Bank Capital
How much is “Enough?”
Anat R. Admati
Stanford University
Federal Reserve Bank of Chicago
48th Annual Conference on Bank Structure and Competition
May 10, 2012
Bank Capital is Loss-Absorbing Funding
Loss Absorbing
Equity
Loans to
productive d ti
enterprises Mortgages
Mortgages and other consumer loans
Other Debt
Trading
Assets
reserves
Deposits
Assets
Funding
Debt is a “Hard
Hard Claim”
Claim
Promised
Payment
Replacing Debt with Equity Reduces Leverage
Loss Absorbing
Equity
Loans to
productive d ti
enterprises Mortgages
Mortgages and other consumer loans
Other Debt
Trading
Assets
reserves
Deposits
Assets
Funding
Banks Don’t
Don t “Set
Set Aside”
Aside their own Equity
• Confusing jargon!
• “Hold” or “set aside” is misleading.
• Equity (“capital”) is not the same as reserves.
• Capital requirements concern funding only.
– No constraints on loans and in
investments.
estments
– A firm does not “hold” securities it issues.
• Confusion implies false tradeoffs with lending.
• “Hold capital” = borrow less, use more equity.
% Debt Financing by Industry D/(E+D)
% Debt Financing by Industry D/(E+D)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
Co
ons. Fin. Svcs.
Investment Services
Auto & Truck Manufacturerss
Real Estatee Operations
Electric Utilitiess
Airline
A
Inssurance (Life)
Insuraance (Prop. & C
Casualty)
Naturaal Gas Utilities
Hotels & Motels
Forestry &
& Wood Produ
ucts
Conglomeerates
Paper & Paper Products
Railroads
Casinos & G
Gaming
Auto & Trucck Parts
Average (Alll U.S. Firms)
Construction
n Services
Retail (Grocery)
Communicatio
ons Services
Motion Picturees
M
Fo
ood Processingg
Healthcare Facilities
Broadcasting & C
Cable TV
Advertising
Beverrages (Alcoholiic)
Retail (Department & Discount)
Tobacco
Beveragges (Nonalcoho
olic)
Printing & Publishing
Restauran
nts
Computerr Services
Oil & Gas ‐‐ Integrated
Retail (Speccialty)
Major Drugs
Computer Hardware
Medical Equip
pment & Supplies
Biotechnologyy & Drugs
Computer Nettworks
Retail (Appareel)
SSemiconducto
ors
Communicatio
C
ons Equipmentt
Computer Periipherals
C
Computer Storaage Devices
C
Sofftware & Programming
L
Leverage
b
by IIndustry
d t
0%
6
History of Banking Leverage in US and UK
History of Banking Leverage in US and UK Source: US: Berger, A, Herring, R and Szegö, G (1995). UK: Sheppard, D.K (1971), BBA, published accounts and Bank of England calculations.
7
Trends: Total Assets Grew, RWA Not Much Trends:
Total Assets Grew, RWA Not Much
More Trading, Fewer Loans and Deposits
International Monetary Fund Global Financial Stability Report, April 2008
8
5 Arguments
g
Why
y Banks Should have
Much More Equity
1. Reduces likelihood of distress or failure.
2 P
2.
Protects
t t the
th economy from
f
spillover
ill
effects
ff t
of distress or failure of banks.
3. Reduces Too-Big-To-Fail subsidies and
huge distortions they generate
generate.
4 Does not restrict any banking activity
4.
activity.
5 Does not increase banks’
5.
banks funding costs,
costs
except through reduction of subsidies.
Additional Observations
• More equity prevents excessive risk taking.
• More
M
equity
it reduces
d
lik
likelihood
lih d off credit
dit crunch.
h
• Risk weight system is very problematic
problematic.
• “Level p
playing
y g field” argument
g
is invalid.
• Leverage is “addictive” to a borrower.
• The best source of equity: retained earnings.
Greenspan on More Equity
• “Had the share of financial assets funded by equity been
significantly higher in September 2008, it seems unlikely that
the deflation of asset prices would have fostered a default
contagion much, if any, beyond that of the dotcom boom.”
“The Crisis,” Brookings paper, April 15, 2010.
• “.. if capital and collateral are adequate...losses will be
restricted to equity shareholders who seek abnormal returns;
Taxpayers will not be at risk
risk. Financial institutions will no
longer be capable of privatizing profit and socializing losses.”
