ANXIETIES ABOUT THE DECLINING INFLUENCE Of U.S. banks in international markets made headlines, and prompted Congressional inquiries, in the late 1980s and early 1990s. More recently, however, concern about U.S. banks, competitiveness overseas has given way to alarm about the growing market share of foreign banks in U.S. markets.(1) By one estimate, foreign banks hold nearly 50 percent of all existing commercial and industrial loans made to U.S. businesses. Moreover, foreign banks made these gains swiftly, more than doubling their share of the U.S. market in the past ten years.
Despite this impressive growth - or perhaps because of it - foreign banks were not particularly profitable. This study investigates the relative profit efficiency of foreign - owned U.S. banks and U.S.-owned banks between 1985 and 1990, years during which foreign bank market share was expanding rapidly. We employ a profit frontier model similar to the one pioneered by Berger, Hancock, and Humphrey (1993), and modify the model so that it is less sensitive to variations in asset size. Our results suggest that foreign-owned U.S. banks were significantly less profit efficient than U.S.-owned banks during this time period. Although there was little difference between the two sets of banks in terms of output efficiency, foreign-owned banks had a distinct disadvantage in terms of input efficiency, a disadvantage primarily driven by excess expenditures on purchased funds. These results imply that foreign-owned banks may have placed growth ahead of profitability, expanding their portfolio of loans faster than their ability to develop the relationships necessary to maintain an accompanying base of core deposits.
1. RECENT PERFORMANCE OF FOREIGN BANKS IN THE UNITED STATES
The share of commercial and industrial (C&I) loans to U.S. businesses held by foreign banks increased dramatically between 1983 and 1993. Using conventional measures, the combined market share held by foreign banks doubled from 14 to 32 percent during the period.(2) Perhaps the most popular explanation for the increase in foreign banks' lending to U.S. businesses is that foreign banks "followed" clients from their home countries into U.S. markets.(3) Having established a presence in the United States, many foreign banks grew their market share by purchasing loans in the secondary market (Calomiris and Carey 1994), or by acquiring existing U.S. banks (Kraus 1995), rather than by originating new loans.
Foreign banks may also have exploited a variety of cost advantages relative to U.S. banks to gain a competitive edge. Zimmer and McCauley (1991) concluded that foreign banks enjoyed cost of capital advantages over U.S.-owned banks. McCauley and Seth (1992) and Terrell (1993) found similar advantages for foreign banks concerning cost of funds. Frankel and Morgan (1992) and Wagster, Cooper, and Kolari (1994) found that differences in cross-country regulatory requirements may have reduced foreign banks' costs relative to U.S. banks.
Given these supposed cost advantages, foreign banks should have been able to gain market share by underpricing U.S. banks,(4) by producing higher-quality services than U.S. banks,(5) or both, and still have maintained profits roughly in line with those earned by U.S.-owned banks. However, both Seth (1992) and Nolle (1995) I found that profit rates at foreign-owned banks operating in the United States lagged behind profit rates for U.S.-owned banks during the past decade. For every year but 1987 (when U.S. money center banks provisioned for problem loans to LDCs), these studies concluded that return on assets (ROA) and return on equity (ROE) at foreign banks were less than that for a group of similar U.S. banks.
Three recent studies suggest that cost inefficiency may explain the low profitability of foreign-owned U. S. banks. Chang, Hasan, and Hunter (1995) estimated a stochastic cost frontier model for a panel of foreign-owned and U.S.-owned multinational banks operating with U.S. charters between 1984 and 1989. In regressions that control for asset size, bank holding company form, and foreign lending activity, the authors found that cost efficiency is negatively related to foreign ownership. Nolle (1995),using thick frontier estimates of cost efficiency from De Young (1996), concluded that the average foreign subsidiary was less cost efficient than the average U.S. bank in every year but one between 1984 and 1992. In a third study, Elyasiani and Mehdian (1993) used data envelopment analysis to measure cost efficiency for a sample of foreign-owned and U.S.-owned banks in 1988. The authors found that foreign subs were less cost efficient than U.S. banks, although the difference was not statistically significant.
2. MEASURING PROFIT EFFICIENCY
Cost efficiency models assume that banks take current input prices and output quantities as given, then seek to minimize costs by hiring the optimal levels of inputs. As such, any inefficiencies estimated using these models must be attributed exclusively to hiring an excess amount, or a suboptimal mix, of inputs. Alternatively, a bank can be inefficient if it produces too few, or a nonoptimal mix of, outputs given the inputs it employs and the prices that it faces. That is, in addition to being cost inefficient, a bank might also be revenue inefficient. By not recognizing this possibility, cost-based models can misrepresent the nature and extent of inefficiency in banks. For example, in order to produce above-average service quality, a bank will probably have to hire more and for more expensive inputs, and as a result the bank may be mistakenly identified as cost inefficient by a purely cost-based model. But such banks may actually be profit efficient - since the market tends to pay more for higher quality service, these banks may generate additional revenues large enough to offset their relatively high expenses.
This section presents the profit efficiency model that we use to estimate technical inefficiency in foreign-owned and domestically owned U.S. banks. The model is a modified version of a profit efficiency model introduced by Berger, Hancock, and Humphrey ((1993, henceforth BHH), and allows us to generate separate estimates of input and output inefficiency.(6) We assume that banks attempt to maximize variable profits [pi] in the short run by choosing a vector of variable netputs x, given fixed factors z and a vector of …