International Variation in Bank Accounting Information Content
International Variation in Bank Accounting Information Content
International Variation in Bank Accounting Information Content
Abstract
This study explores the cross-country impact of nancial system and banking regulations
on the information content of bank earnings and book value. Test results provide
empirical evidence that nancial system and banking regulations have a joint eect on
the association of equity price with earnings and book value components in Germany,
France, the United Kingdom and United States. This eect is explainable by the objective
bank function, which shows that earnings of the period determine the terminal book
value, thus consistent with the clean surplus accounting approach. Cross-country variation in bank accounting information content calls for caution in interpreting international
bank nancial and operating ratios.
1. Introduction
The accelerating globalization of the world economy and integration of
capital markets has led to a proliferation of research on the value
relevance of accounting information across countries, with a focus on
international diversity of earnings quality (e.g., Pope and Walker, 1999;
Ali and Hwang, 2000; Ball et al., 2000) and earnings management (e.g.,
Brown and Higgins, 2001; Bhattacharya et al., 2003; Leuz et al., 2003).
However, based exclusively on industrial/commercial rms, the extant
literature is not generalizable to nancial institutions including commercial banks, which dier signicantly from industrial/commercial rms in
terms of operating characteristics and reporting requirements.
In contrast to the shortage of international accounting studies on
nancial institutions, comparative works on bank eciency and protability abound in economics and nance literature. For example, Berger
and Humphrey (1997) surveyed 130 studies that applied frontier eciency analysis to nancial institutions in 21 countries, and found that
the various eciency methods did not yield consistent results. We argue
that, to an extent, the mixed results are attributable to cross-country
The authors want to thank Professor Frederick D.S. Choi (the Editor) and two anonymous
reviewers for their constructive comments and helpful suggestions.
r 2008 Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
237
variation in bank accounting information content from which international bank nancial and operating ratios are derived. Hence, the need
for a paper to explore the international variation in bank accounting
information and ratio analysis.
Banking is a regulated industry owing to the important intermediary
role banks play in national nancial systems. Each country has its own
regulations for bank capitalization, risk exposure and information
disclosure to address the three fundamental sources of systematic risks
for bank operations: interest risk, credit risk and liquidity risk (Stigum
and Branch, 1983). However, despite their national diversity, commercial
banks have a relatively simple and homogenous production function and
capital structure because they specialize in transforming deposits into
loans and investments, as compared with industrial/commercial rms
engaged in heterogeneous business activities. This production function
provides a feasible basis for identifying the impact of various institutional factors on bank accounting information content.
Our study investigates the variation in bank accounting information
content for France, Germany, the United Kingdom and United States.
All the four countries are members of both the OECD and G-10 group,
characterized by a compatible level in economic development and
nancial market sophistication.1 Nevertheless, signicant dierences
exist among their nancial market structures and banking regulatory
regime. The United Kingdom and United States (France and Germany)
are examples of shareholder (creditor)-oriented nancial systems, with
equity playing a signicantly more (less) important role than bank loan
as a funding source for rms. Whether a country has a shareholder- or
creditor-oriented nancial system is important because each type of
systems has dissimilar implications for dealing with potential asymmetric
information problems that arise between those providing the funds and
those receiving them. Furthermore, commercial banks have been permitted a varying range of allowable activities in the four countries with
France (the United States) as the most deregulated (regulated) one in
terms of engaging in securities, insurance and real estate activities. The
motivation of the paper is to examine how the dierent sets of institutional factors aect accounting information content across the four
countries, thereby leading to varying value relevance of international
bank accounting information and nancial/operating ratios.
The next section reviews the balance sheet and income statement of
commercial banks and develops testable hypotheses. Section 3 describes
the test sample and reports empirical results. Section 4 concludes the paper.
r 2008 Blackwell Publishing Ltd.
