MEFMI

Macroeconomic & Financial Management Institute of Eastern and Southern Africa

ADEQUACY OF FOREIGN EXCHANGE RESERVES:

THE CASE FOR KENYA

By

Ray Charles Musau

Central Bank of Kenya

Mentor

Mr. Dean D’Sa, CFA

Crown Agents Investment Management

A Technical Paper to be Submitted in Partial Fulfilment of the Award of MEFMI

Fellowship

September 2009

2

TABLE OF CONTENTS Topic Page

ABSTRACT ………………………………………………………………... 4 1. INTRODUCTION……………………….………………….……………... 6

1.1 Background………………………….……............................…….. 6

1.2 Definition of Foreign Exchange Reserves and their Sources …… 7

1.2.1 Definition of Reserves ……………………………………… 7 1.2.2 Sources of Reserves ………………………………………… 8

1.3 Increasing Levels of Reserves …………………………………….. 9

1.4 Definition of Reserves Management ……………………………… 11

1.5 Importance of Reserves Management ……………………………. 13

1.6 Justification of the Study ………………………………………….. 15

1.7 Objective of the Study ……………………………………………… 17

2. ADEQUACY OF OFFICIAL INTERNATIONAL RESERVES: LITERATURE SURVEY ………………………………………………….. 19

2.1 Why Hold International Reserves? ……………………………………. 19 2.1.1 Defend Pegged Foreign Exchange Rates …………………….. 19 2.1.2 Forestall Currency Speculative Attacks ……………………… 20 2.1.3 Meet Transactions Demand for Reserves …………………….. 21 2.1.4 Hard Currency Areas versus Soft Currency Areas …………… 21 2.1.5 Oil Economies versus Non-oil Economies …………………… 22 2.1.6 Politically Vulnerable Economies and Politically Correct Economies …………………………………………… 22 2.1.7 Government Spending & Cost of Borrowing ………………… 23 2.1.8 Access to Capital Markets ……………………………………. 24 2.1.9 Support for Exchange Rates ………………………………….. 24 2.1.10 Reserves and the Confidence Factor ………………………….. 25 2.1.11 Seasonality and Foreign Exchange Adequacy ………………... 25 2.1.12 Reserves and Credit Ratings ………………………………….. 26

3

2.2 Adequacy of Official International Reserves ……………………… 27

2.2.1 Accounts for Shocks …………………………………………. 30 2.2.2 Impact on Other Economies of Holding Surplus Reserves ….. 31 2.2.3 Opportunity Cost of Holding Reserves ………………………. 32

2.3 Reserve Adequacy Indicators ……………………………………….. 33

2.3.1 Months of Import Cover ………………………………………. 33 2.3.2 Reserve Levels and Debt ………………………………………. 34 2.3.3 Reserve Levels and Capital Flows …………………………….. 35 2.3.4 The Guidotti Rule on Foreign Exchange Adequacy …………... 36 2.3.5 Ratio of Short Term External Liabilities to Reserves …………. 37 2.3.6 Ratios of reserves to base money ……………………………… 38

2.4 Importance of Stress Tests …………………………………………… 39

2.5 Reserves Build-Up in Africa: Trends and Specific Motivation ……. 40

2.5.1 Trends and motivation for reserve build-up in Africa ……………. 41 2.5.2 Sources and composition of African foreign exchange reserves …. 43

2.6 Foreign Currency Reserves in Kenya ……………………………….. 43

2.6.1 Foreign Exchange Reserves Holding: Kenya’s Experience ……… 45 3. RESEARCH DESIGN & METHODOLOGY ………………………………… 48

3.1 Model of Integral Measure of Reserve Adequacy ……………………. 48

3.2 Estimated Reserve Adequacy from Jan 1999 to Date ………………… 54

3.3 Overall Appraisal of Reserves Adequacy in Kenya ………………….. 61

4. CONCLUSIONS AND POLICY RECOMMENDATIONS …………………. 66

4.1 Strategic Asset Allocation Process …………………………………….. 67

5. BIBLIOGRAPHY……………………………………...………………………… 68

4

ABSTRACT

The recent global financial crisis has clearly highlighted the importance of an adequate level of

reserves for reducing the costs associated with a foreign exchange liquidity shortage. The issue of

reserve adequacy has been broadly discussed for a number of years and this trend is likely to

continue. Conceptually reserve adequacy is the level of reserves that ensures smooth balance of

payments and macroeconomic adjustment in unpredictably changing economic environment, e.g.

external price shocks, reversals in short-term foreign capital flows. However, there is no common

approach for estimation of reserve benchmark level.

Given the inevitability of increasing internationalization, and the ever changing range of financial

instruments, the expected levels of capital flows into and out of a country have become

particularly important in determining foreign exchange reserve adequacy.

The most versatile approach proposed by Alan Greenspan several years ago was to identify all

balance of payments risks and to develop stochastic tests (stress-tests) measuring balance of

payments losses with a given probability. However, the implementation of this approach is

extremely cumbersome and thus countries continue to use more conventional indicators of

reserve adequacy such as ratios of reserves to imports, reserves to money aggregates and reserves

to measures of external debt. Essentially all these indicators are rules of thumb with certain

economic interpretation.

Until recently the most widely spread indicator of reserve adequacy was reserves expressed in

months of imports of goods and services. In this context the level of reserves covering three

months of imports was deemed as the most appropriate. Reserves equivalent to three months

import cover was considered sufficient for adjusting imports without shocks to the economy. But

5

as the Asian financial crisis of 1998 and the current global crisis proved, this indicator was

insufficient to avoid Balance of Payments problems and should therefore be augmented with

other criteria that take account of changes in demand. This measure of reserve adequacy may

actually look good at times when demand for imports collapses but turns out to be inappropriate

in times of financial stress when the reserves are required most.

Following the increased internationalization of financial markets, especially through the opening

of capital accounts, concern for holding adequate international reserves has transcended current

account considerations. In some emerging markets where the confidence of resident investors in

the domestic currency has not been particularly high or financial markets are underdeveloped,

risks of resident capital flight have been high. For such countries important indicators of reserve

adequacy are ratios of reserves to base money or other money aggregates. These indicators are

also relevant for countries with hard exchange rate arrangements, especially a currency board.

Market watchers, rating agencies and countries are nowadays concerned about their respective

net liability or short-term debt positions as well. As such, external debt servicing needs to also

explain international reserve holdings among countries. This has become much more important

given that the amount of international reserve holdings by a country is a major consideration in

the country’s sovereign credit rating (i.e. a measure of the financial capacity and willingness that

a country has to pay external debts). Therefore, adequacy of international reserves should also be

measured in terms of a country’s short-term foreign currency denominated debt.

This study aims to analyse the indicators for reserves adequacy and their application in the

determination of optimal reserve adequacy for Kenya.

6

1. INTRODUCTION

1.1 Background

As the scope of globalization increases and countries’ economies become increasingly open,

there has been increased apprehension among central bankers regarding the adequacy of official

international reserves that they hold. The International Monetary Fund (IMF) whose

responsibility it is to stabilize the world financial system and to whom member countries that are

hit by financial crises turn for bailouts exhibits an equal amount of concern. In its guidelines to

central bankers about the management of official international reserves, the IMF recommends

stress-testing official international reserves as a basis of building up adequate international

reserve levels (IMF, 2003).

Economic integration is a positive development in the global financial markets, but it has

undesirable effects such as increasing the speed at which financial crises spread beyond the

originating countries due to inter-linkages in the financial system. To cushion against such

adverse consequences, inadequate international reserve holdings among the affected countries

need to be checked. Fischer (2001) could not agree more with Pringle and Carver (2003) and

Naameh (2003) when quoted by Williams (2003) arguing that:

“The growth of international capital flows has prompted a rethinking of the way we assess

the adequacy of reserves… In an era when crises are more likely to arise from the capital

rather than the current account, it makes more sense to argue - as a first approximation -

that countries need reserves sufficient to cover their short-term debt. Indeed, the ratio of

short term external debt to reserves is the best crisis indicator, and the IMF now uses it as

a basic indicator of reserve adequacy.”

7

Additionally, Horst Kohler (2002), the former managing director of the IMF contends that more

flexible exchange rates and stronger net international reserves are crucial in a countries’ ability to

contain external shocks. While most of the Far East countries that were hit by financial crises in

the late 1990s are yet to fully liberalize their foreign exchange regimes, they have responded to

the late 1990s crisis by building up huge official international reserves. The build up of these

international reserves was motivated by the need to forestall a repeat of such crises by building

what was considered to be a credible amount of reserves that would give little room for

speculative currency attacks.

1.2 Definition of Foreign Exchange Reserves and their Sources

1.2.1 Definition of Reserves

Foreign Exchange Reserves are official public sector foreign assets that are readily available to

and controlled by the authorities for meeting a defined range of objectives for a country or

monetary union.

Reserves include gold, Special Drawing Rights (SDR), the reserve position in the IMF, and

foreign exchange. The foreign exchange component of reserves includes currency plus deposits

with monetary authorities and banks plus securities (foreign government securities, equity, bonds

and notes, money market instruments and derivatives).