Quoted in “Greenspan
Greenspan Defends Legacy
Legacy, Urges
Higher Capital, Collateral Standards,” WSJ, April 7, 2010.
1. Equity
q y Absorbs Losses
Solvent?
A loss
Equity
Equity
Asset
Value
Debt
Promises
Too Much
Too
Much
Leverage
Asset
Value
Debt
Promises
More Equity
1. Equity
q y Absorbs Losses
Equity
Asset
Value
Debt
Promises
Too Much
Too
Much
Leverage
Asset
Value
Debt
Promises
More Equity
1. & 2. Equity
q y Reduces Likelihood of
Distress and Systemic Risk
• The insolvency and bankruptcy of Lehman Brothers led to
– Enormous ripple effects, financial system meltdown,
guarantees, bailouts, Fed windows, TARP.
– “Out of … 13 of the most important financial institutions
in the US, 12 were at risk of failure within a period of a
week or two.” (Bernanke to FCIC)
– “Everyone
Everyone got hurt
hurt. The entire economy has suffered
from the fall of Lehman Brothers… the whole world.”
(Anton Valukas, Lehman court-appointed investigator
to “60 minutes.”)
i
”)
2 Equity Reduces Deleveraging Multiples
2. Equity Reduces Deleveraging Multiples
A 1% Asset Decline with 3% equity
A 1% Asset Decline
3% equity …
 33% Balance Sheet Contraction
33% Balance Sheet Contraction
•
•
•
Equity
Asset Fire Sales
Illiquidity / Market Failure
Uncertainty / Bailouts
1%
Asset Liquidation
Loans & Invest
ments
Assets
Debt
Liabilities
33%
Loans & Invest
ments
Assets
Equity
Debt
Liabilities
15
2. Equity
q y Reduces Fragility
g y
and “Systemic Risk”
• Solvency concerns are key to system fragility.
• More equity attacks all contagion mechanisms.
mechanisms
• C
Contractual
t t l cascades
d
• Information contagion
• Deleveraging spirals
• Liquidity problems are less likely and easier to
solve without solvency concerns
concerns.
3. Equity
q y Reduces the TBTF Problem
• Fear of “Lehman moment” is evident.
• Excessive growth and concentration trends.
– Top 60 global banks groups held $64 trillion in
2010, larger than global GDP; alarming trends.
– Evidence this is related to TBTF.
• Moral hazard, excessive risk and leverage.
• Large distortions in allocation of resources
(including human).
• Excessive political power for large banks.
3 More Equity Reduces Need for Bailouts
3.
Bailout
Equity
Equity
Equity
Equity
Assets
Before
Debt
Assets
After
TToo Much
M h
Leverage
Assets
Before
Assets
After
Debt
More Equity
3 More Equity Reduces Need for Bailouts
3.
Bailout
Equity
Equity
Debt
Assets
After
TToo Much
M h
Leverage
Assets
After
Debt
More Equity
4. More Equity Does Not Restrict
A B
Any
Banking
ki A
Activity
i i
• Three
ee ways
ays to reduce
educe leverage.
e e age
• Same loans in Balance Sheet B.
• Same loans and debt in Balance Sheet C: Add equity!
Initial Balance Sheet
((10% Capital)
p )
Balance Sheets with Reduced Leverage (higher equity to assets)
((20% Capital)
p )
New Assets: 12.5
Equity: 10
Loans & other Deposits & Other Assets: 100
Liabilities: 90
Equity: 22.5
Equity: 20
Equity: 10
Loans & other Deposits & Other Liabilities: 80
Assets: 100
Loans & other Deposits & Other Assets: 100
Liabilities: 90
Loans & other Deposits & Other Assets: 50
Liabilities: 40
A: Asset Sales
A: Asset Sales
B R
B: Recapitalization
it li ti
C A tE
C: Asset Expansion
i
20
5 More Equity Makes ROE Less Risky
5.
• Higher capital
– Lower ROE in good
times
– Higher ROE in bad times
–  Risk to equity reduced
25%
ROE (Earnings Yield)
Initial
10% Capital
20%
15%
10%
Recapitalization
to 20% Capital
5%
• With lower risk,
required return on
equity is lower
0%
3%
4%
5%
‐5%
‐10%
‐15%
R t
Return on Assets
A t
(before interest expenses)
6%
7%
5. More Equityy Lowers
the Required Return on Equity
• In financial markets, “required” return on any
security depends on its risk.
