## _wp1150

## Source details

**Canonical URL:** [_wp1150](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1150.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1150.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1150.pdf.json)

---

### Background and context
- Liberalization of the Indian banking sector began in 1991; despite reforms "the basic ownership structure of the existing banks remained broadly similar even many years after liberalization had first started."
- The banking system remained predominantly state-owned: private banks’ share of total banking sector assets increased from 3.5 percent in 1991 to about 21 percent in 2007, while nearly 70 percent of banking sector assets belonged to state owned banks as of 2007.
- Fiscal context:
  - The combined fiscal deficit of the federal and state governments averaged nearly 7.8 percent of GDP between 1992 and 2007.
  - Public debt totaled about 80 percent of GDP in 2007.
  - "The gross combined fiscal deficit of federal and state governments declined steadily between 2002 and 2007, from nearly 10 percent in 2002 to about 5 ½ percent in 2007."

### Empirical question, data, and identification
- Research question: Using bank-level data for 1991–2007, did banks redeploy resources freed by reduced state-preemption (statutory requirements to hold government securities and cash) to increase credit to the private sector?
- Data and sample:
  - Scheduled commercial banks data from RBI banking statistics for 1991–2007.
  - Bank-level panel covering SBI and associates, other public sector banks, domestic private banks; foreign banks excluded from main analysis.
  - Sample size for many regressions: Number of Banks = 52; Observations = 787 (panel regressions) and 294 (short-sample comparisons).
- Key regulatory variables:
  - Cash Reserve Requirement (CRR) varied between 5 and 15 percent since 1990; CRR progressively reduced (e.g., from a peak of 15 percent to about 5 percent as liberalization progressed).
  - Statutory Liquidity Requirement (SLR) varied between 25 to 40 percent since 1990; SLR reduced from 38.5 percent to 25 percent between 1992 and 1998 and remained at 25 percent until November 2008.

### Cash holdings and CRR — main empirical findings
- Baseline specification: dependent variable Cash/Assets_it (percent); key regressor CRR_t and interaction CRR_t * Dummy for Public banks.
- Table 3 (Cash/Assets regressions) — selected coefficients (Cash/Assets in percent):
  - CRR:
    - Column I: .94***  [23.07]
    - Column II: .90*** [22.76]
    - Column III: .95*** [23.19]
    - Column IV: .90*** [21.00]
    - Column V: .90*** (reported)
  - CRR * Public Banks:
    - Column I: -0.26***  [4.88]
    - Column II: -0.27*** [4.99]
    - Column III: -0.20*** [3.84]
    - Column IV: -0.24*** [4.27]
    - Column V: -0.20*** [3.32]
- Interpretation:
  - Private banks’ cash ratio responds to CRR with coefficient in range 0.90–0.95 (statistically close to 1).
  - Public banks’ response is α+β ≈ 0.70 (statistically less than 1), indicating public banks reduce cash ratio by less than private banks when CRR declines.
- GDP growth interaction:
  - Column II: GDP Growth Lag coefficient -0.21***  [2.83].
  - Column III: GDP Growth Lag * Public Banks = 0.35***  [3.53]; implies:
    - Private banks reduce cash holdings by 0.21 percent for every one percentage point increase in GDP growth.
    - Public banks increase cash holdings by 0.14 percent for an equivalent increase in GDP growth (0.35 - 0.21 = 0.14).
- Short-sample (1994–1999) comparison (Table 4):
  - CRR dummy (1997–1999 vs 1994–1996): CRR dummy coefficients in Columns I–IV: -4.36***, -4.32***, -3.99***, -4.37*** ([9.20], [9.13], [-.03], [9.25]).
  - CRR dummy * Public Banks: 1.94***, 2.06***, 1.51*, 2.14*** ([3.05], [3.01], [1.93], [3.11]).
  - Average cash ratio declined by about 4 percentage points for private banks between 1994-1996 and 1997-1999; declined by about 2 percentage points for public banks over same periods; difference statistically significant.
- Robustness: results hold with year fixed effects and with interactions controlling for size and profitability.

### Investment in government and approved securities, SLR, and fiscal deficit
- Baseline specification: dependent variable Investment in Govt Sec/Assets (percent); key regressors SLR_t and Combined Fiscal Deficit_t, with interactions for Public banks.
- Table 5 (Column I): decline in SLR associated with an increase in investment in government securities for both private and public banks; increase larger for public banks (initial counterintuitive finding).
- Table 5 (Column II benchmark) — selected coefficients:
  - SLR: 0.08 [1.65]
  - SLR*Public Banks: 0.03 [0.55]
  - Fiscal Deficit: 0.98*** [4.82]
  - Fiscal Deficit* Public Banks: 0.60*** [2.65]
  - Observations: 787; R-squared: .59; Number of Banks: 52
- Quantitative interpretation:
  - Banks increase their share of assets in government and approved securities by 1 percentage point for a 1 percentage point increase in the gross fiscal deficit to GDP ratio.
  - Public banks increase investment by 1.6 percentage points for the same fiscal deficit increase (1 + 0.6).
- Period comparison (Table 6):
  - 1995-1997: SLR averaged 33.2 percent; Combined Fiscal Deficit averaged 6.6 percent of GDP; Investment in Govt and Approved Securities (% of assets): Private Banks 42.5; Public Banks 53.3.
  - 1999-2001: SLR averaged 25 percent; Combined Fiscal Deficit averaged 9.3 percent of GDP; Investment: Private Banks 46.8; Public Banks 58.4.
  - 2007: SLR 25; Combined Fiscal Deficit 5.6; Investment: Private Banks 44.3; Public Banks 47.3.
- Interpretation: changes in fiscal deficit, not SLR, drive banks’ decisions to invest in government securities; public banks respond more strongly than private banks.

### Credit to the private sector, SLR, CRR, and fiscal deficit
- Dependent variable: Credit to Private Sector/Assets (percent); Table 7 key results:
  - Fiscal Deficit coefficients:
    - Fiscal Deficit: -1.08***, -0.97***, -1.03***, -1.11*** ([4.12], [3.25], [4.05], [4.19])
  - Fiscal Deficit* Public Banks: -0.69**, -0.63**, -0.88**, -0.71***, -0.62** ([2.46], [2.30], [2.56], [2.68], [2.22])
  - CRR*Public Banks (in specifications including CRR): -0.28***, -0.43*** ([2.94], [2.72]).
  - Observations: 787; R-squared around 0.64–0.66; Number of Banks: 52
- Quantitative interpretation (Column I):
  - For a 1 percentage point increase in the fiscal deficit, private banks decrease the share of credit to private sector in their assets by 1 percentage point.
  - Public banks decrease by 1.77 percentage points (1 + 0.77) for a 1 percentage point increase in the fiscal deficit (Column I interpretation).
- Overall finding: public banks reduce credit to the private sector by a larger amount in response to an increase in the fiscal deficit than do private banks.

### Robustness and supplementary tests
- Robustness checks reported include:
  - Separate effects when fiscal deficit increasing vs decreasing (1992–2002 and 2003–2007): fiscal deficit coefficients positive for investments and larger for public banks.
  - Central vs state government deficits: Central Govt Deficit * Public Banks = 0.86*** [2.95] (stronger effect).
  - Different coefficients for new private banks vs old private banks: public sector results persist.
  - Inclusion of bond yields: insignificant; does not alter main results.
  - Inclusion of fiscal deficit in cash holdings regressions: insignificant effect on cash holdings.
  - Inclusion of capitalization (capital injection as percent of bank’s assets) in investment regressions: positive but insignificant; does not affect other coefficients.

