This is an extract of our Oxford Bcl Principles Of Financial Regulation Notes document, which we sell as part of our Principles of Financial Regulation Notes collection written by the top tier of Oxford students.
The following is a more accessble plain text extract of the PDF sample above, taken from our Principles of Financial Regulation Notes. Due to the challenges of extracting text from PDFs, it will have odd formatting:
PRINCIPLES OF FINANCIAL REGULATION
IMPORTANCE OF BANKS
TRANSFORMATIONLiquidity, maturity and credit transformation:
Liquidity transformation: allows liquid assets of depositors to be converted into investments in illiquid assets, while keeping the deposits sum certain.
Maturity transformation: converts short-term deposits into long-term investments, thus offering depositors part of the benefit of higher returns that long-term investments earn.
Credit transformation: transform low-risk liabilities into high-risk investments do this by screening, monitoring and controlling the risks of investments, pooling risks across a large number of investments, and holding capital to provide protection against failure of investments.
UNIQUENESS OF BANKSAny of the transformation functions might be performed by another institution - BUT there are no other institutions which performs these functions together.
If we look at the assets and liabilities sides of bank balance sheets at the same time that they look unique - BUT there are substitutes for assets and liabilities if looked at individually.Banks are important for economic growth - eg supply of credit to the economy - particularly SMEs (Davies).
MONEY CREATIONBoE argues that money creation by banks is a positive externality.
Ricks agrees with BoE conception of money creation instantaneous with loan creation.
Deposits are simply a record of how much the bank itself owes its customers they are a liability of the bank, not an asset that could be lent out (BoE).The failure of a bank is of particular social, political and economic significance due to the impact it may have on the country's payment system (Vickers Report).
Primary difference between banks and non-bank financial institutions is that bank liabilities are part of the monetary system and can be used as payment (PFR).
The disruption to the payment system is immediately apparent and costly (PFR).
FRAGILITY OF BANKSFragility arises from liquidity, maturity and credit transformation the function of the carry trade at the heart of the business model of fractional reserve banking.
Relies on the law of large numbers but creates a risk of bank runs.
Balance sheet insolvency
Cash flow insolvency
The value of assets no longer match the value of liabilities.
Unanticipated / substantial short-term funding withdrawal creating liquidity problem.
BANK RUNSThe Diamond and Dybvig coordination problem coupled with the credit, maturity and liquidity mismatches between a bank's assets and liabilities leaves banks vulnerable to runs. Diamond and Dybvig coordination problem: if depositors think bad loans or other assets are undermining bank's balance sheet solvency they may rationally decide to withdraw their deposits,
resulting in cash flow insolvency.
Diamond and Dybvig presents depositors' position as having two principal equilibria: (1)
stay; or (2) run the inability to coordinate their actions means that the stay equilibria is precarious and may shift at any time to a run equilibria.
Risk of bank runs can lead to a liquidity problem where a bank is unable to convert its long-term illiquid assets into cash to meet its depositor liabilities (except at a discount to hold to maturity values).
Resulting fire sale of assets creates a feedback loop where the bank was initially facing a shortage of liquidity - BUT then becomes insolvent on a balance sheet basis as it attempts to meet its liquidity needs (Davies).
RATIONALE FOR REGULATING BANKS
IMPLICIT SUBSIDIESBanks are able to raise finance more cheaply than other financial institutions due to DGS and bail-outs creating implicit subsidies (PFR).
ALSO larger banks are able to borrow more cheaply than smaller banks (Skeel) - eg in 2009 paid
0.78% less for funds than their smaller peers.
MORAL HAZARDThe implicit guarantee for too big to fail banks has led to excessive risk-taking (Salz Review) may have been a contributing cause of the GFC (Armour).
Regulatory strategy to prevent balance sheet insolvency by increasing loss-absorbing capacity.
ALSO to prevent excess leverage in capital structure arising from moral hazard associated with provision of under-priced liquidity insurance (Admati et al).
Internalises systemic costs which are not fully borne by the bank (Kashyap and Stein).
BUT - Ricks argues that the liquidity event in the GFC (rather than inadequate capital) bore primary responsibility for the credit crunch.
PROBLEMS WITH IRB APPROACH IN BASEL I AND BASEL II-
Definition of T1 capital in Basel I and II wasn't truly loss absorbing.
RWA approach to IRB led to capital arbitrage (Jones et al) risk weighting varied widely across banks.
Shift from Basel I standardised approach to RWA to Basel II IRB approach to RWA was crossing a regulatory rubicon (Haldane).
By permitting banks to calculate risk weights on their own value-at-risk (VaR) or other models the regulators passed the baton (Haldane).
BCBS notes a number of weaknesses with using VaR for determining capital requirements, including its inability to capture tail risks - THUS procyclical.