• Borrowing magnifies risk (leverage effect).
• More equity reduces the risk of equity.
• Redistributing risk among investors within
balance sheet does not by itself affect total
f di costs.
funding
t
• Impact of funding mix only through changes in
the total funds available to investors.
ROE Focus is Flawed and Dangerous
• ROE
ROE, unadjusted for risk and leverage
leverage, does not
measure shareholder value.
• Leverage increases risk and thus required ROE.
• Any firm or manager can increase average ROE by
increasing leverage or risk.
• Reaching “target ROE” by increasing risk and
l
leverage
endangers
d
the
h b
bank
k and
d the
h economy.
Is Equity “Expensive?”
Expensive?
• If equity is “expensive” because it has higher
required return than debt, and if ROE
measures shareholder value, then
• Why would Apple use 100% equity? Why not
borrow and create leverage?
– Apple could borrow very cheaply!
– Leverage would increase its ROE!
• Bank stocks trade in same markets as others
others,
are held by same or similar end investors.
5. Leverage Lowers Funding Costs
5
onlyy Because of Debt Subsidies
• Underpriced safety net means
– Borrowing
g costs do not fully
y reflect risk.
– Creditors don’t monitor.
• Additional tax subsidy.
• Loss of subsidies is not a social cost!
Moody’s
Moody
s Announcement: June 2,
2 2011
•
•
SUPPORT FOR BOFA,
BOFA CITI,
CITI AND WELLS FARGO EXCEEDS
PRE-CRISIS LEVELS
Moody's government support assumptions for Bank of
America Citigroup
America,
Citigroup, and Wells Fargo are higher than what
similarly rated institutions would have received prior to the
crisis. For example, Bank of America N.A.'s and Citibank N.A.'s C(C minus) unsupported BFSRs translate to a Baa2 rating on
M d ' llong-term
Moody's
t
debt
d bt scale;
l prior
i tto the
th crisis
i i a similarly
i il l rated,
t d
systemically important bank would typically have benefited from no
more than three notches of uplift, meaning its ratings would be no
higher than A2
A2. Currently,
Currently Bank of America receives five and
Citibank four notches of uplift from government support
assumptions, bringing their senior ratings to Aa3 and A1,
respectively. Wells Fargo's unsupported BFSR of C+ (C plus)
t
translates
l t to
t an A2 rating
ti on M
Moody's
d ' llong-term
t
debt
d bt scale;
l prior
i tto
the crisis a similarly rated, systemically important bank would
typically have received no more than two notches of uplift, to Aa3.
Currently Wells Fargo's
Currently,
Fargo s Aa2 senior rating benefits from three
notches of uplift
Safetyy Net Subsidyy
Lowers Borrowing Costs for Banks
• Rating agencies give uplifts.
• Subsidies are substantial
substantial,
– TBTF subsidies explain “scale effect.”
– Subsidies
S b idi reflected
fl t d iin hi
higher
h ROE
ROE.
Reduced Cost in
Basis Points
2 50
2.50
5.00
7.50
10.00
Extra Return
Extra
Return
on Equity (with 3% equity)
0 81%
0.81%
1.62%
2.43%
3.23%
Debt
(high levels of (hi
hl l f
leverage
create systemic risk
and distort risk
taking incentives) Funding
Financial
Markets
And
Greater Economy
Loans
Equity
(provides
cushion that absorbs
risk and limits incentives
limits incentives
for taking socially inefficient risk) Government Subsidies to Debt:
1. Tax shield (interest paid is a deductible expense but not dividends)
2. Subsidized safety net lowers borrowing costs; bailouts in crisis. Debt
Debt
Funding
Financial
Markets
And
Greater Economy
. Equity
Equity
Higher Stock Price
Happy Banker, Gains are private
Losses are social.
Loans
Lower Loan Costs ?
Additional Benefits to Lower Leverage:
g
Reduces Moral Hazard
• Heavy borrowers may take excessive risk, “heads
I win, tails creditors lose.”
• Guarantees exacerbate the problem.
• More equity shifts downside risk to managers and
shareholders; better incentives to manage risk.
Additional Benefits to Lower Leverage:
Helps Prevent Credit Crunch
• Credit freeze due to too much debt in place.
p
• Debt overhang
g leads to underinvestment.
• Inefficient “deleveraging”
g g can be managed
g to
avoid impact to lending (retain earning!).
• Better capitalized banks make better lending
decisions.
