits to minimize default risk.

The research findings have indicated also that, none of the EAC Central Banks introduced new

external fund managers during the crisis. For instance, where as the Bank of Tanzania was in an

engagement process, the process was halted in 2008 due to the deepening of the crisis.

Nevertheless, those Central Banks which had already engaged external fund managers such as

BNR, BoU and CBK continued holding these managers by tightening their reserves management

guidelines and/or benchmarks to make sure they invests in names and assets that are in

investment grade. This is an indication that these Central Banks became more risk averse.

42

Table 14: East African Central Bank investments during crisis


Increased fixed income securities 60

Increased holding of US treasuries 60

Increased holding of US Agencies 60

Invested more in AAA 80

Invested more in AA 60

Invested more in A 80

Invested more in BBB 20

Introduced external fund managers 0

Allocated part of reserves to Gold 0

Allocated part of reserves to Equites 0


5.4.5 Adjustments in the currency allocation

As indicated in table 15 below, during the crisis 80% of respondents made significant adjustment

in the currency allocation. Results show that, BRB introduced allocation to AUD. Likewise,

CBK introduced allocation to AUD, CAD and RMB. The Bank of Uganda also, introduced

allocation to AUD and CAD. On the other hand, BOT increased allocation to USD, decreased

allocation to GBP and EUR and introduced allocation to AUD and RMB. This is confirmed by

Bernad (2009) who pointed out that, the low level of yields on fixed income markets led Central

Banks to consider diversification towards new currencies and asset classes. The currency

allocation decision was done with a cautious stance. That is, the Central Banks invested in high

quality names and assets denominated in that new currencies which are issued by the Sovereign

Government and in their risk torelance limits.

Table 15 : Adjustments in the currency allocation

Choice Total %


In recent years, the bank has made significant

adjustment in the currency allocation of reserves 80

The bank increased allocation to USD 20

43

The bank increased allocation to EUR, GBP &

JPY 0

The bank decreased allocation to EUR and

GBP 20

The bank introduced allocation to AUD 80

The bank introduced allocation to CAD 40

The bank introduced allocation to NOK and

NZD 0

The bank introduced allocation to RMB 40

5.4.6 Diversification and search for yield

Like wise, the study was set to determine whether EAC Central Banks invested in countries

other than developed economies like emerging markets and advanced economies given the

search for yield and diversification. Results indicate that some Central Banks invested in

emerging markets as depicted in table 16. For example, BOT and CBK invested in money

markets and fixed income in Chinese Yuan. Yet, BOT had also invested in Australian dollar

during the crisis of 2008. The total investment ranged between 0-10% of the total portfolio in the

emerging markets. Moreover, in an interview conducted with some respondents particulary the

BoU and BRB both confirmed to have never invested in emerging market before and during

GFC.

44

Table 16: Allocation to Emerging Markets during the crisis

Choice

Total %


The bank invested in emerging markets 40

The bank invested in fixed income and money

markets of emerging markets 40

The bank invested between 0-10 % of the total

portfolio in emerging markets 20

The bank invested between >10-50 % of the total


The bank invested above 50 % of the total



5.4.7 Determinants of currency composition

In this study, the main considerations in determining reserves currency composition in Central

Banks are defined and classified into eight criteria as summarized in table 17. Several authors

including Fisher and Lie (2004) have indicated that these factors are the most common

determinats for currency composition. In lieu of this, most East Africa Central Banks, consider

the composition of; government’s short foreign currency liabilities, central government short

term external liability, currency of imports, reporting currency and depth & liquidity of the

underlying asset markets as the most important factors in determing currency composition.

Suprisingly, the Central Bank of Kenya did not respond on this matter.

Table 17: Considerations for currency composition

Choice

Total %


Currency composition of Central Bank’s overall

foreign currency liabilities 40

Currency composition of central government’s short

term foreign currency liabilities 60

Currency composition of the economy’s (short term) 80

45

external liability

The currency composition of imports 80

The currency/currencies to which the reporting

currency is pegged or closely related 80

Depth and liquidity of the underlying asset markets 80


Notwithstanding, during GFC, more than a half (60%) of the East African Central Banks

considered adjusting reserves currency composition ( see also table 15) and took action to shift

composition between traditional reserves currencies (USD,GBP,EUR & JPY)5 and other

advanced country currencies such as CHF, AUD,CAD, NZD, DKK, NOK and SEK.

Meanwhile, the findings disclosed that during the global financial crisis of 2008, almost 80% of

respondents were considering investing in other emerging markets out of which, 40% invested

their reserves in Chinese Renminbi.

Other Central Banks in the region, except BNR whose response was missed, had considered

investments of reserves in other emerging market currencies. These results are also depicted in

table 18. One of the possible reasons for this paradox is that respondents consider holding

currencies other than traditional reserve currencies for diversification and yield enhancement.

