## _wp09171

## Source details

**Canonical URL:** [_wp09171](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09171.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09171.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09171.pdf.json)

---

### I. INTRODUCTION
- Financial liberalization can provide many benefits: better pricing of capital and risk, improved allocation and efficiency of investment (Abiad, Oomes, and Ueda, 2004).
- Potential gains from liberalization:
  - Larger share of intermediation by more efficient banks and those better at assessing risk.
  - More credit to sectors previously underserved.
  - Expanded range of financial products for risk management and portfolio diversification.
- Risks of premature liberalization: exposure of financial sector weaknesses, inhibited intermediation, and economic volatility.
- Central question: likely impact of interest rate liberalization on China’s banking system.
- China’s status:
  - Substantial progress in financial market liberalization; remaining floor on bank lending rates and ceiling on deposit rates.
  - Deposit rate ceiling appears to bind: deposit rates clustered at benchmark; real deposit rates on average zero.
- Distortions from interest rate controls:
  - Households not appropriately compensated for savings; financial income share of total income remains one of the lowest in the world.
  - Large administratively determined interest margin reduces incentives for banks to improve efficiency and risk pricing.
  - Credit constraints channel funds to large/well-connected enterprises, away from SMEs and households.
  - Lack of market-determined interest rates deprives the People’s Bank of China (PBC) of critical information on macroeconomic and liquidity conditions.
- Approaches used:
  - Develop and calibrate an oligopoly model of the banking system (Section II) and study interest rate deregulation effects (Section III).
  - Use case studies of other countries’ experiences to infer preconditions and implications for China (Section IV).

### II. A MODEL OF CHINA’S BANKING SECTOR
- Model structure and assumptions:
  - Banks differ in size (deposits and loans), returns to scale (increasing and decreasing), and profitability.
  - Banks modeled as oligopolists choosing deposit (Di) and loan (Li) volumes; required reserves share α.
  - Aggregate deposit supply curve rD(D) and loan demand curve rL(L); aggregate sums defined with standard notation.
  - Deposits can be lent at lending rate rL or placed in interbank market at rate r (reserves and government bonds remunerated at r).
  - Central bank may target interbank rate or allow endogenous determination; banks take r as given.
  - Cost functions separable: deposit management cost di(Di), loan management cost li(Li), fixed cost Fi.
- Profit maximization and constraints:
  - Banks maximize profits defined in equation (1) subject to reserve constraint (1−α)DLi ≥ 0 (multiplier λi).
  - First-order conditions yield equations (2a), (2b), and (2c) (exact functional forms preserved in source).
- Effects of policy instruments:
  - Binding reserve requirement (λi>0) acts as a tax on lending (size λi) and subsidy on deposit taking (1−α)λi.
  - Binding regulated deposit rate rD replaces equation (2a) with (3a); regulated lending rate rL replaces (2b) with (3b).
  - Interbank market clearing condition (equation (4)): DL + B = Nr, where B is net bank holdings of government and central bank paper, and Nr is net demand for funds by nonbanks.

### II.B CALIBRATION
- Targets and sample:
  - Calibration targets: loans, deposits, profits, and costs of 20 of China’s largest banks (in 2007).
  - Agricultural Bank of China excluded due to ongoing restructuring.
  - These banks represent around 60 percent of total loans and deposits in the banking system.
- Key calibration assumptions:
  - Deposit rate ceiling binding; lending rate floor is not binding.
    - Supporting detail: less than one-third of loans are offered at the lending rate floor.
  - Deposit supply elasticity = 0.2 (based on Ho (2008)).
  - Loan demand elasticity = −1.
  - Interbank elasticity IB ε = 2.5.
- Cost-function identification:
  - Five cost parameters per bank (ai, bi etc.) but only four data series per bank → restriction: 0 di a = for each bank.
  - Deposit-side: di b chosen to match deposit share using equation 2 Ddi diD SHi rr a bD D − − =.
  - Lending-side: li a and li b chosen to minimize distance between bank-by-bank lending and non-interest variable costs (Nelder-Mead Simplex algorithm).
  - Fixed cost i F chosen to equate modeled total profits with actual 2007 profits.
- Calibration performance:
  - Model replicates distribution of loans, deposits, profits, and costs across banks fairly well.
  - Bank-by-bank deposits and profits matched exactly.
  - Distributional properties of loans (multiple mass points and highly skewed) closely replicated.
  - Model captures variable costs for most banks; difficulty with highest cost banks.

### III. HOW MAY INTEREST RATE LIBERALIZATION CHANGE CHINESE BANKING?
- Central bank operating assumptions considered:
  (i) central bank targets the interbank (IB) rate (holding it constant);
  (ii) central bank targets a spread over the now endogenously determined deposit rate;
  (iii) central bank is passive and allows interbank rate to clear endogenously.
- Four core questions addressed:
  1. Impact of liberalization on intermediation?
  2. Does central bank operating target affect reform impact?
  3. Implications for monetary policy effectiveness?
  4. How does impact depend on extent of competition?