8

1.2.2 Sources of Reserves

Reserve accumulation is the outcome of explicit decisions by the monetary authorities, rather

than a result of exogenous events such as commodity prices, debt forgiveness, or external factors

such as foreign investors’ appetite for domestic assets. Sources of reserves can be defined using

the following standard Balance of Payments (BoP) identities:

CA + KFA = ΔRES

Where CA is the current account balance, KFA the capital and financial account balance, and

ΔRES change in reserves. Net errors and omissions are generally added on the left hand side to

account for statistical discrepancies.

Further, the current account balance may be defined as:

CA = �GSA+ ��IA+ ��TA

Where:

GSA is the balance on goods and services, IA the balance on income and TA the balance on

transfers.

The capital and financial account balance is given by:

KFA = �KA+ ��FA

Where KA is the capital account balance, which includes debt forgiveness; FA the financial

account, which is equal to the sum of FDI plus portfolio investment and other investments.

9

Evidence indicates that the origins of reserve accumulation differ by country. In Latin America, a

persistent current account deficit was balanced for most of last decade (1990s) by a current

account surplus. Larger capital account surpluses helped some countries such as Brazil and

Venezuela to accumulate reserves. Since the 1997 crisis, East Asia has run capital account

deficits and continuous current account surpluses, except for China that maintained twin

surpluses (UN-DESA 2007).

For African countries, most of the movement has been in the GSA (imports and exports) and with

increasing importance in the TA account due to, among other things, worker remittances from

Diaspora. Kenya is no exception to the African norm and has mostly relied on the GSA given that

her economy is driven by exports from the tourism, agricultural and horticultural sectors. The

export of manufactured goods to countries in the surrounding region has also boosted this sector.

Kenya has remained heavily reliant on energy, manufacturing and agricultural input imports

which have continuously surpassed the supply of foreign exchange from the export sector

resulting in a current account deficit. Kenya’s economy therefore remains very sensitive to CA

developments but since economic and political liberalization in 1994, the capital, financial and

transfers accounts have become equally important.

1.3 Increasing Levels of Reserves

Without doubt, global reserves have increased rapidly over the past decade, principally as a

means of combating financial market turbulence. Between 2001 and the end of 2004, global

foreign exchange reserves grew by over US$1,600 billion, reflecting reserve accumulation by

emerging market economies in Asia. In a survey on “How Countries Manage Reserve”, Pringle

10

and Carver (2003), note that between 1975 and 2001, just after the move to generalized floating

exchange rates by major industrial countries, official holdings of reserve assets as measured by

the IMF increased by more than ten fold (or about 10.2% per year).

The source of foreign exchange accumulation has also shifted over the past two decades. The

pattern in the 1990s when aggregate current account balances for the emerging world as a whole

were negative was for capital inflows to be large and foreign exchange accumulation to be based

on these inflows. Over the period 2000-2004 the accumulation of foreign exchange reserves in

emerging markets was instead mostly through the generation of current account surpluses.

Reserves built from current account surpluses tend to be less volatile than those built by either

borrowing or short-term capital flows, as was the case in the 1990s.

Globalization has led to an increased demand for outward foreign direct investment, as investors

seek competitive markets as well as markets which offer a vast consumer base. This is good

business sense for the investor but can result in sudden capital outflows with little expectation of

short-term repatriation of profits and dividends when investors’ attitudes and risk tolerance

change.

Since independence in 1963, Kenya instituted exchange controls and has mostly run a current

account deficit from that period to date. During this time the CA gap was supported by donor

flows from the IMF, World Bank and bilateral flows from various governments. However Kenya

liberalized her economy in 1994, and since then, the capital and financial accounts have

registered tremendous growth covering the CA deficit and allowing for growth in reserves from

USD 500 million in 1996 to USD 2,600 million in March 2009. Foreign Direct Investment and

government privatisation proceeds have also played their part in contributing to this growth.

11

Since 2003, the Transfers account has also registered impressive growth with Kenyans from the

Diaspora sending money home for family support and investments in real estate and the stock

market. It is important to note that since 1997, Kenya has not received much in the form of

support from multilateral agencies due to governance issues which has also led to bilateral donors

reducing or cutting off their support altogether.

1.4 Definition of Reserves Management

As defined in the IMF Reserves Management guidelines, reserve management is a process that

ensures that adequate official public sector foreign assets are readily available to and controlled

by the authorities for meeting a defined range of objectives for a country or union. In this context,

a reserve management entity is normally made responsible for the management of reserves and

associated risks.

Typically, official foreign exchange reserves are held in support of a range of objectives which

include:

To support and maintain confidence in the policies for monetary and exchange rate

management including the capacity to intervene in support of the national or union

currency;

To limit external vulnerability by maintaining foreign currency liquidity to absorb shocks

during times of crisis or when access to borrowing is curtailed and in doing so;

o provide a level of confidence to markets that a country can meet its external

obligations;

12

o demonstrate the backing of domestic currency by external assets;

o assist the government in meeting its foreign exchange needs and external debt

obligations; and,

o maintain a reserve for national disasters or emergencies.

Reserve management adequacy should therefore seek to ensure that ample foreign exchange

reserves are available for meeting the above defined range of objectives.

African countries have accumulated substantial foreign currency reserves in recent years, mostly

from higher commodity exports as well as aid flows. In the context of macroeconomic

stabilization, which remains at the forefront of national economic policymaking and aid

conditionality, these countries are induced to hold reserves to allow monetary authorities to

intervene in markets to influence the exchange rate and inflation. Adequate reserves would also

allow the country to borrow from abroad and to hedge against instability and uncertainty of

external capital flows as has been the case for Ghana in the recent past. Kenya also intends to

issue Euro paper as soon as the global economy recovers from the current crisis.

However, reserve accumulation can have high economic and social costs, including a high

opportunity cost emanating from low returns on reserve assets, losses due to reserve currency

depreciation, and forgone gains from investment and social expenditure that could be financed by

these reserves. Therefore, central banks need to have a better understanding of the determinants

and economic costs of reserve accumulation and to design optimal reserve management strategies

to minimize these costs.

13

1.5 Importance of Reserves Management

Sound reserve management practices are important because they can increase a country's or

region's overall resilience to shocks. Through their interaction with financial markets, reserve

managers gain access to valuable information that keeps policy makers informed of market

developments and views on potential threats. The importance of sound practices has also been

highlighted by experiences where weak or risky reserve management practices have restricted the

ability of the authorities to respond effectively to financial crises, which may have accentuated

the severity of these crises. Moreover, weak or risky reserve management practices can also have

significant financial and reputational costs. Several countries, for example Russia, have incurred

large losses that have had direct, or indirect, fiscal consequences. Accordingly, appropriate

portfolio management policies concerning the currency composition, choice of investment

instruments, and acceptable duration of the reserves portfolio, and which reflect a country's

specific policy settings and circumstances, serve to ensure that assets are safeguarded, readily

available and support market confidence.

Reserve management should therefore seek to ensure that:

(i) adequate foreign exchange reserves are available for meeting a defined range of

objectives;

(ii) liquidity, market, and credit risks are controlled in a prudent manner; and

(iii) subject to liquidity and other risk constraints, reasonable earnings are generated over

the medium to long term on the funds invested.

As highlighted by Dr. Marion V. Williams (2005), Governor, Central Bank of Barbados foreign

exchange reserves are characterized primarily as a last resort stock of foreign currency for the

14

financing of foreign currency outflows. Reserves help to maintain confidence in the currency, and

allow central banks to intervene in the market in order to influence the exchange rate. They also

permit central banks to limit the vulnerability of the country to external shocks, give confidence

to the public and reassure credit rating agencies and international financial institutions about the

soundness of the economy.

Traditionally, the control and management of foreign exchange reserves has been the preserve of

central banks. However, the accumulation of foreign exchange reserves depends not on the

central bank directly, but on the structure and vibrancy of the economy, the split between the

share in GDP of the traded and non-traded sectors, the level and rate of capital inflows and

outflows, and on the attractiveness of returns offered in other currencies. Reserve accumulation

can be built from strong performance of the economy and from current and capital account

surpluses, or it can be supplemented by external borrowing.

To effectively manage foreign exchange reserves, policy makers therefore need to understand the

major determinants of their countries’ reserves in a globalized world. This is essential in

determining the optimal reserves level that provides them with necessary security at minimum

cost. In this context, some slowdown in the rate of reserve accumulation is likely to be justifiable

for commodity-rich countries that need to finance high-yield domestic investments instead of

locking up the reserves in low-yield foreign assets. Designing a successful reserve management

system requires important institutional and policy reforms at the national and regional levels.

These include policies to enhance domestic demand and regional trade, improved exchange rate

management and above all a new regional or global reserve system.

15

The composition of African reserves highlights high exposure of reserve holders to global

financial risks. Over the last few years, more than 95 per cent of African non-gold reserves were

held in foreign exchange (mainly the US dollar) plus deposits with monetary authorities and

banks and securities (US/foreign government securities, equity, bonds and notes, money markets,

derivatives). Thus the value of African reserves can change with fluctuations in the reserve

currency (especially the US dollar) or wider global financial market fluctuations. The safest

reserve asset, treasury bills, pays the lowest rates of return. Again this makes efficient reserve

management a top priority for reserve holders.