Those Central Banks that invested in an emerging currency such as offshore Chinese RMB

indicated that the offshore market is very accessible with less restrictions when compared to the

onshore market. Therefore, those who did not invest in an emerging currencies opinioned that

convertibility, low credit rating information and high volatility of the currency were the main

hindrace.

5 Traditional reserves assets are referred to those contained in the IMF busket.

46

Table 18: Other currencies considered for reserves investments during the crisis


Shifting composition between traditional reserve

currencies ( those included in the SDR

busket,USD,GBP,EUR,JPY) 60

Holding other advanced country currencies, such

as CHF,AUD,CAD, NZD,DKK,NOK,SEK 60

Holding Chinese Renminbi (onshore/offshore) 40

Holding other emerging market currencies 80


5.4.8 The Policy and Benchmark review

Table 19 below indicates frequency of investment policy review. The BNR and BoU reviews

policy each year while BRB has no specific time to review the policy. On the other hand, BOT

reviews once in every two years while a review for CBK is indicated to be an ongoing activity.

However during the global financial crisis, BoU, CBK and BRB followed the same frequency of

review while Tanzania and Rwanda did not follow the same frequency. That is, the Bank of

Tanzania reviewed its policy annually based on the crisis trend instead of once per 2 years and

BNR reviewed semi annually instead of annually.

Table 19: Frequency of investment policy review

CB Yearly no specific time Ongoing once in 2 years

Burundi

Kenya

Rwanda

Tanzania

Uganda Source: Field findings, August 2014

Notwithstanding, the Central Banks also revised benchmarks and altered the portfolio duration.

Given the market volatility, default risk and rating downgrades, all respondents confirmed

changing their portfolio duration. Besides, findings in an open ended question indicate that, all

respondents stopped allocation to a specific country. For example, the BOT, BoU and BNR

reduced Euro zone specific country allocation while CBK and BRB did not. All the same,

47

findings indicate that, the Bank of Uganda liquidated Italian and Spanish bonds when they fell

out of the investment grade and thus dropped an allocation to those countries. Importantly, the

Bank of Tanzania used to invest in Ireland, Italy and Spanish bonds and when they fell out of the

investment grade the Bank reduced allocation to those countries. The Bank of Rwanda also

reduced allocation to Italy due to credit risk aversion.

Additionally, there are certain countries which were among the constituents of public indices

before the crisis. Yet, during the GFC, some countries like France were downgraded by the

rating agencies and were dropped from the benchmark. In view of this, the Bank of Tanzania

was forced to treat French assets as spread products after being dropped out of its investable

universe. This suggest that, any country, counterparty and or assets that did not meet Central

Banks limits in their reserves management guidelines was consequently dropped out of

investment universe.

Results as depicted in table 20 show that, BRB, BOT and BNR revised their benchmarks during

the crisis while CBK and BoU did not. BRB and BOT revised benchmark by altering portfolio

duration and currency composition. BNR revised benchmark by altering portfolio duration only.

As stipulated earlier, during GFC, deposits with commercial banks were reduced. Given the

downgrades, many commercial banks in which Central Banks were investing, found themselves

out of the rating limits and hence their credit worthiness was questionable. Since capital

preservation is key among the Central Banks objectives, CBs pulled out all deposits with

commercial banks and invested fund in the Central Banks like BIS, BDF and FED.

Table 20: Central Banks benchmark revision

Choice

Total %


The bank revised the benchmarks 60

The benchmark was revised by altering portfolio

duration 60

The benchmark was revised by altering portfolio

asset allocation 40

The benchmark was revised by altering currency

composition 40


48

5.4.9 Risk management

As presented in Table 21, 60% of respondents revised their risk management framework. While

CBK and BOT measured risk using VAR and TE during and after the crisis; BRB and BNR

measured risk using TE and VAR only and respectively during and after the crisis. This informs

that, East African Central Banks managed risks of their reserves actively.

To summarize, statistical tests supports the hypothesis that there is relationship between asset re

allocaction and change in risk tolerance, however findings from the sample results does not

support this conclusion because in all scenerios, the study found that Central banks’ ability and

willingness to tolerate adverse outcomes on their reserves did not change.

Table 21: Risk measurement

Choice

Total %


The bank revised risk management framework 60

The bank measured risk using VAR during and

after crisis 60

The bank measured risk using TE during and

after crisis 60


5.5 East Africa Central Banks investments after the crisis

Hypothesis 3

There is no difference between asset allocaction prior to and after the financial crisis

5.5.1 Allowable Asset classes in the Banks’ Investment policy

Given that the crisis has stabilised6, table 22 below details assets in which EAC Central Banks

were allowed to deal with. Among the assets approved in their investment policies include;

government bonds, supranationals, agencies, inflation protected bonds, covered bonds,

corporates, MBS/ABS, and Asian (ex Japan) bonds. Results show that, approved policy allows

all respondents (100%) to invest in government bonds. With the exception of BRB, other EAC 6 This is the time believed that the financial stability in the eurozone had improved significantly-from mid-2012 to late 2013, due to successful fiscal consolidation and implementation of structural reforms in the countries being most at risk and various policy measures taken by EU leaders and the ECB. However not that,as of May 2014 only two countries (Greece and Cyprus) still needed help from third parties.

