(1989).
2. Literature Review
The determinants of the interest margin have been explored by a vast body of literature. The concentration of the industry is the most examined variable. The industrial organization literature predicts that an oligopolistic market structure may contribute to a higher interest margin (Samuel & Valderrama, 2006). Broadly, the salient determinants of the margin include: (i) the industry’s market structure, (ii) bank-specific variables, (iii) macroeconomic variables, and (iv) financial regulations.
Bain’s (1951) structure-conduct-performance (SCP) hypothesis holds that firms in a concentrated market are able to collude to pay relatively less on their liabilities and charge more on their assets, thereby increasing the margin. Ho and Saunders (1981) view the bank as ―a dealer‖ that demands depositors and suppliers of loans, and argue that the bank’s interest margin depends on four factors: the degree of the bank management’s risk aversion, the market structure of the industry, the average size of bank transactions, and the variance of interest rates. In addition, the authors reckon that a number of imperfections and regulatory restrictions have an impact on spread, and consider the probability that loan defaults and the opportunity cost of holding mandatory reserves are additional variables that influence the spread. Neumark and Sharpe (1992) implicitly confirm the SCP hypothesis for the US. They find that banks in concentrated markets are slower to raise deposit rates in response to rising market
interest rates, but faster to reduce these in response to declining market rates, thereby maintaining higher spreads. Corvoisier and Gropp (2001) examine the hypothesis for Euroarea countries, and confirm that the SCP hypothesis holds for loan and demand deposit rates, but not for savings and time deposit rates. Hannan and Liang (1993) and Barajas, Steiner, and Salazar (1999), using data for the US and Colombia, also suggest that industry concentration might lead to a higher spread. Prager and Hannan (1998) examine the price effects of US bank mergers that led to a substantial increase in local market concentration, and find that, during 1991–94, the deposit rates offered in local markets where mergers took place declined proportionately more than in markets without mergers. Sapienza (2002) examines the effect of banking consolidation on banks’ credit policies in Italy, and reports that, in the case of in-market mergers, the interest rates charged by consolidated banks decrease if the merger involves the acquisition of banks with a small market share.
However, as the local market share of
the acquired bank increases, the decline is offset by market power. Edwards (1965) examines the impact of concentration and competition in the US banking industry, and finds that mergers have a greater negative impact on performance in less concentrated markets. Regulators should, therefore, be wary of mergers in less concentrated markets as much as they are in more concentrated markets.
3. Methodology and Data
We draw on the dealership model of the interest margin used by Martinez-Peria and Mody (2004) to specify a model to examine the determinants of the interest margin in Pakistan. The model predicts that the market structure of the banking sector, its operating cost, the cost of regulating the sector, and various macroeconomic variables may affect the interest margin. Here, we also include the short-term government debt held by banks as a determinant of the interest margin, which previous studies have not done. The motivation is that banks find it beneficial to invest in government paper on several counts: investment in government paper is less risky, it does not entail a large administrative and analytical cost to make an investment decision, and above all, if the monetary stance is tight, the return on government paper might be such that the investment becomes an attractive opportunity on its own. The cost of investing in government paper is smaller and the return relatively larger; such investment is likely to raise the margin if banks do not pass on the higher returns to their customers, i.e.,
borrowers and depositors. The inclusion of government bonds in the model is all the more important given the anecdotal evidence that government bonds crowd out lending to the private sector. Taking our lead from Khawaja and Din (2007), we also include in our model the share of interest-insensitive deposits held by banks. The rationale is that banks need not offer a higher return on such deposits, the flows of which are insensitive to the interest rate offered. Thus, a larger share of interest-insensitive deposits could lower the overall cost of funds for banks, and thereby raise the interest margin. Around 50 percent of the business volume—both deposits and lending—is held by five or six major banks. Given that a significant portion of these deposits, which we term interest-insensitive deposits, comes to the bank on its own (i.e., without any effort on the bank’s part), the cost to the bank of deposit mobilization is likely to be rather low. For example, government departments bank mostly with the state-owned National Bank of Pakistan, while the armed forces
bank only with Askari, which is owned by the Army Welfare Trust that was set up to promote the welfare of retired army employees.
Similarly, many people deposit money in banks for the sole purpose of safety, and are not concerned about the rate of return. Deposit products typically offered by banks in Pakistan include current deposits, savings deposits, and fixed deposits. Deposits held in current accounts are zero-rated, i.e., they do not earn any interest and are therefore insensitive to changes in the interest rate. We treat the deposits held in savings accounts as interest-insensitive because they are held typically by small depositors and salaried persons who maintain these accounts to fulfill everyday banking needs rather than to earn interest. The category of ―other deposit accounts‖ constitutes a negligible percentage of total deposits, and its inclusion on either side is not likely to influence results. These deposits are also interest-insensitive. Theoretically, changes in the policy rate—proxied here by the six-month treasury-bill rate—should be passed on, in a competitive environment, one for one, to deposit rates. However, interestinsensitive deposits enable a bank to keep to itself all or part of a favorable change in interest rates and pass on to agents the entire burden or even more if the change in policy rate is adverse for the bank. Thus, the greater the interest-insensitivity of deposits, the higher is the interest margin. Given that current deposits (checking accounts) and saving deposits, which we consider insensitive, constitute a sizable portion (66 percent) of total industry deposits in 2009, the inclusion of interest-insensitive deposits in the model is all the more important.
Here, it is the interest margin, defined as the difference between the interest earned on average assets and the interest paid on average liabilities; (α, β) are vectors of parameters, it is the stochastic error term, and it is a vector of explanatory variables which includes the following.