Banks/Bankers Prefer to Borrow and Resist More Equity.
1
3
2
DEBT
1. Subsidies (taxes and safety net)
2. ROE fixation
3 Debt overhang
3.
Debt overhang
EQUITY
For Society, Excessive Bank Leverage is “Expensive!”
2
DEBT
1. Subsidies (taxes and safety net)
2. ROE fixation
3 Debt overhang
3.
Debt overhang
1
3
EQUITY
1. Reduces systemic risk
2. Reduces deadweight cost of distress, default crisis
default, crisis
3. Reduces inefficiencies of high leverage (excessive risk, debt overhang)
“Level
Level Playing Field”
Field Argument is Invalid
• Banks can endanger an entire economy
(Ireland, Iceland).
• Banks compete with other industries for inputs
(i l di ttalent);
(including
l t) subsidies
b idi di
distort
t t markets.
k t
• It is not a national priority that “our” banks are
y impose
p
risk and cost on us.
successful if they
• Argument
g
creates “race to the bottom.”
Basel Capital Requirements
• Tier 1 capital Ratio: Equity to risk
risk-weighted
weighted assets:
– Basel II: 2%
– Basel
B
l III
III: 4.5%
4 5% - 7%,
7% up tto 9.5%
9 5% ffor SIFIs.
SIFI
– Definitions changed.
• Leverage Ratio: Equity to total assets:
– Basel II: NA
– Basel III: 3%.
• Numbers are based on flawed analyses of tradeoffs.
• Risk weights hide risks, are manipulable & distortive.
Balance Sheet Realities
• Contingent and other liabilities (and assets)
live off balance sheet.
– SPVs,
SPV Money
M
Market
M k t Funds,
F d etc
t .
– Ca
Can show
s o up suddenly
sudde y o
on ba
balance
a ce ssheet.
eet
• Loan accounting is highly problematic.
• IFRS vs GAAP: derivatives netting must be
meaningful when it matters,
matters i.e.,
i e in default
default.
• Accounting tricks (Repo 105).
Debt-Like “Capital”
Capital is Ineffective Substitute
• No subordinated debt or hybrid lost in crisis
crisis.
• Equity dominate co
co-cos
cos and bail
bail-in
in debt
debt,
– Straightforward,
Straightforward less complex
complex,
– More reliable to absorb losses.
• Hybrids, bail-in can create instability around
triggers, it matters who holds them.
“Shadow
Shadow Banking”
Banking and
Enforcement Challenge
• Crisis exposed ineffective enforcement.
– Must watch the system.
– Regulated banks sponsor entities in the shadow
banking system.
• Enforcement issues are not a valid argument
g
regulation:
g
Give up
p tax collection?
against
How Much Is “Enough”
Enough Bank Capital?
• Much more than Basel III levels.
• Order of magnitude 20-30% of total assets.
– Benchmark: eliminate TBTF
TBTF, easier than resolution
resolution.
– Significant social benefits; what is the relevant cost?
• Retained earnings easiest source of equity.
• Viable banks can raise equity at appropriate prices
prices.
– “Dilution” only from equity bearing more downside.
– Inability
I bilit tto raise
i equity
it fl
flags iinsolvency.
l
• Risk weights are very problematic
problematic, distort lending
decisions, hide risk, are manipulable.
The BIG Picture
Mutual
Funds
A
Equity
C
Bankingg
Sector
Assets
All the Assets
All
th A t
In the Economy
•
•
•
B
Investors
B
Deposits
And
Other
“Liquid”
Debt
Banking Sector
Mutual
Funds
A
Investors
Equity
C
All the Assets
All
th A t
In the Economy
Banking
Sector
Assets
Deposits
And
Other
“Liquid”
Debt
Banking Sector
All risks are held by final investors. Rearranging claims aligns incentives better.
s s a e e d by a
es o s ea a g g c a s a g s ce
es be e
Key question: Are all productive activities taken? Is risk spread efficiently? A lot of funding in the economy not through banks. Fallacies, Irrelevant Facts, and Myths
in the Discussion of Capital Regulation:
Why Bank Equity is Not Expensive
August, 2010, revised March 2011
Debt Overhang and Capital Regulation
March 2012
Anat R. Admati
Peter M. DeMarzo
M ti F.
Martin
F Hellwig
H ll i
Paul Pfleiderer
http://www.gsb.stanford.edu/news/research/Admati.etal.html