49

Central Banks (80%) are allowed to invest in both supranationals and agencies each. While BoU

has policies that allow investment in inflation protected bonds, covered bonds, corporates,

MBS/ABS, Asian (ex Japan) bonds and equities, CBK is allowed to invest in inflation protected

bonds and BNR is allowed to invest in covered bonds. Therefore, commodities were not

approved in the policy of any East Africa Central Bank. In addition, as pointed out in table 18

above, since 2013, the Bank of Uganda started investing in other emerging markets such as

Brazil and Mexico through their external fund managers.

Table 22: Asset classes allowed for investment after the crisis


Sovereign bonds/government bonds 100

Supranational 80

Agencies 80

Inflation protected bonds 40

Covered bonds 40

Corporates 20

MBS / ABS 20

Asian (excl. Japan) bonds 20

Equities 20

Commodities 0


From table 23, analysis indicate that before crisis, 60% of Central Banks had government bonds

as their eligible instruments, however, after crisis all Central Banks allocated large percentage to

government bonds in their investment policy.

50

Table 23: Comparison of eligible Government bonds before and after crisis

After the crisis the bank allowed greater percentage allocation to Government

bonds in the investment policy

Total

Yes

Burundi 1 1

Tanzania 1 1

Kenya 1 1

Uganda 1 1

Rwanda 1 1

Total 5 5


Like wise, before crisis, 60% of the respondents invested large percentage in supranational

bonds and after crisis 80% of Central Banks increased allocation to supranational bonds in their

list of eligible instruments stipulated in the investment policy. These results are summarized in

table 24.

51

Table 24: Comparison of Supranational bonds before and after crisis

After crisis the Bank allowed large percentage allocation to Supranational

bonds in the investment policy

Total

Yes No

Burundi 0 1 1

Tanzania 1 0 1

Kenya 1 0 1

Uganda 1 0 1

Rwanda 1 0 1


As summarized in table 25, before crisis, only 20% of Central Banks invested large percentage

of their reserves in agencies. However, after crisis 80% of Central Banks allowed large

percentage allocation to agencies in their investment policy.

52

Table 25: Comparison of eligible Agency bonds before and after crisis


Results in table 26 indicate that all East Africa Central Banks had no investment in inflation

protected bonds, however, after crisis only CBK and BoU investment policies approved

investment in that asset class.

Table 26: Comparison of approved inflation protected bonds before and after crisis

53

The bank had no investment in Inflation protected

bonds before crisis

Total

Yes

Burundi 1 1

Tanzania 1 1

Kenya 1 1

Uganda 1 1

Rwanda 1 1


To analyse whether there is no difference between asset allocation prior and after the financial

crisis, the Chi square statistical test was conducted and revealed that relationship was significant

at 5% level (χ2= 5.00; 4 d.f. or p=0.287).

To summarize, statistical tests supports the hypothesis that there is no difference between asset

allocaction prior to and after the financial crisis, however findings from the survey results does

not support this conclusion because in all tests, the study found that Central banks’ allocated

large percentage of their reserves to Government bonds and Agencies and even approved

Inflation protected bonds which was not in the list of eligible instrument before the crisis.

5.5.2 Future changes in the asset allocation

Given the changes and the level of uncertainities in the global economy, table 27 depicts that,

60% of the respondents would like to implement changes in asset allocation in the next 2 years.

The Central Bank of Burundi desires to add new allocation to: major currencies i.e USD,

EUR,GBP & JPY. For diversification and yield enhancement, the bank is now considering

investment in certain currencies like AUD, CAD and Yuan. Likewise, the Central Bank of

Kenya wishe to increase allocation to CAD and AUD and decrease allocation to EUR.

Moreover, the Bank of Tanzania expects to add new allocation in advanced economies

government debt and emerging market debt. BOT also is seeking to increase allocation to; major

currencies (USD,EUR,GBP &JPY), government agencies and corporate agencies. Whereas the

Bank of Uganda responded that they are interested to increase allocation to major currencies

(USD,EUR,GBP &JPY) and decrease investment in CAD and AUD, the Bank of Rwanda was

surprisingly silent on this matter.