A. The Impact of Liberalization — Intermediation and Banking System Soundness
- Modeled outcomes of deposit rate liberalization:
  - Smaller banks offer higher deposit rates and expand; larger banks shrink deposit bases to reduce costs.
  - Deposit market concentration falls.
  - Distributional impact on lending is smaller than on deposits (lending market largely liberalized already).
  - On average, profitability (as a share of loans) changes little; lower profits at large banks offset higher profits at smaller ones.
  - Net effect: deposit rate rises sufficiently to bring additional deposits into the system; loan rates and cost of capital increase; lending volume declines.
  - Share of deposits at banks making losses increases slightly (never above 3¾ percent); number of loss-making banks falls.
- Robustness across central bank procedures:
  - Trends consistent across operating procedures.
  - Larger increase in interest rates if central bank targets a constant spread of interbank over deposit rate.
- Limit on deposit rate increases:
  - Cost of interbank funds places an effective cap on deposit rate rises; deposit rate typically below interbank rate due to deposit management costs.
  - Simulated scenario assumes little change in desired monetary conditions; deposit rate rise limited to up to 37 basis points.
  - Liberalization links interbank fund costs, lending and deposit rates, and volumes more directly.
- Selected baseline and scenario values (preserved exactly):
  - Calibration: Deposit Rate 8.31; Share of Lending 13.36; Deposit HHI 63.73; IB Rate 2.28; Profit/Loans 3.04; Unprofitable Banks 0.19; Deposit HHI (lending) 0.19
  - Target IB rate: Deposit Rate 8.65; Share of Lending 3.62; Deposit HHI 63.73; IB Rate 2.19; Profit/Loans 3.53; Unprofitable Banks 0.19; Deposit HHI 0.12
  - Target IB spread: Deposit Rate 8.72; Share of Lending 3.64; Deposit HHI 64.00; IB Rate 2.44; Profit/Loans 3.73; Unprofitable Banks 0.19; Deposit HHI 0.10
  - Endog. IB rate: Deposit Rate 8.67; Share of Lending 3.61; Deposit HHI 63.72; IB Rate 2.23; Profit/Loans 3.66; Unprofitable Banks 0.19; Deposit HHI 0.11
  - Note: 1/ In percent of loans. 2/ Measured by the Herfindahl-Hirschman index.

B. Monetary Policy
- Liberalization effects on policy transmission:
  - Enhances effectiveness of indirect monetary policy and interest rates as price signals.
  - Given liberalization, a given change in IB target results in larger feed-through to lending and deposit rates.
  - Central bank must undertake larger interventions in the money market to achieve its target when deposit rate is free.
  - Quantity-based monetary policy remains relatively more effective under regulation when deposits unresponsive; indirect policy impact greater when deposit rate can fall.
- Exact scenario deviations (percentage point deviation from baseline; preserved):
  - Target IB rate: +50 bps:
    - Lending rate: 0.34 (Liberalized) vs 0.31 (Regulated)
    - Deposit rate: 0.50 vs 0.23
    - IB rate: 0.00 vs 0.50
  - Target IB rate: −35 bps:
    - Lending rate: −0.34 vs −0.40
    - Deposit rate: −0.35 vs −0.26
    - IB rate: 0.00 vs −0.35
  - Target IB rate: −50 bps:
    - Lending rate: −0.48 vs −0.53
    - Deposit rate: −0.50 vs (not reported in provided extract)
  - RR reduced (15% to 7.5%):
    - Lending rate: −0.30 vs −0.18
    - Deposit rate: 0.00 vs −0.12
    - IB rate: 0.00 vs 0.00
  - Note: Table assumes deposit rate level is regulated in the comparison; central bank cannot lower its target rate more than 37 bps if deposit rate held constant.

C. Market Behavior (Competition vs Collusion)
- Two behavioral extremes modeled and results (exact deviations preserved):
  - Table 3 (percentage point deviation from baseline):
    - Competition, Target IB rate: Lending Rate 0.3; Deposit Rate 0.9; Interbank Rate 0.0; Margin −0.6
    - Competition, Target IB spread: Lending Rate 0.4; Deposit Rate 1.3; Interbank Rate 1.0; Margin −0.9
    - Competition, Endog. IB rate: Lending Rate 0.4; Deposit Rate 1.2; Interbank Rate 2.6; Margin −0.8
    - Cooperation (collusion), Target IB rate: Lending Rate −0.3; Deposit Rate −1.0; Interbank Rate 0.0; Margin 0.7
    - Cooperation, Target IB spread: Lending Rate −1.7; Deposit Rate −1.8; Interbank Rate −1.9; Margin 0.1
    - Cooperation, Endog. IB rate: Lending Rate −1.0; Deposit Rate −1.4; Interbank Rate 0.2; Margin 0.4
    - Note: Margin = lending − deposit interest rates.
  - Table 4 (percentage point deviation from baseline; concentration and profitability):
    - Competition, Target IB rate: Unprofitable Banks −1.4; Profit/Loans 0.2; (Deposit share) Lending 0.05; Deposits 0.11
    - Competition, Target IB spread: Unprofitable Banks −1.7; Profit/Loans 0.5; (Deposit share) Lending 0.05; Deposits 0.11
    - Competition, Endog. IB rate: Unprofitable Banks −1.7; Profit/Loans 1.4; (Deposit share) Lending 0.05; Deposits 0.11
    - Cooperation, Target IB rate: Unprofitable Banks 3.8; Profit/Loans −3.5; (Deposit share) Lending −0.04; Deposits −0.06
    - Cooperation, Target IB spread: Unprofitable Banks 1.4; Profit/Loans −3.7; (Deposit share) Lending −0.03; Deposits −0.03
    - Cooperation, Endog. IB rate: Unprofitable Banks 3.8; Profit/Loans −3.7; (Deposit share) Lending −0.04; Deposits −0.05
    - Note: Concentration measured by Herfindahl-Hirschman index.
- Behavioral implications:
  - Competition → large expansion in lending and deposit taking; margins squeezed; profitability falls slightly; only one bank fails (less than 1 percent of system deposits).
  - Collusion → banks scale back operations to raise margins; lending and deposit demand reduced; profitability much higher; no banks make losses.
  - Under collusion, liberalization can lead to lower deposit rates due to inelastic deposit supply.