### Explanations for differential behavior of public banks (evidence-based conclusions)
- Profitability motive:
  - Regressions of return on assets (Table 9) show public banks are less profitable after controlling for size:
    - PSBs Dummy: -.47***, -.29***, -.26*** (Columns I–III) with t-stats [-6.79], [-3.59], [-3.28].
  - Investment in Government Securities associated with lower profitability:
    - Investment in Govt Securities: -.03***, -.19*** (Columns I–II) [-4.30], [-5.30].
  - Investment in Govt Securities * Fiscal deficit: .02*** [4.95] — investments associated with relatively higher returns during periods of high fiscal deficits.
  - Interpretation: evidence contradicts hypothesis that public banks hold government securities primarily to maximize profits.
- Demand-side constraints:
  - Table 10 includes private capital formation and other demand proxies:
    - Private capital formation/GDP, Lagged: -0.86***, -0.68*** (investment regressions) and 0.33*, 0.16 (private credit regressions) with t-stats [-5.58], [-4.85], [1.93], [0.92].
    - External Commercial Borrowings/GDP, Lagged: -2.76*** (investment) [-12.1].
    - Stock Market Capitalization/GDP, Lagged: -0.04*** and .06*** (investment and private credit) [-3.85], [3.57].
  - Demand-side variables matter, but fiscal deficit * public bank interaction remains significant, indicating supply-side/ownership effects beyond demand constraints.
- Institutional and incentive explanations supported by evidence and narrative:
  - Asymmetric incentive structures in public banks: "Greater credit expansion or higher profit making is not rewarded but a loan decision that turns bad can be punished."
  - Moral suasion / political pressure: public banks "are perhaps coaxed to hold government securities when the deficit is high."
  - Synthesis: overinvestment by public banks in government securities likely due to a combination of "lazy" behavior, moral suasion, and prioritization of government financing needs rather than profit maximization or risk mitigation.

### Contribution, policy implications, and conclusions
- Main contributions:
  - Demonstrates that reduced statutory preemption (CRR and SLR reductions) did not translate into a proportional increase in private sector credit at the bank level during 1991–2007.
  - Establishes ownership as a key determinant: public banks reallocated less toward private credit and more toward government securities compared to private banks.
  - Shows fiscal deficits, not regulatory SLR reductions, are the primary driver of banks’ investments in government securities; public banks react more strongly to fiscal deficits.
- Policy implications:
  - In developing countries with large government deficits and substantial public ownership of banks, financial liberalization may yield limited gains in private credit provision.
  - Reductions in state preemption (CRR and SLR) are not sufficient by themselves to increase private sector credit if ownership and incentive structures channel freed resources toward government financing.
- Empirical caution: results robust across multiple specifications and robustness checks documented in the paper.

### Key descriptive and sample statistics (selected values preserved)
- Sample sizes and panel metrics:
  - Panel regressions: Number of Banks = 52; Observations = 787 (typical).
  - Short-sample comparisons: Observations = 294; Number of Banks = 52.
- Appendix C summary statistics (selected entries preserved exactly):
  - Return on Assets:
    - Number of Observations: 7870; Mean: .521; Std. Dev.: .10; Minimum: -7.51; Maximum: 2.34
  - Cash/Assets:
    - Number of Observations: 7879; Mean: .044; Std. Dev.: .13; Minimum: 2.47; Maximum: 26.26
  - Credit (Other)/Assets:
    - Number of Observations: 7872; Mean: 3.95; Std. Dev.: 7.47; Minimum: 0.00; Maximum: 59.51
  - Investment in Government Securities/assets:
    - Number of Observations: 7872; Mean: 5.11; Std. Dev.: 6.05; Minimum: 11.29; Maximum: 44.74
  - Investment in Approved Securities/Assets:
    - Number of Observations: 7872; Mean: 8.05; Std. Dev.: 5.86; Minimum: 11.29; Maximum: 46.18
  - Statutory Liquidity Ratio:
    - Number of Observations: 172; Mean: 9.47; Std. Dev.: 5.73; Minimum: 25; Maximum: 38.5
  - Combined Fiscal deficit:
    - Number of Observations: 177; Mean: .91; Std. Dev.: 1.37; Minimum: .58; Maximum: 9.94

*Source: Excerpted content from IMF working paper _wp1150 (selected sections and appendices provided in the content unit).*

### 1992. However, despite these measures, the basic ownership structure of the existing banks

### _wp1150 - 1992. However, despite these measures, the basic ownership structure of the existing banks

### Background and context
- Liberalization of the Indian banking sector began in 1991; many measures were introduced but "the basic ownership structure of the existing banks remained broadly similar even many years after liberalization had first started."
- Despite increases in private and foreign bank shares of total assets, the Indian banking system "has remained predominantly state-owned."
- Fiscal context:
  - The combined fiscal deficit of the federal and state governments averaged nearly 7.8 percent of GDP between 1992 and 2007.
  - Public debt totaled about 80 percent of GDP in 2007.
  - Note: "The gross combined fiscal deficit of federal and state governments declined steadily between 2002 and 2007, from nearly 10 percent in 2002 to about 5 ½ percent in 2007."

### Empirical question and data
- Research question: Using bank-level data from 1991-2007, did banks redeploy resources freed by reduced state-preemption (statutory requirements to hold government securities) to increase credit to the private sector?
- Key sample/timeframe: bank-level data covering 1991-2007.

### Main findings
- Banks did not increase credit to the private sector commensurate with the decline in statutory requirements to hold government securities that accompanied liberalization.
- Public banks and private banks behaved differently in redeploying resources freed by reduced state preemption:
  - "Public banks appear to voluntarily allocate relatively more resources to finance the fiscal deficit."
- Ownership appears to be a determinant of effective crowding out of private sector credit at the bank level.
- The greater willingness of public sector banks to hold public debt over private banks:
  - "Does not appear to be due to the objective to maximize profits or to lower the risk profile of their assets, or indeed due to the lower demand for credit by the private sector."

### Plausible explanations consistent with evidence
- Incentive structure in public banks: "Greater credit expansion or higher profit making is not rewarded but a loan decision that turns bad can be punished."
- Moral suasion: public banks "are perhaps coaxed to hold government securities when the deficit is high."

### Relation to existing literature
- Financial liberalization and financial deepening:
  - Tressel and Detragiache (2008) find context-dependent effects; positive short-run effects for developing countries and stronger effects where political institutions provide stronger protection of property rights.
  - The paper suggests the effect of liberalization on private credit availability may also depend on state ownership of banks and the size of the fiscal deficit.
- Allocation-of-investment literature:
  - Studies like Galindo et al (2007) and Chari and Henry (2008) show liberalization improves allocation across firms, but they do not analyze allocation between government and private sectors.
- Government-owned banks literature:
  - La Porta, Lopez-de-Silanes and Shleifer (2002): contrasts "development" vs "political" views; find higher government ownership associated with slower financial development and lower productivity growth.
  - Hauner (2008, 2009): contrasts "safe asset" view and "lazy bank" view; public debt can both lend safety and impart laziness to banks.
  - The paper's results support La Porta et al. and Hauner by suggesting government ownership and fiscal deficit size may together limit liberalization's impact on financial development.
- Indian banking sector studies:
  - Banerjee, Cole and Duflo (2004) and Cole (2004) characterize Indian public banks as "lazy"—not lending adequately to the private sector and basing lending decisions on past turnovers and outlays rather than profitability. This lazy behavior is also manifested in overinvestment in government securities, "especially in the states where, for various institutional reasons, it might be particularly difficult and costly to scrutinize the private sector applications for credit."
  - This paper adds by focusing on changes in regulations pertaining to asset allocation and comparing responses of public and private banks.

