(Paper2) Presentation by Lary Wall

“On Deposit Stability in Failing Banks”
Christopher Martin, Manju Puri and
Alexander Ufier
Comments by Larry D. Wall
Federal Reserve Bank of Atlanta
Disclaimer
The views expressed here is the
discussant’s and not necessarily those
of the Federal Reserve Bank of Atlanta,
or the Federal Reserve System.
2
Overview of the paper
• Looks at the failure of a $2 billion bank
• Failure primarily due to credit losses on exotic residential
mortgage products (including adjustable rate mortgages?)
• Data available on
•
•
•
•
End-of-day, deposit account level balances
Age of deposit
Whether deposit was held by an institution
Proportion of days where account had deposit or
withdrawal
• Time line of events including supervisory intervention
3
Overview of the paper
• Empirical results
1.
2.
3.
4.
5.
Deposit insurance proved credible
Uninsured transactions accounts were more stable
Uninsured term deposits more likely to run
Older accounts less likely to run
Deposit runoff by account type did not exceed LCR
assumptions but did sometimes exceed NSFR
assumptions
6. Bank could replace runoff with new institutional term
deposits that paid a 65 basis point premium
4
Comment: Paper is overly focused on
deposits
• Paper does not discuss why depositors run
• Depositors care about PD and LGD of their bank
• PD depends upon the regulators
•
•
Solvency and liquidity of the bank
Other factors, including available resources
• LGD depends on
•
•
•
Value of bank’s capital and assets
Whether the claim is insured
Other relationships with the bank
5
Comment: Paper should discuss nondeposit factors
• The value of the bank’s assets and capital
•
•
Public information
Private information maybe available to some depositors
• Nondeposit sources of funds
•
•
FHLB funding both as substitute deposits and
increasing deposit LGD by taking collateral
Capital Purchase Program (TARP)
6
Comment: Term versus transaction
deposit run-off
• Standard view is that uninsured transaction deposit
are more subject to deposit run-off.
• My understanding of the reasoning behind this view
• One day a previously healthy bank is hit with adverse
shock and becomes insolvent
• Transactions deposits run immediately because they can
• Term deposits run as they mature because they must
wait
• However, the paper finds term deposits are more
likely to run
7
Comment: Term versus transaction
deposit run-off
• What happened in this case (and most U.S. failures)
• Solvent bank is hit with a series of shocks that ultimately
cause it to become insolvent
• Forbearance given so reported losses < economic losses
• Transactions depositors can run any time and have larger
switching costs so they can afford to wait
• Term deposits can only run at maturity and have low
switching costs so are more likely to run
• Implications
• LCR and NSFR based on particular scenarios
• But which deposits will run depends upon the scenario.
8
Comment: Market discipline versus
“Improving funding stability”
• Paper concludes that banks can undo disciplining
effect of market by raising new insured deposits
• But also says “we find that deposit insurance is
effective in improving banks’ funding stability.”
• These are two ways of describing the same
phenomenon
• Deposit insurance helps banks retain and attract funding
9
Comment: Market discipline versus
“Improving funding stability”
• Market did provide some discipline
• Bank paid 2.8 percentage point premium before the crisis
• Premiums (and deposit competition) reduced through time
Likely average maturity of new/roll-over term deposits also
decreased
• But authorities acted to mute that discipline
• Congress increased DI from $100,000 to $250,000
• FDIC provided optional coverage for transactions accounts
• If authorities are going to suppress market discipline
then disciplining banks is up to supervisors
10
Comment: Implications for the future
1. This is not a “TBTF” bank. Depositor behavior
may well be different at systemically important
banks
2. The implementation of current expected credit loss
(CECL) for recognition of losses (IFRS 9 and
changes to US GAAP) will likely reduce banks’
ability to understate credit losses
11
Comment: Speculation about the
future
• The timing of depositor withdrawals will depend in
part on the costs of switching to another bank
• Development of FinTech may change switching
costs for transactions accounts
• Some changes may lower the cost of moving
accounts
• EU PSD2 mandates information sharing
• Some changes may raise cost of moving accounts
• Some US banks reaching separate agreement
with individual FinTech firms
12
Further discussion
On Deposit Stability in Failing Banks
By
Christopher Martin, Manju Puri and Alexander Ufier
13