### IV. ROBUSTNESS
- Baseline elasticity assumptions:
  - Deposit supply elasticity = 0.2.
  - Loan demand elasticity = −1.
- Findings robust to sizable changes in these elasticities (section truncated in extract).
- Additional simulated elasticity impacts (preserved):
  - Deposit-rate elasticity:
    - Larger deposit-rate elasticity → banks need to offer smaller interest rate increases to achieve a given change in deposit volumes.
    - With deposit-rate elasticity of 0.5 or 1, deposit rates increase by less than under deregulation (statement preserved verbatim).
  - Loan demand elasticity:
    - If loan demand becomes less elastic (to −0.5): banks become significantly more profitable after reform; banks could reduce the deposit rate.
    - If loan demand becomes more elastic (to −2): smaller increase in lending rate, marginally higher deposit rate; share of deposits at loss-making banks = 5.5 percent (none fail); market concentration increases.
  - Scenario where lending rate initially binding:
    - Dual liberalization → much tighter margins: deposit rate rises, lending rate falls; both deposit and lending volumes rise above regulated levels; lower market concentration; banks remain sound (none fail).

### IV. WHAT HAS BEEN THE EXPERIENCE WITH LIBERALIZATION ELSEWHERE?
- Common cross-country outcomes:
  - Liberalization often led to higher real interest rates, shifted surplus from borrowers to savers, and allowed previously crowded-out borrowers access to credit.
  - Liberalization often increased interest rate volatility and the likelihood of banking crises unless institutional environment and supervision were strong.
- Case study summaries and lessons:
  - Nordic Countries (1978–91):
    - Rapid increases in loans to GDP; loan-to-deposit ratios surged; real interest rates increased.
    - Banking crises driven by failures in macroprudential policy, housing booms, external borrowing.
    - Key implication: active supervision and regulation crucial immediately after liberalization.
  - U.S. Savings and Loan (1980s):
    - Deposit rate liberalization combined with rising funding costs and S&Ls’ risky expansion caused failures.
    - More than 1,000 thrift institutions failed; total government/taxpayer cost > $150 billion.
    - Root causes: regulatory forbearance, weak supervision, inadequate capital enforcement.
  - Turkey (1980s):
    - Initial liberalization led to rapid rate increases, fierce competition, Ponzi schemes; crisis within two years; controls reinstated.
    - Underlying cause: weak financial infrastructure; later improvements allowed successful liberalization on second attempt.
  - Korea:
    - First attempt informal controls persisted; controls reinstated in 1989.
    - Second attempt (1991–1997) staged liberalization; Asian crisis attributed to inadequate supervision, poor prudential enforcement, and rapid external liberalization rather than interest rate liberalization per se.
- Historical context:
  - Interest rate controls and credit allocation restrictions used to support development and lower government borrowing costs (Caprio, Hanson, and Honohan, 2001).
  - Financial repression led to inefficient intermediation and lower growth (Caprio, Atias, and Hanson, 1994).

### Policy lessons and supporting measures
- Pace and timing:
  - Both too rapid and too slow reform can be damaging; ideal removal timing is when constraints are less binding (e.g., slowing economy).
  - China began reforms in 1996; final lift of deposit ceilings and lending floors remained to be done.
- Essential complements to liberalization:
  - Bank regulation and supervision: most important determinants of success; inadequate supervision underlies post-liberalization crises.
  - Monetary policy: must monitor and guard against excess demand and credit expansion inconsistent with macro stability.
  - Policy tools: strengthen indirect monetary instruments; develop money markets as interest rate restrictions are phased out.