### Contribution and implications
- Demonstrates that reduced statutory preemption did not translate into a proportional increase in private sector credit at the bank level during 1991-2007.
- Highlights the role of bank ownership and incentive structures in determining whether liberated bank resources finance private activity or the fiscal deficit.
- Suggests policy relevance: the impact of financial liberalization on financial development depends on ownership structure of banks and the fiscal stance.

*Source: IMF working paper content (excerpt) provided in the content unit.*

### Section V explains the findings, and the last section concludes.

### _wp1150 - Section V explains the findings, and the last section concludes.

### Financial sector liberalization and bank ownership in India
- At independence in 1947 the Indian banking sector consisted of domestic private and foreign banks.
- Public ownership introduced and expanded in three waves:
  - 1955: Imperial Bank of India taken over and renamed State Bank of India (SBI); 7 banks taken over as subsidiaries by SBI — resulting in 8 public sector banks, 53 private sector banks and 15 foreign banks.
  - 1969: Fourteen of the largest private banks (each with deposits greater than Rs. 50 crores or Rs. 500 million) nationalized.
  - 1980: Six more banks with deposits above Rs. 2 billion nationalized.
- By 1982 private and foreign banks accounted for less than 10 percent of banking sector assets.
- Reserve Bank of India priority sector lending guidelines:
  - 1974: At least one-third of aggregate advances to the priority sector.
  - 1980: Quota increased to 40 percent with sub-targets for agriculture and weaker sectors.
- Two main regulatory instruments governing asset allocation:
  - Cash Reserve Requirement (CRR): prescribed as a percentage of net demand and time liabilities; has varied between 5 and 15 percent since 1990.
  - Statutory Liquidity Requirement (SLR): requires banks to maintain between 25 to 40 percent of demand and time liabilities in cash, gold, or approved government securities; ratio has varied between 25 percent and 38.5 percent since 1990.
- These restrictions contributed to severe financial repression and set the stage for reforms beginning in 1991.

### A. Financial liberalization in India
- Reforms in early 1990s included removal of interest rate controls, reductions in reserve and liquidity ratios, entry deregulation, relaxation of credit controls, introduction of inter-bank money market and auction-based repos and reverse repos.
- Abiad, Detragiache and Tressel (2010) financial liberalization index:
  - Index aggregates seven dimensions: credit controls and reserve requirements, interest rate controls, entry barriers and state ownership, policies on securities markets, banking regulations, restrictions on capital account.
  - Normalized between zero and one.
- India’s financial sector was highly repressed until late 1980s; liberalization gathered pace in early to mid 1990s and by late 1990s India had substantially narrowed the gap with emerging Asia and the world average.
- Note on index extension: values for India in 2006 and 2007 are assumed to be at the same level as in 2005.

### B. Ownership structure in the Indian banking sector
- Liberal entry allowed in the 1990s led to changes in ownership structure:
  - Number of private banks increased in mid-1990s, then declined due to mergers or closures.
  - Number of foreign banks increased through 1980s and mid-1990s, then declined.
  - Total number of banks peaked at 105 in mid-1990s and declined to 82 by 2007.
- Table 1: Number of Banks and Share in Assets in 2007, by ownership groups
  - State Bank of India and its 7 associates banks: No of Banks 8; Share in Assets (%) 23.3
  - Other Public Sector Banks: No of Banks 20; Share in Assets (%) 47.2
  - Private Banks: No of Banks 25; Share in Assets (%) 21.5
  - Foreign Banks: No of Banks 29; Share in Assets (%) 8.0
- Market share changes:
  - Private banks’ share of total banking sector assets increased from 3.5 percent in 1991 to about 21 percent in 2007.
  - Nearly 70 percent of banking sector assets belong to state owned banks as of 2007.

### C. Statutory requirements on cash and liquidity and allocation of credit
- CRR reductions:
  - CRR brought down from a peak of 15 percent in early 1990s to about 5 percent as liberalization progressed.
  - Public and private banks reduced actual cash holdings as percent of assets commensurate with decline in CRR.
- SLR reductions:
  - SLR progressively reduced from 38.5 percent to 25 percent between 1992 and 1998; remained at 25 percent until November 2008, lowered to 24 percent in November 2008, and increased back to 25 percent in November 2009.
- Public banks invested a larger share of assets in government securities; share of government securities in public bank assets increased post liberalization.

### III. Competition and efficiency indicators post liberalization
- Concentration:
  - Herfindahl Index (based on asset shares for all banks including foreign banks) shows a sharp decline in concentration since early 1990s, with another decline from early 2000s.
- Profitability and efficiency:
  - Public sector banks made remarkable progress in profitability; by 2007 public banks were broadly at par with private banks.
  - Charts and table evidence show convergence driven mainly by cost-side improvements (decline in wage costs, decline in interest paid, and dramatic decline in loan loss provisions for public banks).
- Persistent differences in allocation:
  - Public banks continued to allocate a lower share of assets to private sector credit compared with private banks, and a higher share to government securities, even after CRR and SLR reductions.

### IV. Statutory requirements and allocation of assets post liberalization — analysis approach
- Research question: To what extent do public and private banks change cash holdings when CRR is reduced, and reallocate assets between government securities and private sector credit when SLR is reduced?
- Data:
  - Scheduled commercial banks data from RBI banking statistics for 1991–2007.
  - Bank-level data for SBI and its associates, other public sector banks, domestic private banks, and foreign banks.
  - Analysis limited to public banks and domestic private banks (foreign banks excluded due to limited branch expansion and different business model).
  - Bank mergers and name changes recorded to build panel data set.

### B. Cash holdings and the Cash Reserve Requirements (empirical specification)
- Dependent variable: Cash/Assets_it (cash holdings of bank i in year t as percent of assets).
- Right-hand-side controls: bank fixed effects, GDP growth (proxy for demand for bank credit and macro environment), bank size (share of bank’s assets in total banking sector assets), bank health (return on assets), CRR_t, and interaction CRR_t * Dummy for Public banks.
- Interpretation:
  - Coefficient α on CRR estimates effect of CRR change on private banks.
  - α+β (sum of CRR and interaction coefficient) estimates effect on public banks.
  - Negative and significant β implies public banks change cash ratio by a smaller amount than private banks in response to CRR changes.
- Equation (1):
  (Cash/Assets)_it = γ_i Bank Dummies_i + α CRR_t + β CRR_t * Dummy for Public banks_i + δ GDP Growth_t + λ (Bank Characteristics: Size_it, Return on Assets_it) + ε_it

### Empirical results (Table 3 summary)
- Sample: 52 banks; Observations 787; R-squared ranges 0.66–0.68 across specifications.
- Key coefficients (dependent variable: Cash/Assets in percent):
  - CRR:
    - Column I: .94***  [23.07]
    - Column II: .90*** [22.76]
    - Column III: .95*** [23.19]
    - Column IV: .90*** [21.00]
    - Column V: .90*** (reported in table)
  - CRR * Public Banks:
    - Column I: -0.26***  [4.88]
    - Column II: -0.27*** [4.99]
    - Column III: -0.20*** [3.84]
    - Column IV: -0.24*** [4.27]
    - Column V: -0.20*** [3.32]
  - Interpretation:
    - Private banks’ cash ratio responds to CRR with coefficient in range 0.90–0.95 (statistically close to 1).
    - Public banks’ response is α+β ≈ 0.70 (statistically less than 1), indicating public banks reduce cash ratio by less than private banks when CRR declines.
- Other findings across columns:
  - GDP Growth Lag coefficients:
    - Column II: -0.21***  [2.83]
    - In Column III an interaction GDP Growth Lag * Public Banks = 0.35***  [3.53], implying:
      - Private banks reduce cash holdings by 0.21 percent for every one percentage point increase in GDP growth.
      - Public banks increase cash holdings by 0.14 percent for an equivalent increase in GDP growth (0.35 - 0.21 = 0.14).
  - Size (share in assets) and interactions:
    - Size coefficients vary across specifications; interaction CRR * Size is negative and significant in Columns IV and V: -0.01** and -0.01*** respectively ([2.45], [2.78]).
    - CRR * Return on Assets = 0.06** [2.43] in one specification.
  - Return on Assets entered alone in some specifications with coefficient -0.56* [1.81] (indicating more profitable banks hold lower cash ratios and reduce less sharply).
- Robustness:
  - Results hold when including year fixed effects and when interacting GDP growth with public bank dummy.
  - Interaction of size and returns with CRR included to ensure ownership dummy is not proxying other bank characteristics; core result persists that public banks reduce cash ratio less than private banks.