Industry variables(i) Concentration.(ii) Interest-insensitivity of deposits. Firm (bank) variables(i) Market share.(ii) Liquidity.(iii) Administrative cost. (iv) Nonperforming loans.(v) Equity. Macro-variables
(i) Short-term debt (the government’s floating debt).(ii) Real output.(iii) The real interest rate.
The literature on industrial organization offers two competing hypotheses regarding the market structure of the industry. The SCP hypothesis holds that market concentration leads to collusion among firms. With the cost of collusion being smaller
in a concentrated market, firms are able to collude and thereby reap rents. Given market power, a bank would earn more on assets than is possible in a competitive market, and pay relatively less on liabilities, thereby raising the interest margin. If the SCP hypothesis holds, then the concentration variable should carry a positive sign.
The efficient-structure hypothesis asserts that the efficient operation of leading firms in the industry drive out the less efficient ones, the market becomes concentrated, and firms earn Ricardian rents. To the extent that efficiency is represented by the lower marginal cost of producing output of a given quality, banks in concentrated markets should find it advantageous to charge lower interest on loans and offer higher interest on deposits, thereby decreasing the margin. Thus, if the efficient-structure hypothesis holds, then the concentration variable should carry a negative sign. Given the conflicting predictions of the two hypotheses, we use an ambiguous sign for the concentration variable. The two hypotheses have been tested extensively for the banking industry (Berger & Hannan, 1989). Besides industry concentration, the two variables of primary interest are the volume of short-term government bonds (floating debt) and interest-insensitive deposits. We hypothesize that both carry a positive sign. Floating debt is expected to bear a positive sign because, as the debt increases, interest rates are likely to go up, thereby increasing the yield on government paper and advances for banks. Ceteris paribus, this raises the margin. Typically, a bank would pay less on interest-insensitive deposits. Therefore, the larger the share of interest-insensitive deposits in a bank’s total deposits, the smaller the average cost of funds. This would raise the margin and, therefore, justify the positive sign on interestInterest Margins and Banks’ Asset-Liability Composition 261 insensitive deposits. The remaining variables in Equation (1) are control variables. The coefficient on liquidity is hypothesized to be negative because liquidity has an opportunity cost, i.e., the cost of not investing in high-yield assets such as ―advances.‖ Therefore, the increase in liquidity should make a dent in the interest margin.
The equity held by a bank also carries an opportunity cost. If the bank manages to pass on this cost to its depositors and borrowers, then the spread would vary positively with equity. Failure to do so would decrease the bank’s interest margin. Given the conflicting expectations, we posit an ambiguous sign on equity. If the bank has to incur a greater intermediation cost when mobilizing deposits or lending funds, it would attempt to recover the cost by paying less on deposits and charging more on loans. Therefore, the interest margin should vary positively with the intermediation
cost. Nonperforming loans—loans that a bank fails to recover—inflict a cost on the bank, and should cause the margin to shrink. The market share of a bank in the industry reflects the former’s market power and influences the margin positively. However, the scale economies that accompany a larger market size may allow the bank to charge its borrowers less and offer its depositors more. If this happens, the margin would vary negatively with market share. This conflict leads us to posit an ambiguous sign on the market share of an individual bank. The macroeconomic environment has the potential to influence a bank’s interest margin. Thus, we control for the impact of real output and the real interest rate on the interest margin. Real output growth captures the impact of a business cycle on the interest margin. In this context, Bernanke and Gertler (1989) argue that a slowdown in economic activity adversely affects borrowers’ net worth and hence reduces the interest margin (positive effect). The coefficient on the real interest rate would depend on the extent to which changes in the rate are passed on by the bank to its customers. A one-for-one pass-through would result in a positive sign on the real interest rate coefficient, while a smaller passthrough would result in a different sign.
4. Further Discussion
The question that begs an answer is, what can be done to raise the interest margin? The answer is, not much. However, to answer the question more clearly, one must first ask what the objective of containing the interest margin is. Is it depositors’ welfare or an effort to encourage savings and investment? If the former, perhaps continuing with a floor rate on savings deposits would be the correct policy to pursue. However, although the floor rate on savings deposits has yielded greater returns to depositors, the policy has merely raised the return on those deposits that were meant to be in the banking system even if the rate was not enhanced. Researchers and the authorities must grapple with the question as to how to decrease the interest margin to increase savings and investment, and enhance the effectiveness of the interest rate channel of monetary policy. With the source of a high interest margin being the interest-insensitivity of a significant percentage of depositors and the short-term debt of the government, finding a way to cause a dent in the interest margin in a manner that encourages savings and investment will prove difficult. Curbing the former source calls for a behavioral change—Economic growth might also contribute to increasing the share of interest-sensitive deposits, as there would be more money to save. Tackling the latter—a reduction in government debt—calls for containing the
fiscal deficit, which boils down to improving the whole set of macroeconomic variables and macroeconomic management.
5. Conclusion
Banks’ interest margin has remained on the higher side throughout the previous decade. Policy efforts initiated in January 2008 to contain the margin have yet to yield dividends. Moreover, this policy effort seems directed more at improving the welfare of depositors than encouraging savings and investment. The study finds that the (i) government’s short-term debt—which is an attractive investment avenue for banks—and (ii) the share of interest-insensitive deposits are the two primary determinants of the high interest margin. Thus, the present structure of banks’ assets and liabilities keeps the margin on the higher side. Given the determinants of and the ways in which this margin might be contained, the process is likely to prove a tall order.