### Conclusions specific to China (as presented)
- Modeling results suggest reform should:
  - raise the cost of capital;
  - increase the return on savings;
  - allow smaller, more efficient banks to increase their role in intermediation.
- Anticipated effects:
  - Some marginal lending discouraged, improving investment efficiency.
  - Smaller banks expanding could increase lending to underserved sectors (SMEs and households).
  - Benefits likely achievable without widespread bank unprofitability or failure; market structure likely to retain a few large banks at the core with a range of smaller competing banks.
- Policy imperatives:
  - Maintain vigilant banking supervision and monetary policy to monitor rising credit risks.
  - The People’s Bank of China (PBC) will likely need to expand its policy toolkit and rely more on indirect monetary instruments.
- Modeling caveat:
  - Exercise selective; did not focus on lending quality dynamics or post-development of new financial products; nonetheless, liberalization would likely improve aggregate loan quality by reducing marginal lending and enhancing competition for deposits.

### Data and sample (methodological notes)
- Data: balance sheet and profitability data for 20 of the largest banks for 2007.
  - These banks represent around 60 percent of total loans and deposits in the banking system.
- Bank-level series used:
  - bank-by-bank deposits and loans
  - net interest income (interest revenue less interest expenditure)
  - total and “other” non-interest expenditure
- Proxies and definitions:
  - non-interest variable costs proxied by total less “other” non-interest expenditure
  - total profits proxied by net interest income less non-interest expenses
- Deposit and loan functional forms used:
  - 1 L LL ArL ε− =
  - 1 D DD ArD ε− =
- Interbank relation specified using 2007 data and average 12 month SHIBOR:
  - ( ) 1 IB IB ArDL ε− = −

*Source: IMF staff estimates and findings as presented in _wp09171 - References (excerpts).*

### References..............................................................................................................

### _wp09171 - References..............................................................................................................

### I. INTRODUCTION
- Financial liberalization can provide many benefits: better pricing of capital and risk, improved allocation and efficiency of investment (Abiad, Oomes, and Ueda, 2004).
- Potential gains from liberalization:
  - Larger share of intermediation by more efficient banks and those better at assessing risk.
  - More credit to sectors previously underserved.
  - Expanded range of financial products for risk management and portfolio diversification.
- Risks of premature liberalization: exposure of financial sector weaknesses, inhibited intermediation, and economic volatility.
- Paper’s central question: likely impact of interest rate liberalization on China’s banking system.
- China’s status:
  - Substantial progress in financial market liberalization; remaining floor on bank lending rates and ceiling on deposit rates.
  - Deposit rate ceiling appears to bind: deposit rates clustered at benchmark; real deposit rates on average zero.
- Distortions from interest rate controls:
  - Households not appropriately compensated for savings; financial income share of total income remains one of the lowest in the world.
  - Large administratively determined interest margin reduces incentives for banks to improve efficiency and risk pricing.
  - Credit constraints (in place because interest rates do not limit credit growth) channel funds to large/well-connected enterprises, away from SMEs and households.
  - Lack of market-determined interest rates deprives the People’s Bank of China (PBC) of critical information on macroeconomic and liquidity conditions.
- Approaches used:
  - Develop and calibrate an oligopoly model of the banking system (Section II) and study interest rate deregulation effects (Section III).
  - Case studies of other countries’ experiences to infer preconditions and implications for China (Section IV).

### II. A MODEL OF CHINA’S BANKING SECTOR
- Model features:
  - Banks differ in size (deposits and loans), returns to scale (increasing and decreasing), and profitability.
  - Banks modeled as oligopolists choosing deposit (Di) and loan (Li) volumes; required reserves share α.
  - Aggregate deposit supply curve rD(D) and loan demand curve rL(L); aggregate sums defined with standard notation.
  - Deposits can be lent at lending rate rL or placed in interbank market at rate r (reserves and government bonds remunerated at r).
  - Central bank may target interbank rate or allow endogenous determination; banks take r as given.
  - Cost functions separable: deposit management cost di(Di), loan management cost li(Li), fixed cost Fi.
- Profit maximization problem (unregulated rates):
  - Banks maximize profits defined in equation (1) subject to reserve constraint (1−α)DLi ≥ 0 (multiplier λi).
  - First-order conditions yield equations (2a), (2b), and (2c) (exact functional forms preserved in source).
- Effects of reserve requirements:
  - Binding reserve requirement (λi>0) acts as a tax on lending (size λi) and subsidy on deposit taking (1−α)λi, reducing lending and increasing deposit volumes.
- Binding rate regulation:
  - Binding regulated deposit rate rD replaces equation (2a) with equation (3a); marginal revenue from deposits no longer varies with deposit volume.
  - Binding regulated lending rate rL replaces equation (2b) with equation (3b).
  - If both deposit and lending rate regulations bind, (3a) and (3b) replace analogous conditions in (2).
- Interbank market clearing condition (equation (4)):
  - DL + B = Nr, where B is net bank holdings of government and central bank paper, and Nr is net demand for funds by nonbanks.
  - If regulations not binding, deposit and loan levels reflect banks’ decisions from (2a) and (2b); if binding, supply/demand fixed at regulated rates.

### II.B Calibration
- Calibration targets: loans, deposits, profits, and costs of 20 of China’s largest banks (in 2007).
  - Note: Agricultural Bank of China excluded due to ongoing restructuring.
- Assumption: deposit rate ceiling binding; lending rate floor is not binding.
  - Supporting detail: less than one-third of loans are offered at the lending rate floor.
- Calibration success:
  - Model replicates distribution of loans, deposits, profits, and costs across banks fairly well (Figure 1).
  - Bank-by-bank deposits and profits matched exactly.
  - Distributional properties of loans (multiple mass points and highly skewed) closely replicated.
  - Model captures variable costs for most banks; difficulty with highest cost banks.