### Experimental comparison approach (brief)
- To isolate CRR effect further the authors compare cash ratios across nearest time periods when CRR declined significantly:
  - Compare years 1994–1996 (CRR ranged between 14 ¼ -14 ¾ percent) with years 1997–1999 (CRR ranged between 10- [text truncated in source]).

*Italic source attribution: Content derived from _wp1150 - Section V explains the findings, and the last section concludes.*

### 11.64 percent. Since we now are looking at a shorter time period, when perhaps other

### _wp1150 - 11.64 percent. Since we now are looking at a shorter time period, when perhaps other

### Cash holdings, CRR, and ownership (alternative approach using 1994–1999)
- Data sample: 1994-1999 with a dummy = 0 for 1994-1996 and 1 for 1997-1999; dummy interacted with ownership dummy.
- Main result:
  - Average cash ratio declined by about 4 percentage points for private banks between 1994-1996 and 1997-1999.
  - Average cash ratio declined by about 2 percentage points for public sector banks over the same periods.
  - The difference in response between private and public banks is statistically significant.
- Specification checks (Table 4 columns):
  - Column II: include interaction of CRR dummy with size to rule out ownership as proxy for size.
  - Column III: include interaction of returns on assets with CRR dummy to rule out ownership as proxy for profitability.
  - Column IV: include GDP growth (proxy for demand conditions) and interact GDP growth with public banks dummy; find that, controlling for CRR, private banks lower cash holdings in years of higher GDP growth while public banks do not.
- Regression details (selected coefficients from Table 4):
  - Lagged, GDP Growth: -0.06, -0.06, -0.08, -0.33** (with corresponding t-statistics [0.56], [0.57], [0.77], [2.11])
  - Size (share in assets): 0.47, 0.18, 0.60, 0.14 ([0.85], [0.25], [0.83], [0.19])
  - Return on Assets: 0.32* [1.93]
  - CRR dummy: -4.36***, -4.32***, -3.99***, -4.37*** ([9.20], [9.13], [-.03], [9.25])
  - CRR dummy*Public Banks: 1.94***, 2.06***, 1.51*, 2.14*** ([3.05], [3.01], [1.93], [3.11])
  - Observations: 294 (all columns)
  - R-squared: 0.54, 0.54, 0.55, 0.55
  - Number of Banks: 52
- Interpretation: evidence points to relatively passive behavior of public banks in reallocating assets towards earning assets even when required to hold less cash.

### Investment in government and approved securities, SLR, and fiscal deficit
- Framework: Equation (2) — dependent variable = Investment in Govt Sec/Assets (in percent); key regressors include SLR, fiscal deficit (combined federal and state), GDP growth, bank characteristics, and interactions with public banks dummy.
- Counterintuitive initial result (Table 5, Column I): decline in SLR associated with an increase in investment in government securities for both private and public banks; increase larger for public banks.
- When fiscal deficit included (Table 5, Column II):
  - SLR coefficients become not significantly different from zero.
  - Fiscal deficit coefficients are positive and significant.
  - Interaction Fiscal Deficit*Public Banks positive and significant.
  - Quantitative interpretation: banks increase their share of assets in government and approved securities by 1 percentage point for a 1 percentage point increase in the gross fiscal deficit to GDP ratio; public banks increase investment by 1.6 percentage points (i.e., 1 + 0.6).
- Selected coefficients from Table 5 (Column II as benchmark):
  - Lagged, GDP Growth: 0.06 [0.65]
  - Size (share in assets): -0.32 [-1.33]
  - SLR: 0.08 [1.65]
  - SLR*Public Banks: 0.03 [0.55]
  - Fiscal Deficit: 0.98*** [4.82]
  - Fiscal Deficit* Public Banks: 0.60*** [2.65]
  - Observations: 787; R-squared: .59; Number of Banks: 52
- Robustness checks (Table 5 other columns and narrative):
  - Include interaction of GDP growth and ownership (Column III).
  - Include year fixed effects (Column IV).
  - Include bank characteristic interactions (Column V).
  - Include capitalization (capital injection as percent of bank’s assets) — positive but insignificant coefficient (last column).
- Period comparison (selective years):
  - Two periods compared: 1995-1997 (SLR averaged 33.2 percent) vs 1999-2001 (SLR averaged 25 percent).
  - Combined fiscal deficit increased from an average of 6.6 percent of GDP during 1995-1997 to 9.3 percent during 1999-2001.
  - Investment in government securities increased and the increase was sharper for public banks (Table 6):
    - I: 1995-1997 — SLR 33; Combined Fiscal Deficit 6.6; Investment in Govt and Approved Securities (% of assets): Private Banks 42.5; Public Banks 53.3
    - II: 1999-2001 — SLR 25; Combined Fiscal Deficit 9.3; Investment: Private Banks 46.8; Public Banks 58.4
    - III: 2007 — SLR 25; Combined Fiscal Deficit 5.6; Investment: Private Banks 44.3; Public Banks 47.3
- Chart 10 (described): plots relationship between fiscal deficit and investment in government securities for public and private banks; relationship positive for all banks and steeper for public banks.
- Interpretation: changes in fiscal deficit, not SLR, drive banks’ decisions to invest in government securities; public banks respond more strongly than private banks.

### Credit to the private sector, SLR, CRR, and fiscal deficit
- Dependent variable: Credit to Private Sector/Assets (in percent); results summarized in Table 7.
- Main findings:
  - Column I: share of assets extended to private sector declines for all banks when fiscal deficit increases; decline is sharper for public banks.
  - Quantitative interpretation (Column I): for a 1 percent increase in the fiscal deficit, private banks decrease the share of credit to private sector in their assets by 1 percentage point; public banks decrease by 1.77 percentage points (i.e., 1 + 0.77).
- Selected coefficients from Table 7:
  - Size (share in assets): 1.56* [1.72] (other columns vary)
  - Lagged, GDP Growth: 0.06 [0.49] (various columns)
  - SLR: -0.11 [0.85] or small positive/insignificant in other columns
  - SLR*Public Banks: -0.14, -0.08, -0.19* (t-stats [1.60], [0.89], [1.85])
  - Fiscal Deficit: -1.08***, -0.97***, -1.03***, -1.11*** ([4.12], [3.25], [4.05], [4.19])
  - Fiscal Deficit* Public Banks: -0.69**, -0.63**, -0.88**, -0.71***, -0.62** ([2.46], [2.30], [2.56], [2.68], [2.22])
  - CRR*Public Banks: -0.28***, -0.43*** ([2.94], [2.72]) in columns including CRR interactions.
  - Observations: 787; R-squared around 0.64–0.66; Number of Banks: 52
- Specification variations:
  - Column II: include year dummies.
  - Column III: include GDP growth*public banks interaction.
  - Column IV: include CRR and CRR*public banks.
  - Column V: include SLR, fiscal deficit, CRR and their interactions with public bank dummy.
- Overall result: public banks reduce credit to private sector by a larger amount in response to an increase in the fiscal deficit than do private banks.
- Additional note: when using public bank dummy instead of bank fixed effects, the public bank dummy is positive and significant for investment in government securities and cash holdings and negative for credit to private sector.