54

Table 27: The likely review of strategic asset allocation

Choice Total %


The bank wish to implement any changes in your asset

allocation in the next 24 months 60

The bank wishes to add new allocation in advanced

economies governemnt debt 20

The bank wishes to add new allocation in emerging markets

governemnt debt 20

The bank wishes to add new allocation to major currencies

(USD,EUR,GBP, JPY) 20

The bank wishes to invest in AUD or CAD 20

The bank wishes to increase allocation to government

agencies 20

The bank wishes to increase allocation to AUD & CAD

The bank wishes to increase allocation to corporate agencies 20

The bank wishes to increase allocation to major currencies

(USD,EUR,GBP, JPY) 40

The bank wishes to decrease investment in AUD or CAD 20

The bank wishes to decrease investment in EUR


55

CHAPTER SIX CONCLUSION AND RECOMMENDATIONS The study investigated the effects of the global financial crisis of 2008 on East Africa Central

Bank’s asset allocation. Specifically the study evaluated the extent to which priorities of CB’s

objectives (safety, liquidity and return) changed during the financial crisis, assesed whether it

reflects the change in risk torelance and examined the asset allocation prior to and after the

financial crisis. A semi structured questionnaire was used as a data collection tool where the

effects of the global financial crisis of 2008 on Central Bank’s asset allocation were identified

from the five Central Banks which constitute to the East Africa Community. The Statistical

Package for Social Scientists was used for analysis.

The findings, analysis and discussions led to the conclusion that, given the effects of the global

financial crisis of 2008; all East African Central Banks had choosen to reallocate their financial

assets without changing priorities of the key reserves management objectives of preserving

capital, providing liquidity and generating reasonable return. The East Africa Central Banks

limited their risk exposures by: stopping from placing money markets deposits with commercial

banks, increasing investment in fixed income, placing all redeemed money markets in major

Central Banks like BIS, FED and BDF and changing the currency composition. These findings

found to be supported by Cardon and Coche (2004) who said that, “a general requirement might

be that changes in the investment universe should be made in connection with a review of

strategic asset allocation”.

Furthermore, Central Banks changed their currency composition towards non traditional reserves

currencies such as Australian Dollar, Canadian Dollar and Chinese Renmnbi as a way to

diversify from developed economies and to enhance return. In a way, this suggested that Central

Banks risk torelance changed and statistical tests supported the hypothesis that there was

relationship between asset re allocaction and change in risk tolerance at 5% significant level.

Nonetheless, since 2008, the share of alternative currencies in reserves assets has tripled. The

IMF has started showing these currencies especially CAD and AUD separately in the Cofer

release, in recognition of the growing role they play in foreign exchange reserves management

(IMF Cofer data, 2014). The justification for the diversification of Central Bank’s reserve

holdings has been due to a reassessment of the relative credit risk of holding Euro zone and US

government papers as well as a reassessment of the same at their current historically low yield

56

levels. The enthusiasm for RMB exposure, in particular, is linked to the desire to diversify out of

developed markets and also due to the recognition that China is indeed rising as a global

economic power. In lieu of this, findings from the sample results did not support the stated

hypothesis because Central banks justified that their ability and willingness to tolerate adverse

outcomes on their reserves did not change.

When examining the asset allocation prior to and after the crisis, the findings from the research

showed that, prior to the crisis, Central Banks reserves were held in traditional reserves

currencies of USD, EUR and GBP where eligible assets were deposits and highly liquid

government bonds. However after the crisis, Central Banks diversified their currency

composition portfolios to non traditional currencies; increased percentage allocation to

government bonds, supranationals, agencies and approved Inflation protected bonds and

MBS/ABS as eligible instruments. These findings was found to be inline with Bernad (2009)

that, the low level of yields on fixed income markets which still represents the bulk of Central

Bank’s portfolios led them (CBs) to consider diversification towards new currencies and asset

classes as well as more actively managing their asset allocation. Statistical tests did not support

these findings because at 5% significant level, it was found that, there was no difference between

asset allocaction prior to and after the financial crisis.

Experiences across EAC Central Banks in dealing with the challenges of the global environment

have been quite similar, although some have perhaps been more courageous than others in terms

of the asset allocation decisions and avenues in which they have chosen to diversify. What is

needed is caution because , Central Bank reserve managers are custodians of public funds, and

at all times have to be guided by prudent risk management strategies. The global financial

market environment has created many new risks and challenges for reserve managers. It has been

observed from literature that shortcomings in risk management were the major factors

underlying the crisis. With the increase in the level of reserves and changes in the global

economy there is much higher level of interest and scrutiny in how these reserves are managed.

Hence this paper recommends the following;

Improve risk management practices and procedures

The global financial crisis highlighted the need for improved risk management procedures.

According to Hull (2007), risk management is nowadays considered a key activity for all

57

companies. Many of the disastrous losses7 of 2008 due to Lehman's bankruptcy filling would

have been avoided if good risk management practices were in place. It is worth to re call that, in

2008, Lehman faced an unprecedented loss as a result of having held on to large positions in

subprime and other lower-rated mortgage tranches. Like wise, the improvements in the risk

management practices and procedures should consider the correlations of currencies and a

vigilant approach to risks emanating from unseen avenues.

Strengthen risk management culture

Sustainable risk and control framework must be one of the top agenda items for senior

management. Strengthening risk roles and responsibilities, enhancing communication and

training, and reinforcing accountability are the key initiatives to strengthen risk culture. To

improve risk-management practices, controls have to be set at various levels, exposure limit and

deviations should be monitored closely.