### III. HOW MAY INTEREST RATE LIBERALIZATION CHANGE CHINESE BANKING?
- Central bank operating assumptions considered:
  (i) central bank targets the interbank (IB) rate (holding it constant);
  (ii) central bank targets a spread over the now endogenously determined deposit rate;
  (iii) central bank is passive and allows interbank rate to clear endogenously.
- Four core questions addressed:
  1. Impact of liberalization on intermediation?
  2. Does central bank operating target affect reform impact?
  3. Implications for monetary policy effectiveness?
  4. How does impact depend on extent of competition?

A. The Impact of Liberalization — Intermediation and Banking System Soundness
- Key modeled outcomes of deposit rate liberalization:
  - Smaller banks offer higher deposit rates and expand; larger banks shrink deposit bases to reduce costs.
  - Deposit market concentration falls.
  - Distributional impact on lending is smaller than deposits (lending market largely liberalized already).
  - On average, profitability (as a share of loans) changes little; lower profits at large banks offset higher profits at smaller ones.
  - Net effect: deposit rate rises sufficiently to bring additional deposits into the system; loan rates and cost of capital increase; lending volume declines.
  - Share of deposits at banks making losses increases slightly (never above 3¾ percent); number of loss-making banks falls.
- Robustness across central bank procedures:
  - Trends consistent across central bank operating procedures.
  - Larger increase in interest rates if central bank targets a constant spread of interbank over deposit rate.
- Limitation on rate increases:
  - Cost of interbank funds places an effective cap on deposit rate rises; deposit rate typically below interbank rate due to deposit management costs.
  - Simulated scenario assumes little change in desired monetary conditions; deposit rate rise limited to up to 37 basis points.
  - Liberalization creates a more direct link between interbank funds cost, lending and deposit rates, and volumes.
- Table 1 (selected baseline and scenarios; values preserved as in source):
  - Calibration: Deposit Rate 8.31; Share of Lending 13.36; Deposit HHI 63.73; IB Rate 2.28; Profit/Loans 3.04; Unprofitable Banks 0.19; Deposit HHI (lending) 0.19
  - Target IB rate: Deposit Rate 8.65; Share of Lending 3.62; Deposit HHI 63.73; IB Rate 2.19; Profit/Loans 3.53; Unprofitable Banks 0.19; Deposit HHI 0.12
  - Target IB spread: Deposit Rate 8.72; Share of Lending 3.64; Deposit HHI 64.00; IB Rate 2.44; Profit/Loans 3.73; Unprofitable Banks 0.19; Deposit HHI 0.10
  - Endog. IB rate: Deposit Rate 8.67; Share of Lending 3.61; Deposit HHI 63.72; IB Rate 2.23; Profit/Loans 3.66; Unprofitable Banks 0.19; Deposit HHI 0.11
  - Notes: 1/ In percent of loans. 2/ Measured by the Herfindahl-Hirschman index.

B. Monetary Policy
- Liberalization enhances effectiveness of indirect monetary policy and interest rates as price signals.
- Comparison of monetary policy transmission with and without deposit rate liberalization (Table 2 preserves exact scenario results in percentage point deviation from baseline):
  - Target IB rate: +50 bps:
    - Lending rate: 0.34 (Liberalized) vs 0.31 (Regulated)
    - Deposit rate: 0.50 vs 0.23
    - IB rate: 0.00 vs 0.50
  - Target IB rate: −35 bps:
    - Lending rate: −0.34 vs −0.40
    - Deposit rate: −0.35 vs −0.26
    - IB rate: 0.00 vs −0.35
  - Target IB rate: −50 bps:
    - Lending rate: −0.48 vs −0.53
    - Deposit rate: −0.50 vs (not reported in snippet)
  - RR reduced (15% to 7.5%):
    - Lending rate: −0.30 vs −0.18
    - Deposit rate: 0.00 vs −0.12
    - IB rate: 0.00 vs 0.00
  - Note: Table assumes deposit rate level is regulated in the comparison; central bank cannot lower its target rate more than 37 bps if deposit rate held constant.
- Key implications:
  - Given liberalization, a given change in IB target results in larger feed-through to lending and deposit rates.
  - Tighter policy (e.g., +50 bps) results in larger increases in borrowing and lending rates when deposit rate is liberalized because deposit-demand response raises deposit rates further.
  - Central bank will need to undertake larger interventions in the money market to achieve its target when deposit rate is free.
  - Quantity-based monetary policy remains more effective under regulation when deposits unresponsive; indirect policy impact greater when deposit rate can fall.