### Robustness tests (Table 8 and additional analyses)
- Column I: separate effects when fiscal deficit increasing vs decreasing by including separate dummies for 1992-2002 and 2003-2007.
  - Coefficients of fiscal deficit positive and significant in both periods; larger for public banks.
- Column II: include state government and central government deficits separately (with year fixed effects); results stronger for interaction of central government deficit with public sector banks (Central Govt Deficit*Public Banks 0.86*** [2.95]).
- Column III: allow different coefficients for new private banks and old private banks; public sector results hold but coefficients for new vs old private banks differ somewhat.
- Other robustness checks:
  - Including bond yields: coefficient insignificant and does not affect other results.
  - Including fiscal deficit in cash holdings regressions: coefficient insignificant and does not affect other coefficients.
  - Including capitalization (capital injection as % of bank’s assets) in investment regressions: positive but insignificant; does not affect other coefficients.

### Explanations for differential behavior of public banks
- Hypotheses examined and evidence:
  - Profitability motive (lower operating costs and no default risk on government securities; declining interest rates provide trading profits):
    - Regressions of return on assets (Table 9) show:
      - PSBs Dummy: -.47***, -.29***, -.26*** (Columns I–III) with t-stats [-6.79], [-3.59], [-3.28] — public banks less profitable after controlling for size.
      - Investment in Govt Securities: -.03***, -.19*** (Columns I–II) [-4.30], [-5.30] — investment in government securities associated with lower profitability.
      - Investment in Govt Securities*Fiscal deficit: .02*** [4.95] — investments associated with relatively higher returns during periods of high fiscal deficits.
    - Interpretation: evidence contradicts hypothesis that public banks invest more in government securities to earn higher profits; instead, larger investments in government securities are associated with lower profitability.
  - Risk mitigation due to priority sector lending:
    - Including priority sector lending (lagged and current) yields negative coefficient; lending to priority sectors not correlated with larger investment in government securities.
  - Demand-side explanation (low demand for private credit):
    - Table 10 includes private capital formation, external commercial borrowings, stock market capitalization.
    - Findings:
      - Private capital formation/GDP, Lagged: -0.86***, -0.68*** (investment regressions) and 0.33*, 0.16 (private credit regressions) with t-stats [-5.58], [-4.85], [1.93], [0.92].
      - External Commercial Borrowings/GDP, Lagged: -2.76*** (investment) and 0.10 (private credit) [-12.1], [0.36].
      - Stock Market Capitalization/GDP, Lagged: -0.04*** and .06*** (investment and private credit) [-3.85], [3.57].
      - Fiscal Deficit*Public Banks remains significant in many specifications, whereas fiscal deficit coefficient can become insignificant for private banks in some specifications.
    - Interpretation: demand-side factors matter, but public banks’ decisions also appear driven by other objectives.
  - Other potential explanations (not directly testable here):
    - Asymmetric incentive structure for public bank managers—greater penalties for bad loan decisions and limited reward for high profits, leading to lower lending risk-taking.
    - Moral hazard from perceived government support (implicit/explicit), reducing incentives to maximize profits — supported by Acharya et al (2010) evidence cited in the text.
    - Moral suasion or prioritization to assist government financing needs.
- Synthesis from authors: overinvestment by public banks in government securities likely due to a combination of “lazy” behavior, moral suasion, or prioritization of government financing needs; not primarily driven by desire to earn higher returns, overhead cost control, or to improve asset quality.

### Conclusion and policy implications
- Main conclusions:
  - Financial liberalization and increased entry of private banks increased competition and significantly improved the efficiency and profitability of public banks, making them comparable to private banks in many respects.
  - However, reductions in state preemption of resources via CRR and SLR appear less effective in changing public banks’ behavior compared to private banks.
  - Public banks invest a larger share of assets in government securities, hold larger cash balances, and lend less to the private sector than private banks.
  - Fiscal deficits, not changes in SLR, are a key driver of banks’ investments in government securities; public banks increase such investments more in response to fiscal deficits than private banks do.
- Broader implication:
  - In developing countries with limited alternative financing channels, government ownership of banks combined with high fiscal deficits may limit the gains from financial liberalization and contribute to crowding out of private sector credit.
- Empirical caution: results are robust across multiple specifications and robustness checks documented in the paper.

*Italic: Source — Excerpted content from the supplied IMF PDF chapter/section.*

### References

### References

### Bibliographic citations
- Abiad Abdul, Enrica Detragiache, Thierry Tressel, 2010. "A New Database of Financial Reforms," IMF Staff Papers, Palgrave Macmillan Journals, vol. 57(2), pages 281-302.
- Acharya Viral, Anukaran Agarwal and Nirupama Kulkarni, 2010, State Ownership and Systemic Risk: Evidence from the Indian Financial Sector during 2007-09, mimeo.
- Banerjee, Abhijit V., Shawn Cole, and Esther Duflo, 2004. "Banking Reform in India," Brookings Papers on Economic Activity, Economic Studies Program, The Brookings Institution, vol. 1(1), pages 277-332.
- Chari, Anusha and Peter Henry, 2008, “Firm Specific Information and Efficiency of Investment,” Journal of Financial Economics, 87(3), pp. 636-655.
- Cole, Shawn, 2004, bank ownership, bank lending behavior and political capture, evidence from India, mimeo.
- Galindo, Arturo, Fabio Schiantarelli, and Andrew Weiss, 2007, “Does Financial Liberalisation Improve the allocation of Investment? Micro Evidence from Developing Countries,” Journal of Developing Economics, Vol. 83, pp. 562-87.
- Gupta, P., R. Hasan, and U. Kumar. 2008. “What constrains Indian manufacturing?” ICRIER. Working Paper, No. 211.
- Gupta, P., R. Hasan, and U. Kumar. 2009. “Big Reforms but Small Payoffs: Explaining the Weak Record of Growth in Indian Manufacturing” in S. Bery, B. Bosworth, and A. Panagariya (eds), India Policy Forum, volume 5, pp 59-108.
- Hauner, David., “Credit to Government and banking sector performance”, Journal of Banking and finance, Vol. 32 (2008), pp. 1499-1507.
- Hauner, David., “Public debt and Financial Development”, Journal of Development Economics, Vol. 88 (2009), pp. 171-183.
- Koeva, Petya, 2003, The Performance of Indian Banks during Financial Liberalization, IMF Working Paper WP/03/150.
- La Porta Rafael, Florencio Lopez-de-Silanes, and Andrei Shleifer, 2002. “Government Ownership of Banks.” Journal of Finance, 57:265-301.
- Mohan, R., 2004, Financial Sector Reforms in India: Policies and Performance Analysis, RBI Bulletin, 2004, pp. 851-877.
- Panagariya, A. 2008. India: The Emerging Giant. Oxford University Press. New York.
- Tressel, Thierry and Enrica Detragiache., “Do Financial Sector Reforms Lead to Financial Development? Evidence from a New Dataset”, IMF Working Paper, WP/08/265.

