## 1. One important legacy of the global financial and Euro Area (EA) crises is their impact on the funding models of internationally active banks

## Source details

**Canonical URL:** [1. One important legacy of the global financial and Euro Area (EA) crises is their impact on the funding models of internationally active banks](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1471.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1471.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1471.pdf.json)

---

### Legacy: shift from retail deposits to wholesale funding and resulting vulnerabilities
- Before the crisis many large global banks had difficulty attracting core deposits and increasingly supplemented retail deposits with wholesale market funding to finance credit growth during the credit boom.
- Growing reliance on short term wholesale finance to fund long term assets created:
  - currency and maturity mismatches;
  - increased liquidity risk.
- Banks more reliant on customer deposit funding fared better during the crisis; evidence suggests their market value exceeded banks funded predominantly through wholesale markets (Beltratti and Stultz, 2011).

### Funding market paralysis, 2007–12
- The start of the global financial crisis in August 2007 coincided with worldwide liquidity shortages at major banks and turmoil in interbank funding markets:
  - interbank lending slowed to a halt;
  - wholesale funding markets froze and interest rates on unsecured term loans between banks rose sharply and remained unusually volatile for an extended period.
- Funding conditions remained tight as investors shunned bank debt and new issuance fell to historic lows.
- The period includes the sub-prime crisis and the EA crisis, contributing to sustained paralysis of bank funding markets through 2012.

### Decline in market funding and bank responses
- By 2007 growth in customer loans had exceeded growth in customer deposits by far for Nordic, EA and UK banks; banks bridged the gap with short term and collateralized term wholesale funding.
- Peak issuance in March 2007: major European countries and the US issued some US$415 billion in secured and unsecured debt, compared to some US$8bn just over a decade earlier.
- Since 2007:
  - capital market issuance has fallen sharply for the majority of global banks;
  - customer funding gaps have narrowed.
- Banks responded to scarce and costly wholesale funding by:
  - shifting towards more customer deposit funding where possible;
  - relying on official financing sources;
  - reducing non-core assets.

### Focus of the paper: factors underlying the cost of international banks’ funding
- The paper investigates why funding costs varied across banking systems and individual banks during the global financial and EA sovereign debt crises.
- Three sets of potential determinants are examined:
  - bank-specific balance sheet variables (including changes in level and quality of capital);
  - country-specific macro-financial factors (domestic economic conditions and changes in short term interest rates);
  - global financial market factors (implied market volatility, shocks such as the global financial crisis and EA sovereign risks, and the global growth outlook).
- The analysis disentangles effects of banks’ resilience measures (e.g., higher capitalization) on funding costs and assesses crisis impacts on long run funding costs.

### Marginal funding cost measure and empirical scope
- Focus: a bank’s marginal cost of funding (the cost of raising an additional unit of funding), excluding costs from secured funding (e.g., covered bonds or repo) and deposits.
- Rationale: prior to the crisis, banks increasingly used wholesale unsecured funding as their marginal source because such markets provided substantial, relatively cheap, short-notice funds.
- The paper assumes changes in banks’ marginal unsecured wholesale funding costs can be inferred from five-year CDS premia.
- Marginal funding cost estimated as the sum of a bank’s five-year CDS premia plus three-month Libor (reflecting cost of raising fixed rate senior unsecured bonds and entering an interest rate swap).

### Key empirical findings (short run and long run)
- Short run changes in bank unsecured funding costs are associated with:
  - bank-specific characteristics such as an institution’s credit worthiness and changes in the level and quality of capital;
  - country-level factors such as domestic economic conditions and changes in short term interest rates;
  - global risk factors such as implied market volatility, shocks to financial markets (notably the global financial crisis and EA sovereign risks), and the global growth outlook.
- Larger, systemically important banks enjoy a funding advantage, and this advantage has risen since the onset of the crisis.
- Long run evolution:
  - banks’ funding costs appear to have risen over the sample period, driven mainly by deterioration in banks’ asset quality; decline in domestic economic conditions; the impact of successive financial crises (global financial crisis and EA sovereign debt crisis).
- Capital effects:
  - Increased capital buffers can lead to a decline in a bank’s funding costs in the long run;
  - if the increase is due to higher quality Tier 1 capital, there is a short run reduction in funding costs as well.

### Policy implications (high level)
- Results counter the view that higher capital requirements necessarily increase bank funding costs and reduce lending to the real economy.
- Supporting evidence cited:
  - higher bank equity financing raises overall funding costs only modestly (Miles (2011)), and the long-run impact on borrowing costs is likely small;
  - increasing capital buffers may not raise banks' cost of capital and could reduce long-run funding costs, potentially supporting bank lending growth.
- Balance sheet repair may reduce bank lending in the short run, but higher capital levels contribute to lower bank funding costs in the long run.

### Literature review highlights (selected)
- ECB (2009) and CGFS document pre-crisis balance sheet growth, shorter maturities of money market funding, securitization and debt issuance that exposed EA banks to maturity and currency mismatches.
- Studies find deposit-funded banks performed better during the crisis, continued lending, and were less risky (Ivashina and Scharfstein (2010); Dermirguc-Kunt and Huizinga (2010); Raddatz (2010); Cornett et al. (2011); Beltratti and Stulz (2012); Dagher and Kazimov (2012); Vazquez and Federico (2012); IMF (2013)).
- Research on “too-big-to-fail” funding cost subsidy shows some evidence that G-SIBs have lower total funding costs attributed to perceived bailouts.
- Empirical studies of bank CDS indicate common factors in CDS premia movements and increased co-movement after the onset of the crisis (Eichengreen et al. (2009); Ballester Miquel, Lukac and González-Urteaga (2013)).

### Empirical model (overview)
- Changes in bank funding costs are modeled in a panel Error Correction Model to disentangle long- and short-run effects of bank-specific and macro variables and to estimate the long-run equilibrium of funding costs for major international banks.
- Dynamic linear regression specification includes:
  - FCi,t: bank i’s marginal cost of funding (three-month Libor plus five-year CDS premia) at time t;
  - Kt-2: vector of capital variables at time t-2 (total capital ratio and capital quality);
  - Mt-1: vector of macro variables at time t-1 (real GDP growth, short term interest rate, yield curve slope).
- Financial crisis dummies:
  - C1t = 1 from Q3 2007 to allow for a change in relationships after the onset of the global financial crisis in August 2007;
  - C2t = 1 from Q2 2011 to allow for a change following market disruption from Q2 2011 triggered by concerns about a potential break-up of the EA.
  - The constant is allowed to change during crisis periods; when C2 = 1 its effect adds to C1 so impacts are cumulative.

### Model specification and short-run dynamics (Z)
- Zt = [Bt-1, Ft] stacks a vector of bank-specific variables at time t-1 (Bt-1) (provision ratio and equity returns) and macro-financial variables at time t (Ft) (market implied volatility and an index of notional weighted CDS premia for periphery European countries).
- The Zt variables have only short-run effects and therefore are included only in changes.
- The model is stable and has a long run relationship if -1< α1 < 0.
- Short run estimates refer to an effect lasting around 2–4 quarters before fading away; long run estimates refer to a more permanent effect on funding costs.