C. Market Behavior
- Two extreme behavioral cases analyzed: aggressive competition vs collusion.
- Competition scenario:
  - Greater competition → large expansion in lending and deposit taking.
  - Deposit rate bid substantially higher; lending rate increases as cost of funds rises.
  - Margins squeezed; profitability falls; share of deposits in loss-making banks rises slightly.
  - Only one bank fails in open competition (less than 1 percent of system deposits).
  - Market concentration measures rise because largest banks expand most.
- Collusion scenario:
  - Banks scale back operations to increase margins; lending and deposit demand reduced.
  - Because deposit supply inelastic, deposit rate can be forced down significantly with small deposit loss.
  - Under collusion, liberalization can lead to lower deposit rates, indicating current market behavior likely far from collusive.
  - Largest banks cut lending and deposit taking most; measures of market concentration fall despite collusion.
  - Profitability much higher; no banks make losses.
- Tables preserving exact deviations (Table 3 and Table 4):
  - Table 3 (In percentage point deviation from baseline):
    - Competition, Target IB rate: Lending Rate 0.3; Deposit Rate 0.9; Interbank Rate 0.0; Margin −0.6
    - Competition, Target IB spread: 0.4; 1.3; 1.0; −0.9
    - Competition, Endog. IB rate: 0.4; 1.2; 2.6; −0.8
    - Cooperation (collusion), Target IB rate: −0.3; −1.0; 0.0; 0.7
    - Cooperation, Target IB spread: −1.7; −1.8; −1.9; 0.1
    - Cooperation, Endog. IB rate: −1.0; −1.4; 0.2; 0.4
    - Note: Margin = lending − deposit interest rates.
  - Table 4 (In percentage point deviation from baseline; Market Concentration and profitability measures):
    - Competition, Target IB rate: Unprofitable Banks −1.4; Profit/Loans 0.2; (Deposit share) Lending 0.05; Deposits 0.11
    - Competition, Target IB spread: −1.7; 0.5; 0.05; 0.11
    - Competition, Endog. IB rate: −1.7; 1.4; 0.05; 0.11
    - Cooperation, Target IB rate: 3.8; −3.5; −0.04; −0.06
    - Cooperation, Target IB spread: 1.4; −3.7; −0.03; −0.03
    - Cooperation, Endog. IB rate: 3.8; −3.7; −0.04; −0.05
    - Note: Concentration measured by Herfindahl-Hirschman index.

### IV. ROBUSTNESS
- Baseline elasticity assumptions:
  - Deposit supply elasticity assumed small: 0.2 (based on Ho (2008) and limited individual investment options).
  - Loan demand elasticity assumed somewhat larger: −1.
- Broad findings do not change materially for sizable changes in these elasticities (section cuts off in provided extract).

*Source: IMF staff estimates and findings as presented in _wp09171 - References (excerpts).*

### 0.5 or 1) then deposit rates increase by less than under deregulation. The larger

### IV. WHAT HAS BEEN THE EXPERIENCE WITH LIBERALIZATION ELSEWHERE?

### Impact of deposit-rate elasticity on reform outcomes
- Larger deposit-rate elasticity implies banks need to offer a smaller increase in interest rates to achieve a given change in deposit volumes.
- With larger elasticity, deposit and lending markets end up being less concentrated as smaller banks expand their presence by more than in the baseline.
- When deposit-rate elasticity is 0.5 or 1, deposit rates increase by less than under deregulation.

### Loan demand elasticity effects
- If loan demand becomes less elastic (falling to -0.5):
  - Banks become significantly more profitable after reform, reflecting greater market power.
  - Given this market power and their lending decisions, banks would even be able to reduce the deposit rate (meaning that at this elasticity, the deposit cap was not binding).
- If loan demand becomes more elastic (rising to -2):
  - Loss of market power translates into a smaller increase in the lending rate and a marginally higher deposit rate after reform.
  - The share of deposits at loss-making banks is higher, at 5.5 percent, although none fail after reform.
  - Measures of market concentration increase.

### Scenario: lending rate initially binding
- When the lending rate is initially binding, dual liberalization of deposit and lending markets results in much tighter margins:
  - Deposit rate rises.
  - Lending rate falls.
  - On balance, both deposit and lending volumes rise above their regulated levels.
- Simulations (not reported) suggest lower market concentration in both markets, and despite lower profits, banks generally remain sound (and none fail).

### Sequencing and broader international experience
- Sequencing of interest rate liberalization is a recurring issue in country experiences; most governments have controlled interest rates and credit allocation at some point.
- In China, most necessary steps toward liberalization have already been taken (Box 1), but the final step—lifting the ceiling on deposit rates and floor on lending rates—still remains.
- A key concern during liberalization is that aggressive competition for deposits and market share among banks could drive down margins, lead to bankruptcies, and create conditions for financial instability; country case studies pay particular attention to this aspect.

### Historical context and consequences of financial repression
- Interest rate controls and credit allocation restrictions were often used to maintain financial stability and support development, and to reduce governments’ borrowing costs by imposing low interest rate ceilings (Caprio, Hanson, and Honohan, 2001).
- Financial repression led to inefficient intermediation and lower growth (Caprio, Atias, and Hanson, 1994).
- Most countries removed these restrictions, often after experiencing macroeconomic crises.