Kenya’s reserves are entirely held in short term cash deposits as the law previously prohibited

investment in instruments with tenures in excess of one year. Upon amendment of the Central

Bank Act in 2007, the CBK has now started investing in fixed income to diversify its portfolio

and manage risk.

1.6 Justification of the Study

In commenting on the accumulation of foreign exchange reserves in recent years, former

Chairman of the US Federal Reserve Board Alan Greenspan in an article “Current Policy Issues

in Reserve Management” (1999), observed that while the stock of foreign reserves held by

industrial countries has increased over time, those increases have not kept pace with the dramatic

increases in foreign exchange trading or gross financial flows. Thus in a relative sense, the

effective stock of foreign exchange reserves held by industrial countries has declined. This

suggests that despite the high level of global foreign exchange reserves, reserve growth is likely

to continue. This is based on the argument that as transactions become more international, there

16

will be a greater need for higher levels of foreign exchange reserves. What is very clear is there

has already been, and there is likely to be, an acceleration in international transactions over the

next decade. However if former Chairman Greenspan is correct when he said the level of reserves

relative to transactions is falling, and that there is likely to be continued efforts to increase foreign

exchange reserves in the years ahead.

This suggests that any new measurement systems for determining foreign exchange adequacy

will need to include not only some of the traditional factors, such as equivalence of three months

of imports, debt service ratios, reserves relative to debt amortization, and estimates of current and

capital account balances but also the size, nature and volatility of international transactions

relative to reserves. This suggests also that future formulae will not be as simple as the 3-months

import rule or the Guidotti/Calvo Rule but will become increasingly complex so as to assess the

cumulative effect of all the indicators. One way to do this would be to create a calculation based

on all the above indicators weighted by the countries’ reserves sensitivity to the respective

effects.

The Central Bank of Kenya Act stipulates that the minimum level of international reserve holding

by the Bank is a minimum of four months’ worth of imports averaged over the preceding three

years. Going by global trends, emerging and developing countries have continued to build up

their respective international reserve holdings according to the IMF. Considering the post

financial crisis responses among emerging countries in the Far East where they have built official

international reserves to levels deemed adequate to forestall financial crises as well as dealing

decisively with such crises in the event that they recurred, Kenya’s official reserve holdings

currently at 3.4 months cover are apparently inadequate. Emerging markets and developed

17

countries consider reserves adequate at the level of twenty-two and eighteen months of import

cover, respectively.

In view of the foregoing pertinent questions can be raised concerning Kenya’s experience with

international reserve holdings:

Is the four months of import cover in the form of official unencumbered international

reserves held with the Central Bank of Kenya adequate in view of the depletion effects

likely to arise from unanticipated shocks?

1.7 Objective of the Study

The purpose of this study is to determine the adequacy of official international reserves held by

the CBK. This would be done by defining and assessing the various indicators of Kenya’s foreign

exchange exposure and thereafter determining their cumulative effect to arrive at the adequate

level of official international reserves.

Upon introducing the paper in section 1 whereby the sources of reserves and the importance of

reserves management are highlighted, section 2 provides a selected review of the literature on the

adequacy of official international reserves among countries. Three main questions are dealt with

in this case. These are:

Why do countries choose to hold international reserves?

What are the commonly used reserve adequacy indicators?

18

What are the trends and motivation for reserves build-up in Africa and Kenya in

particular?

What level of international reserves would be considered adequate?

The analytical framework for determining the adequacy of official international reserves in

Kenya is then formulated in section 3 specifying a reserve demand equation. Estimation and

discussion of results is also presented in Section 3 and consistent with these results, Section 4

concludes with policy recommendations that would guide the development of policy regarding

the management of official international reserves in Kenya. Ultimately, the framework could also

be used to assess the adequacy of reserves in other MEFMI member countries that have similar

economies as that of Kenya.

19

2. ADEQUACY OF OFFICIAL INTERNATIONAL RESERVES: LITERATURE

SURVEY

2.1 Why Hold International Reserves?

Though traditionally, international reserves were used to intervene in the foreign exchange

markets, in recent times the frequency of intervention by central banks in developed currency

markets has waned. Few central banks in developed countries have actively intervened in the past

decade, since for the most part, financial markets in developed countries are sufficiently deep to

absorb and manage most shocks. During the current severe global crisis though, central banks

such as the Swiss National Bank and the Reserve Bank of Australia have had to intervene in the

foreign exchange markets. However, intervention aimed at providing foreign exchange liquidity

still frequently occurs in emerging market economies such as Tanzania and Uganda, but

intervention in developed countries, if and when it takes place, is often limited to ensuring

orderly conditions for the movement in exchange rates.

2.1.1 Defend Pegged Foreign Exchange Rates

Historically, international reserves were held to help defend the pegged foreign exchange rates.

However, 25 years down the line, following the general floating of the major currencies in 1975,

international reserves holdings among countries have risen ten fold; an average of 10.2% per

year. By the end of 2001, developing countries held 62% of the total foreign exchange reserves

(Williams, 2003).

20

Central banks in fixed exchange rate regimes find it necessary to hold higher levels of foreign

exchange than their counterparts in floating rate regimes where the exchange rate is not

sacrosanct.

Kenya pursued a fixed exchange rate under the Exchange Control Act which was repealed in

1995. Thereafter it has pursued a free floating regime where demand and supply of foreign

exchange determines the price of the Kenya shilling to major currencies.

2.1.2 Forestall Currency Speculative Attacks

Following the Asian financial crisis, countries such as Taiwan, Malaysia, South Korea and

Indonesia have built up their respective international reserve holdings in order to forestall

repetition of similar financial crises in the future. Feldstein (2002) points out that China, Taiwan

and Korea do not worry much about speculative attacks because they have built up large amounts

of foreign exchange reserves to serve as a deterrent. The amounts involved are on the higher side

of US$150 billion, US$100 billion and US$100 billion, respectively. Gin and McDonald (2003)

have also documented international reserve holdings among the Far East countries and among

other developing countries confirming the trends.

Since opening up the foreign exchange market, Kenya has not experienced a direct attack on its

currency by speculators; however sudden reversals of foreign exchange flows have been

witnessed during that time. When the IMF withdrew its support to the government in July 1997

the CBK had to intervene and sell foreign exchange to the tune of USD 150m, a substantial figure

during a time when the interbank market was very shallow and only turned over USD 10m daily.

21

2.1.3 Meet Transactions Demand for Reserves

International reserves are required to meet the day-to-day transaction in international markets.

These include the need for reserves to import goods and services as well as servicing external

debts of both the government and sometimes the private sector.

In Kenya, the private sector covers all its foreign exchange transaction needs from the interbank

market which has mostly met this requirement without recourse to the CBK. The CBK transacts

on behalf of government for all its foreign exchange requirements and for this purpose regularly

sources foreign exchange liquidity from the interbank by way of Open Market Operation (OMO).

However domestic liquidity injected into the financial system from this action is normally

addressed by way of mop up operations by the OMO desk.

2.1.4 Hard Currency Areas versus Soft Currency Areas

Some central banks intervene (purchase foreign exchange) through open market operations aimed

at preventing their currency from appreciating and consequently build substantial reserves. For

many years Japan engaged in a programme of buying US dollars in order to keep the Yen/US

dollar rate down so as to make its exports more competitive. This therefore led to a massive build

up in foreign exchange reserves at the Bank of Japan. In these circumstances, the level of foreign

exchange reserves is a secondary consideration. Canada did the same in the 1980s and its

reserves rose steadily, reaching US$18 billion in 1990 compared to an average of US$2.5 billion

22

between 1975 and 1986. This motive for holding reserves is referred to as the mercantilist motive

(Aizenman and Lee 2005).

In small developing countries the reverse is the case. Intervention is usually intended to stop the

rate from falling rather than to stop it from rising, so open market operations usually take the

form of selling foreign exchange; and therefore the level of foreign exchange reserves to support

interventions is a much more important factor. This is true for Kenya.

2.1.5 Oil Economies versus Non-oil Economies

Oil exporting countries are able to build reserves much more easily as they normally run current

account surpluses, while non-oil exporting countries have to continuously concern themselves

with replenishment of foreign exchange reserves. This gives oil exporters much more flexibility

in adopting a floating exchange rate regime, as they are able more easily to defend the rate due to

availability of reserves.

Kenya is a non-oil exporting country and therefore has to continuously concern itself with

replenishment of foreign exchange reserves to cater for payment of oil imports which constitute

its largest outflow.

2.1.6 Politically Vulnerable Economies and Politically Correct Economies

The political correctness of an economy influences what it considers to be an adequate level of

foreign exchange reserves. Politically correct economies have nothing to hide and would

23

normally open up their markets thereby experiencing high volumes of foreign exchange inflows

and rarely any outflows. Politically incorrect economies on the other hand see themselves as

being vulnerable to demands requiring them to change their regimes if they were dependent on

donor agencies for support. They would therefore need to avoid placing themselves in a position

where they may be forced to change their regimes thus the need to stoke up on reserves for

independence. There is evidence from the case of China that substantially high levels of foreign

exchange reserves add political clout (Robert Pringle 2003).

Kenya considers itself a politically correct economy despite various governance issues that have

constrained its relations with multilateral agencies and various governments. For this reason it

has not received much in the form of budget and development assistance in the last 10 years. It

would therefore need to stock up more in the form of reserves since no other flows would be

received from other quarters during periods of financial crisis. However it may be stated that

Kenya’s non-reliance on foreign assistance may make its economy more resilient and

independent of assistance.