### Data and sample
- Sample composition and frequency:
  - 52 banks in 14 advanced economies from 2001–12.
  - Quarterly CDS premia data are used (balance sheet variables are available only on a quarterly basis).
  - Primary data sources are Bloomberg and Datastream.
- Core sample:
  - 25 major international banks from four regions: US, UK, EA and Nordic countries (Denmark, Sweden and Norway).
  - Two criteria for core banks: (i) systemically important within its economy at end-2012; and (ii) availability of CDS premia over 2001–12 on a quarterly frequency.
  - 19 of the 25 banks in the core sample meet the Financial Stability Board’s classification as global systemically important banks at end-2012.
  - Equation (3) is estimated on the core sample of 25 banks.
- Extended sample:
  - Augmented dataset with an additional 30 banks (mainly European, a few US, some Australian and Japanese banks) for tests on funding cost advantages for systemically important banks.
- Key data features:
  - Lagged bank balance sheet variables are used to avoid simultaneity bias.
  - The analysis uses changes in provisioning and non-performing loans to proxy balance sheet ill-health, and capital ratios (total capital/total assets and Tier 1 capital) to proxy resilience.
  - Provisions and fitted values, total capital, Tier 1 capital, real GDP growth, volatility, and yield curve spread are used in scatter plots to illustrate long run relationships.

### Descriptive patterns in CDS premia (2003–2013)
- Common episodes and dates noted:
  - August 2007—US subprime crisis and global bank liquidity shortages worldwide.
  - September 2008—Collapse of Lehman Brothers.
  - April 2010—Greece requests a “troika” bailout (April 23, 2010 marks the beginning of the EA sovereign debt crisis).
  - April 2011—Portugal requests a “troika” bailout.
  - December 2011—ECB initiates LTRO.
  - September 2012—ECB announces OMT.
- Cross-country observations:
  - US banks experienced the sharpest rise in premia following the Lehman collapse in Q3 2008, diverging significantly from other major banking systems.
  - US, UK and EA bank CDS rose in tandem during the EA crisis.
  - UK and EA bank CDS premia diverged around 2012, with UK bank spreads falling faster than EA bank spreads.
- Range and examples of average spreads over the period:
  - Rabobank low of 3.4 bps.
  - Ally Financial high of 2911 bps.
  - Morgan Stanley’s CDS peaked at 1240bp after Lehman.
- Evolutionary phases:
  - 2001 to early-2007: relative calm with little volatility in CDS premia.
  - mid-2007 onward: sharp widespread rises in bank CDS tied to crisis events.
  - April 23, 2010 onward: start of EA sovereign debt crisis with further increases in bank CDS.
  - April 2011 (Portugal bailout): a sharper widespread increase in bank CDS.
  - December 2011 (LTRO) and September 2012 (OMT): preceded sharp falls in bank CDS premia.
  - Since September 2012: steady decline across regions; as of May 2013 bank CDS had settled at levels higher than their pre-crisis peaks; US, Nordic and Japanese banks roughly more than 50 bps higher than at end-2006.

### Long-run relationships and variables used
- Bank-specific variables:
  - Provisions (proxy for asset quality / balance sheet ill-health).
  - Equity returns (proxy for profitability; short-run effect).
  - Total capital ratio and Tier 1 capital ratio (proxy for capital quality and resilience).
  - CDS bid-ask spread (proxy for CDS liquidity / market illiquidity).
- Macro-financial variables:
  - Actual domestic real GDP growth and expected global economic growth.
  - Equity implied volatility and global market illiquidity indicators.
  - Short term policy interest rate (monetary policy stance).
  - Yield curve slope (10yr–3m) to capture term spread effects.
  - EA sovereign risk index: weighted index of sovereign CDS premia for Greece, Italy, Ireland, Portugal and Spain (weighted by notional amounts).
  - Two crisis dummy variables:
    - Crisis dummy one: onset of the sub-prime crisis in Q3 2007.
    - Crisis dummy two: begins in Q2 2011 capturing EA breakup-related extreme financial market stress.

### Estimation results — bank capital and fundamentals
- Capital effects:
  - Short-run: an increase in total bank capital increases bank funding costs (positive short-run coefficient).
  - Long-run: an increase in bank capital reduces funding costs; on average, a 1pp increase in total bank capital reduces funding costs by 0.26bp.
  - The long-run reduction in equilibrium funding costs from additional capital is highest for US banks and lowest for EA banks.
  - Higher capital quality (Tier 1 / total capital) is associated with lower equilibrium bank funding costs.
- Asset quality and liquidity:
  - Bank provisions: positive and statistically significant — deterioration in asset quality increases funding costs.
  - US and Nordic bank funding costs are especially sensitive to changes in provisions in the short run.
  - CDS bid-ask spread: positive and statistically significant — market illiquidity is associated with higher CDS premia.
- Equity returns:
  - Negative impact on bank funding costs (higher returns imply lower probability of default).

### Estimation results — crisis effects and interactions
- Crisis dummy impacts:
  - First crisis dummy (sub-prime onset): positive and statistically significant; largest coefficient for US banks.
  - Second crisis dummy (EA breakup risk): positive and significant; largest coefficient for EA banks and smallest for US banks; UK banks most affected by the second crisis.
- Interaction effects:
  - Interaction of the sub-prime crisis dummy with capital variables yields a positive and significant coefficient — funding costs became more sensitive to capital since the onset of the crisis (investors differentiate more by capitalization post-crisis).

### Estimation results — macro-financial variables (selected coefficients preserved exactly)
- Yield curve slope (short run and long run):
  - Short run: 12.36*** (2.360) — Global
  - Short run: 4.991 (7.999) — US
  - Short run: 8.463*** (2.495) — EA
  - Short run: 29.37*** (8.512) — UK
  - Short run: -4.138 (4.201) — Nordics
  - Long run: 7.981*** (1.931) — Global
  - Long run: 5.774 (10.57) — US
  - Long run: 6.230*** (2.253) — EA
  - Long run: 36.95*** (8.134) — UK
  - Long run: 0.931 (3.658) — Nordics
- Selected short-run coefficients and standard errors (exactly as reported):
  - Total capital, Short run: 0.202* (0.106) — Global; 0.155* (0.080) — US; 0.227*** (0.113) — EA; 0.113* (3.660**) — UK.
  - Capital quality, Short run: -0.456* (0.256) — Global; -0.787** (0.358) — US; -1.301 (1.663) — EA; -0.633** (0.320) — UK; -3.317*** (1.083) — Nordics.
  - Provisions, Short run: 6.641*** (1.341) — Global; 7.651*** (2.658) — US; 2.780 (2.147) — EA; 3.964 (2.525) — UK; 4.419* (2.545) — Nordics.
  - CDS liquidity, Short run: 0.714*** (0.120) — Global; 10.20*** (1.199) — US; 4.696*** (0.475) — EA; 0.357*** (0.109) — UK; 2.183*** (0.770) — Nordics.
  - Equity Returns, Short run: -0.308*** (0.0584) — Global; -0.613*** (0.154) — US; -0.165*** (0.0613) — EA; -0.0261*** (-0.007) — UK; -0.521*** (0.145) — Nordics.
  - Implied market volatility, Short run: 0.502*** (0.0409) — Global; 0.249** (0.0992) — US; 0.429*** (0.0476) — EA; 0.731*** (0.0900) — UK; 0.110 (0.0746) — Nordics.
  - Euro area sovereign risk, Short run: 0.850*** (0.0797) — Global; 1.406*** (0.236) — US; 0.166*** (0.0227) — EA; 0.283* (0.144) — UK; 0.198*** (0.0368) — Nordics.
  - Short term interest rate, Short run: 11.16*** (3.103) — Global; 11.78 (12.79) — US; 5.484 (4.132) — EA; 49.62*** (9.974) — UK; 3.208 (4.478) — Nordics.
- Adjusted R2 reported for the model: 0.57 and 0.622 (table formatting preserved).