*Source: Excerpt from the provided IMF document content.*

### Box 1. Key Dates in Interest Rate liberalization in China

### Box 1. Key Dates in Interest Rate liberalization in China

### Timeline of key reforms
- 1996: Abolished the upper limit on interbank lending rates
- 1997: Liberalized repo rates
- 1998–2004: Gradually increased the upper limit on lending rates
- 1999: Begun to gradually allow different institutions to negotiate rates on over Y30 million deposits with above 5-year maturity
- 2000: Liberalized foreign currency lending rates
- 2000: Liberalized foreign currency deposit rates for deposits over $3 million
- 2003: Removed floor on foreign currency deposit rates
- 2003: Liberalized deposit rates in pound, franc, Swiss Franc, and Canadian Dollar
- 2004: Liberalized all foreign currency deposit rates with maturity above 1 year
- 2004: Removed ceiling on all lending rates (except for urban and rural credit cooperatives, which have a cap of 130 percent over reference rates)
- 2004: Removed floor on all deposit rates

*Source: The People’s Bank of China (2005).*

### Cross-country outcomes and general findings from interest rate liberalization
- In most cases, interest rate liberalization:
  - led to higher real interest rates;
  - shifted surplus from borrowers to savers;
  - allowed those previously crowded out to gain access to credit.
- Liberalization often increased interest rate volatility and the likelihood of banking crises unless institutional environment and supervision were strong.
- Examples:
  - India: medium-sized enterprises gained access to credit as interest rates and credit allocation were liberalized and regulation and supervision improved significantly (Hanson, 2001).
  - Generic risk: volatility of rates and crisis likelihood rose after liberalization (Demirguc-Kunt and Detragiache, 2001).

### Case studies — major lessons and outcomes
- A. Nordic Countries
  - Pre-deregulation similarities to China: low/stable controlled interest rates, credit rationing, limited competition, dominated by banks.
  - 1978–91: most interest rates liberalized; quantitative controls initially used then eased; most external capital account restrictions abolished.
  - Rapid outcomes:
    - Ratio of loans to GDP increased (example: from 40 percent of GDP to close to 60 percent of GDP within four years in Sweden).
    - Loan-to-deposit ratio in Norway and Finland surged significantly above 1.
    - Real interest rates increased; net interest margins remained stable in Sweden but declined 60–80 basis points in Norway and Finland between 1980 and 1990.
  - Crisis drivers:
    - Banking crisis reflected failures in macroeconomic and prudential policies: excessive demand financed by external borrowing, housing booms, banks’ role in housing finance.
    - Policy failures included:
      - Prudential regulations not strengthened;
      - Borrowing encouraged via tax deductibility of interest expenses;
      - Monetary policy not tightened to rein in excess demand.
  - Key implication: active supervision and regulation are critical immediately after liberalization; scrutinize bank balance-sheet expansion and wholesale funding reliance.

- B. Savings and Loan (S&L) in the U.S.
  - 1980 law gradually liberalized deposit rates for S&Ls (within 6 years).
  - Mismatch: S&Ls had long-term low-rate 30-year mortgages but faced sharply rising funding costs as the Fed raised rates.
  - Response: aggressive expansion into real estate lending with insufficient risk control.
  - Outcomes:
    - More than 1,000 thrift institutions—with total assets more than $500 billion—went bankrupt.
    - Total cost to the government and taxpayers amounted to more than $150 billion (Curry and Shibut, 2000).
  - Root causes: regulatory forbearance and moral hazard, weak supervision, extension into unfamiliar businesses, inadequate enforcement of capital requirements.

- C. Turkey
  - 1980: lifted restrictions on deposit and lending rates; interest rates surged and fierce competition ensued, including Ponzi schemes by brokerage firms (Saracoglu, 1996).
  - Within two years the financial system faced a crisis: several banks and brokerage houses collapsed; authorities reinstated interest rate controls.
  - Underlying cause: unusually weak financial infrastructure (no capital adequacy requirement, inadequate classification/provisioning, undeveloped interbank market).
  - Government response: improved supervision, established interbank market; reforms allowed successful liberalization during 1984–88 on a second attempt.

- D. Korea
  - First attempt (1980s): formal liberalization but rates effectively controlled via informal window guidance; increased competition encouraged by lowering entry barriers; reinstated controls in 1989 after economic shocks.
  - Second attempt (1991–1997): four-step program liberalizing short-term lending rates, remaining lending rates, long-term deposit rates, and most remaining deposit rates (by 1997).
  - The Asian crisis hit as final stage was completed. Analyses attribute the crisis to inadequate supervision, poor enforcement of prudential rules, inappropriate monetary and exchange rate policies, and overly rapid external financial liberalization rather than interest rate liberalization per se.