Ensure robust portfolio and risk management infrastructure is in place

Following unpredictable market conditions and inclusion of some non-traditional assets in

investment portfolios for various Central Banks, it has meant that reserves management has

become more complex. This calls upon building capacity to Central Bank reserves managers

and investing in sophisticated IT and risk management systems that are able to handle complex

transactions which CBs are now engaging in.

Over reliance on single credit rating models should be avoided

Literature indicates that, during the global financial crisis era, the banks had similar risk

management strategies, used the same models which led them to the same conclusion and similar

decisions. Single risk methodologies and over reliance on specific models should be avoided. In

order to address this risk of having similar decision, Central Bank should use more stress testing

and scenario analysis to help measure and manage risks.

7 According to Bloomberg news, the Dow Jones closed down just over 500 points, several money funds and institutional cash funds had significant exposure to Lehman and fell below $1 per share, following losses on their holdings of Lehman assets and close to 100 hedge funds who used Lehman as their prime broker were also affected.

58

Credit risk management systems in East Africa Central Banks are still heavily reliant on external

credit rating agencies such as the Fitch, S&P and Moody’s as the key instrument for assessing

credit risk. It is recommended for Central Banks to monitor and report Credit Default Swap

(CDS) spreads of the countries they invest in to asses the threat of sovereign default and

complement the use of ratings from rating agencies. According to the survey by Mulder and

Morahan (2013), 80 percent of asset reallocation decisions by many Central Banks were

triggered by rating downgrades and that credit risk management systems remain tightly linked to

credit ratings. To supplement ratings from credit rating companies, an in-house credit risk system

may need to be established. However, it requires heavy investment in human, information

technology and financial resources.

Diversification of assets

While it is not possible to predict when global financial markets will return to a higher interest

rate environment, some financial analysts expect interest rates to remain at historically low levels

for an extended period. Therefore, East Africa Central Banks should weigh the opportunity cost

associated with investing in low risk, low yielding strategies against a higher probability of

preserving capital to invest in a higher interest rate environment. Meanwhile in a world of

deteriorating credit, currencies and non-traditional countries with high credit rating stand out as

potential reserve investment alternatives. However, any increase in the range of allowable

instruments and markets that includes non-traditional investment areas should be done within the

permissible risk parameters and tolerance limits.

59

REFERENCES

Aryeetey, E., and Ackah, C. (2011); The Global Financial Crisis and African economies:

Impacts and Transmission Channels. African Development Review, Vol. 23, No. 4, pp.407- 420

Bakker A. F. P., and Ingmar R.Y., (2007). Central Bank reserves management: New trends from

liquidity to return

Bernad., T (2009), how to define your asset allocation, Amundi; Balanced and allocation

solution: accessed during September 2013.

BIS papers No 58 pg 19, 2009: Managing foreign exchange reserves in the crisis and after.

Blumberg B., Cooper, R.D., and Schindler, P.S., (2011); Business Research Methods, McGraw-

Hill, 3rd ed

Borio C., Galati G., and Heath A., (2008), FX reserve management: trends and challenges, BIS

Papers No.40.

Bryman.A, and Bell.E., (2011). Business Research Methods (Vol. 3rd ). New York: Oxford.

Cardon P. and Coche J., (2004), "Strategic asset allocation for foreign exchange reserves", ECB

publication on Reserve Management for Central Bank Foreign Reserves, May 2004, pp. 13-28,

Churchil, G. (1996). Basic Marketing Research; (3rd ed.), Dryden Press.

Davis J., Roger A., Patterson A., (2011), Asset allocation in a low yield and volatile

environment: The Vanguard Group.

ElSayeed and Hegazi (2013); Contagion in MENA countries during the 2008 Financial

meltdown. Journal of Business and Economic Research. Vol. 11 No. 7

Fisher, S. J. and Lie, Min C. (2004), "Asset allocation for Central Banks: optimally combining

liquidity, duration, currency and non-government risk", ECB publication on Reserve

Management for Central Bank Foreign Reserves, May 2004.

Hull, J.C. (2007), Risk Management and Financial Institutions, Pearson Education, New Edition

Ibbotson, R. and P. D. Kaplan (2000), “Does Asset Allocation Policy Explain 40, 90, or 100 Per

Cent of Performance?”, Financial Analysts Journal, 56 (1), pp. 26-33.

Idzorek. T. (2006), “Strategic asset allocation and commodities”, Ibbotson Associates

investment management services, pg 3.

IMF (2005a), Guidelines for Foreign Exchange Reserve Management (Part I: Summary of

Country Practices, April 2005). The document can be found on the IMF website.

IMF (2005b), Guidelines for Foreign Exchange ReserveManagement (Part II: Country Case

Studies, April 2005). The document can be found on the IMF website.

60

IMF (2013), Revised Guidelines for Foreign Exchange Reserve Management.