### Long-run evolution and regional differences in funding costs
- Overall long-run funding costs rose for all banks in the sample over the sample period, driven mainly by:
  - Deterioration in individual bank credit worthiness (asset quality deterioration).
  - Yield curve steepening due to increased sovereign and banking sector risk.
  - Risks from the global financial crisis and EA sovereign debt crisis.
- Offsetting factors: increases in total bank capital and tier 1 capital helped offset some of the rise in long-run funding costs.
- Regional observations by end-2012:
  - US and Nordic banks: long-run funding costs fell to lowest levels, mainly due to increased capitalization and reduction in asset riskiness.
  - EA periphery banks: long-run funding costs remained elevated; roughly one-half of all EA banks (mainly Italian and Spanish banks) had funding costs above long-run equilibrium.
  - EA core banks (German and French): funding costs had fallen to below long-run equilibrium.
  - UK banks: funding costs declined to below long-run equilibrium, with much of the decline in the last few quarters of 2012.
  - Among five US banks in core sample: funding costs for two remained above long-run equilibrium.
  - Among Nordic banks in sample: funding costs remained above long-run equilibrium for only one bank out of five at end-2012.
- Drivers of sharp increases:
  - US: sharp rise mid-2007 mainly driven by deterioration in asset quality; government capital injections in 2008 and improved capital quality mitigated funding cost increases.
  - EA periphery (early 2011): sharp rise tied to escalation of EA sovereign debt crisis — sovereign and bank perceived riskiness rose, investor outflows accelerated, term spreads rose (short and long rates rising in tandem), asset quality deteriorated and capital levels remained insufficiently low.

### Domestic Systemically Important Banks (D-SIBs) — funding cost differentials
- Definition used in analysis: dummy = 1 if book assets of the bank exceed US$500 billion in that quarter (SIB).
- Main empirical findings:
  - SIBs had lower funding costs relative to peers over the sample period.
  - Over the full sample period the coefficient on the SIB dummy suggests the cost of marginal unsecured funding was on average 16 bps lower for SIBs relative to non-SIBs.
  - After 2011 the funding advantage rose by more than 40 percent (particularly for EA SIBs); post crisis, SIBs’ cost of funding was roughly 22 bps lower compared to non-SIBs.
  - This funding advantage likely reflects market perceptions that large banks are “too-big-to-fail” and enjoy an implicit government support advantage.
- Reported coefficients from the SIB regression table (exact values and standard errors):
  - SIB dummy: -15.81*** (4.580) — Full sample; -18.24 (16.10) — US; -8.530** (3.103) — EA; -3.875 (13.49) — UK; -0.696 (8.536) — Nordics.
  - Crisis1 dummy: 12.54*** (3.934) — Full sample; 20.853* (18.08) — US; 4.483*** (1.020) — EA; 6.09* (3.65) — UK; 10.90** (5.061) — Nordics.
  - Crisis2 dummy: 8.116** (3.319) — Full sample; 3.29** (1.56) — US; 17.84*** (6.26) — EA; 10.131*** (4.11) — UK; 1.777 (7.988) — Nordics.
  - SIB*crisis1: -15.08** (6.103) — Full sample; -12.33** (5.042) — US; -11.14*** (3.027) — EA; 1.051 (0.812) — UK.
  - SIB*crisis2: -22.30*** (6.737) — Full sample; -9.42** (3.990) — US; -29.94*** (10.30) — EA; 0.827 (7.553) — UK.
- Note: crisis dummies defined as Crisis1 = 2007Q3; Crisis2 = 2011Q2.

### Policy implications and conclusions (concise)
- Importance: funding costs affect bank profitability, capital and ultimately the banking system’s supply of credit to the real economy.
- Key conclusions:
  - Short-run increases in unsecured funding costs are associated with bank-specific creditworthiness and capital quality, country-level growth and short-term rates, and global risk factors including market volatility and sovereign risks.
  - Long-run funding costs increased over the sample period, mainly driven by deterioration in asset quality, lower near-term domestic growth prospects, continued decline in short-term interest rates, and the impact of successive financial crises.
  - Larger, systemically important banks enjoy a funding advantage; this advantage has risen since the onset of the crisis.
- Policy-relevant finding:
  - Regulatory efforts to strengthen banks' capital buffers may not necessarily raise banks' cost of capital and therefore reduce lending; while balance sheet repair may reduce lending in the short run, higher capital levels lead to lower bank funding costs in the long run, potentially supporting bank lending growth.

### Annex I — decomposition, sample lists and funding trends (selected highlights)
- Decomposition components in Figure 1: Total capital; Policy rate; Tier 1 capital; EA sovereign risk; Yield curve slope; Provisions; Crisis; Long run funding cost. Vertical axis scale shown in Figure 1: from -400 to 500 (bp).
- Core sample table (selected entries and shares, End-2012): Danske A/S — 0.57; BNP Paribas SA — 0.75; Deutsche Bank AG — 0.34; Intesa Sanpaolo SpA — 0.61; Banco Santander SA — 0.51; Handelsbanken — 1.36; Nordea — 0.31; Credit Suisse — 0.75; UBS AG — 0.32; Barclays Plc — 0.84; Bank of America Corporation — 0.89; Citigroup — 0.21; Wells Fargo & Co. — 0.07.
- Trends in cost and volume of funding (key points):
  - From 2007 to 2012 total US banking system deposits rose by US$27trillion.
  - Deposits as a share of total liabilities in the US: close to 85 percent in 1991; 67 percent in 2008; 76 percent of total liabilities at end-2012.
  - Total US banking system liabilities peaked at US$13 trillion at end-2012.
  - Major disruptions in August 2007 led to sharp rises in interbank funding spreads across regions.
  - Quantity of wholesale funding issued by international banks fell sharply since mid-2007 and remained low by historical standards through end-2013.
  - US banks sharply increased reliance on retail deposits; repo funding, interbank funding and debt issuance as a proportion of total assets fell substantially.
  - EA banks restructured financing: market funding was replaced by central bank funding and, in some cases, customer deposits.
  - Nordic banks maintained a strong reliance on collateralized term funding, notably covered bonds.

*Source: _wp1471 - 1. One important legacy of the global financial and Euro Area (EA) crises is their impact*

### 1. One  important legacy  of  the  global  financial  and  Euro  Area  (EA)  crises  is their impact

### 1. One important legacy of the global financial and Euro Area (EA) crises is their impact on the funding models of internationally active banks

### Legacy: shift from retail deposits to wholesale funding and resulting vulnerabilities
- Before the crisis many large global banks had difficulty attracting core deposits and increasingly supplemented retail deposits with wholesale market funding to finance credit growth during the credit boom.
- Growing reliance on short term wholesale finance to fund long term assets created:
  - currency and maturity mismatches;
  - increased liquidity risk.
- Banks more reliant on customer deposit funding fared better during the crisis; evidence suggests their market value exceeded banks funded predominantly through wholesale markets (Beltratti and Stultz, 2011).

### Funding market paralysis, 2007–12
- The start of the global financial crisis in August 2007 coincided with worldwide liquidity shortages at major banks and turmoil in interbank funding markets:
  - interbank lending slowed to a halt;
  - wholesale funding markets froze and interest rates on unsecured term loans between banks rose sharply and remained unusually volatile for an extended period.
- Funding conditions remained tight as investors shunned bank debt and new issuance fell to historic lows.
- The period includes the sub-prime crisis and the EA crisis, contributing to sustained paralysis of bank funding markets through 2012.

### Decline in market funding and bank responses
- By 2007 growth in customer loans had exceeded growth in customer deposits by far for Nordic, EA and UK banks; banks bridged the gap with short term and collateralized term wholesale funding.
- Peak issuance in March 2007: major European countries and the US issued some US$415 billion in secured and unsecured debt, compared to some US$8bn just over a decade earlier.
- Since 2007:
  - capital market issuance has fallen sharply for the majority of global banks;
  - customer funding gaps have narrowed.
- Banks responded to scarce and costly wholesale funding by:
  - shifting towards more customer deposit funding where possible;
  - relying on official financing sources;
  - reducing non-core assets.

