Economic &
Social Affairs
DESA Working Paper No. 60
ST/ESA/2007/DWP/60
November 2007
What does excess bank liquidity say about
the loan market in Less Developed Countries?
Tarron Khemraj
Abstract
Evidence about developing countries’ commercial banks’ liquidity preference suggests the following
about their loan markets: (i) the loan interest rate is a minimum mark-up rate; (ii) the loan market
is characterized by oligopoly power; and (iii) indirect monetary policy, a cornerstone of financial
liberalization, can only be effective at very high interest rates that are likely to be deflationary.
e minimum rate is a mark-up over a foreign interest rate, marginal transaction costs and a risk
premium. A calibration exercise demonstrates that the hypothesis of a minimum mark-up loan rate
is consistent with the observed stylized facts.
JEL Classification: O10; O16; E52; G21; L13
Keywords: Excess bank liquidity, oligopoly banking, loan market, monetary policy
Tarron Khemraj is Assistant Professor of Economics at the New College of Florida (the Honours
College of Florida State University). E-mail: tkhemraj@ncf.edu. e author gratefully acknowledges
the valuable comments and criticisms by Professor Duncan Foley of the New School for Social
Research. Any remaining errors are, of course, the responsibility of the author.
Comments should be addressed by email to the author.
UN/DESA Working Papers are preliminary
documents circulated in a limited number of
copies and posted on the DESA website at
http://www.un.org/esa/desa/papers to stimulate
discussion and critical comment. e views
and opinions expressed herein are those of the
author and do not necessarily reflect those of
the United Nations Secretariat. e designations
and terminology employed may not conform to
United Nations practice and do not imply the
expression of any opinion whatsoever on the part
of the Organization.
Editor: Donald Lee
Typesetter: Valerian Monteiro
United Nations
Department of Economic and Social Affairs
2 United Nations Plaza, Room DC2-1428
New York, N.Y. 10017, USA
Tel: (1-212) 963-4761 • Fax: (1-212) 963-4444
e-mail: esa@un.org
http://www.un.org/esa/desa/papers
Contents
Stylized Facts ................................................................................................................................... 2
Oligopoly Banking and Monetary Policy ......................................................................................... 5
e loan market ................................................................................................................. 7
Treasury bill market............................................................................................................ 8
Indirect monetary policy and the loan market .................................................................... 8
Quasi-Calibration Exercise .............................................................................................................. 9
Conclusion ...................................................................................................................................... 12
References ....................................................................................................................................... 13
What does excess bank liquidity say about
the loan market in Less Developed Countries?
Tarron Khemraj
e financial liberalization hypothesis holds that allowing the market determination of real interest rates
would mobilize savings and increase deposits (Fry, 1997a). Commercial banks—that are able to select good
from bad borrowers, diversify risks, minimize transaction costs, etc—would then channel these savings to
the best investors who earn the highest rate of return. Performing such roles of intermediation, banks not
only increase the rate of capital accumulation but also increase productivity, thereby boosting the economy’s
steady-state growth (Bencivenga and Smith, 1991).
However, in many less developed countries, banks hold large quantities of excess liquidity—a large
part of which is non-remunerative—in their asset portfolio (Saxegard, 2006; Khemraj, 2006; Fielding and
Shorthand, 2005). For the remainder of this paper, excess liquidity is defined as total bank liquidity minus
required bank liquidity. e required liquidity (or reserve) ratio is set by the central bank in the individual
country. Excess liquidity is usually non-remunerated.