IMF 2009, Foreign Exchange Reserve Adequacy in East African Community Countries.

IMF 2009, Spillover effects and the East African Community: Explaining the slowdown and

recovery.

IMF, 2009, The Global Financial Crisis and Adjustment to Shocks in Kenya, Tanzania, and

Uganda: A Balance Sheet Analysis Perspective

IMF, 2010, Procyclicality in Central Bank Reserve Management: Evidence from the Crisis

IMF, 2011. "Long-Term Investors and Their Asset Allocation: Where Are They Now?" In:

IMF(ed.) Global Financial Stability Report.

Karl E, Stephane M and Thierry R, (2011), “white paper strategic asset allocation”, Lyxor asset

management research, issue # 6.

Kothari, C.K. (1990). Research methodology: Methods and Techniques, 3rd edition, Wiley

Eastern Limited: New Delhi

Lee W, Risk-Based Asset Allocation: A New Answer to An Old Question? Published in The

Journal of Portfolio Management Spring 2011.

Louis M., (2010) Recent trend in reseves management

Lunogelo B, Mbilinyi A and Hangi M, (2009), The grobal financial crisis and Tanzania: Effects

and policy responses.

Marczyk, G., DeMatteo, D., & Festinger, V. (2005). Essentials of Research Design and

Methodology. New Jersey:: John Wiley & Sons, Inc.

Morahan, A and Mulder , 2013, “Survey Of Reserve Managers: Lessons From The Crisis,” IMF

Working Paper WP/13/99.

Nakaso,H 2013, Financial Crisis and the Central Banks: Impact of financial crisis on Central

Bank function.

Nichlas E., and Niklas R., The Financial Crisis effects on asset allocation: A study regarding

the individuals in Umeå financial behaviour in response to the financial crisis of 2008.

Njoroge J, (2009), Effects of the global financial crisis on corporate social responsibility in

multinational companies in Kenya.

Pfleiderer, P and Marsh, T (2013), Flight to Quality and Asset allocation in financial Crisis,

Financial Analyst journal Vol 69. Number 4

Reilly D & Ebsworth W (2013), Strategic Asset Allocation: An effective risk management

approach for long term investors.

61

Saunders, M., Lewis, P., & Thornhill, A. (2003). Research Methods for Business Students (3rd

ed.). Harlow, UK: Prentice Hall.

Sekaran, U. (2003). Research methods for business: A skill-building approach. New York: John

Wiley and Sons.

Shane O., (2013), Asset allocation is critical for investors, AMP CAPITAL

Srimany A.K., Gayen S., and Ranjeev, (2011), Effect of recent financial cirisis on optimal asset

allocation of foreign exchange resereves, RBI working paper series.

Warwick, J.M and Stoeckel,A (2009), The global financial crisis: causes and consequences, a

working paper in International economics

Were M., (2012), Central Bank’s Response to Economic Crises from a Developing African

Economy Perspective: Lessons from Kenya’s Experience

World Gold council (2011), Central Bank diversification strategies:Rebalancing from the dollar

and Euro, accessed on 19th July 2013.

Yin, R. ( 1994 ). Case Study Research: Design and Methods (2nd ed.). Califonia: Sage

Publications, Inc.

Zikmund, W. (2000). Business Research Methods (6th ed.). Harcourt College, Fort worth: The

Dryden Press

62

ANNEX 1

Chart 1a: The Level of reserves for Advanced and Emerging economies including BRICs

Source: IMF Cofer data.

Chart 1b: The Lehman Brothers’ bankruptcy and Risk Premia

63

ANNEX 2

A QUESTIONNAIRE UNDER MEFMI FELLOWSHIP PROGRAMME

I, Kabula Irene Mulihano is a MEFMI Fellow who is conducting a research on “The effect of The Global Financial Crisis on Asset allocation in East African Central Banks”. This is done in partial fulfillment of the Fellowship programme. Therefore, the questionnaire aims to gain insight into how East African Central Banks reacted to the global financial crisis to date. It also aims at understanding how reserve managers arrive(d) at their strategic asset allocation and how they operate their risk management frameworks in practice. Kindly mark/indicate (X) to the response/answer of your choice and you may select more than one response where appropriate.