2.1.7 Government Spending & Cost of Borrowing

High levels of foreign exchange reserves allow the borrowing country to be less circumspect in

assuming debt where they are a good credit risk, as financing is always available. Research

suggests (Moreno and Turner, 2004) that higher foreign reserves can reduce sovereign spreads as

well as improve credit ratings. Thus the cost of holding reserves can be narrowed as reserve

holdings rise. Accordingly, the greater the quantum of reserves the lower the cost of borrowing,

and the greater the benefit of holding reserves.

24

Kenya has in the recent past considered issuing a sovereign bond in the European market with the

objective of financing part of the budget gap, diversifying its debt portfolio and lowering its cost

of debt. However this was complicated by the current global financial crises.

2.1.8 Access to Capital Markets

Uncertainty about access to capital markets also influences the level of reserves which Central

Banks hold. Access to capital markets can be constrained because a country’s credit rating

declines or because the market itself is short of liquidity. In addition, access may also be so costly

as to be prohibitive. In these circumstances, central banks can be pressed into building and

frequently drawing on international reserves when need arises.

Kenya has had no previous access to the international capital markets and would therefore require

substantial reserves to draw from in the event of crisis.

2.1.9 Support for Exchange Rates

Foreign exchange intervention aimed at preventing currency depreciation requires high levels of

foreign exchange reserves. In some floating rate economies regimes, authorities often worry that

a steep depreciation of the currency could have political and credit risks. They would therefore

intervene in order to support the rate or to slow its depreciation. In Kenya the CBK has

intervened on various occasions during 2008 to slow down the depreciation of the Kenya shilling

by selling USD 130m to the interbank market.

25

2.1.10 Reserves and the Confidence Factor

Where there is some weakness in underlying fundamentals, it is even more important that foreign

exchange levels are healthy. Some will argue that foreign exchange reserves have their own

dynamic, both in terms of encouraging inflows when stocks are high and encouraging outflows

when stocks are low.

The growth of Kenya’s reserves levels from 1 month imports cover in 1996 to 4 months cover in

2008 Kenya has resulted in investor confidence, increased Diaspora transfers and capital account

inflows.

2.1.11 Seasonality and Foreign Exchange Adequacy

Invariably, every economy has periods when foreign exchange outflows are heavy and periods

when foreign exchange inflows are thin. It is important to convey the presence of seasonal

movements so that the public will not change their opinions about economic performance

because of large outflows during such seasonal periods of high reserve use. Of importance is that

seasonality is planned for, so that the reserves are accumulated during periods of good inflows

and subsequently released to the market during dry spells.

Kenya is an economy driven by seasonal sectors such as agriculture, horticulture and tourism.

Immediately after the liberalization of the foreign exchange market in 1995, the interbank market

was thin and lacked depth (turnover USD 10m daily). The CBK constantly bought foreign

26

exchange in times of surplus and supplied the market when there was a deficit. However the

market has grown (turnover is now at USD 80m daily) and currently manages the seasonality

patterns itself without recourse to the CBK. However the CBK stands ready to intervene during

periods of extreme volatility but not to defend any rate.

2.1.12 Reserves and Credit Ratings

In a survey by the BIS on the impact of reserve accumulation on credit ratings and external

vulnerability of central banks in emerging markets, all 16 central banks who responded believed

that reserve accumulation has some positive impact on sovereign credit ratings. Several central

banks, in both fixed and floating rate regimes, commented that higher reserves gave them greater

confidence and credibility in foreign exchange markets. They observed that this helped to

improve the sustainability of their external positions and hence their credit ratings. Higher

reserves implied greater capacity to redeem external debt and reduced the risk of speculative

attacks on the currency. It also reduced international funding costs in a number of emerging

economies. Most central banks also noted that rating agencies generally viewed the steady trend

of reserve accumulation as an indication of the underlying strength of the economy.

In a rating exercise conducted in 2008, Kenya was rated by Standard & Poors at BB-, a grade that

was considered favourable in facilitating Kenya’s access to the international capital markets. The

country intends to use this rating to negotiate a favourable spread in the sovereign bond issue that

is planned after the global economy recovers.

27

2.2 Adequacy of Official International Reserves

Conceptually reserve adequacy is the level of reserves that ensures smooth balance of payments

and macroeconomic adjustments in an unpredictably changing economic environment, e.g.

external price shocks, reversals in short-term foreign capital flows. However, there is no common

approach for the estimation of a benchmark reserves adequacy level. The most versatile approach

proposed by Alan Greenspan several years ago is to identify all balance of payments risks and to

develop stochastic tests measuring balance of payments losses with a given probability.

Adequacy of reserves has emerged as an important parameter in gauging a country’s ability to

absorb external shocks. With the changing profile of capital flows, the traditional approach of

assessing reserve adequacy in terms of import cover has been broadened to include a number of

parameters which take into account size, composition and risk profiles of various types of capital

flows as well as the types of external shocks to which the economy is vulnerable.

Countries generally maintain reserves in order to effectively manage exchange rate volatility and

to reduce adjustment costs associated with fluctuations in international payments. Accordingly,

the demand for international reserves increases with global trade. Empirical research shows that

both the variance and level of trade (current account and openness to trade or the propensity to

import) are important determinants of demand for reserves (see Mendoza 2004). In practice,

however, most countries follow the “rules of thumb” in determining the optimal level of reserves,

including maintaining reserves equivalent to at least three months of imports (Mendoza 2004).

This would suggest that they would only be able to deal with a run on the currency for three

months equivalent of imports and would thereafter have no defence.

28

However, the recent accumulation of reserves in developing countries has been largely

interpreted as a form of self-insurance precipitated by the high level of global economic and

financial instability and the absence of an adequate international system for crisis management.

The 1997 East Asian financial crisis is a good example in this regard (Stiglitz 2006). Moreover,

many countries see reserve accumulation not only as a means for effective exchange rate

management, but also as a tool for maintaining low exchange rates in order to promote trade and

international competitiveness.

Maintaining adequate reserves can also boost investors’ confidence and hence enhance

investment and growth. This can be more clearly seen in the case of Asian countries that recorded

the highest rate of increase in international reserves following the 1997 financial crisis coupled

with high growth in external trade and output. A study by the IMF (2003) shows that reserve

build-up in emerging Asia between 1997 and 2002 was large both in absolute terms and relative

to imports and short-term external debt. The study also points out that the reserve build-up in

emerging Asia has been similar across exchange rate regimes, including countries with limited

exchange rate flexibility as well as countries with managed floating exchange rates. More

importantly the reserve build-up in 2002 has been in excess of that warranted by the economic

fundamentals of the region.

Increases in international reserves in Latin America have been largely in line with rising imports

and commercial transactions in general and was only partially driven by demand for insurance

against financial shocks (Eichengreen 2006).

The import cover notion of reserves appears inadequate when there is high capital mobility and

costly financial crisis’ caused by a sudden reversal in capital flows such as those experienced

29

during the last two decades. Therefore, there is need for developing countries to hold reserves at

least equal to their short-term foreign currency debt. Evidence suggests that higher reserves

reduce both the likelihood of a crisis and the depth of a crisis, should one occur (IMF 2003). IMF

(2003) suggests that “a ratio of reserves to short-term external debt above one marks an important

reduction in crisis vulnerability, as long as the current account is not out of line and the exchange

rate is not misaligned”. At the global level, there is still a need for a more effective global

insurance framework or an international monetary system that can help prevent financial shocks

or mitigate their costs. Asian economies have created a reserve pool to deal with foreign

exchange intervention …………………..

While many economies struggle to maintain adequate reserve levels, a growing number of

countries have accumulated vast stocks of reserves; in some cases the reserve stocks are so large

as to bring into question their necessity. The most common motivation for holding large reserves

is to insure against currency crises by enabling authorities to support their own currency.

Table 1 shows that seven of the top ten reserve holders are countries typically considered

emerging economies. Other reasons discussed in this paper to hold reserves will not likely require

the holding of such large quantities of reserves.

Table 1: Top 10 Holders, Total Reserves Minus Gold

Level USD

bln as of

Percent

Increase

Percent

Increase

June 2006 From June 2005 from June 2002

China 943.6 32% 283%

30

Japan 849.8 2% 94%

Taiwan 262.0 3% 77%

Russia 243.2 64% 510%

Korea 225.6 10% 101%

India 156.8 17% 183%

Singapore 127.3 10% 59%

Hong Kong SAR 126.6 4% 13%

Mexico 84.9 29% 86%

Malaysia 78.4 5% 143%

Source: IMF

2.2.1 Accounts for Shocks

Financial crises are contagious and whenever they set in, the IMF gets concerned, as it is the

institution charged with the responsibility of ensuring international financial stability. During

times of financial crises, the demand for bailouts / support from the International Monetary Fund

tends to increase (Gin and McDonald (2003)). Concerned about the enormous cost implications

of financial crises among member countries, the IMF places a lot of emphasis on the efficient

management of international reserve holdings whilst minimising exposure to market risks and

shocks. Consequently, it has published some guidelines to facilitate such an endeavour among

member countries (IMF, 2003).

