國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
I. Introduction
The traditional role of a commercial bank is mainly to serve as an intermediary of funds through accepting deposits and granting loans. People deem the
intermediation role of banks as the key contribution to the economic growth of a country, since timely and efficient allocation of loanable funds enables firms to expand their business activities. Nevertheless, both the emergence of nonbank creditors and the prosperity of direct finance markets decrease the volume of core deposits and demand for loans of banks, accompanied by the keen competition in the banking industry. These facts demonstrate that relying solely on traditional activities, such as loans and investments, would deteriorate the profitability of commercial banks. Even worse, the decreasing interest rate, caused by an economic downturn, further squeezes out the benefits of traditional activities for banks.
In light of the declining net interest margin, advocacy of financial liberalization becomes the main stream. Until 1999, the US governments repeal the Glass-Steagall Act, limitation on commercial bank securities activities and affiliations between commercial banks and securities firms. Later, other developed and developing countries follow suit. Since then, commercial banks ailed by confined businesses beheld the bright side of revenue growth, and commenced benefiting from involving non-traditional activities such as loan commitment, credit lines, letter of credit, swap and hedging transactions, securities underwriting, off-shore financial services, brokerage, cash management, export/import services etc. Most of them are
off-balance sheet activities and chiefly will not be classified as assets or liabilities on balance sheet. These non-traditional activities contribute commissions and fee income and other non-interest income. However, as far as risk management is concerned, some of these are still required to be disclosed through notes in financial reports to
1
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
fulfill due diligence obligation. As a financial market matures with time, the proportion of non-traditional activities is on rising. Banks hence are likely to gain from diversifying income sources and simultaneously, solidifying customer relationship.
Now that the increasing importance of non-traditional activities is foreseeable, the imperativeness to examine the relationship between traditional and non-traditional activities is beyond description, either for the authority or bank managers. Would allocating more resources on non-traditional activities advantage traditional ones and ameliorate performance or, inversely, damage traditional activities and aggravate net interest margin? Some literatures advocate business concentration. The main concern lays emphasis on cross selling issue, which banks are inclined to loosen protocol on loan application scrutiny and subsequently accept clients with worse credit to
establish customer relationship, targeting the profits afterwards by undertaking more non-traditional activities with those clients. However, the corresponding higher default risk further erodes the core capital of bank and serves as a negative signal to investors. On the other hand, some authors are in favor of business diversification.
They claim that appropriate engagement in non-traditional activities can effectively reduce the idiosyncratic risk of bank operation and carry solid client relationship via more active interaction.
Relevant literatures intend to find the relationship between traditional and non-traditional activities of banks mainly in European and American countries, yet similar works on Asian countries are relatively scarce. Developing countries as most Asian countries are, they are growing importance in terms of the international trade and capital markets. Their thriving markets and promising future make them worth thoroughly investigated. This paper focuses on the banking industries of some Newly Industrialized Economies (NIEs) and a few developed economies in Asia, including
2
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
China, Hong Kong, Indonesia, Japan, Malaysia, Philippines, Singapore, Taiwan and Thailand.1 Aside from examining the association between traditional and
non-traditional activities, we further attempt to observe the determinants that significantly contribute to the volatility of the net interest margin in these countries.
Specifically, we will construct a model of simultaneous equations to examine the foregoing issues, rather than relying on a single equation model that suffers from the endogeneity problem of some explanatory variables.2 This problem usually leads to inconsistent parameter estimates derived by the employment of the ordinary least squares (OLS). Doing so, the so-called simultaneous equations bias results and the identification problem remains to be solved. We herewith apply the econometric method of Lewbel (2012) to jointly deal with the endogeneity problem, as well as the identification problem, using the generalized method of moments (GMM).
Since the degree of market competition affects both the conduct and profitability of banks, which in turn impacts the net interest margin, production efficiency and the resource allocation on traditional and non-traditional activities. We are particularly selecting the Lerner index as the measure of market competition.3 The index is defined as the ratio of the gap between the output price and the marginal cost (MC) of that output to the output price. All previous works first estimate a cost function and then obtain the MC by taking the first derivative of the cost function with respect to the output; meanwhile the output price is derived from taking the ratio of total revenues to total assets. Unfortunately, the so-derived Lerner index is not guaranteed to be non-negative and a negative value of the index lacks economic
1 South Korea is excluded from our sample countries due to serious missing data problem. We instead replace it by China and Japan that do not belong to NIEs but locate on the neighboring region to the remaining six economies.
2 See Stiroh (2006), Laeven and Levin (2007), Baele et al. (2007).
3 The concentration ratio and the Herfindahl-Hirschman index are two conventional measures of market power. Panzar and Rosse (1987) develop the popular H-statistic to reflect the degree of market concentration. Numerous papers have employed the H-statistic to investigate the market power of various industries in many countries.
3
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
implications.
This paper employs the new econometric approach, proposed by Huang et al.
(2013), to evaluate the Lerner index in the context of the simultaneous equations, which requires estimating the translog cost frontier, along with a price equation that relates the output price to the MC of the output. The gap between the price and the MC is embedded as if an inefficiency term. Similar to the translog cost frontier, the price equation is also modeled in the context of the stochastic frontier approach (SFA).
In this manner, the output price must be greater than or equal to the MC that is
simultaneously determined by the translog cost frontier. The implied Lerner index will be non-negative as a result. The derivation of the likelihood function counts on the use of the copula method. With this method, the estimation bias on ignoring cost
inefficiencies can be resolved.
Our main finding is that net interest margin and noninterest income display significantly positive relationship, indicating overall benefits for banks from income diversification. Lucrative as nontraditional activities are, banks in these countries generally lack evidence for loss-leader behaviors, which asserts that banks attract new customers and establish long-term relationships by setting lower interest margins or charging lower lending rates. Pivotal variables include management cost, opportunity cost of reserves, degree of risk aversion, risk factors, etc. Our empirical evidence is consistent with previous works. Most importantly, banks with stronger market power are inclined to specialize their business instead of engaging more nontraditional activities. Therefore, the loss-leader behavior doesn’t prevail in our sample countries.
The rest of this paper is organized as follows. Section 2 briefly reviews prior literatures on the fields of the Lerner index and net interest margins. Section 3 introduces our methodology. Section 4 describes the dataset and empirical variables.
Section 5 presents empirical results, while the last section concludes the paper.
4
‧
國立 政 治 大 學