Arguments for IRBBanks use own internal data and actual historical experience with losses reflected in the results.
Standardised approach gives no value to spread of portfolio of assets across different classes (may be uncorrelated).
Arguments against IRBBig banks welcomed IRB as it produced lower capital charges than demanded by standardised approach.
2 -IRB worked poorly from the perspective of financial stability banks were able to build up excessive leverage while still showing strong RWA capital ratios (BCBS).
The IRB approach was too reliant on banks' internal models that reflect a private view of risk
The assets of the 10 largest banks approx. doubled in decade before the crisis, but their RWAs increased by only one third core activities of banks (taking deposits and making loans) both fell as a proportion of total assets, while securities holdings increased (PFR).
There was an excessive variation of RWA results between banks (BCBS).
FSA study of 13 banks found some banks with estimates 3-6x higher than others.
Opacity of RWA, inconsistency between firms and granularity making it nearly impossible to account for differences between banks provides a near limitless scope for arbitrage (Haldane).
Responses to Problems with IRB
- Sophistication of BCBS approach to market risk substantially increased (BCBS Review of the Trading
Criteria for access to IRB approach more stringent specifies qualities IRB must display (BCBS).
- Setting floor below which IRB can't reduce capital set by standardised approach (DFA Collins Amendment).
- Still generates capital requirement formulated on RWA basis - BUT is a major qualification of IRB approach as it calculates amount of capital needed to survive a hypothetical stress test - THUS not just relying on historical loss data (PFR).
BCBS highlights the importance of ensuring sufficient capital during significant market stress as this is when it is most likely to absorb losses.
- Rely on IRB approach with supplementary stress testing potentially uncrosses the rubicon - putting power back in the hands of regulators.
- BCBS backstop LR = 3%.
US has gone further with the DFA Collins Amendment with 6% on the deposit-taking member of the group and a 5% leverage buffer at group level (supplemental leverage approach).
Haldane argues that LR of 7% needed to guard against failure in the GFC of the world's largest banks.
Vickers argues that 3% is a lot of leverage way above corporate norms in the rest of the economy.
- Potentially the most radical approach to remedy problems with RWA and IRB designed as a backstop against distortions created by RWA and IRB approach.
Three Primary Functions of LR
- Three primary functions:
1. constraining activity of banks in economic upswings when historical internal data likely to indicate less chance of borrower default;
2. combatting model risk (that the banks' internal models are defective); and
3. dealing with uncertain events in the Knightian sense (BCBS).
- Haldane argues that in responding to risks, a fine turned response is required - BUT in responding to uncertainty we need broad brush heuristics.
Haldane argues that IRB models are unlikely to be robust for many decades, perhaps centuries - it is close to impossible to tell whether their results are prudent - THUS proposes removal of IRB models and applying simple LR.
Haldane proposes that a formulation that would avoid regulatory arbitrage (eg LR encouraging banks to increase their risk per units of assets) and which preserves robustness would be to place leverage and capital ratios on more equal footing (rather than LR as a backstop) - eg BoE has given leverage ratios equal billing.
Twin-track Approach to LR and RWA
- LR doesn't displace RWA it supplements it via a twin-track approach (as LR can easily be manipulated by banks increasing riskiness of activities but not value of associated assets).
Firm believers in heuristics might argue that risk-weighted elements could be removed entirely from the regulation (Haldane) BUT study by Aikman et al found that the RWA approach works well and IRB substantially improved if number of prior year data in the model increased.
Stress testing can be seen as functional equivalent of increasing data - THUS Aikman et al's study shows support for twin-track approach.
PROCYCLICAL ISSUES WITH BASEL I AND II RWASRWA can be procyclical during expansionary and contractionary phases of business and credit cycle (PFR).
Drumond argues that the RWA requirement amplifies procyclicality.
If it is difficult for banks to raise capital in downtimes, banks may be induced to reduce lending and sell assets to meet their capital requirements - THUS amplifying the initial downturn (Drumond).
Basel III Capital Conservation Buffer (CCB)-
Designed to serve as a buffer that banks build up during upswings for use during downswings aims to address procyclicality during downswings (Drumond).
Applies to all banks all of the time.
Mandatory with the ability for banks to distribute dividends to shareholders and bonuses to managers linked to levels of compliance 2.5% above minimum 4.5% CET1 allows 100% of earnings distributable.
Supervisory pressure on banks to behave differently will be needed if rationale of CCB is to be realised
Private interests of managers and shareholders will probably still favour reducing assets in a downturn
early empirical study by IMF suggests that operation of CCB has the potential to exacerbate the problems its supposed to cure (PFR).