31

2.2.2 Impact on Other Economies of Holding Surplus Reserves

The argument against countries holding too large a stock of reserves has another dimension. It is

argued that struggling economies with high levels of foreign exchange reserves are really

financing the development of other large economies by holding their government securities and

may become susceptible to risks and costs emanating from adjustments in reserve currency

countries e.g. China holding substantial amounts of US Treasuries financing the US budget

deficit and by extension the US government’s excessive spending and consumption. These risks

and costs include inflationary pressures, over-investment, asset bubbles, complications in the

management of monetary policy, potentially sizable capital losses on monetary authorities’

balance sheets, sterilization costs, segmentation of the public debt market and misallocation of

domestic banks’ lending (ECB 2006)

This argument has not been sufficiently persuasive as to stop reserve accumulation, simply

because most developing countries generally do not find themselves in a position where they hold

such excessive stocks of reserves. Even in Botswana, a diamond rich country, where reserves are

equivalent to over 3 years of imports, it is debatable whether this is excessive given how rapidly

fortunes can change.

The greatest risk of holding excessive reserves is depreciation of the currency in which the

reserves are held. Where large reserves are held then the impact of this depreciation is greater.

Central banks can diversify away depreciation risk by holding multiple currencies, since the

depreciation of one hard currency is often matched by appreciation of others. This highlights the

benefits of currency diversification.

32

G7 countries argue that the counterpart to the Asian Bloc’s current account surpluses and

acquisition of dollar reserves is the large current account deficit in the United States. This

continued deficit could cause investors to worry about the US’ ability to repay their debt. This, it

is suggested, could then contribute to potential instability of the international financial system.

However, the international system is already taking corrective measures. Data already show that

the share of US dollars in the official holding of global foreign exchange reserves has declined

from 71% in 1999 to 65.9% in 2004, mostly reflected in a shift to the euro.

2.2.3 Opportunity Cost of Holding Reserves

The benefits of building large reserves should be carefully weighed against potentially high

economic and social costs. The costs of maintaining reserves comprise the opportunity cost of

foregone domestic consumption and investment as well as financial costs and the strain on

monetary policy arising from efforts to sterilize the effects of excessive monetary expansion

through higher domestic interest rates. This can increase fiscal pressure (control of government

spending and deficits) and make reserve accumulation inconsistent with fiscal policy objectives

(Stiglitz 2006). In addition, reserve build-up can pose challenges to the macroeconomic policy

framework. It is impossible for government to reconcile policy objectives vis-à-vis exchange rate

stability through a fixed exchange rate, monetary independence, and free capital mobility. While

it is possible to combine selective capital controls with occasional interventions in the exchange

rate market, still inconsistencies may arise e.g. if the economy is overheating, accumulating

reserves and keeping domestic currency from appreciating might be inconsistent with a tight

monetary policy. Higher domestic interest rates resulting from sterilization may also be

33

inconsistent with a tight fiscal policy (as higher interest payments put pressure on the fiscal

balance), leading to exchange rate and output volatility (UNDESA 2007).

Research however suggests (Moreno and Turner, 2004), that higher foreign reserves can reduce

sovereign spreads as well as improve credit ratings. Thus the cost of holding reserves can be

narrowed as reserve holdings rise. Accordingly, the greater the quantum of reserves, the lower the

cost of borrowing and the greater the benefit of holding reserves. Therefore it is important to

identify the optimal position which matches the trade-off between these costs and benefits.

2.3 Reserve Adequacy Indicators

The emerging market crises of the 1990s resulted in growing literature on the level of reserves

necessary to adequately insure against shocks. In that literature, several basic benchmarks for

emerging economies were suggested:

2.3.1 Months of Import Cover

For a long time the ratio of reserves to imports of goods and services measured in terms of weeks

of imports has been used to measure the adequacy of international reserves (Williams, 2003). A

measure based on imports, it was felt, could determine how long a country could continue to

import if all other inflows of foreign exchange dried up.

For many years the guideline of reserves equivalent to 12 weeks (3 months) imports was used as

the accepted measure by the International Monetary Fund. In recent years, particularly after the

34

South East Asian crisis this figure has been questioned. For emerging markets and developing

countries this ratio has averaged between eighteen (4.5 months) and twenty-two weeks (5.5

months) and respectively since 2003.

The currently held view is that this measure is of limited value, since it focuses solely on the

external current account, which was relevant in the years of limited capital flows to developing

and emerging market countries. Indeed, these countries have more open capital accounts and are

receiving greater amounts of capital flows. Therefore, in a world of highly mobile capital, reserve

adequacy should be based on rules that also focus on the vulnerabilities to the capital account.

This is the only indicator that has been used by the CBK to date. However in view of increased

internationalization of international finance, especially through the opening of capital accounts,

adequate international reserves have transcended current account considerations with risks of

resident capital flight becoming high. It is important for Kenya therefore to consider the use of

other indicators of reserve adequacy in computing her required reserves i.e. ratios of reserves to

base money, debt, capital flows.

2.3.2 Reserve Levels and Debt

Debt based indicators have been developed as a means of measuring reserve adequacy. One

measure, the level of reserves to short term debt by remaining maturity, is deemed to be a useful

indicator. The level of debt service repayment obligations as a ratio of export of goods and

services is another which attempts to link earnings to payments. The ratio of debt service to gross

domestic product is probably the most heavily used. The latter does not take into account foreign

35

exchange earnings; it is related more to the growth of the overall economy. However, high levels

of growth of GDP do not guarantee foreign exchange growth. The structure of production must

be geared towards export earnings in order for economic growth and foreign exchange

accumulation to be in sync.

Generally speaking, however, it is the foreign component of debt that is of most interest to capital

market watchers and rating agencies, though international financial institutions like the IMF are

also concerned about the overall debt burden and its sustainability.

Traditionally, government debt service obligations have been the major consideration, but with

increasing globalization, central banks can indirectly be put under pressure to provide foreign

exchange for the servicing of private sector foreign debts as well.

2.3.3 Reserve Levels and Capital Flows

Given the inevitability of increasing internationalization, and the increasing range of financial

instruments, the expected levels of capital outflows and inflows are particularly important in

determining foreign exchange reserve adequacy. Alan Greenspan (1999) suggests a “liquidity at

risk” rule which moves beyond simple balance sheet rules and works toward a standard that takes

into account the foreseeable risks that countries face. One approach, he suggests would be to

calculate a country’s liquidity position for relevant financial variables such as exchange rates,

commodity prices, and credit spreads. It might be possible to express a standard in terms of the

probabilities of different outcomes. He was of the view that such a “liquidity at risk” standard

36

could handle a wide range of innovative financial instruments – contingent credit lines with

collateral, options on commodity prices and put options on bonds etc, in an appropriate manner.

This idea has never been developed into an implementable formula, although value-at risk

models have become more important in measuring overall portfolio risks.

2.3.4 The Guidotti Rule on Foreign Exchange Adequacy

Pablo Guidotti, a former Deputy Finance Minister of Argentina, developed a simple rule for

policy makers in emerging market economies. It was that usable foreign exchange should exceed

scheduled amortizations of foreign currency debts (assuming no rollovers during the following

year). The Calvo rule is similar in that it requires a government’s external debt repayments falling

due in the next 12 months not to exceed its foreign exchange reserves. This was meant to lead to

a measure that could serve as an indicator of how long a country could sustain external imbalance

without resorting to foreign borrowing.

However, for small developing countries this rule is highly risky and is not likely to instil

confidence in fixed exchange rate regimes since their annual debt repayment levels tend to be

relatively small when compared to their current account requirements. Foreign exchange reserves

equivalent to debt repayments in the 12 months ahead would be woefully inadequate in small

African and Caribbean countries with fixed exchange rates and possibly even in small countries

with floating exchange rates. In such regimes, experience suggests that the pool of foreign

exchange reserves needs to be much higher.

37

For the most part, holding foreign reserves in excess of short-term debt obligations could help to

reduce the incidence of financial crisis. The level held in excess of short-term debt, however,

depends on country-specific circumstances. Issues such as the prevailing macroeconomic

conditions, particularly whether there is a large and persistent external current account deficit that

needs to be financed by international capital inflows, or an overvalued exchange rate that can

lead to capital outflows, high levels of short-term public debt, and weak banking systems that can

contribute to capital flight, are critical in this regard, and must be factored into the analysis. This

underscores the need for a sound understanding of the interaction between reserve adequacy,

vulnerability, and country-specific factors.

For instance, in Hungary (Shcherbakov 2002), the targeting of reserve adequacy takes into

account the amount of reserves needed to cover short-term debt as well as a mark-up (a kind of

assurance) to reflect unfavourable macroeconomic fundaments – i.e., the potential for an

increasing external current account deficit or real exchange rate appreciation.

2.3.5 Ratio of Short Term External Liabilities to Reserves

Following the increased internationalization of the financial markets, especially through the

opening of capital accounts, concern for holding adequate international reserves has transcended

current account considerations. Countries are now concerned about their respective net liability or

short-term debt positions. As such, international reserve holdings among countries need to also

account for external debt servicing needs. This is much more so considering that the amount of

international reserve holding by a country is a major ingredient in the country’s sovereign credit

rating (i.e. a measure of the financial capacity and willingness that a country has to pay external

38

debts). Therefore, adequacy of international reserves should also be measured in terms of a

country’s short-term foreign currency denominated debt. Short-term is in this case interpreted in

terms of remaining debt maturity.