In spite of efforts to liberalize and modernize financial institutions, markets and instruments in
LDCs, the banking sector is the most important source of financing in these economies and it is likely to
continue to be that way indefinitely (Stiglitz, 1989; Singh, 1997). erefore, the investment choice of banks
can either retard finance’s role in growth or augment that role. Hence, examining banks’ liquidity preference
in LDCs will emphasize important information regarding their behaviour in such economies.
is paper posits the hypothesis that banks in LDCs require a minimum rate of interest in the
loan market before they make a specific loan. A bank must receive a minimum loan rate that compensates
for risks, marginal transaction costs and the rate of return on a safe foreign asset before it makes a loan to a
particular borrower. If the marginal borrower is unwilling to pay the minimum rate, then the banks accumu-
late non-remunerative excess liquidity
1
. is phenomenon is depicted by a liquidity preference curve that is
flat at a relatively high loan rate. erefore, non-remunerative excess liquidity and loans can become perfect
substitutes at a very high rate of interest in the loan market. e paper will demonstrate that such behaviour
is consistent with a loan market that is oligopolistic. Moreover, to present its case, the paper utilizes the in-
dustrial organization banking model of Klein (1971) and Freixas and Rochet (1999). e model will also be
modified to suit the institutional characteristics of underdeveloped economies.
A key implication of this study for policy is the postulation that commercial banks set the loan rate
exogenously via a mark-up over the marginal transaction costs and the exogenous safe rate of interest
2
. It there-
1 e obvious question would be why banks still hold unremunerated liquidity given that the foreign rate is the
opportunity cost. ere are likely to be several factors that preclude banks from doing so. e first is a shortage of
foreign exchange in the domestic foreign currency market. is is a foreign currency constraint. e constraint can be
unofficially maintained by the central bank when it accumulates international reserves by buying up US dollars in the
local currency market. is behaviour was found in Khemraj (2006) in the Guyana case. A further research agenda
would be to find out whether the foreign currency constraint is a key factor accounting for excess unremunerated
liquidity in other developing countries. Explicit capital controls would be another reason why banks would demand
the unremunerated liquidity.
2 e same argument can be made about the deposit market and market for government bonds and Treasury bills. at
issue, however, is beyond the scope of this paper and is the subject of further research.
2 DESA Working Paper No. 60
fore means that a liquidity shock emanating from the central bank will not elicit a response in the interest rate
over the flat section of the banks’ liquidity preference curve. is is important for LDCs that have been imple-
menting indirect (or market-based) monetary policy as a means of influencing bank credit—and ultimately
consumption and investment decisions—by managing excess reserves and/or a short-term interest rate
3
.
Indirect monetary policy is often seen as a precondition for the adoption of inflation targeting—or
at least a milder version of inflation targeting known as inflation targeting ‘lite’ (Stone, 2003)—in LDCs.
Monetary policy shocks—characterized by shifts in the excess reserves supply curve through open market
operations—are only likely to be effective at even higher loan rates (above the minimum rate at which the
excess liquidity preference curve is horizontal) when the liquidity preference curve is downward sloping.
High interest rates, however, can contribute to economic stagnation even after significant efforts have been
made in liberalizing and developing financial systems in developing countries.
e paper is structured as follows. e next section presents the stylized facts that depict the liquid-
ity preference curves in the eight countries. e following section develops a banking model that helps to
explain the stylized facts. e model is used to make theoretical statements regarding banks’ response to
central bank monetary policy shocks. e penultimate section performs a calibration exercise to buttress the
core proposition of the research. e fifth section concludes.
Stylized Facts
is section utilizes the technique of locally weighted polynomial regressions (LOESS) of degree one in
order to extract bank liquidity preference curves for eight less developed countries
4
. e sample of countries
is based on primarily two factors. Firstly, the country must be known to have a very liquid banking sector.
Secondly, the relevant data for the analysis must be available over a long enough time period. e African
countries chosen in this paper were also included in the sample of Saxegaard (2006). Egypt is also included
because Fielding and Shorthand (2005) analyzed excess bank liquidity and political violence in that country.
e Caribbean countries are all included in the work of Khemraj (2006).
Each figure below plots bank liquidity (in local currency) against the loan rate. However, for three
of the countries—Egypt, Uganda and Namibia—data on the level of excess bank liquidity is unavailable.
However, bank liquidity is represented by total bank reserves (the summation of required and excess) taken
from the International Financial Statistics (IFS). e loan rate series for each country was also sourced from
the IFS. e level of excess liquidity for the other five countries comes directly from the respective central
bank statistical report. In each case, the nominal interest rate is utilized
5
.