### Focus of the paper: factors underlying the cost of international banks’ funding
- The paper investigates why funding costs varied across banking systems and individual banks during the global financial and EA sovereign debt crises.
- Three sets of potential determinants are examined:
  - bank-specific balance sheet variables (including changes in level and quality of capital);
  - country-specific macro-financial factors (domestic economic conditions and changes in short term interest rates);
  - global financial market factors (implied market volatility, shocks such as the global financial crisis and EA sovereign risks, and the global growth outlook).
- The analysis disentangles effects of banks’ resilience measures (e.g., higher capitalization) on funding costs and assesses crisis impacts on long run funding costs.

### Marginal funding cost measure and empirical scope
- The analysis focuses on a bank’s marginal cost of funding (the cost of raising an additional unit of funding), excluding costs from secured funding (e.g., covered bonds or repo) and deposits.
- Rationale:
  - prior to the crisis, banks increasingly used wholesale unsecured funding as their marginal source because such markets provided substantial, relatively cheap, short-notice funds.
- The paper assumes changes in banks’ marginal unsecured wholesale funding costs can be inferred from five-year CDS premia.
- Marginal funding cost estimated as:
  - the sum of a bank’s five-year CDS premia plus three-month Libor (reflecting cost of raising fixed rate senior unsecured bonds and entering an interest rate swap).

### Key empirical findings
- Short run changes in bank unsecured funding costs are associated with:
  - bank-specific characteristics such as an institution’s credit worthiness and changes in the level and quality of capital;
  - country-level factors such as domestic economic conditions and changes in short term interest rates;
  - global risk factors such as implied market volatility, shocks to financial markets (notably the global financial crisis and EA sovereign risks), and the global growth outlook.
- Larger, systemically important banks enjoy a funding advantage, and this advantage has risen since the onset of the crisis.
- In the long run, banks’ funding costs appear to have risen over the sample period, driven mainly by:
  - deterioration in banks’ asset quality;
  - decline in domestic economic conditions;
  - the impact of successive financial crises (global financial crisis and EA sovereign debt crisis).
- Increased capital buffers:
  - can lead to a decline in a bank’s funding costs in the long run;
  - if the increase is due to higher quality Tier 1 capital, there is a short run reduction in funding costs as well.

### Policy implications
- Results counter the view that higher capital requirements necessarily increase bank funding costs and reduce lending to the real economy.
- Supporting evidence:
  - higher bank equity financing raises overall funding costs only modestly (Miles (2011)), and the long-run impact on borrowing costs is likely small;
  - increasing capital buffers may not raise banks' cost of capital and could reduce long-run funding costs, potentially supporting bank lending growth.
- Balance sheet repair may reduce bank lending in the short run, but higher capital levels contribute to lower bank funding costs in the long run.

### Literature review highlights
- ECB (2009) and CGFS contributions document pre-crisis balance sheet growth, shorter maturities of money market funding, securitization and debt issuance that exposed EA banks to maturity and currency mismatches.
- Studies find deposit-funded banks performed better during the crisis, continued lending, and were less risky (Ivashina and Scharfstein (2010); Dermirguc-Kunt and Huizinga (2010); Raddatz (2010); Cornett et al. (2011); Beltratti and Stulz (2012); Dagher and Kazimov (2012); Vazquez and Federico (2012); IMF (2013)).
- Research on “too-big-to-fail” funding cost subsidy shows some evidence that G-SIBs have lower total funding costs attributed to perceived bailouts.
- Empirical studies of bank CDS indicate common factors in CDS premia movements and increased co-movement after the onset of the crisis (Eichengreen et al. (2009); Ballester Miquel, Lukac and González-Urteaga (2013)).

### Empirical model (overview)
- Changes in bank funding costs are modeled in a panel Error Correction Model to disentangle long- and short-run effects of bank-specific and macro variables and to estimate the long-run equilibrium of funding costs for major international banks.
- Dynamic linear regression specification includes:
  - FCi,t: bank i’s marginal cost of funding (three-month Libor plus five-year CDS premia) at time t;
  - Kt-2: vector of capital variables at time t-2 (total capital ratio and capital quality);
  - Mt-1: vector of macro variables at time t-1 (real GDP growth, short term interest rate, yield curve slope).
- Financial crisis dummies:
  - C1t = 1 from Q3 2007 to allow for a change in relationships after the onset of the global financial crisis in August 2007;
  - C2t = 1 from Q2 2011 to allow for a change following market disruption from Q2 2011 triggered by concerns about a potential break-up of the EA.
  - The constant is allowed to change during crisis periods; when C2 = 1 its effect adds to C1 so impacts are cumulative.

*Italic: Source: _wp1471 - 1. One  important legacy  of  the  global  financial  and  Euro  Area  (EA)  crises  is their impact*

### 19. Finally, Z

### 19. Finally, Z

### Model specification and short-run dynamics
- Zt = [Bt-1, Ft] stacks a vector of bank-specific variables at time t-1 (Bt-1) (provision ratio and equity returns) and macro-financial variables at time t (Ft) (market implied volatility and an index of notional weighted CDS premia for periphery European countries).
- The Zt variables have only short-run effects and therefore are included only in changes.
- The model is stable and therefore has the following error correction representation.
- The model is stable and has a long run relationship if -1< α1 < 0.
- Short run estimates refer to an effect lasting around 2–4 quarters before fading away; long run estimates refer to a more permanent effect on funding costs.

### Data and sample
- Sample composition and frequency:
  - 52 banks in 14 advanced economies from 2001–12.
  - Quarterly CDS premia data are used (balance sheet variables are available only on a quarterly basis).
  - Primary data sources are Bloomberg and Datastream.
- Core sample:
  - 25 major international banks from four regions: US, UK, EA and Nordic countries (Denmark, Sweden and Norway).
  - Two criteria for core banks: (i) systemically important within its economy at end-2012; and (ii) availability of CDS premia over 2001–12 on a quarterly frequency.
  - 19 of the 25 banks in the core sample meet the Financial Stability Board’s classification as global systemically important banks at end-2012.
  - Equation (3) is estimated on the core sample of 25 banks.
- Extended sample:
  - Augmented dataset with an additional 30 banks (mainly European, a few US, some Australian and Japanese banks) for tests on funding cost advantages for systemically important banks.
- Key data features:
  - Lagged bank balance sheet variables are used to avoid simultaneity bias.
  - The analysis uses changes in provisioning and non-performing loans to proxy balance sheet ill-health, and capital ratios (total capital/total assets and Tier 1 capital) to proxy resilience.
  - Provisions and fitted values, total capital, Tier 1 capital, real GDP growth, volatility, and yield curve spread are used in scatter plots to illustrate long run relationships.