2.3.6 Ratios of reserves to base money

In some emerging markets the confidence of resident investors in the domestic currency is not

very high or financial markets are underdeveloped. This creates risks of resident capital flight.

For these countries important indicators of reserve adequacy are ratios of reserves to base money

or other money aggregates. These indicators are also relevant for countries with hard exchange

rate arrangements, especially a currency board. However if financial markets are well tuned,

money based indicators seem to be not very important.

Table 2: Emerging Market Adequacy Reserve Ratios, 2005

Reserves/short-

term debt

Reserves/M2 Reserves/months

of imports

China 11.58 0.22 15.72

Taiwan 5.95 0.35 15.65

South Korea 2.63 0.21 7.93

39

Russia 4.43 0.93 16.40

India 4.29 0.80 13.17

Mexico 2.71 0.18 3.78

Malaysia 3.09 0.43 7.49

Benchmark 1.00 0.05 - 0.20 3.00

Source: IMF, BIS

2.4 Importance of Stress Tests

Globally, financial institutions are increasingly relying on statistical models to measure and

manage the financial risks to which they are exposed. These models are gaining credibility

because they provide a framework for identifying, analysing, measuring, communicating and

managing these risks.

As part of the effective management of a country’s international reserve portfolio, it is

recommended that the risks and vulnerability of the reserves portfolio deriving from both

anticipated and unanticipated changes in relevant macroeconomic and financial variables both

within and outside the country should be assessed regularly (IMF, 2003).

The purpose of stress tests is to determine the quantitative and/or qualitative impact of market

risk on the value or size of a country’s international reserve portfolio. Based on the stress test

findings, the adequacy of the current level of international reserve holdings can be evaluated and

if need be, revised.

40

Since central banks cannot wait until the adverse market developments occur in order to assess

their impact, the nature of stress tests is that they are forward looking and therefore anticipatory.

They are carried out with a view to understanding better the extent to which the current

international reserve holdings are adequate and if need be take precautionary measures.

The IMF has proposed that in addition to the benchmarking of foreign reserves to short-term

debt, countries should undertake stress testing of the balance of payments. The results of these

stress tests in different scenarios can serve as key inputs in the determination of reserve

adequacy. For example, a ratio of less than one (< 1) in the reserves to short-term debt adequacy

measure would be a simple stress test indicating the impact of the lack of access to international

markets for a single year, and represents a useful point of departure for more complex tests.

More complex tests should reflect different market scenarios for each type of capital inflow, be it

FDI, portfolio investment or trade credits, the risk of capital flight and the need to finance an

external current account deficit. However, in the field of finance and currency crises, the

experience of the past has not necessarily been a good predictor of the future and therefore stress

tests should incorporate the maximum historical variations as a basis for potential or future

variations in each balance of payments flow. It is important to note that stress tests are only as

good as the scenarios they are based on.

2.5 Reserves Build-Up in Africa: Trends and Specific Motivation

The build-up of reserves in Africa and emerging economies has accelerated over the last decade

with the bulk of the increase occurring in oil-exporting countries. The accumulation of reserves

41

has occurred at a time of generally stable or slightly appreciating exchange rates, particularly

against the US dollar.

For Africa, recent commodity price increases have allowed reserves accumulation among

exporters, while draining reserves among importers. Macroeconomic stabilization remains at the

forefront of national economic policymaking and aid conditionality in Africa (Mckinley 2007).

This induces countries to hold reserves to allow monetary authorities to intervene in markets to

influence the exchange rate and inflation. As with other economies, adequate reserves are

perceived to be of importance as they allow African countries to borrow abroad, attract foreign

capital and promote domestic private investment as a result of strengthened external position and

reduced vulnerability to external shocks.

2.5.1 Trends and motivation for reserve build-up in Africa

Accumulation of foreign exchange reserves by African countries may best be understood in the

context of reserves behaviour in developing regions in general. Global official foreign exchange

reserves rose from US$1.2 trillion in January 1995 to US$5.04 trillion in December 2006, and the

share of developing countries in world reserves increased from 50 to 72 per cent over the same

period. This share is relatively large especially in view of the fact that developing countries

accounted for only 41 per cent of world trade in 2005. Developing countries are therefore seen to

accumulate relatively more reserves than developed countries with Asia leading Latin America

and Africa in this surge in growth.

42

The East Asia and Pacific region witnessed one of the most severe financial crises in the last two

decades and as a result three factors, beside high oil prices, account for the build-up of reserves in

developing countries (ECB 2006). The first factor is the need for insurance against future

financial crisis similar to the Asian crisis. The second factor relates to the strong export-led

growth in Asia following large exchange rate depreciation in the region as a result of the financial

crisis. Finally certain features of the domestic financial markets of emerging economies, and the

Asian markets in particular, have stimulated the unprecedented accumulation of reserves. These

features include weak financial intermediation between domestic savers and investors and

inefficient hedging markets; the tendency towards dollarization of international assets; and excess

domestic savings over investment.

Thus the accumulation of official reserves as an outcome as well as a means of integration into

global financial market has been a common factor behind the recent reserve build-up in emerging

markets (ECB 2006). Emerging economies, especially in Asia, have the need to accumulate

reserves to protect their economies against financial market fluctuations because while they are

major players in international trade they still lag behind in terms of financial market

development.

The trend of foreign currency reserves relative to imports and external short-term debt confirms

the strong influence of the factors discussed above in relation to reserves accumulation in Africa.

In all the developing regions, reserves have generally increased as a ratio of imports of goods and

services during the last ten years. On average official reserves in Africa rose from the equivalent

of about two months of imports in 1990 to about 5 months in 2004. This suggests that on average

reserve holdings of African countries are just adequate (Elhiraika & Ndikumana 2007).

43

For the 40 African countries with available data, reserves represented 10 months of imports in

2006. The respective import cover for oil-exporting and oil-importing countries was 15 months

and 5 months respectively. This raises the question of why some African countries are

accumulating excess reserves in recent years and what policies they should adopt to avoid this in

the future.

Literature suggests that a ratio of reserve to short-term external debt of more than one indicates

adequate capacity of the country to service its external liabilities and face unexpected financial

risk in case of deteriorating external financial conditions.

Conversely a ratio of less than one indicates vulnerability to capital account risks (IMF 2003).

Since 2004 this ratio has risen above 2 for African countries. By 2005, on average, African

reserves as a ratio of short-term external debt exceeded that of all other developing regions with

the exception of South Asia. The higher ratios for relatively poorer developing regions may

reflect greater desire for self insurance against external shocks as well as their inability to tap the

foreign debt markets.

2.5.2 Sources and composition of African foreign exchange reserves

African countries for which detailed data exist have recorded very high rates of growth of foreign

exchange reserves since the turn of this century. Reserve flow is the sum of the current account

balance and the capital and financial account balance. A sample of 40 African countries as a

group recorded current account surpluses for most of the period under review, mainly because of

44

high current account surplus in resource-rich countries and net transfers; the income balance has

always been in deficit (Ndikumana 2007).

Regarding the capital and financial account balance, while the capital account switched between

surplus and deficit over the years, the financial account balance has shown net financial inflows

to Africa since 1990. These financial flows, including ODA and increasing remittances by

African nationals working abroad, contributed to the high rate of reserves accumulation during

this period. Overall sustained current and capital account surpluses in mainly resource-rich

countries are behind the high growth in reserves in Africa.

2.6 Foreign Currency Reserves in Kenya

As highlighted earlier, one of the functions of the CBK is to hold and maintain foreign currency

reserves. The CBK builds reserves through interest earnings from foreign exchange reserves,

purchase in the domestic interbank market, bilateral and multilateral aid flows into the country

and net forex inflows. The reserves are essentially used for:

a) servicing Central government external debt;

b) paying non-debt government external obligations;

c) intervention to smoothen erratic movements of the exchange rate

d) CBK external payments.

e) acts as a cushion against external crises.

f) The size of official reserves sends confidence signal to potential investors, rating agencies

and those contemplating capital flight.

45

Section 26 of the CBK Act states that “The Bank shall at all times use its best endeavours to

maintain a reserve of external assets at an aggregate amount of not less than the value of four

months’ imports as recorded and averaged for the last three preceding years.” Since late 1990s,

Kenya developed significantly stronger fundamentals (monetary, fiscal, trade agreements,

political) and during the transition to low inflation, international reserves built up substantially

from USD 500 million in 1996 to USD 2,600 million in March 2009 (equivalent to 3.13 months

of import). Therefore while the reserves levels have registered consistent growth over the years,

the CBK is still under the legal minimum of 4 months cover. Kenya’s economy has grown at an

average 3.5% in the last ten years and the monthly import average has consequently grown

thereby consistently hindering the attainment of the four months’ cover.

2.6.1 Foreign Exchange Reserves Holding: Kenya’s Experience

Chart 1

46

Chart 1 shows official international reserve holdings in Kenya between 1999:Q1 through

2009:Q1. It is evident there is a marked acceleration in the build up of official international

reserves during this period.

Chart 1 shows the shortfalls in the reserves from the mandatory minimum four months of imports

cover during the last ten years.