3 Alexander, et al. (1995: 2) define direct versus indirect monetary policy instruments. Direct instruments set or limit
prices (interest rates) or quantity (credit). e quantity-based direct instruments often place restrictions on commercial
banks’ balance sheet. Hence, they are associated with financial repression. Indirect instruments, in contrast, operate
through the market by influencing the demand and supply conditions of commercial bank reserves.
4 e local regressions are extremely useful for deciphering underlying nonlinear relationships. e technique was first
proposed by Cleveland (1979) and further developed by Cleveland and Devlin (1988). Only a subset of observations
within a neighbourhood of the point to fit the curve is used. e regression is weighted so that observations further
from the given data point are given less weight. e subset of data used in each weighted least squares fit is
αN, where
α = the smoothing parameter and N = number of data points. A higher parameter, α, gives a smoother fit.
roughout the exercise a smoothing parameter of 0.3 is used.
5
e nominal interest rate is used throughout the analysis. is is because the real interest rate does not
change the conclusion since if inflation is important for the banks, they would want to get rid of all non-
remunerative assets. is is clearly not the case since the data confirm that the banking sectors in these
economies have always been liquid.
What does excess bank liquidity say about the loan market in LDCs? 3
ere are two clear tendencies in the figures: (i) the fitted liquidity preference curves tend to become
flat
6
; and (ii) the flatness occurs at a relatively very high rate of interest. is means the elasticity of demand
for bank excess liquidity is perfectly elastic (or approaches perfect elasticity) at a loan rate significantly above
zero. Banks in these economies view loans and unproductive excess liquidity as perfect (or near perfect)
substitutes at very high loan rates.
In the case of Guyana, the bank’s liquidity preference curve becomes flat at approximately 16 per
cent. is implies that a bank will not lend, on average, to the marginal borrower if that borrower cannot
pay at least 16 per cent. e same can be said for Barbados where the curve becomes flat at around 8.5 per
cent. In the case of Jamaica, commercial banks will not lend to the marginal borrower who wishes to pay a
rate below 18 per cent. e marginal borrower in Uganda will find credit difficult to come by if he or she is
unwilling to borrow at around 20 per cent. In Trinidad and Tobago the curve becomes flat at 10.5 per cent;
while in Egypt the minimum rate seems to be approximately 13 per cent. e Bahamas has the lowest mini-
mum loan rate at approximately 6 per cent.
Under a perfectly competitive loan market—an assumption that is implicitly made in the financial
liberalization literature (see Arestis and Demetriades, 1999)—excess liquidity and bank loans should become
substitutes at a zero loan rate. e fact that they are substitutes at a very high rate implies the banking sector
in our selected economies (and very likely other underdeveloped economies also) is far from the case of com-
petition. is paper makes the realistic assumption that the banking sector is oligopolistic and not competi-
tive. As oligopolies, banks are able to mark-up the loan rate over an exogenous benchmark rate, transaction
costs, and also take into consideration any risk of default associated with a specific borrower.
6 It might be tempting to view the flat liquidity preference curves as indicative of a liquidity trap. However, that is not
the case for two reasons: (i) the analysis uses the loan rate rather than the safe government bond/Treasury bill rate;
and (ii) the curves tend to become flat at a very high rate of interest. ere is a liquidity trap when money and bonds
become perfect substitutes at zero bond/Treasury bill rate.
Figure 1.
Guyana: Bank liquidity and the loan rate
(Quarterly data: 1988Q1—2006Q4)
Figure 2.
Barbados: Bank liquidity and the loan rate
(Quarterly data: 1990Q1—2006Q4)
12
16
20
24
28
32
36
40
-2000 0 2000 4000 6000 8000 10000
Liquidity (G$mill)
Guyana LR
LOESS Fit (degree = 1, span = 0.3000)
Data source: IFS; Bank of Guyana
7
8
9
10
11
12
13
14
15
16
-100000
100000
300000 500000
Liquidity (B$000)
Barbados LR
LOESS Fit (degree = 1, span = 0.3000)
Data source: IFS; Bank of Barbados