Institution: Department: Kindly please click below in the sheet and fill in all sections: Section 1. Crisis Response

Section 2. Asset Allocation

Section 3. Currency Allocation

Please return this completed questionnaire to [email protected] Copy to [email protected] Copy to [email protected]

64

Section 1. Crisis Response

1. What type of difficulties did you experience in managing your reserves during the crises episodes of the (Please mark 'X')

a. Level of reservesb. Liquidity of reservesc. Increased needs for certain Currencies’ as part of the currency compositiond. Concerns related to specific asset classese. Did not experience any difficultiesf. Other

If other, please specify

2. What actions did you take to address those difficulties?(Please mark 'X')

a. Immediate changes to asset allocationb. Strengthening risk management systemc. Use of reserves to provide liquidity or for market interventiond. Seeking other sources of liquiditye. No measures takenf. Other


3 If you changed your asset allocation immediately,what types of assets did you actively reduce/ increase as a share of total reserves duringthe crisis(Please mark 'X')

a. Deposits with commercial banksb. Bonds without government guaranteec. Agency, ABS, MBS, Corporate bondsd. Government bonds that were downgradede. Non-Core European government bondsf. Equityg. Other


4. what triggered the asset re allocation?(Please mark 'X')

a. Increased volatilityb. Rating downgradesc. Government bonds that were downgradedd. Change in reserves management objectives prioritiese. Change in risk appetitef. Reputational risk g. Other


65

5. When you made immediate changes to asset allocation, did you consider any of the following?(Please mark 'X')

a. Risk reduction/flight to quality in reservesb. Rating downgradesc. Change in risk appetited. Reputational risk e. Other


6. If you have selected any response in question 5 above, what policy actions would you consider appropriate?(Please mark 'X')

a. Greater allocation of reserves to investment grade assetsb. Holding more safe assetsc. Change in reserves management objectives prioritiesd. Other


7. What are the main consideartions in determining your currency composition?(Please mark 'X')

a. Currency composition of central bank’s overall foreigncurrency liabilities

b. Currency composition of central government’s short termforeign currency liabilities

c. Currency composition of the economy’s (short term)external liabilities??

d. The currency composition of importse. The currency/currencies to which your currency is

pegged or closely relatedf. Depth and liquidity of the underlying asset marketsg. Depth and liquidity of the FX marketse. Other


8. Are you currently considering adjusting currency composition of your reserves ?(Please mark 'X')

a. Yesb. Noc. Other


9. If yes to above, what types of actions/currencies are you considering?(Please mark 'X')

a. Shifting composition between traditional reserve currencies ( those included in the SDR busket,USD,GBP,EUR,JPY)

b. Holding other advanced country currencies, such as CHF,AUD,CAD, NZD,DKK,NOK,SEK

c. Holding Chinese Renminbi (onshore/offshore)d. Holding other emerging market currenciese. Other


10. If you have not invested in emerging currencies, what have been the main reasons for not doing so?(Please mark 'X')

a. The currencies are not convertibleb. The currencies cannot be counted as ‘Official

Reserves’ in IMF datasetsc. The currencies are too volatile to invest currentlyd. The FX markets in these currencies are not deep and

liquid enoughe. Credit risk considerations (ratings that are too low)f. The underlying government bond markets are not

sufficiently liquidg. Quality of institutional framework, governance issuesh. Peers have not yet entered into these marketsi Other


66

Section 2. Asset Allocation

1. Given the search for yield and diversification of reserves during crisis period, how was your priority of inv(Please rank the objectives by marking 'X' in the selected numbers)

1 2 3a. Capital preservationb. Liquidityc. Maximising returnd. Other


2. What is your investment horizon per tranche?(Please mark 'X')

Liquidity Investment Stablea. Less than 1 yearb. 1 to 3 yearsc. 2 to 5 yearsd. Over 5 to 10 yearse. None

3. How often is your investment policy reviewed?(Please mark 'X')

a. Yearlyb. No specific timec. Ongoingd. Other


4. Did you follow the same frequency to review investment policy during the crisis?(Please mark 'X')

a. yesb. No C. Did not review the policy

5. If NO to above how often was your investment policy reviewed during the financial crisis?(Please mark 'X')

a. semi annualb. Yearlyc. Other


6. Did you revise your benchmarks?(Please mark 'X')

a. Yesb. No

7. If yes, please specify in what manner?(Please mark 'X')

a. Altered durationb. Altered asset allocationc. Altered currency compositiond. Other


TRANCHE

67

8. Were some countries added or dropped due to rating downgrades?(Please mark 'X')

a. Increased existing Euro zone country allocationIf so, please specify county or countries

b. Reduced existing Euro zone country allocationIf so, please specify county or countries

c. Introduced new country allocation If so, please specify county or countries

d. Stopped a particular country allocation If so, please specify country or countries

e. OtherIf other, please specify

9. Did the financial crisis cause the bank to revise or change risk management framework?(Please mark 'X')

a. Yesb. Noc. Other


10. If yes to above, how did you measure risk( before, during and after crisis) in your portfolios?(Please mark 'X')

During crisis After crisisa. VaRb. Tracking errorc. Other


11. Which of the following products were you investing more or less before crisis? (Please mark 'X')

lessa. Government bondsb. Supranationalsc. Sovereign eurobondsd. Agenciese. Inflation protected bonds f. Corporatesg. MBS / ABSh Asian (excl. Japan) bondsi Equitiesj Commoditiesk Other


12. Did you alter your strategic asset allocation during financial crisis period?(Please mark 'X')

a. Yesb. No

13. If yes, how exactly(Please mark 'X')

a. Increased fixed income b. Reduced fixed income c. Increased equitiesd. Reduced equitiese. Increased MBS/ABSf. Reduced MBS/ABSg. Introduced other asset classh. Introduced external manageri. Other