Chart 2

47

Chart 2 shows that Kenya’s position with regard to the generally accepted three months import

cover criteria of reserves adequacy has been mostly positive with a few exceptions. However it

has not mostly met the four months of import cover requirement introduced in 2002 in the

Central Bank of Kenya Act.

48

3. RESEARCH DESIGN & METHODOLOGY

3.1 Model of Integral Measure of Reserve Adequacy

The determination of the appropriate level of reserves for a particular country should primarily

focus on the most vulnerable items in the balance of payments as well as the main potential

drains on reserves.

In Kenya's case, the determinants of financial uncertainty in the balance of payments

are:

external debt payments scheduled within forthcoming years;

limited access to international financial markets;

high variability of export prices and low diversification of exports in conditions

when they drive economic growth and ensure employment and budget revenues;

open capital and money markets with unpredictable magnitude of permanent

capital flight/capital outflow

Main drains on Kenya’s international reserves are government external debt payments.

The government obtains foreign exchange for the payments by buying it from the CBK

for Kenya shillings at market rates. It has emerged from literature that in present market

conditions the appropriate level of reserves depends more and more on the size and

structure of external debt rather than on the current balance of payments transactions.

There is evidence that countries with uncertain access to international financial markets

49

tend to use the Guidotti rule for assessing reserves adequacy (Shcherbakov 2002).

According to this rule, reserves should hedge a predominant part of forthcoming external

debt payments more specifically they must cover all short-term debt with the remaining

maturity of one year or less (all external debt payments scheduled in the coming year).

The CBK builds reserves mainly from the interbank foreign exchange market sterilising

the Kenya shilling injections through the regular sale of open market operation

instruments. Other sources of foreign exchange have been donor inflows which have

been rather limited in recent times. The Kenya government is exploring the possibility of

international financial market borrowing through sovereign bond issuance but such plans

have been shelved in the wake of the current global financial crisis.

Kenya pursues a floating exchange rate regime and therefore the forces of demand and

supply dictate the value of the shilling. However the CBK has the mandate to ensure

smooth balance of payments adjustment in case of potential internal or external shocks

that may cause extreme volatility in the exchange rate and is therefore ready to

intervene in the market when necessary. For this purpose reserves should cover

correspondingly several months of imports of goods and services and a part of base

money.

The above mentioned drains on reserves appear rather independent and the joint

probability of their occurrence may be deemed as small. However, the Asian financial

crisis in 1998 showed that additional uncertainty induced by any balance of payments

problem may trigger off a chain reaction of other balance of payments problems. This

would imply the use of a conservative approach in determining Kenya's reserve

50

adequacy measure which should ideally be defined as the cumulative effect of different

potential drains on foreign exchange liquidity rather than as a weighted combination of

different criteria i.e.

RA D I M + L

Where:

RA - Integral measure of reserve adequacy,

D - Debt-based measure of reserve adequacy, denotes all external debt payments

scheduled in the coming year.

I - Import-based measure of reserve adequacy, denotes the three months’ worth

of

imports of goods and services.

M - Money-based measure of reserve adequacy, denotes the part of base money

that will be exchanged for foreign assets if residents loose their confidence in the

domestic currency.

L- Liquidity at Risk measure of reserve adequacy, denotes the estimated

depletion

of international reserves arising from shocks that would precipitate short-term

foreign currency capital outflows.

The determination of the Integral measure of reserve adequacy (RA) would guide CBK

in the establishment of a reserves target level for each succeeding quarter. Each

51

component of the integral reserve adequacy criterion is defined below as well as the

approach used in this study to determine its contribution to the integral measure.

For the purposes of this study the ten year period from January 1999 to the current

period will be considered.

3.1.1 Debt-based measure of reserve adequacy

This measure denotes all external debt payments scheduled in the coming year. However the

main drawback of Guidotti rule is that reserves are allocated to cover external debt payments in

the coming year irrespective of the remaining external debt repayment profile. If the external debt

repayment profile has considerable fluctuations, as is the case in Kenya, Guidotti rule will

produce highly volatile estimates of reserve adequacy over time. Such unstable estimates would

not be convenient for policy-making purposes and they would therefore need to be smoothed

using the following modification of the Guidotti rule:

t + 365

D(t0) = max ∑ P(t) τ≥ t0 t = τ

Where:

D(t) – debt-based measure of reserves adequacy on day t,

P(t) – external debt payments scheduled in day t,

t0 – reporting date

52

Thus, the debt-based measure of reserve adequacy is defined as the maximum amount

of forthcoming external debt payments, scheduled during any twelve month period (but

not necessarily calendar year). If the debt repayment schedule is more intensive at the

end than at the beginning then a part of reserves will be allocated for payments due in

the long run. As a result, this method will overestimate reserves subject to allocation for

debt service purposes.

3.1.2 Import - based measure of reserve adequacy

This measure denotes the generally accepted three months’ worth of imports of goods and

services. The main drawback of the traditional reserve adequacy ratio with respect to

imports is that it includes all goods and services irrespective of their impact on economic

growth. For instance, it is clear that expenditure on travel abroad, imports of luxury

items, etc. may be easily cut without any effect on domestic production. However, with

unlimited access to the foreign exchange market by importers it would be impossible to

constrain imports that are not essential for economic growth i.e. those not affecting

crucial imports. The only possible way to do this would be to introduce certain import

taxes and restrictions which in itself may jeopardise the balance of payments position

further.

Crucial imports can however be approximated by the minimum three month imports

registered during a certain period of time, say in Kenya’s case, the last ten years using

the following:

53

τ

I(t0) = min ∑ IM(t) τ≥ t0 t = τ-3

Where:

I(t) - import - based measure of reserve adequacy in month t,

IM(t) - monthly imports of goods and services

t0 - month corresponding to the reporting date.

3.1.3 Money - based measure of reserve adequacy

This measure denotes the part of base money that will be exchanged for foreign assets if residents

loose their confidence in the domestic currency. However, since Kenya has no fixed

exchange rate regime, there would be no need to cover all base money with reserves

since a part of base money issued by the CBK will never be exchanged for foreign

assets because there will always exists some demand for Kenya shillings as a means of

transactions.

In order to define a money-based measure of reserve adequacy, it is necessary to

determine what part of base money will be exchanged for foreign assets if residents

loose their confidence in the domestic currency. For a country that has experienced a

currency crisis a possible solution would be to look at the monetary contraction at that

period. The money-based measure of reserve adequacy would then be estimated by

comparing the ratios of M2 to GDP during the crisis and during the reporting periods.

54

However Kenya has not had a currency crisis and therefore a different approach would

be to identify the most liquid part of the base money and assume that this would easily

be changed to foreign exchange in the event of a crisis. Reserve Money would proxy the

liquid part of base money and generally consists of:

(i) balances on banks' correspondent accounts in the CBK (excess reserves)

(ii) balances of Non Bank Financial Institutions in the CBK

(iii) cash in circulation.

Only a part of cash in circulation would be exchanged for foreign currency cash by

households. The share of household incomes spent on acquisition of foreign currency is

not readily available from national statistics data but can be proxied by the lowest ratio

that was observed during the ten year period of this data i.e. 61.7%

Accordingly this measure of reserve adequacy with respect to money supply can be

approximated as follows:

M(t0 ) RM(t0)

Where:

RM(t) – Reserve Money on day t,

- share of purchases of foreign currency cash from reserve money.

3.1.4 Liquidity at Risk measure of reserve adequacy

55

This measure denotes the estimated depletion of international reserves arising from shocks on the

capital account. By providing for this component in the reserves, it is assumed that Kenya would

deal effectively with any reversals in short-term net foreign capital inflows. Of importance is

whether all short-term foreign capital inflows should be hedged with reserves because

reserves can only hedge a part of net foreign currency position of the private sector. In

this measure of reserve adequacy the hedged part shall be set to 50%.

3.2 Estimated Reserve Adequacy from Jan 1999 to Date

The first part of the study analyses the various criteria to determine the estimated level

of adequate reserves for the period January 1999 to date and, compares this with actual

reserves for the same period.

3.2.1 Debt-based measure of reserve adequacy

Chart 3

56

Chart 3 shows that annual debt payments in Kenya have fluctuated between USD 350m

and USD 700m during the last ten years. These annual payments according to the

Guidotti rule need to have been covered by reserves as a minimum requirement.

Chart 4

57

The chart shows that with the exception of the early part of 1999, international reserves

in Kenya have continued to adequately cover annual forthcoming debt and have

therefore been sufficient according to the debt-based measure.

3.2.2 Import-based measure of reserve adequacy

Chart 5

58

Kenya’s monthly imports have risen steadily since 1999from USD 200m to the current

USD 850m in line with the country’s growing economy.

Chart 6

59

The chart shows that international reserves presently cover more than three months of

imports of goods and services and from the view point of import-based alone the level of

reserves has been quite adequate. However the legal mandate of four months cover

introduced in 2002 in the Central Bank of Kenya Act has not been met for most of the period.

3.2.3 Money-based measure of reserve adequacy

Chart 7

60

As the chart shows, international reserves in Kenya have adequately covered all base

money since 2001 and therefore from the viewpoint of the money based measure alone,

the level of reserves in Kenya is adequate.