### Descriptive patterns in CDS premia (2003–2013)
- Common episodes and dates noted:
  - August 2007—US subprime crisis and global bank liquidity shortages worldwide.
  - September 2008—Collapse of Lehman Brothers.
  - April 2010—Greece requests a “troika” bailout (April 23, 2010 marks the beginning of the EA sovereign debt crisis).
  - April 2011—Portugal requests a “troika” bailout.
  - December 2011—ECB initiates LTRO.
  - September 2012—ECB announces OMT.
- Cross-country observations:
  - US banks experienced the sharpest rise in premia following the Lehman collapse in Q3 2008, diverging significantly from other major banking systems.
  - US, UK and EA bank CDS rose in tandem during the EA crisis.
  - UK and EA bank CDS premia diverged around 2012, with UK bank spreads falling faster than EA bank spreads.
- Range and examples of average spreads over the period:
  - Rabobank low of 3.4 bps.
  - Ally Financial high of 2911 bps.
  - Morgan Stanley’s CDS peaked at 1240bp after Lehman.
- Evolutionary phases:
  - 2001 to early-2007: relative calm with little volatility in CDS premia.
  - mid-2007 onward: sharp widespread rises in bank CDS tied to crisis events.
  - April 23, 2010 onward: start of EA sovereign debt crisis with further increases in bank CDS.
  - April 2011 (Portugal bailout): a sharper widespread increase in bank CDS.
  - December 2011 (LTRO) and September 2012 (OMT): preceded sharp falls in bank CDS premia.
  - Since September 2012: steady decline across regions; as of May 2013 bank CDS had settled at levels higher than their pre-crisis peaks; US, Nordic and Japanese banks roughly more than 50 bps higher than at end-2006.

### Long-run relationships and variables used
- Bank-specific variables (long-run and short-run roles):
  - Provisions (proxy for asset quality / balance sheet ill-health).
  - Equity returns (proxy for profitability; short-run effect).
  - Total capital ratio and Tier 1 capital ratio (proxy for capital quality and resilience).
  - CDS bid-ask spread (proxy for CDS liquidity / market illiquidity).
- Macro-financial variables:
  - Actual domestic real GDP growth and expected global economic growth.
  - Equity implied volatility and global market illiquidity indicators.
  - Short term policy interest rate (monetary policy stance).
  - Yield curve slope (10yr–3m) to capture term spread effects.
  - EA sovereign risk index: weighted index of sovereign CDS premia for Greece, Italy, Ireland, Portugal and Spain (weighted by notional amounts).
  - Two crisis dummy variables:
    - Crisis dummy one: onset of the sub-prime crisis in Q3 2007.
    - Crisis dummy two: begins in Q2 2011 capturing EA breakup-related extreme financial market stress.

### Estimation results — bank capital and fundamentals
- Capital effects:
  - Short-run: an increase in total bank capital increases bank funding costs (positive short-run coefficient).
  - Long-run: an increase in bank capital reduces funding costs; on average, a 1pp increase in total bank capital reduces funding costs by 0.26bp.
  - The long-run reduction in equilibrium funding costs from additional capital is highest for US banks and lowest for EA banks.
  - Higher capital quality (Tier 1 / total capital) is associated with lower equilibrium bank funding costs.
- Asset quality and liquidity:
  - Bank provisions: positive and statistically significant — deterioration in asset quality increases funding costs.
  - US and Nordic bank funding costs are especially sensitive to changes in provisions in the short run.
  - CDS bid-ask spread: positive and statistically significant — market illiquidity is associated with higher CDS premia.
- Equity returns:
  - Negative impact on bank funding costs (higher returns imply lower probability of default).

### Estimation results — crisis effects and interactions
- Crisis dummy impacts:
  - First crisis dummy (sub-prime onset): positive and statistically significant; largest coefficient for US banks.
  - Second crisis dummy (EA breakup risk): positive and significant; largest coefficient for EA banks and smallest for US banks; UK banks most affected by the second crisis.
- Interaction effects:
  - Interaction of the sub-prime crisis dummy with capital variables yields a positive and significant coefficient — funding costs became more sensitive to capital since the onset of the crisis (investors differentiate more by capitalization post-crisis).

### Estimation results — macro-financial variables
- Market indicators:
  - Equity returns and implied volatility have only short-run effects on funding costs.
  - Equity returns: negative and significant (higher returns -> lower funding costs).
  - Implied equity market volatility: positive and significant (higher volatility -> higher funding costs).
  - EA sovereign risk index: positive association with higher funding costs.
- Growth:
  - Actual real GDP growth: negative sign (improved domestic growth -> lower funding costs).
  - Expected global growth: negative impact on funding costs, but statistically significant only for EA and UK banks.
- Interest rates and term spread:
  - Changes in short term interest rates: increases in central bank policy rates have a positive and statistically significant impact on bank funding costs.
  - Yield curve slope (term spread) included to control for term-structure effects; theoretical channels allow for either positive or negative relationships depending on net effects on bank profitability and credit risk.

*Source: _wp1471 - 19. Finally, Z*

### 40. We  find  that  the  yield  curve spread is  positively  related  to  bank  funding  costs,

### _wp1471 - 40. We find that the yield curve spread is positively related to bank funding costs

### Yield curve effects on bank funding costs
- The yield curve spread is positively related to bank funding costs, suggesting that in the short run bank funding costs rise with a steepening in the yield curve.
- Mechanisms:
  - Direct impact via the risk-free rate: a steeper yield curve stemming from a rise in interest rates along the curve is expected to translate into a direct increase in banks’ funding costs, with the size of the impact depending on the maturity structure of a bank’s liabilities.
  - Indirect impact via bank credit risk: a steeper curve may reflect weaker recent economic conditions (leading monetary authorities to lower short rates) and thus raise banks’ credit risk in real time.
  - Sovereign-driven channel: a steeper yield curve caused by investor outflows from government bonds due to reduced sovereign creditworthiness can raise bank funding costs (example: EA sovereign debt crisis).
- Empirical coefficients for the slope of the yield curve (preserving reported estimates and standard errors exactly):
  - Short run: 12.36*** (2.360) — Global
  - Short run: 4.991 (7.999) — US
  - Short run: 8.463*** (2.495) — EA
  - Short run: 29.37*** (8.512) — UK
  - Short run: -4.138 (4.201) — Nordics
  - Long run: 7.981*** (1.931) — Global
  - Long run: 5.774 (10.57) — US
  - Long run: 6.230*** (2.253) — EA
  - Long run: 36.95*** (8.134) — UK
  - Long run: 0.931 (3.658) — Nordics

### Short-run and long-run drivers of bank funding costs
- Short-run drivers identified:
  - Bank-specific: institution credit worthiness, changes in level and quality of capital, provisions.
  - Country-level: domestic real GDP growth, changes in short term interest rates.
  - Global risk factors: implied market volatility, CDS liquidity, shocks from financial crises and EA sovereign risks, global growth expectations.
- Selected reported short-run coefficients and standard errors (exactly as reported):
  - Total capital, Short run: 0.202* (0.106) — Global; 0.155* (0.080) — US; 0.227*** (0.113) — EA; 0.113* (3.660**) — UK (table formatting preserved as in source).
  - Capital quality, Short run: -0.456* (0.256) — Global; -0.787** (0.358) — US; -1.301 (1.663) — EA; -0.633** (0.320) — UK; -3.317*** (1.083) — Nordics.
  - Provisions, Short run: 6.641*** (1.341) — Global; 7.651*** (2.658) — US; 2.780 (2.147) — EA; 3.964 (2.525) — UK; 4.419* (2.545) — Nordics.
  - CDS liquidity, Short run: 0.714*** (0.120) — Global; 10.20*** (1.199) — US; 4.696*** (0.475) — EA; 0.357*** (0.109) — UK; 2.183*** (0.770) — Nordics.
  - Equity Returns, Short run: -0.308*** (0.0584) — Global; -0.613*** (0.154) — US; -0.165*** (0.0613) — EA; -0.0261*** (-0.007) — UK; -0.521*** (0.145) — Nordics.
  - Implied market volatility, Short run: 0.502*** (0.0409) — Global; 0.249** (0.0992) — US; 0.429*** (0.0476) — EA; 0.731*** (0.0900) — UK; 0.110 (0.0746) — Nordics.
  - Euro area sovereign risk, Short run: 0.850*** (0.0797) — Global; 1.406*** (0.236) — US; 0.166*** (0.0227) — EA; 0.283* (0.144) — UK; 0.198*** (0.0368) — Nordics.
  - Short term interest rate, Short run: 11.16*** (3.103) — Global; 11.78 (12.79) — US; 5.484 (4.132) — EA; 49.62*** (9.974) — UK; 3.208 (4.478) — Nordics.
- Adjusted R2 reported for the model: 0.57 and 0.622 (table formatting preserved).