14. Have you altered your allocations to US Treasuries and Agencies during the crisis?(Please mark 'X')

a. Yesb. No

15. If yes, how exactly? (Please mark 'X')

Increased decreaseda. Holdings of US Treasuries b. Holding of Agenciesc. Duration of US Treasuriesd. Duration of Agenciese. Other


16. Did you change your asset allocation due to effects of downgrades?(Please mark 'X')

a. Yesb. No

17. If yes to above, how did the bank limited its investments in the below selected rating grades?(Please mark 'X')

Own more Own lessa. AAA/Aaab. AA/Aac. Ad. BBB/Baaf. BB/Bag. Bh CCC/Caa or belowi Other


18. Have you changed, increased or decreased your allocation to gold?(Please mark 'X')

a. Increased allocation b. Decreased allocationc. No allocationd. Allocation stayed the samee. Other


19. In searching for yield and diversification, did you invest in emerging markets?(Please mark 'X')

a. Yesb. Noc. Other


68

20. If yes to above, what was the structure of your investments?(Please mark 'X')

In local currency In hard currencya. Money markets onlyb. Fixed income onlyc. Money markets and fixed incomed. Other


21. What percentage of your total portfolio was invested in emerging market debt?(Please mark 'X')

a. 0 - 10%b. <10% - 20%c. <20% - 30%d. <30% - 40%e. <40% - 50%f. >50%

22. If you answered 'yes' in question 19 above, which country did you invest in?(Please mark 'X')

a. Koreab. Indonesiac. Singapored. Malaysiae. China offshoref. China onshore

g. OtherIf other, please specify

23. Did you invest in equities after the crisis?(Please mark 'X')

a. Yesb. No

24. If yes, is it?(Please mark 'X')

a. Passiveb. Active

25. Do you invest in commodities? (Please mark 'X')

a. Yesb. No

26. If yes, what percentage of your total portfolio are invested in commodities?(Please mark 'X')

a. 0 - 10%b. >10%

69

Given that the crisis has stabilized 27. Which of the following asset classes are approved in the policy?(Please mark 'X')

a. Sovereign bonds/government bondsb. Supranationalsc. Agenciesd. Inflation protected bondse. Covered bondsf Corporatesg MBS / ABSh Asian (excl. Japan) bondsi Equitiesj Commoditiesk Other


28. Do you wish to implement any changes in your asset allocation in the next 24 months?(Please mark 'X')

a. Yesb. No

29. If yes, how exactly ?(Please mark 'X')

Add new Remove Increase Decreasea. Advanced economies government bondsb. Emerging market government bonds Xc. Government Agenciesd. Corporate Agenciese. Advanced economy corporate bondsf. Emerging economy corporate bondsg. Emerging market debth. Major currencies (USD, EUR, GBP, JPY)i. Invest in AUD or CADj. Other currenciesk. Equitiesl. Goldm. Other


30. Which of the selected products would you want to own more and/or less? (Please mark 'X')

Own more Own lessa. Sovereign bonds/government bonds b. Supranationalsc. Agenciesd. Inflation protected bondse. Covered bondsf. Corporatesg. MBS / ABSh. Asian (excl. Japan) bondsi Equitiesj Commoditiesk Other


70

Section 3. Currency Allocation

1. Have you made significant adjustments in the currency allocation of your reserves in recent y(Please mark 'X')

a. Yesb. No

2. If yes, please specify(Please mark 'X')

a. Increased existing currency allocationIf so, please specify currency or currencies

b. Reduced existing currency allocationIf so, please specify currency or currencies

c. Introduced new currency If so, please specify currency or currencies

d. Stopped a particular currency allocation If so, please specify currency or currencies

3. If yes to having added new currency, please answer:(Please mark 'X')

Increase DecreaseDollarEuroYenPoundCAD dollarAUD dollarNOKRMBRUBOther European exposureOther Asian exposure

4. Going forward, how do you expect your currency allocation will be adjusted?(Please mark 'X')

Increase DecreaseDollarEuroYenPoundCAD dollarAUD dollarNOKRMBRUBOther European exposureOther Asian exposure

5. Please describe your attitude to RMB(Please mark 'X')

a. Invested, or consider investingIf so, please specify perceived benefits

b. Not considering investingIf so, please specify major concerns

6. Please describe your attitude to AUD(Please mark 'X')



7. Please describe your attitude to CAD(Please mark 'X')



8. How do you obtain currency exposure in your portfolio?(Please mark 'X')

a. Bondsb. Deposits

c. Currency outlayd. Other


Kabula Irene Mulihano Bank of Tanzania Mentor: …mefmi.org/mefmifellows/wp-content/uploads/2016/10/The...1 Macroeconomic and Financial Management Institute of Eastern and Southern

Documents