3.2.4 Liquidity At Risk - measure of reserve adequacy

Chart 8

61

The chart shows that Kenya’s capital account has moved from a deficit position in early 1999 to a

surplus of USD 150m in 2009.

The capital account comprises the medium / long-term investments as well as the short-term

investments which in Kenya include what is referred to as net errors and omissions that capture

all unidentified items.

Chart 9

62

As indicated in chart above, medium to long term investments/flows have been relatively low and

were actually in deficit during the early part of the survey period. The short term net position has

posted surpluses for the last ten years and is mostly responsible for the current status of capital

account surplus in the country.

As indicated previously not all the short term reserves need to be hedged but only 50%.

3.3 Overall Appraisal of Reserves Adequacy in Kenya

From the foregoing, we can now estimate what the adequate reserve level should have been

during the last 10 years and compare the same with the actual reserves during that period.

63

RA D I M + L

Table 3

Actual Reserves and the estimated Adequate Level 

USD M  Debt 

Measure Imports Measure 

Money Measure 

Liquidity at Risk 

Estimated Adequate Reserves  

Actual Reserves   Ratio 

   D  I  M  L  RA    (%) 

1999‐Q1  723  776  689  ‐14  2174   699  32.2 

1999‐Q2  759  647  611  26  2042   632  31.0 

1999‐Q3  725  655  575  28  1983   674  34.0 

1999‐Q4  664  700  590  88  2042   722  35.4 

2000‐Q1  585  776  625  61  2046   790  38.6 

2000‐Q2  502  761  598  18  1879   794  42.2 

2000‐Q3  415  924  580  94  2012   828  41.1 

2000‐Q4  448  879  576  55  1958   846  43.2 

2001‐Q1  481  816  557  11  1864   905  48.6 

2001‐Q2  491  1,019  562  169  2240   935  41.8 

2001‐Q3  553  927  559  156  2195   997  45.4 

2001‐Q4  570  744  587  43  1943   1,016  52.3 

2002‐Q1  570  932  586  84  2172   1,040  47.9 

2002‐Q2  604  791  600  11  2005   718  35.8 

2002‐Q3  596  778  625  7  2006   1,100  54.8 

2002‐Q4  610  856  637  ‐14  2089   1,055  50.5 

2003‐Q1  625  914  667  0  2206   1,161  52.6 

2003‐Q2  574  979  700  29  2282   1,244  54.5 

2003‐Q3  552  932  668  18  2170   1,268  58.4 

2003‐Q4  569  965  711  63  2308   1,383  60.0 

2004‐Q1  507  1,074  687  ‐16  2251   1,411  62.7 

2004‐Q2  561  1,112  702  ‐10  2364   1,365  57.7 

2004‐Q3  531  1,179  694  17  2422   1,325  54.7 

2004‐Q4  452  1,268  739  134  2592   1,407  54.3 

2005‐Q1  429  1,301  764  ‐76  2417   1,440  59.6 

2005‐Q2  351  1,721  758  167  2996   1,523  50.8 

2005‐Q3  380  1,457  802  111  2750   1,684  61.2 

2005‐Q4  417  1,471  866  77  2832   1,739  61.4 

2006‐Q1  463  1,655  891  126  3135   1,923  61.3 

2006‐Q2  514  1,835  923  273  3546   2,222  62.7 

64

2006‐Q3  504  1,838  944  83  3368   2,401  71.3 

2006‐Q4  557  2,005  1,050  111  3723   2,398  64.4 

2007‐Q1  538  2,189  1,080  230  4037   2,498  61.9 

2007‐Q2  668  2,173  1,174  90  4105   2,676  65.2 

2007‐Q3  702  2,323  1,216  208  4450   2,781  62.5 

2007‐Q4  739  2,384  1,365  415  4902   3,038  62.0 

2008‐Q1  769  2,691  1,391  135  4986   3,446  69.1 

2008‐Q2  691  2,678  1,468  120  4956   3,401  68.6 

2008‐Q3  552  3,269  1,370  323  5514   3,263  59.2 

2008‐Q4  329  2,854  1,275  178  4636   2,871  61.9 

2009‐Q1  202  2,363  1,187  174  3926   2,722  69.3 

The table above shows the derivation of the adequate level of reserves of reserves during the last

ten years which comprises the following:

D – the debt-based measure which takes into account actual quarterly external payments that

are then used to determine the annual debt payments for use in the Guidotti rule.

I - the import-based measure that is determined by the actual for the quarter i.e. three month

cover

M - the money-based measure that is determined by 61.7% of reserve money to estimate the

amount of base money that would be converted to foreign exchange in the event of a

shock.

L - the liquidity at risk measure that is estimated by 50% of the short term capital flows that

are deemed vulnerable to external shocks.

65

Chart 10

As the chart shows the level of international reserves in Kenya has consistently underperformed

the estimated adequate level for the last ten years. The ratio between the two reserve levels as

shown on the table has however shown improvement from a low of 31% to a high of 70% and

indicates a move towards attainment of adequacy in reserves as defined in the above model.

Chart 11

66

As shown in the chart, the level of adequate reserves as estimated with the adequacy model

throughout the ten year period has been over the four month threshold CBK is legally required to

maintain as a minimum. This means therefore that the CBK would be within its legal mandate

while maintaining reserve levels estimated by this model.

67

4.0 CONCLUSIONS AND POLICY RECOMMENDATIONS

In real life situations, international reserves holdings are susceptible to change following

corresponding changes in global market developments such as changes in external terms of trade,

exchange rate fluctuations, interest rate changes, insecurity, changes in climatic conditions and

the occurrence of natural disasters.

Though the changes may result in increments or reductions in international reserve holdings,

concern is normally about the adverse implications and whether or not adequate international

reserves are held to withstand the most extreme of shocks. For this reason, official international

reserves are deemed adequate when the present level of reserve holdings account for the

predetermined months of import cover, coverage of external debt, balance of payments liquidity

at risk, and other shocks whose effect is to deplete the level of international reserves held.

As has been witnessed during the current global financial crisis everything that could possibly go

wrong went wrong (Murphy’s Law). The argument against holding a very sizeable amount of

reserves may not therefore hold as the recent crisis clearly highlighted. It is critically important to

maintain an adequate level of reserves for reducing the costs associated with a foreign exchange

liquidity shortage and to ensure smooth balance of payments and macroeconomic adjustment

during reversals in foreign capital flows.

Going forward, the Central Bank of Kenya will need to use the model above to periodically

determine the appropriate level of reserves for every forthcoming year that will then guide its

accumulation of reserves. Other MEFMI countries would also find this useful for reserve

adequacy determination.

68

4.1 Strategic Asset Allocation Process

For CBK, the principle investment management objectives remain safety, liquidity and return and

to attain these, the Strategic Asset Allocation (SAA) process is the single most important

investment policy decision that the board should make i.e. define portfolio size, portfolio

tranches, acceptable risk and acceptable instruments.

The process of determining reserves adequacy in Kenya as defined in the model above would

highlight the various uses and sources of reserves in Kenya namely external debt, imports, capital

flows. This process would therefore support portfolio management especially in the computation

of foreign exchange amounts and currency required for the various components outlined in the

model and would be useful in the determination of reserves tranches (working capital, liquidity

and investment), currency composition, investment objectives and investment horizons.

In identifying the various reserve adequacy indicators and their respective measures for Kenya,

this study would therefore support the SAA process and consequently define the tranches that the

CBK would need to have and their respective sizes.

69

5. BIBLIOGRAPHY

Aizenman, J. and Marion N. P. (2002), “The High Demand for International Reserves in the Far

East: What is Going on?” NBER Working Paper No. 9266.

Bird, G. and R. Rajan (2003). “Too Much of a Good Thing? The Adequacy of International

Reserves in the Aftermath of Crises”.

De Beaufort Wijnholds, J.O. and A. Kapteyn (2001). “Reserve Adequacy in Emerging Market

Economies”, Working Paper No.01/143, IMF.

Elhiraika A. & Ndikumana L. (2007) Reserves Accumulation in African Countries: Sources,

Motivations, and Effects

Government of Kenya (GOK), 2002, The Central Bank of Kenya Act, Cap. 491.

International Monetary Fund (IMF), 2003, “Risk Management”, in IMF, Guidelines for Foreign

Exchange Reserve Management.

Kohler, H. (2002), Speech, at IMF Website, (July 5th).

Naameh, M., 2003, “Reserve Management in Developing Countries”, in Pringle R. and N. Carver

(eds.), How Countries Manage Reserve Assets. Central Banking Publications of the Royal

Bank of Scotland, Pg.137-151.

Olivier J. and Rancière R., 2005 “The Optimal Level of International Reserves for Emerging

Market Economies: Formulas and Applications,” IMF Working Paper No. WP/06/229.

70

Pringle R. and N. Carver (2003), “How Countries Manage Reserve Assets”, Central Banking

Publications of the Royal Bank of Scotland.

Shcherbakov S. G. (2002) Foreign Reserve Adequacy: Case of Russia, Bank of Russia

Tweedie, 2002, “The Demand for International Reserves – A Review of the Literature”.

Williams, D. (2003), “The Need for Reserves”, Publications of the Royal Bank of Scotland.

Williams M. V. (2005), Governor, Speech, Central Bank of Barbados