### Policy lessons and supporting measures
- Pace of reform:
  - China began earnest liberalization in 1996 and took gradual steps; over the five years prior to publication progress stalled with deposit rate floors and lending rate ceilings remaining.
  - Both too rapid and too slow reform can be damaging; China’s total reform span could reach 13 years if finalized.
- Timing:
  - Ideal time to remove constraints is when they are less binding (e.g., during a slowing economy).
  - Removing constraints can expose macroeconomic and structural imbalances (as in the Nordic example).
- Supporting measures — essential complements to interest rate liberalization:
  - Bank regulation and supervision:
    - Proper regulation and supervision are the most important determinants of success; inadequate supervision underlies all banking crises after liberalization.
    - Prevent destructive competition where banks enter excessively risky businesses or support excess demand.
  - Monetary policy:
    - Must be vigilant against excess demand when constraints are removed; loosened restrictions often spur credit expansion inconsistent with macro stability.
    - In China’s context, current economic slowdown may allow some credit expansion but growth should be watched to ensure risk control and consistency with macro goals.
  - Policy tools:
    - Strengthen indirect monetary policy instruments as direct controls (interest rate and credit quantity) are phased out.
    - Removing interest rate restrictions will catalyze further development of money markets; indirect instruments become more important as money demand becomes less predictable.

### Conclusions specific to China (as presented)
- Modeling results suggest reform should:
  - raise the cost of capital;
  - increase the return on savings;
  - allow smaller, more efficient banks to increase their role in intermediation.
- Anticipated effects:
  - Some marginal lending discouraged, improving investment efficiency.
  - Smaller banks expanding could increase lending to currently underserved sectors (e.g., SMEs and households).
  - Benefits achieved without widespread bank unprofitability or failure; market structure likely to retain a few large banks at the core with high liquidity and a range of smaller competing banks.
- Policy imperatives:
  - Maintain vigilant banking supervision and monetary policy to monitor rising credit risks.
  - The People’s Bank of China (PBC) will likely need to expand its policy toolkit and rely more on indirect monetary instruments rather than direct rate or quantity controls.
- Additional modeling caveat:
  - The modeling exercise was selective and did not focus on lending quality dynamics or post-development of new financial products; nonetheless, liberalization would likely improve aggregate loan quality by reducing marginal lending and enhancing competition for deposits.

*Italic: Source — The People’s Bank of China (2005); content as presented in the provided excerpt.*

### 2007. We use balance sheet and profitability data for 20 of the largest banks for 2007.

### _wp09171 - 2007. We use balance sheet and profitability data for 20 of the largest banks for 2007.

### Data and sample
- Balance sheet and profitability data used for 20 of the largest banks for 2007.
- These banks represent around 60 percent of total loans and deposits in the banking system.
- Bank-level series used:
  - bank-by-bank deposits and loans
  - net interest income (interest revenue less interest expenditure)
  - total and “other” non-interest expenditure
- Proxies and definitions:
  - non-interest variable costs proxied by total less “other” non-interest expenditure
  - total profits proxied by net interest income less non-interest expenses
- Exclusions:
  - The Agricultural Bank of China excluded from the sample because its restructuring was ongoing during 2007 and did not end until late-2008.

### Demand and supply curve parameterization
- Parameters chosen using past elasticity estimates and aggregate market information.
- Known quantities used to solve for constant terms: aggregate lending and deposits and average interest rates from the 20 banks in the sample during 2007.
- Elasticity choices:
  - deposit elasticity = 0.2 (based on Ho (2008))
  - loan demand elasticity = -1
- Functional forms (as stated):
  - 1 L LL ArL ε− =
  - 1 D DD ArD ε− =
- Deposit supply and loan demand constants chosen to be consistent with average implied interest rates from 2007 activities of the 20 banks.

### Cost function calibration
- Identification constraint:
  - Five cost function parameters per bank ( , , , , lidilidii aa bbF) but only four data series per bank available → one restriction imposed.
  - Restriction chosen: 0 di a = for each bank.
- Deposit-side calibration:
  - Deposit rate assumed to bind, so equation (3a) implies di b can be chosen to match the deposit share:
    - 2 Ddi diD SHi rr a bD D − − =
    - where D is a parameter representing how binding the deposit rate cap is, and SHi D is the share of deposits at bank i.
- Lending-side calibration:
  - Two variable cost terms for lending (li a and li b) chosen to minimize the distance between bank-by-bank lending and non-interest variable costs (as a share of loans).
  - Estimation method: Nelder-Mead Simplex algorithm.
- Fixed cost calibration:
  - With estimates of cost parameters, bank-specific fixed cost i F is calculated to equate total profits of the banks in the model with actual profits in 2007.

### Interbank rate determination and reform experiments
- Reform experiments include simulation when the central bank does not have a specific interbank rate target and the interbank rate is determined endogenously.
- Assumption: there is a net demand curve by nonbank participants for interbank funds which includes the central bank; given the central bank is included, demand is more elastic than the demand for aggregate loans.
  - IB ε = 2.5
- Using 2007 data on deposit, loan and the average 12 month SHIBOR rate, the interbank relation is specified as:
  - ( ) 1 IB IB ArDL ε− = −

### Key methodological notes
- Four data series per bank used to identify cost parameters: lending, deposits, non-interest variable costs (proxied), and total profits (proxied).
- Calibration strategy balances matching deposit shares, minimizing deviations in lending variable costs, and equating modeled profits to observed 2007 profits.

*Source: _wp09171 - 2007. We use balance sheet and profitability data for 20 of the largest banks for 2007.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09171.pdf_