Basel III Counter-Cyclical Capital Buffer (CCyB)-
Designed to constrain excessive lending during upswings.
May be applied to all or some banks and applies on discretion of supervisors also linked to ability to distribute dividends and bonuses if applied.
Supervisors may impose CCyB anywhere between 0% and 2.5% CET1.
Unclear whether supervisors will have the foresight to see a crisis / bubble coming or whether they have the incentive to put the breaks on lending.
Out of the Core Basel Committee - only three have set CCyB first mover disadvantaged (increased financial security benefits all banks, but big banks lose out from the rules - eg in UK most banks area really big so would have a big disadvantage if imposed).
In the case of the US, the current (2016) rate is 0%, in the UK it is 1% but going down to 0% soon,
and in Sweden it is 2%.
BASEL III CAPITAL REFORMS
What Counts as Capital under Basel III
Types of capital (Basel III)
Liabilities included 4 Core Tier 1 capital (CET1)
Additional Tier 1 capital (AT1)
Tier 2 capital (T2C)
Ordinary common shares
Perpetual preferred stock
Perpetual subordinated debt capable of being written off or converted into equity before insolvency (cocos)
- Subordinated debt capable of being written off or converted into equity before insolvency with a maturity of greater than 5 years (cocos)Haldane is critical of the regulatory complexity in the definition of capital increases opacity.
Haldane argues that simpler measures of accounting capital based on equity capital (CET1)
outperform more complex measures simple market-measures of banks' equity dominate accounting measures in their crisis-predictive performance.
Hanson et al argue all of T1 capital should be equity / cocos preferred stock shouldn't count.-
Should the Basel III Capital Requirements be Increased?
Types of capital (Basel III)
Core Tier 1 capital (Common Equity T1 or CET1)
Additional Tier 1 capital (AT1)
Tier 2 capital
CET1 + AT1 >6%
CET1 + AT1 + T2 >8%
Total possible capital requirements for G-SIB: CET1 4.5% + CCB 2.5% + CCyB 2.5% + G-SIB 3.5% = 13%This is assuming that the discretionary CCyB is applied by supervisors and that the bank is deemed a GSIB in highest bracket but JP Morgan only at 2.5% of G-SIB surcharge and no CCyB = 9.5%
PLUS LR of 3%.Capital required under Basel III iss small compared to the actual losses suffered by banks in the crisis (PFR).
Reason for the modest changes = higher levels of capital require long transition periods / there will be competitive disadvantages if countries move to CET1 implies that policymakers don't accept the proposition that equity isn't more expensive (PFR) - or due to political constraints / lobbying.
Is Equity More Expensive than Debt?
Arguments that Equity isn't More Expensive than Debt
- Admati et al argue equity is not more expensive than debt for society (cf. private costs).
Replacing a portion of long-term debt with equity will increase capital without reducing productive lending and deposit-taking activity banks can meet capital requirements without reducing amount of deposits or assets (Admati et al).
Miles et al support this equity is a form of financing and all other things equal, if a bank raises more equity it has more money to lend - not less.
- Holding excessive HQLA (eg LCR) is costly and inefficiency can be interpreted as a social cost - BUT
additional equity doesn't need to be invested in cash and can be put towards profitable lending (Admati et al).
- Modigliani and Miller theorem shows (under strong assumptions) that the value of a firm is not increased or decreased by its funding structure.
Miles et al note that the MM theorem is unlikely to hold exactly but that recent empirical research for the US suggests that it might not be a bad approximation even for banks.
ALSO - Hanson et al argue that the Modigliani and Miller theorem isn't meant to be an accurate depiction of reality the value of the framework is that it allows for more disciplined analysis of effects associated with changes in capital structure.
5 Miles et al argue that, even if we assume a substantial departure from MM theorem and that any extra tax paid by banks is a loss to society the costs of stricter capital requirements are fairly small.
Hanson et al find that the displacement of debt finance (long-term or short-term) with equity will only result in modest cost increases.
Arguments that Equity is More Expensive than Debt
- Admati et al note that equity is more expensive than debt for four reasons - BUT the first three can be remedied by regulation and the fourth is only a private cost:
1. Tax deductibility of interest on debt and not for dividends.
Increased costs may be pushed onto borrowers, with the result that marginal projects are no longer funded - THUS potentially social cost (Firestone et al).
BUT Admati et al argue that this could be remedied by using extra tax revenues (from dividends) to subsidise financing of marginal projects OR altering tax system.
2. Signalling effect of equity issuance (Myers and Majluf).
Admati et al argue that prohibiting dividend and equity payouts for banks is a prudent and efficient way to have banks build up capital - BUT may lead to a negative inference of the health of the bank.
BUT Admati et al argue could be remedied by mandatory issues if bank can't raise equity at any price - may be insolvent / non-viable without subsidies - THUS should be unwound.