*Appendix A — Chronology, Appendix B — Variable Definitions, Appendix C — Summary Statistics follow.*

---

### Appendix A — Chronology of Reforms

### Interest rate deregulation
- Deregulation of rupee-denominated deposit rates
  - 1992: First steps toward rate liberalization; partial deregulation of rupee-denominated term deposit rates; substitution of single interest rate per term deposit category with a ceiling below which banks free to fix rates.
  - 1995-96: RBI gradually eliminated ceilings on domestic and Non-Resident External (NRE) rupee deposits with maturities over one year.
  - 1997: Banks allowed to determine interest rates on domestic term deposits of 30 days and above; interest rates on NRE deposits with maturities over 6 months deregulated.
  - 1998: Minimum lock-in period reduced from 30 days to 15 days; banks permitted differential interest rates on domestic term deposits above Rs. 1.5 million and to determine penalties for early withdrawal of domestic and NRE deposits and loans against fixed deposits.
  - 2000: Restrictions preventing banks from charging differential rates on NRE deposits (depending on deposit size) relaxed.
  - 2002: All banks encouraged to put a flexible interest rate system on deposits (with a fixed rate option for depositors).
  - 2003: Interest rate on savings account offered by banks reduced to 3.5 per cent per annum from 4.0 per cent per annum.

- Deregulation of foreign-currency denominated deposit rates
  - 1993: Foreign Currency (Non-Resident) Deposit Scheme replaced; initially interest rates stipulated by RBI; exchange rate risk shifted from RBI to commercial banks.
  - 1997: Banks allowed to set interest rates on FCNR(B) deposits, subject to RBI ceiling; certain ceilings linked to LIBOR; interest rate for FCNR(B) deposits with maturity over one year stipulated within ceiling of swap rates.
  - 1998: Banks allowed to establish their own penalties for early withdrawal of FCNR(B) deposits.
  - 2000: Differential rates on FCNR(B) deposits introduced depending on deposit size, subject to overall ceiling; banks given option to choose current swap rates when offering FCNR(B) deposits.

- Deregulation of lending rates
  - 1992-94: Number of lending categories reduced from six to three.
  - 1998: Lending structure rationalized further; another category eliminated.
  - 1998: Interest rates on loans against term deposits liberalized; April 1998 stipulation that interest rate on loans against domestic and NRE deposits not exceed bank-specific PLR.
  - 1999: Banks given freedom to charge own interest rates on advances against domestic/NRE deposits without reference to PLR ceiling if certain deposit rate conditions satisfied. Since October 1999, interest rates on loans against domestic/NRE/FCNR(B) term deposits could be determined without reference to PLR.
  - 2000: Restrictions on interest rates on advances up to Rs. 200,000 against third party deposits removed.
  - 2003: Banks allowed to determine rates of interest on loans and advances for purchase of consumer durables, to individuals against shares and debentures/bonds, and other non-priority sector personal loans without reference to PLR and regardless of loan size.

- Deregulation of Prime Lending Rate (PLR) restrictions
  - 1994: Banks permitted to establish their own Prime Lending Rates (PLR) for advances over Rs. 200,000 beginning October 1994.
  - 1997: October 1997 rules relaxed to allow separate Prime Term Lending Rates (PTLR) for term loans with at least 3-year maturity.
  - 1998: Bank-specific PLR became the ceiling for loans below Rs. 200,000; each bank had to announce its PLR and maximum spread.
  - 1999: Scheduled commercial banks given freedom to offer fixed rate term loans if adhering to Asset Liability Management (ALM) guidelines; since April 1999 different PLRs could be used for loans with different maturities; banks allowed to charge interest rates without reference to PLR for certain categories (refinancing schemes, intermediary agencies, discount of bills).
  - 2000: April 2000 allowed banks to offer all loans on fixed or floating rate basis, subject to PLR stipulations.

### Reduction in reserve requirement
- Reduction in Cash Reserve Ratio (CRR)
  - 1992-1993: Incremental CRR of 10 percent eliminated.
  - 1992-2001: Average CRR fall from 15 percent to 5.5 percent.
  - 2000: Minimum daily requirement of CRR balances lowered from 85 percent to 65 percent.
  - 2001: Minimum daily requirement of CRR balances reduced from 65 percent to 50 percent.
  - 2001: Interest rate on eligible cash balance of banks with the Reserve Bank aligned with the Bank Rate.
  - 2002: CRR reduced from 5.5 per cent to 5 per cent.
  - 2003: Inter-bank term liabilities with original maturity of 15 days to one year exempted from prescription of minimum CRR requirement of 3.0 per cent.
  - 2003: CRR reduced by 50 basis points from 5 percent to 4.5 percent.
  - 2004: CRR raised by 50 basis points to 5 percent.
  - 2006: CRR raised by 50 basis points to 5.5 percent.
  - 2007: CRR raised by 200 basis points in four stages (each 50 bps) to 7.5 percent.
  - 2007: Banks required to maintain a minimum of 70 per cent of the required amount of average daily CRR.
  - 2007: No interest payable on CRR balances of banks with effect from March 31, 2007.
  - 2008: CRR raised by 150 basis points (50 bps each in April, June and July 2008) to 9 percent, reduced by 250 bps in Oct and 100 bps in Nov to 5.5 percent.
  - 2009: CRR reduced by 50 basis points from 5.5 percent to 5 percent.

- Reduction in Statutory Liquidity Ratio (SLR)
  - 1992-1994: SLR on incremental deposits cut down; base date in SLR computation pushed forward; base level SLR decreased to 33.8 percent; statutory liquidity requirement for any increase in NDTL above their level as of September 30, 1994 stipulated to be 25 percent.
  - 1997: Uniform SLR of 25 percent came into effect.
  - 2007: SLR reduced by 100 basis points to 24 per cent.
  - 2008: SLR increased to 25 percent in Nov 2009.

### Entry deregulation and ownership
- Competition / Entry of new banks
  - 1993: Rules for establishing new private sector banks introduced; main provisions included minimum capital requirement of Rs. 100 million; limited foreign bank participation up to 20 percent, maximum overall non-resident participation of 40 percent; public listing; computerized environment.
  - 1994-1996: Nine new private banks founded between 1994 and 1996 (Bank of Punjab Ltd, Centurion Bank Ltd, Global Trust Bank, HDFC Bank, ICICI Bank, IDBI Bank, Indusland Bank Ltd, UTI Bank Ltd, and Times Bank Ltd).
  - 2001: Guidelines for licensing new private sector banks revised in January 2001; minimum capital requirement raised and private bank ownership of large industrial houses restricted.
  - 1990-2001: New foreign banks entered market and existing foreign banks allowed to open additional branches; number of foreign banks increased from 21 to 42; foreign banks acquired 51 additional offices, increasing branches from 151 in 1992 to 202 in 2001.

- Ownership changes
  - 1993: SBI Act amended; SBI became first public bank to raise capital from the public in December 1993.
  - 1994: Nationalized banks allowed to raise up to 40 percent of their capital from the market in 1994.
  - 1994-2001: Eleven public sector banks accessed the market.

### Credit policies
- Credit controls (1992-2001)
  - Reforms focused on: giving banks more freedom to set credit requirements for borrowers; relaxing conditions for consortium lending; withdrawing regulations on Maximum Permissible Bank Finance (MPBF); allowing banks to use their methods to assess working capital requirements; allowing discretion in levying commitment charges; deciding on levels of inventory and receivable holdings of industries.