### Long-run evolution and regional differences in funding costs
- Overall long-run funding costs rose for all banks in the sample over the sample period, driven mainly by:
  - Deterioration in individual bank credit worthiness (asset quality deterioration).
  - Yield curve steepening due to increased sovereign and banking sector risk.
  - Risks from the global financial crisis and EA sovereign debt crisis.
- Offsetting factors: increases in total bank capital and tier 1 capital helped offset some of the rise in long-run funding costs.
- Regional observations by end-2012:
  - US and Nordic banks: long-run funding costs fell to lowest levels, mainly due to increased capitalization and reduction in asset riskiness.
  - EA periphery banks: long-run funding costs remained elevated; roughly one-half of all EA banks (mainly Italian and Spanish banks) had funding costs above long-run equilibrium.
  - EA core banks (German and French): funding costs had fallen to below long-run equilibrium.
  - UK banks: funding costs declined to below long-run equilibrium, with much of the decline in the last few quarters of 2012.
  - Among five US banks in core sample: funding costs for two remained above long-run equilibrium.
  - Among Nordic banks in sample: funding costs remained above long-run equilibrium for only one bank out of five at end-2012.
- Drivers of sharp increases:
  - US: sharp rise mid-2007 mainly driven by deterioration in asset quality; government capital injections in 2008 and improved capital quality mitigated funding cost increases.
  - EA periphery (early 2011): sharp rise tied to escalation of EA sovereign debt crisis — sovereign and bank perceived riskiness rose, investor outflows accelerated, term spreads rose (short and long rates rising in tandem), asset quality deteriorated and capital levels remained insufficiently low.

### Domestic Systemically Important Banks (D-SIBs) — funding cost differentials
- Definition used in analysis: dummy = 1 if book assets of the bank exceed US$500 billion in that quarter (SIB).
- Main empirical findings:
  - SIBs had lower funding costs relative to peers over the sample period.
  - Over the full sample period the coefficient on the SIB dummy suggests the cost of marginal unsecured funding was on average 16 bps lower for SIBs relative to non-SIBs.
  - After 2011 the funding advantage rose by more than 40 percent (particularly for EA SIBs); post crisis, SIBs’ cost of funding was roughly 22 bps lower compared to non-SIBs.
  - This funding advantage likely reflects market perceptions that large banks are “too-big-to-fail” and enjoy an implicit government support advantage.
- Reported coefficients from the SIB regression table (exact values and standard errors):
  - SIB dummy: -15.81*** (4.580) — Full sample; -18.24 (16.10) — US; -8.530** (3.103) — EA; -3.875 (13.49) — UK; -0.696 (8.536) — Nordics.
  - Crisis1 dummy: 12.54*** (3.934) — Full sample; 20.853* (18.08) — US; 4.483*** (1.020) — EA; 6.09* (3.65) — UK; 10.90** (5.061) — Nordics.
  - Crisis2 dummy: 8.116** (3.319) — Full sample; 3.29** (1.56) — US; 17.84*** (6.26) — EA; 10.131*** (4.11) — UK; 1.777 (7.988) — Nordics.
  - SIB*crisis1: -15.08** (6.103) — Full sample; -12.33** (5.042) — US; -11.14*** (3.027) — EA; 1.051 (0.812) — UK.
  - SIB*crisis2: -22.30*** (6.737) — Full sample; -9.42** (3.990) — US; -29.94*** (10.30) — EA; 0.827 (7.553) — UK.
- Note: crisis dummies defined as Crisis1 = 2007Q3; Crisis2 = 2011Q2.

### Policy implications and conclusions
- Importance: funding costs affect bank profitability, capital and ultimately the banking system’s supply of credit to the real economy.
- Key conclusions:
  - Short-run increases in unsecured funding costs are associated with bank-specific creditworthiness and capital quality, country-level growth and short-term rates, and global risk factors including market volatility and sovereign risks.
  - Long-run funding costs increased over the sample period, mainly driven by deterioration in asset quality, lower near-term domestic growth prospects, continued decline in short-term interest rates, and the impact of successive financial crises.
  - Larger, systemically important banks enjoy a funding advantage; this advantage has risen since the onset of the crisis.
- Policy-relevant finding:
  - Regulatory efforts to strengthen banks' capital buffers may not necessarily raise banks' cost of capital and therefore reduce lending; while balance sheet repair may reduce lending in the short run, higher capital levels lead to lower bank funding costs in the long run, potentially supporting bank lending growth.

*Source: IMF working paper content provided in the supplied excerpt.*

### ANNEX I

### ANNEX I

### Decomposition of Long-run Bank Funding Costs
- The components in Figure 1 add up to the estimated long run funding cost over the period 2001-2012 and are based on average values for factors.
- Components displayed: Total capital; Policy rate; Tier 1 capital; EA sovereign risk; Yield curve slope; Provisions; Crisis; Long run funding cost.
- Regions illustrated in Figure 1: US; UK; EA Periphery; EA Core; Nordics.
- Vertical axis scale shown in Figure 1: from -400 to 500 (bp).  
- Time horizon shown: 2001, 2003, 2004, 2006, 2007, 2009, 2010, 2012.

### Core sample of banks used in regressions (selected entries and shares)
- Denmark: Danske A/S — 0.57 (Share of banking system total assets, End-2012); G-SIB: No.
- France: BNP Paribas SA — 0.75; G-SIB: Yes.
- France: Group Credit Agricole — 0.29; G-SIB: Yes.
- France: Societe Generale SA — 0.28; G-SIB: Yes.
- Germany: Deutsche Bank AG — 0.34; G-SIB: Yes.
- Germany: Commerzbank AG — 0.26; G-SIB: No.
- Italy: Intesa Sanpaolo SpA — 0.61; G-SIB: No.
- Italy: Unicredit SpA — 0.25; G-SIB: Yes.
- Spain: Banco Santander SA — 0.51; G-SIB: Yes.
- Spain: BBVA — 0.34; G-SIB: Yes.
- Sweden: Handelsbanken — 1.36; G-SIB: No.
- Sweden: Nordea — 0.31; G-SIB: Yes.
- Sweden: Skandinaviska Enskilda Banken AB (SEB) — 0.75; G-SIB: No.
- Switzerland: Credit Suisse — 0.75; G-SIB: Yes.
- Switzerland: UBS AG — 0.32; G-SIB: Yes.
- UK: Barclays Plc — 0.84; G-SIB: Yes.
- UK: Lloyds Banking Group Plc — 0.23; G-SIB: No.
- UK: Royal Bank of Scotland Group Plc — 0.14; G-SIB: Yes.
- UK: HSBC Holdings Plc — 0.20; G-SIB: Yes.
- US: Bank of America Corporation — 0.89; G-SIB: Yes.
- US: Citigroup — 0.21; G-SIB: Yes.
- US: Goldman Sachs Group, Inc — 0.17; G-SIB: Yes.
- US: JP Morgan Chase & Co. — 0.09; G-SIB: Yes.
- US: Morgan Stanley — 0.22; G-SIB: Yes.
- US: Wells Fargo & Co. — 0.07; G-SIB: Yes.

