What do banks do? Measuring traditional and non

Download Report

Transcript What do banks do? Measuring traditional and non

What do banks do?
Measuring traditional and nontraditional bank output
Robert Inklaar
Groningen Growth and Development Centre,
University of Groningen and
The Conference Board
with Christina Wang (FRB Boston)
Why banks?
 Importance for monetary policy
 Not the topic here
 Substantial sector of the economy
 5% of GDP in U.S., 4% in EU
 Important investor in new technology
 7% of private U.S. ICT investment
What do banks do: the short answer
 Resolve asymmetric information
problems of borrowers
 Provide transaction services for
depositors
 Fund management
Main argument
 Existing output measures of bank
output are mostly ad-hoc
 GE model of Wang-Basu-Fernald
(WBF, 2004) provides a coherent
framework
 Output measure presented here
consistent with model
 Choice of measure makes substantial
difference
Outline WBF-model (1)
 Banks provide transaction services and
information services
 Implicit payment, mostly aspart of
interest income & expenses
 Other services paid for through fees &
commissions
 But trading gains/losses should be
excluded
Outline WBF-model (2)
 Risk-adjusted opportunity cost of funds is
key concept:
Y L  r L  r L L  r L  rp L  r F L

 

 Banks provide funds plus services, so only
income in excess of cost of funds is output
 Similar for deposits: depositors accept
below-market interest rates in return for
transaction services => this is an output!


Y D  rF  rD D
Shift towards securitization and
away from deposits-funded lending
Output at current prices of U.S. commercial banks, 1990 and 2004
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1990
Deposits
Loans
Fiduciary activities
2004
Securitization activities
Investment banking & insurance
Other activities
Securitized loans crucial growth
factor
Bank loans: balance sheet and securitized, 1987-2005
(trillions of dollars)
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Balance sheet loans
Securitized loans
Real growth: the case for counting
 Bank efficiency literature uses CPI-deflated
loans and deposits as output indicators

 

Y L  r L  r L L  r L  r L PCPI LQ
 Example: residential mortgages
 Change in housing prices relative to
average prices
 Constant services per dollar or per loan?
 Compare car dealers: counting # of cars sold
is an improvement over using CPI to deflate
sales, even if it’s not perfect
Choice between deflation and counts
is very relevant
Average annual real growth of loan and depositor services, 1987-2004 (%)
6
1987-1995
1995-2004
5
4
3
2
1
0
-1
-2
-3
Deflated loans
Loan counts
Deflated deposits
Transaction counts
Non-traditional activities: what do
banks get paid for?
 Bank efficiency literature
 Credit equivalent of OBS items
 Asset equivalent of OBS items
 Net non-interest income
(CPI-deflated)
 Suggestion:
 Treat like other industries
 Separately deflate different activities
How does separate deflation work?
 Fund management
 Count of the number of trust accounts
 Investment banking & insurance
 Output price of relevant industry
 Securitization
 Services similar to on-balance sheet
lending
 Other activities
 Use misc. bank charges deflator from CPI
Again: the measure matters a lot
Average annual real growth of non-traditional activities, 1987-2004 (%)
25
20
15
10
5
0
Credit equivalent
Asset equivalent
1987-1995
Net non-interest income
1995-2004
Total fee-generating
activities
Conclusions
 Measurement affects the results of
industry-level and bank-level analysis
 Use a theoretical framework to inform
measurement
 Framework can be partly implemented
at bank level (in U.S.)
 Still many improvements possible