- Priority sector lending
  - 1992-2001: Definition of priority sector expanded to include bank investments in designated bonds (NABARD, SIDBI, NHB), contributions to Rural Infrastructure Development Fund, irrigation, agriculture machinery, food and agro-based processing, traditional plantation loans, advances to housing, retail trade, software, transport operator industries (subject to loan-size restrictions), venture capital, micro-credit, and credit to NDFCs for small road, water transport operator, and tiny sector lending.
  - 1992-2001: Overall target for Indian banks was 40 percent of net bank credit, with sub-targets of 18 percent for agriculture and 10 percent for weaker sections. Early 1992 foreign bank export credit target of 15 percent revised to 32 percent with sub-targets; 1993 indirect loans of 4.5 percent allowed as part of agriculture target; 1996 export credit target revised to 12 percent. Currently, overall priority sector lending targets for domestic and foreign banks remain 40 percent (18 percent for agriculture and 10 percent for weaker sections) and 32 percent (12 percent for export credit and 10 percent for SSI), respectively.
  - 2009: RBI revised guidelines on lending to priority sector. Under new guidelines, priority sector lending target and sub-targets for all banks linked to adjusted net bank credit (ANBC=Net Bank Credit plus investments made by banks in non-SLR bonds held in HTM category) or credit equivalent of off-balance sheet exposure, whichever is higher. Outstanding FCNR(B) and NRNR Deposits balances no longer be deducted for computation of net bank credit for priority sector lending purposes.

### Bank finance / lending markets and market access
- Call Money
  - 2001: Phasing out of non-bank participation in call money market to be done in four stages starting in 2001.
  - 2002: Prudential limit on SCBs exposure in call money market: SCBs fortnightly average lending in the call/notice money market not to exceed 25 per cent of their owned funds; fortnightly average borrowings not to exceed 100 per cent of their owned funds or 2.0 per cent of aggregate deposits as at the end of March of the previous financial year, whichever is higher. They allowed to lend and borrow a maximum of 50 per cent and 125 per cent, respectively, of their owned funds on any day during a fortnight.

- Domestic Stock Market
  - 2001: Revised guidelines on bank financing of equities and investments in shares made ceiling of 5 per cent applicable to total exposure of a bank to stock markets.
  - 2002: RBI liberalised norms for issue and pricing of shares by private sector banks; private sector banks free to issue bonus and rights issues without prior RBI approval; initial public offerings and preferential shares require RBI approval.

- Overseas operations and FDI
  - 2002: Limit on banks to borrow and invest from/in overseas market increased from 15 per cent to 25 per cent of their unimpaired Tier I capital.
  - 2002: Consolidated guidelines on FDI in banking sector: FDI in private banks permitted under automatic route up to 49 percent; FDI and portfolio investment in PSBs, including State Bank of India, permitted up to 20 percent. Maximum limit of 49 percent applicable also to foreign banks having branch presence in India and wishing to make FDI in private banks.
  - 2008: Banks allowed to borrow funds from their overseas branches and correspondent banks up to a limit of 50 per cent of their unimpaired Tier I capital.

### Other actions
- 2001: Guidelines for entry of FIs into insurance business formulated.
- 2002: RBI introduced supervisory rating system based on “CAMELS” model for FIs, on lines similar to banks.
- 2003: Less complex Over the Counter (OTC) interest rate rupee options permitted.
- 2003: Banks/FIs allowed to deal in exchange traded interest rate derivatives in a phased manner.
- 2003: Foreign banks operating in India permitted to remit net profits/surplus (net of tax) earned out of their Indian operations on a quarterly basis without prior RBI approval, provided specified conditions met.
- 2004: Prudential guidelines on banks’ investment in non-SLR debt securities issued to contain risks arising out of non-SLR investment portfolio; guidelines require banks should not invest in non-SLR securities of original maturity of less than one year and also in unrated debt securities and unlisted shares of AIFIs.
- 2004: Banks allowed to raise long term bonds with a minimum maturity of five years to provide boost to infrastructure lending.
- 2005: Introduction of asset-backed commercial paper (ABCP) to further deepen CP market.
- 2005: Banks allowed to extend financial assistance to Indian companies for acquisition of equity in overseas joint ventures.
- 2005: Policy for authorisation of branches of banks in India liberalised and rationalised.
- 2007: Banks permitted to undertake Pension Fund Management (PFM) through subsidiaries set up for the purpose, subject to eligibility criteria prescribed by PFRDA and RBI guidelines.
- 2009: Liberalisation of the FCCBs buyback policy.

---

### Appendix B — Variable Definitions

- Return on Assets: Income minus expenses and provisions as percent of assets.
- Operating Profit/Assets: Return on Assets defined above plus provisions as percent of assets.
- Operating Expenses/assets: Sum of payments to employees, rent, taxes, printing, stationary, advertising, depreciation, post, telephone, insurance, and other expenditure as percent of assets.
- Wages/Assets: Payments to and provisions of employees as percent of assets.
- Nonwage Expenses/Assets: Operating expenses other than wages as percent of assets.
- Cash/Assets: Cash in hand and balances with the RBI as percent of assets.
- Credit (Other)/Assets: Advances made to “others” (i.e. to non government, non priority, and non bank sectors) as percent of assets.
- Investment in Government Securities/assets: Investment in Government Securities as percent of assets.
- Investment in Approved Securities/Assets: Investment in other approved Securities as percent of assets.

- Data source: RBI’s database "Statistical Tables Relating to Banks in India" and "Basic Statistical Returns".

---

### Appendix C — Summary Statistics of the Variables

- Return on Assets
  - Number of Observations: 7870
  - Mean: .521
  - Std. Dev.: .10
  - Minimum: -7.51
  - Maximum: 2.34

- Operating Profit/Assets
  - Number of Observations: 7871
  - Mean: .750
  - Std. Dev.: .93
  - Minimum: -1.83
  - Maximum: 4.49

- Operating Expenses/assets
  - Number of Observations: 7872
  - Mean: .490
  - Std. Dev.: .60
  - Minimum: .93
  - Maximum: 4.70

- Wages/Assets
  - Number of Observations: 7871
  - Mean: .620
  - Std. Dev.: .62
  - Minimum: .12
  - Maximum: 3.89

- Nonwage Expenses/Assets
  - Number of Observations: 7870
  - Mean: .870
  - Std. Dev.: .35
  - Minimum: .41
  - Maximum: 3.49

- Cash/Assets
  - Number of Observations: 7879
  - Mean: .044
  - Std. Dev.: .13
  - Minimum: 2.47
  - Maximum: 26.26

- Credit (Other)/Assets
  - Number of Observations: 7872
  - Mean: 3.95
  - Std. Dev.: 7.47
  - Minimum: 0.00
  - Maximum: 59.51

- Investment in Government Securities/assets
  - Number of Observations: 7872
  - Mean: 5.11
  - Std. Dev.: 6.05
  - Minimum: 11.29
  - Maximum: 44.74

- Investment in Approved Securities/Assets
  - Number of Observations: 7872
  - Mean: 8.05
  - Std. Dev.: 5.86
  - Minimum: 11.29
  - Maximum: 46.18

- Statutory Liquidity Ratio
  - Number of Observations: 172
  - Mean: 9.47
  - Std. Dev.: 5.73
  - Minimum: 25
  - Maximum: 38.5

- Gross Fiscal Deficit
  - Number of Observations: 175
  - Mean: .45
  - Std. Dev.: 1.13
  - Minimum: .45
  - Maximum: 7.84

- Combined Fiscal deficit
  - Number of Observations: 177
  - Mean: .91
  - Std. Dev.: 1.37
  - Minimum: .58
  - Maximum: 9.94