### Additional set of banks included in estimation (selected entries and shares)
- Austria: Erste Group Bank — 0.40; G-SIB: No.
- Austria: Raiffeisen Zentralbank — 0.24; G-SIB: No.
- Australia: Macquarie Bank — 0.97; G-SIB: No.
- Australia: National Australia Bank — 0.04; G-SIB: No.
- Australia: Westpac — 0.25; G-SIB: No.
- Australia: Australia & New Zealand Banking Group — 0.22; G-SIB: No.
- Australia: Commonwealth Bank of Australia — 0.21; G-SIB: No.
- Belgium: KBC Bank — 1.11; G-SIB: No.
- Belgium: Dexia — 0.46; G-SIB: No.
- France: Natixis — 0.08; G-SIB: Yes.
- Germany: Deutsche Postbank — 0.04; G-SIB: Yes.
- Germany: HSH Nordbank — 0.02; G-SIB: No.
- Netherlands: ING Bank NV — 0.91; G-SIB: Yes.
- Netherlands: Rabobank — 0.46; G-SIB: No.
- Netherlands: ABN Amro — 0.30; G-SIB: No.
- Norway: DNB Nor ASA — 3.08; G-SIB: No.
- Spain: Banco Popular Espanol — 0.11; G-SIB: No.
- Spain: Banco de Sabadell — 0.04; G-SIB: No.
- US: Ally Financial — 0.03; G-SIB: No.
- US: American Express — 0.02; G-SIB: No.

### CDS premia for sample of banks (summary points)
- Core sample total (as reported): Total88.552.599.541033 (table shows core sample total CDS statistics with mean, median, standard deviation, min, max).
- Additional banks total (as reported): Total150.8109.6108.53.42910.7 (table shows additional banks total CDS statistics with mean, median, standard deviation, min, max).
- Individual bank CDS statistics are reported in Table 3 with Mean, Median, Standard Deviation, Min, Max (basis points) for each bank in the Core Sample and Additional Banks.

### Empirical literature on bank CDS (high-level)
- Table 4 summarizes empirical studies (examples): Chiaramonte and Casu (2011); Annaert, J., et al. (May 2010); Volz and Wedow (2009); Raunig and Scheicher (2008); Yibin, B., et al. (2005).
- Common controls across studies include bank-specific controls (asset quality, capital ratios, leverage, liquidity, equity volatility) and macro-financial controls (risk free rates, yield curve slope, market volatility indices such as VIX/VSTOXX, market liquidity measures).

### Trends in the cost and volume of bank funding (Annex II — key findings)
- Deposits:
  - During a financial crisis investor risk appetite declines, producing a “flight to liquidity” and a rise in precautionary savings.
  - In the US and the UK, the average cost of deposits fell steadily from 2008; not entirely the case in the EA.
  - The average effective cost of deposits rose for some banks in the EA, most pronounced for Spanish and Italian banks; increases were more moderate for French and especially German banks.
  - The increase in deposit costs in periphery EA countries reflected accelerating deposit flight during 2011-2012.
  - Drivers for increased US deposit availability included: (i) corporate liquidity buildup; (ii) FDIC’s introduction of an unlimited guarantee of deposit balances in 2008 and Dodd-Frank mandated guarantee of non-interest bearing deposits from 2011-2012; (iii) “flight to safety” out of money market funds after the Lehman failure.
  - From 2007 to 2012 total US banking system deposits rose by US$27trillion.
  - Deposits as a share of total liabilities in the US: close to 85 percent in 1991; 67 percent in 2008; 76 percent of total liabilities at end-2012.
  - Total US banking system liabilities peaked at US$13 trillion at end-2012.
  - For UK banks, the average effective cost of deposits fell sharply for relatively stronger banks and only moderately for weaker banks.
- Interbank funding:
  - Major disruptions in August 2007 led to sharp rises in interbank funding spreads across regions (Figures 3-4).
  - Interbank spreads do not necessarily imply accessibility to funding at those costs because many banks were frozen out of the interbank market on fears of counterparty risk.
  - In the US, the overnight federal funds effective rate rose sharply at the start of the crisis but fell quickly after Federal Reserve intervention; 3month LIBOR surged and remained elevated and disconnected from the target federal funds rate for a period.
  - In the EA, interbank funding costs rose owing to increased uncertainty around counterparty creditworthiness; banks became increasingly dependent on ECB liquidity operations and overnight borrowing post-Lehman collapse.
- Secured debt and debt issuance:
  - The cost of bank debt (both secured and unsecured) increased substantially from 2007 across all types of securities, including covered bonds.
  - Quantity of wholesale funding issued by international banks fell sharply since mid-2007 and remained low by historical standards through end-2013 (Figures 6-7).
  - Global issuance of asset and mortgage backed securities collapsed and remained low at end-2013; covered bond issuance also remained well below historical levels.
  - US banks’ secured issuance had been dominated by asset-backed and mortgage-backed instruments; at the peak in early 2007 roughly two thirds of all term debt issued by US banks was in mortgage backed securities.
  - New issuance collapsed across the Euro Area, the US, and the UK with respect to both secured and unsecured funding sources beginning mid-2007; the decline was especially precipitous for US banks in 2008.
  - Since the crisis, there has been some recovery in unsecured debt issuance but US issuance of collateralized debt securities remains negligible; EA and UK new debt issuance has continued downward and does not appear to have bottomed out by the period covered.
- Changes in funding structure and balance sheets:
  - US banks sharply increased reliance on retail deposits; repo funding, interbank funding and debt issuance as a proportion of total assets fell substantially.
  - EA banks reduced funding requirements and disposed of assets: from 2007–12, core EA (France and Germany) and periphery EA (Italy and Spain) total banking system assets fell by US$610bn and US$645bn, respectively.
  - EA banks restructured financing: market funding was replaced by central bank funding and, in some cases, customer deposits.
  - Among core EA banks, customer funding accounted for the bulk of bank liabilities in 2012; in 2007 the bulk of bank assets were funded through issuance of debt securities.
  - Among periphery EA banks, the decline in debt issuance has largely been replaced by increased recourse to central bank funding.
  - UK banks contracted balance sheets beginning in 2007 but, since 2009, total banking system assets have risen again to near pre-crisis levels supported by an increase in customer deposits and central bank financing.
  - Nordic banks maintained a strong reliance on collateralized term funding, notably covered bonds; DNB Nor ASA reported a share of 3.08 (end-2012) in the additional banks table. Nordic banks’ reliance on interbank funding has fallen sharply since the beginning of the crisis.

### Observations on data sources and figures
- Figures and tables cited use Dealogic, US Flow of Funds, SNL Financial, and authors’ estimates.
- Figures referenced: Figure 1 (Decomposition of Long-run Bank Funding Costs), Figure 2 (Average cost of deposits), Figures 3-4 (Overnight and 3month Interbank Rate Spreads), Figure 5 (Cost of Bank Debt), Figures 6-8 (Secured and Unsecured Debt Issuance and Debt Issuance by Region), Figure 9 (Structure of liabilities for major banking systems).
- Notes in figures:
  - For Figure 2: (a) Calculated as Total Interest Expense on Deposits/Total Interest bearing deposits; (b) Banks included in each category are detailed in Annex 1 Table 1. EA refers to France, Germany, Italy and Spain. Nordics refers to Denmark, Sweden and Switzerland.
  - For Figures 3-4: (a) US is the spread between the Fed Funds Effective and Target Rates; (b) EA is the spread between EURIBOR and EUREPO overnight; (c) UK is the LIBOR-OIS overnight spread.
  - For Figure 5: Spread to benchmark rate (10 year US Treasury).
  - For Figures 6-8: (a) Includes major international banks in the EA, the US, UK and Nordic countries. (b) Secured debt includes mortgage and asset backed securities and covered bonds; unsecured debt includes medium term notes, high yield and investment grade bonds.