3. Debt overhang problem (Myers).
May instead prefer to raise capital ratios by reducing assets - THUS reducing the supply of funding to business projects.
Bebchuk and Spamann and Gordon also note that executive compensation tied to shares creates the Fuld problem.
BUT Admati et al argue that this could be remedied by mandatory share issues.
4. Loss of implicit state subsidy - BUT this is a private cost which is socially beneficial (Admati et al).
Higher capital requirements complement other policy initiatives to reduce or eliminate the guarantees (Admati et al).
- BUT - Firestone et al argue that the costs of bank capital are passed along to borrowers.
Cf. Hanson et al note worrying effects of higher capital requirements will be the pressure created for activity to migrate outside of regulated sector (rather than effect on the cost of borrowing) - eg rise in
P2P lending following Basel III.
- ALSO argument that capital is costly for banks because the composition of a bank's balance sheet is essential to its value the business of a bank is financing long-term credit claims with deposits / extending sum certain / continuously redeemable liquidity services to depositors (PFR).
Gorton and Winton suggest that an increase in bank capital is costly as it reduces liquidity creation by banks by decreasing the aggregate amount of deposits - may be esp. significant for small banks.
THUS all equity bank will be less valuable than bank funded at least in part by deposits (PFR).
Trade-off Between Financial Stability and Economic Growth-
Unless the regulatory action is taken debt is at least slightly more expensive than equity - BUT Miles et al estimate that even proportionally large increases in bank capital are likely to result in small long-run impact on the borrowing costs faced by bank customers.
Miles et al note that in the UK and US economic performance was not obviously far worse and rates charged on bank loans were not obviously higher when banks relied on higher equity funding.
ALSO little evidence that investment / growth rate of the economy picked up as leverage moved sharply higher in recent decades (Miles et al).
Admati et al argue the assumption that more lending is always better for economic growth and welfare is absurd.
Allen et al argue that we can have our cake (financial stability) and eat it (higher economic activity) too.
6 --Hanson et al argue that higher capital requirements would have the potential to reduce competition that creates negative externalities and systemic risk.
Jorda et al find that higher capital requirements can lower the costs of a financial crisis even if it can't prevent it Cf. BCBS finds that averaging across a number of studies of financial crises in many countries,
increased capital and liquidity leads to a substantial reduction in the probability of a crisis event.
Jorda et al argue that the role of capital regulation is in mitigating the social and economic costs of a crisis a distinct but arguably more important benefit.
Haldane (using recent bank level data) found that pre-crisis capital ratios performed no better than a coin toss in predicting which institutions would be distressed during the GFC.
Even if the reduced chance of a crisis is only modest the estimated benefits of more rigorous capital requirements can be a large number Vickers Report suggests that CET1 figures between 16-24% RWA
would have enabled 95% of banks to survive recent crises.
Only at very high equity levels (which might interfere with the bank's transformation function) will equity provide something approaching a guarantee of survival (PFR).
THUS need other mechanisms alongside to promote safety and soundness of banks.
ALSO - the trade-off is actually more evident in the Basel III liquidity requirements.
Basel III requirements
Admati et al
Miles et al
Firestone et al
Allen et al
BASEL III CAPITAL REQUIREMENTS vs COMMENTARY
4.5% CET1 against RWAs (13% if incl. all buffers)
20-30% of unweighted assets (LR) and over 50% against RWAs 20% of RWAs - find that equity is slightly more expensive than debt - BUT benefits of substantially higher capital requirements are likely to be great.
- Miles et al find that the benefit of greater equity capital in reducing the chances of a banking crisis tends to decline with additional capital (given the assumed distribution of shocks to bank asset values).
BUT - since it looks like there are very occasionally extremely negative shocks to asset values the benefit of extra capital doesn't fall monotonically.
The costs of having banks finance more of their assets with equity is linear .
13-26% (even though they argue that equity is more expensive than debt - still argue for increase).
NB: Firestone et al study incorporates LCR and NSFR (noting that they have increased the resilience of financial firms since GFC) and still recommend higher than Basel III.
Agree with Admati et al and Miles et al that the economic impact of the changes in capital requirements should be much less than many in the industry fear and suggest that higher capital requirements could actually (for x-efficiency reasons) lower the cost of borrowing to some borrowers.
Suggested that CET1 figures in the range of 16-24% of RWA would have enabled 95%
of banks to survive recent crises.
Proposed TLAC of:
- 17% for retail banks; and
- 20% for investment banks.
Increasing Capital with Contingent CapitalContingent capital as a substitute for equity is highly controversial.
Buy the full version of these notes or essay plans and more in our Principles of Financial Regulation Notes.