PRESCRIPCIONES DE LOS ESTABLECIMIENTOS DE ALIMENTOS Y BEBIDAS SEGÚN EL

Determinants of switching costs

Switching costs mainly arise from asymmetries of information, which are proved to be heterogeneous among different sized banks (Shy, 2002; Kim et al, 2003). Asymmetrical information contains two aspects, which are: information asymmetric in the bank-borrower relationship and lack of information sharing between banks. Large banks tend to have more customers, which raise the asymmetrical information comparative advantages. Symmetrically, small firms that are usually considered as opaque stand more for the asymmetrical information and are less likely to switch banks (Gopalan et al, 2011). Although information sharing is considered to reduce switching costs (Gehriga and Stenbacka, 2007), currently only negative information (bad credit information) sharing is provided in China. In addition, a rational bank is not willing to share their client’s information with their competitors. Hence, bank size plays an important role in determining the level of switching costs.

Operational efficiency is different among banks. It is imaginable that the degree of switching costs are partly dependent on how efficiently the bank takes advantage of asymmetrical information and collect useful information. Berger et al. (2005b) argue that small banks are better able to collect and act on ‘soft’13 _{information than large banks. Hence, operational} efficiency should be included as one determinant of switching costs.

13 _{Soft information is the internal information about the investing project cannot be credibly}

Banks usually create artificial switching costs to set barriers for consumers to change suppliers (Smidt et al., 2006). An effective strategy for banks is to develop other correlations and borrowing to develop a closer bank-firm relationship. Kim et al. (2003) suggested that when switching between credit suppliers, costs related to the loss of capitalized value of the previously established relationship would be involved. Having other business relationships with banks set the lock-in power and results in higher switching costs.

Having enough funding sources enhances the banks’ market power in lending. The total loans in the China banking sector have increased from 2003 to 2010 at an average annual growth rate of 28.8%. Huge and growing loan demand has significantly challenged the banks’ lending capabilities. Many firms have high likelihood to switch to another bank to overcome the financial constraint when they are dissatisfied with their incumbent banks. An ample deposit can guarantee that banks satisfy the growing borrowing requirements of their customers and lower the probability of their customers switching to other banks for financial help.

Based on above analysis, here I use the switching costs as the dependent variable regressed on the bank characteristics (measure the degree of asymmetries, operation efficient, artificial barrier and fund sources) and a set of macro variables (measure economy and industry environment) to explore whether the independent variables have explanatory power for switching costs; that is:

𝑆𝑤𝑖𝑡𝑐ℎ𝑖𝑛𝑔 𝐶𝑜𝑠𝑡_𝑖𝑡 = 𝛼₀+ 𝑋_𝑖𝑡𝛼₁+ 𝑀_𝑖𝑡𝛼₂+ 𝑢_𝑖𝑡 (23)

Determinants of banks’ profit with switching costs

Set N banks in a competitive market, where 𝑖 ∈ {1,2, … , 𝑁}. The customers have been locked- in to their incumbent banks in previous period. When they switch to non-relationship banks, they bear the switching costs. Assume when choosing from which bank to borrow that the firms compare the gain from the difference of prices charged by the various suppliers and the loss from switching costs. Here, as in Kim (2003), switching costs (𝑠) are added to the prices charged by the non-relationship banks. Hence, the probability of a firm staying with its bank is given by 𝑝𝑟_𝑡(𝑖 → 𝑖) = 𝑓{𝑝_𝑖𝑡, 𝑝_𝑖𝑅𝑡+ 𝑠_𝑖𝑡}, where 𝑝_𝑖𝑡 is the price charged by the incumbent bank, i and 𝑝_𝑖𝑅𝑡are the prices charged by the rival banks. Since higher switching costs (𝑠_𝑖𝑡) help bank i to lock-in more customers, there is 𝜕𝑝𝑟𝑡(𝑖→𝑖)

𝜕𝑠𝑖𝑡 > 0. Symmetrically, the probability for firms switching from rival banks to bank i is given by 𝑝𝑟_𝑡(𝑅 → 𝑖) = 𝑓{𝑝𝑖𝑡+ 𝑠𝑖𝑅𝑡, 𝑝𝑖𝑅𝑡},

where 𝑠_𝑖𝑅𝑡 is the rival bank’s switching costs. The switching costs of rival bank’s offer a barrier for their customer switching to bank i, hence 𝜕𝑝𝑟_𝜕𝑠𝑡(𝑅→𝑖)

𝑖𝑅𝑡 < 0. So the total loans of bank i in period t (𝐿_𝑖𝑡) is given as:

𝐿𝑖𝑡 = 𝑝𝑟𝑡(𝑖 → 𝑖) ∗ 𝐿𝑖𝑡−1+ 𝑝𝑟𝑡(𝑅 → 𝑖) ∗ 𝐿𝑖𝑅𝑡−1 (24)

where is the market share of rival banks of bank i.

Denote πit as the profit of bank in period t. 𝑁𝑖𝑡is probability of the non-performing lending.

Meanwhile, 𝑟 is the interest rate of deposit, which is assumed as homogenous for all banks. 𝐶_𝑖𝑡 stands for the non-interest expense. Here, assume that the total amount of lending is equal to the total amount of deposit for bank I in period t (𝐿_𝑖𝑡 = 𝐷_𝑖𝑡). The profit function of bank I in time t is as follows:

By substituting equation (24) into (25), there is:

𝜋_𝑖𝑡 = (𝑝_𝑖𝑡 ∗ (1 − 𝑁_𝑖𝑡) − 𝑟) ∗ [𝑝𝑟_𝑡(𝑖 → 𝑖) ∗ 𝐿_𝑖𝑡−1+ 𝑝𝑟_𝑡(𝑅 → 𝑖) ∗ 𝐿_{𝑖𝑅𝑡−1}] − 𝐶_𝑖𝑡 (26)

In equation (26), it is clear that 𝜕𝜋_𝜕𝑠𝑖𝑡

𝑖𝑡 > 0, which suggests that switching costs have a positive effect on the bank’s profits.

A large amount of previous studies have claimed that the bank’s profits are linked to the bank’s characteristics and macroeconomics variables. Similar to Stephan et al. (2009) and Gopalan et al. (2011), here I set the profits determination model as:

𝜋_𝑖𝑡 = 𝛽₀+ 𝛽₁𝑠𝑤𝑖𝑡𝑐ℎ𝑖𝑛𝑔 𝑐𝑜𝑠𝑡𝑠_𝑖𝑡+ 𝑋_𝑖𝑡𝛽₂+ 𝑀_𝑖𝑡𝛽₃+ 𝜀_𝑖𝑡. (27)