*Content unit: _wp1150 - References*

### Appendix D: Banks and Ownership Details

### Appendix D: Banks and Ownership Details

### Banks and Ownership Status (selected entries)
- Allahabad Bank — PSB — Included with no change
- Andhra Bank — PSB — Included with no change
- Axis Bank — PVT — UTI bank's name changed to Axis bank. UTI was established in 1995. Dropped the data for 1995, 1996.
- Bank of Baroda — PSB — Bareilly Corporation bank merged in 1998 and Benaras State Bank merged in 2002. Added the data for all three banks.
- Bank of India — PSB — Bank of Karad Merged in 1993--use data from 1993 for the merged bank
- Bank of Madura — dropped — Merged with ICICI in 2001. Added the data in ICICI Bank.
- Bank of Maharashtra — PSB — Included with no change
- Bank of Punjab — PVT — dropped — Merged with Centurion bank in 2005. Added the data for the banks.
- Bank of Rajasthan — PVT — Included with no change
- Bareilly Corporation Bank — PVT — dropped — Merged with bank of Baroda in 1998
- Benares State Bank — PVT — dropped — Merged with Bank of Baroda in 2002
- Bharat Overseas Bank — PVT — Included with no change
- Canara Bank — PSB — Included with no change
- Catholic Syrian Bank — PVT — Included with no change
- Central Bank of India — PSB — Included with no change
- Centurion Bank — PVT — dropped — Name changed in 2005 to Centurion Bank of Punjab after merger with Bank of Punjab. The bank was created in 1995
- Centurion Bank of India — PVT — Added the data for Bank of Punjab and Centurion Bank and used from 1997, since Centurion Bank was created in 1995. Merged with HDFC in 2008.
- City Union bank — PVT — Included with no change
- Corporation Bank — PSB — Included with no change
- Dena Bank — PSB — Included with no change
- Development Credit Bank — PVT — Included with no change
- Dhanalakshmi Bank — PVT — Included with no change
- Federal Bank — PVT — Merged with Ganesh bank in 2006. Added the data all through.
- Ganesh Bank of Kurundwad — PVT — dropped — Merged with federal bank in 2006
- Global Trust Bank — PVT — dropped — Merged with oriental Bank of Commerce in 2004. Global Bank was created in 1995. Thus use the data of the merged bank from 1997.
- HDFC Bank — PVT — Times bank merged in 1999. Added the data throughout. HDFC created in 1995, thus use the data from 1997. (Centurion bank merged in 2008)
- ICICI Bank — PVT — Merged Bank of Madura in 2001. Added the data. Merged ICICI personal finance and ICICI capital services in 2002. We do not have data for these entities. We drop 2002 and 2003 from the data. ICICI was created in 1995. Drop 1995, 1996 from the database. Merged Sangli bank in 2007, but since data for Sangli Bank is available till 2007 we do not do any corrections for this merger.
- IDBI Bank — PVT — Merged with IDBI LTD. No data for IDBI LTD prior to 2005. Thus we use the data till 2004. The bank was created in 1995, thus we use the data from 1997.
- IDBI LTD. — PSB — dropped — Created in 2005, merged IDBI Bank. Merged United Western Bank in 2006.
- Indian Bank — PSB — Included with no change
- Indian Overseas Bank — PSB — Included with no change
- INDUSIND Bank — PVT — Bank created in 1995. Use the data from 1997
- ING Vysya Bank — PVT — dropped — Data are available from 2003, when ING took over the Vysya bank and it was renamed as ING VYSYA bank in 2003. We use the data from 2004 as a new bank.
- Jammu and Kashmir Bank — PVT — Included with no change
- Karnataka Bank — PVT — Included with no change
- Karur Vysya Bank — PVT — Included with no change
- Kotak Mahindra Bank — PVT — dropped — Created in 2004.
- Lakshmi Vilas Bank — PVT — Included with no change
- Lord Krishna Bank — PVT — Included with no change
- Nainital Bank — PVT — Included with no change
- Nedungadi Bank — PVT — dropped — Merged with Punjab National Bank in 2003.
- New Bank of India — PSB — dropped — Merged with Punjab national bank in 1993.
- Oriental Bank of Commerce — PSB — Merged Punjab Cooperative bank in 1996; and Global Trust Bank in 2004. Added the data for all three banks all through.
- Punjab and Sind Bank — PSB — Included with no change
- Punjab Cooperative Bank — PVT — dropped — merged with Oriental Bank of Commerce in 1996
- Punjab National Bank — PSB — Merged New Bank of India in 1993 and Nedungadi Bank in 2003. Added the data of all three banks all through.
- Ratnakar Bank — PVT — Included with no change
- Sangli Bank — PVT — Acquired by ICICI in 2007. Included with no change till 2006 (?)
- SBI commercial and International Bank — PVT — dropped — Subsidiary of SBI, data from 1997. Not sure how to treat. Therefore dropped
- South Indian Bank — PVT — Included with no change
- State Bank of Bikaner and Jaipur — PSB — Included with no change
- State Bank of Hyderabad — PSB — Included with no change
- State Bank of India — PSB — Included with no change
- State Bank of Indore — PSB — Included with no change
- State Bank of Mysore — PSB — Included with no change
- State Bank of Patiala — PSB — Included with no change
- State Bank of Saurashtra — PSB — Included with no change
- State Bank of Travancore — PSB — Included with no change
- Syndicate Bank — PSB — Included with no change
- Tamilnad Mercantile Bank — PVT — Included with no change
- Times Bank — PVT — dropped — Merged with HDFC in 1999
- UCO Bank — PSB — Included with no change
- Union Bank of India — PSB — Included with no change
- United Bank of Indian — PSB — Included with no change
- United Western Bank — PVT — merged with IDBI in 2007
- Vijaya Bank — PSB — Included with no change
- Vysya Bank — PVT — Included till 2002. Merged with ING after that
- Yes Bank — PVT — dropped — New bank data from 2005

Note: PSB refers to public sector bank, and PVT refers to private banks.

### Data-cleaning procedure and sample construction
- Initial sample: all banks included in the RBI’s database.
- Rule: drop the banks which do not exist since 1991.
- For banks opened during the sample period: drop the data for the first two years after the banks were opened.
- Nine new banks were opened during 1994-1996: Bank of Punjab limited, Centurion Bank Limited, Global Trust Bank, HDFC Bank, ICICI, IDBI Bank, Indusind Bank, UTI Bank Limited and Times Bank Limited.
- Banks opened only after 2004 were dropped from the sample; these include Yes bank, IDBI Limited, and Kotak Mahindra Bank.
- Name changes: matched name changes of banks and recorded merger information.
- Merger treatment:
  - When a small bank merged into a large bank, the merged (smaller) bank was dropped from the database.
  - For the parent bank, balance sheets of the parent and merged bank were added up and treated as the merged bank from the beginning of the sample.
  - Exception: ICICI’s merger with ICICI personal finance and ICICI capital services in 2002 involved financial companies not in the database; data for ICICI Bank for 2002 and 2003 were dropped.
- Specific data availability notes:
  - Global Trust Bank was created in 1995; therefore use the data of the merged bank from 1997.
  - HDFC was created in 1995; use the data from 1997.
  - ICICI was created in 1995; drop 1995, 1996 from the database.
  - IDBI Bank was created in 1995; use the data from 1997. No data for IDBI LTD prior to 2005, thus use IDBI Bank data till 2004.

### Key statistics and exact rules preserved
- Drop first two years of a bank's data after opening.
- Nine new banks opened during 1994-1996 (listed explicitly above).
- Banks opened only after 2004 that were dropped: Yes bank, IDBI Limited, Kotak Mahindra Bank.
- Years explicitly referenced for data treatment and mergers include: 1991, 1993, 1995, 1996, 1997, 1998, 1999, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008.

*Source: Appendix D: Banks and Ownership Details (IMF working paper).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1150.pdf_