*Source: _wp1471 - ANNEX I*

### References

### _wp1471 - References

### Liquidity, funding patterns, and interbank markets
- Acharya, Viral, V. and Ouarda Merrouche, (2010). “Precautionary Hoarding of Liquidity and Interbank Markets: Evidence from the Sub-prime Crisis,” NBER Working Paper 16395.
- Angelini, P, Andrea Nobili, and Maria Cristina Picillo. (2011). “The Interbank Market After August 2007: What has changed and Why?” Journal of Money, Credit and Banking, Vol.43(5), 923-958.
- Araten, Michel and Christopher Turner (2012), “Understanding the Funding Cost Differences between Global Systemically Important Banks (G-SIBs) and Non-G-SIBs in the United States,” JPMorgan Chase.
- Funding Patterns and Liquidity Management of Internationally Active Banks, Committee on the Global Financial System (CGFS) Papers, No 39, May 2010.
- FithRatings Global Special Report (June 2011). “Trends in Bank Funding Profiles”.
- van Rixtel, Adrian., and Gabriele Gasperini (2013). “Financial crises and bank funding: recent experience in the euro area.” BIS Working Paper No.406.
- Raddatz, Claudio, 2010. "When the rivers run dry : liquidity and the use of wholesale funds in the transmission of the U.S. subprime crisis," Policy Research Working Paper Series 5203, The World Bank.
- Dagher, Jihad and Kazimov, Kazim, Banks' Liability Structure and Mortgage Lending During the Financial Crisis (October 1, 2012). IMF Working Paper, No 12/155.
- Le Lesle, V., (2012). “Bank Debt in Europe: Are Funding Models Broken?” IMF Working Paper WP/12/299.
- Pablo Federico & Francisco F. Vazquez, 2012. "Bank Funding Structures and Risk: Evidence from the Global Financial Crisis," IMF Working Papers 12/29, International Monetary Fund.

### Bank liquidity risk management, lending, and credit supply during crises
- Cornett, Marcia, Jamie McNutt, Philip Strahan, and Hassan Tehranian. (2011). “Liquidity Risk Management and Credit Supply in the Financial Crisis.” Journal of Financial Economics 101(2), pp. 297–312.
- Ivashina, Victoria, and David S. Scharfstein. "Bank Lending During the Financial Crisis of 2008." Journal of Financial Economics 97, no. 3 (September 2010): 319–338.
- Carpenter, Seth B., S Demiralp and J. Eisenschmidt, 2013. "The effectiveness of the non-standard policy measures during the financial crises: the experiences of the Federal Reserve and the European Central Bank," Finance and Economics Discussion Series 2013-34, Board of Governors of the Federal Reserve System (U.S.).
- Ritz, Robert.A., (2012). “How do banks respond to increased funding uncertainty?” Cambridge Working Papers in Economics 1213, Faculty of Economics, Cambridge University.
- Button, Richard, S Pezzini and N Rossiter (2010) “Understanding the price of new lending to households”, Bank of England Quarterly Bulletin, 50:4.

### Bank capital, regulation, and too-big-to-fail subsidies
- Admati, A, DeMarzo, P, Hellwig M, and Pfleiderer P. (2010) “Fallacies, Irrelevant Facts, and Myths in Capital Regulation: Why Bank Equity is Not Expensive”. Stanford University Working Paper No. 86.
- Kashyap, Anil, Jeremy C. Stein, and Samuel G. Hanson (2010), ‘An Analysis of the Impact of ‘Substantially Heightened’ Capital Requirements on Large Financial Institutions’, Mimeo.
- Miles, David, Jing Yang and Gilberto Marcheggiano (2012), “Optimal Bank Capital,” The Economic Journal, 123 (March), 1–37.
- Dean Baker & Travis McArthur, (2009). "The Value of the “Too Big to Fail” Big Bank Subsidy," CEPR Reports and Issue Briefs 2009-36, Center for Economic and Policy Research (CEPR).
- Demirgüç-Kunt, Asli and Huizinga, Harry, Are Banks Too Big to Fail or Too Big to Save? International Evidence from Equity Prices and CDS Spreads (January 2, 2010). CentER Discussion Paper Series No. 2010-59.
- Demirgüç-Kunt, Asli & Huizinga, Harry, 2010. "Bank activity and funding strategies: The impact on risk and returns," Journal of Financial Economics, Elsevier, vol. 98(3), pages 626-650, December.
- Kenichi Ueda and Beatrice Weder di Mauro (2012). ”Quantifying Structural Subsidy Values for. Systemically Important Financial Institutions.” WP/12/128
- Warburton, A. Joseph, Anginer, Deniz and Acharya, Viral V., (2013) “The End of Market Discipline? Investor Expectations of Implicit State Guarantees” Available at SSRN: http://ssrn.com/abstract=1961656 or http://dx.doi.org/10.2139/ssrn.1961656
- Li, Zan, Shisheng Qu, and Jing Zhang (2011), “Quantifying the Value of Implicit Government Guarantees for Large Financial Institutions,” Moody’s Analytics Quantitative Research Group.

### Credit default swaps, sovereign credit risk, and market measures of fragility
- Eichengreen, Barry, Ashoka Mody, Milan Nedeljkovic, and Lucio Sarno (2009), "How the Subprime Crisis Went Global: Evidence From Bank Credit Default Swaps," NBER Working Paper 14904, April.
- Ballester Miquel, Laura, Barbara Casu Lukac, Ana González-Urteaga (2013), “Bank CDS Spreads and Banking Fragility,” Electronic copy at: http://www.efmaefm.org/0EFMAMEETINGS/EFMA%20ANNUAL%20MEETINGS/2013-Reading/papers/EFMA2013_0018_fullpaper.pdf
- Francis A. Longstaff, Jun Pan, Lasse H. Pedersen, and Kenneth J. Singleton, 2007. "How Sovereign is Sovereign Credit Risk," NBER Working Papers 13658, National Bureau of Economic Research.

### Shadow banking, market structure, and models
- Pozsar, Zoltan., T. Adrian, A. Ashcraft and H. Boesky (2012). “Shadow Banking”. Federal Reserve Bank of New York, Staff Report No. 458.
- Stephane Priaulet (2008), “Are credit spreads paying enough? How models can help find the answer.” A theoretical discussion of JP Morgan’s “Rock Bottom Price” CDS pricing model.
- Taylor, John B. and John C. Williams (2009). “A black swan in the money market.” American Economic Journal: Macroeconomics 1, 58.83.

### Precautionary savings and macroeconomic context
- Mody, Ashoka, Franziska Ohnsorge and Damiano Sandri (2012). “Precautionary Savings in the Great Recession”, IMF Working Paper WP/12/42.
- International Monetary Fund, 2013, “Changes in Bank Funding Patterns and Financial Stability Risks,” Global Financial Stability Report, World Economic and Financial Surveys (Washington; October)
- Financial Stability Report, 2011(2), Bank of England.

### Other theoretical and empirical corporate finance foundations
- Myers, Stewart, and Nicholas Majluf (1984). “Corporate Financing and Investment Decisions When Firms Have Information that Investors Do Not Have.” Journal of Financial Economics 13, 187-221.
- Beltratti, Andrea and Rene Stultz (2012) "The credit crisis around the globe: Why did some banks perform better?" Journal of Financial Economics, 2012, v105(1), 1-17.
- Franc is A. Longstaff et al., 2007. "How Sovereign is Sovereign Credit Risk," NBER Working Papers 13658.
- Raddatz, Claudio, 2010. "When the rivers run dry : liquidity and the use of wholesale funds in the transmission of the U.S. subprime crisis," Policy Research Working Paper Series 5203, The World Bank.

*References list as provided in the source document.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1471.pdf_