238
2. Hypotheses development
Liabilities
Cash
Deposits
Borrowed Funds
Securities
Others
Others
Stockholders Equity
239
240
adopted the Basle Accord in 1988 and Basel Accord II in 2005. The Basle
Accords specify denitions of Tier 1 (core) capital (stockholders equity),
Tier 2 capital (additional components of capital), and a general framework for assigning assets and o-balance sheet items to broad risk
categories to enable the calculation of risk-based capital ratio. A
minimum ratio of total capital (Tier 1 plus Tier 2) to risk-weighted
assets of 8 percent (of which at least 4 percent should be in the form of
core capital) was specied in 1992. Although the Basle Accords are
viewed by many as providing uniform minimum capital requirements, it
does not aect banks in each country equally because there are
dierences in the items that may be included as components of capital
for purposes to fulll the requirements. Table 1, panel A lists the
components of regulatory capital for meeting the capital requirement
in the four countries.
Failure to meet capital adequacy would increase banks cost of
capital, thus resulting in lower equity value. Bank managers can
increase regulatory capital through the combination of its components
(Moyer, 1990). Thus, dierences in the composition of regulatory
capital constitute a major source of cross-country variation in the
information content of capital adequacy for commercial banks. The rst
hypothesis is:
H1 : The association of market valuation with asset risk for capital
adequacy of commercial banks dier signicantly between countries.
The following model can be used to test Hypothesis 1:
MVit a b1 CCit b2 RCit b3 Di b4 CCit Di b5 RCit Di
eit
1
where MVit 5 market value of bank is common equity at the end of year
t, CCit 5 contributed capital of bank i at the end of year t, RCit 5
reserved capital of bank i at the end of year t, Di 5 system/country
dummy for bank i, CCit Di 5 interaction of CCit and Di,
RCit Di 5 interaction of RCit and Di, eit 5 disturbance term.
Contributed capital (CC) consists of capital stock and additional paidin capital, which represent the legal capital requirement for an incorporated commercial bank and cannot be distributed to shareholders except
under liquidation. Reserved capital (RC) includes retained earnings,
other distributable and non-distributable reserves (e.g., unrealized securities gains/losses, xed asset revaluation reserve), which reect the
r 2008 Blackwell Publishing Ltd.
Securities
Insurance
Real estate
Investment in nonnancial rms
Nonnancial rm investment in commercial banks
France
No
Yes
No
No
No
Yes, but not prevalent
Yes
Yes, but not issued
Yes
No
Yes
Yes, but limits
France
Yes
No
No
No
Yes,
Yes,
Yes,
No
No
Yes,
Yes,
No
Unrestricted
Restricted
Permitted
Unrestricted
Unrestricted
Germany
but limits
but limits
but limits
but limits
but limits
Germany
Table 1.
Unrestricted
Permitted
Unrestricted
Unrestricted
Unrestricted
UK
Yes
Yes
No
No
N/A
Yes, but limits
Yes, but limits
Yes
Yes, with caution
N/A
Yes, but limits
No
UK
Yes
Yes
No
No
No
Yes,
Yes,
Yes,
No
No
Yes,
No
Restricted
Restricted
Restricted
Restricted
Restricted
US
but limits
but limits
but limits
but limits
US
242
243
where MVit 5 market value of bank is common equity at the end of year
t, NIIit 5 net interest income for bank i in year t, NNIIit 5 net noninterest
income for bank i in year t, ONIit 5 other net income for bank i in year t,
The dummy and interaction terms are as explained in Equation (1).
NII is the dierence between interest income and interest expense,
while NNII is the dierence between non-interest income and noninterest expense. The variable ONI consists of other nonrecurring income
items such as foreign exchange income, exceptional items and other
provisions/reserves. NII and NNII are expected to have positive signs in
relation to market value because they provide primary recurring sources
of earnings for banks functioning as nancial intermediaries. On the
other hand, ONI, which includes various types of provisions and other
r 2008 Blackwell Publishing Ltd.
244
Maximize
Subject to
PtT Si gi Ai Sj dj Lj CAi Lj =WtT1
PtT =WtT1
r 2008 Blackwell Publishing Ltd.
245
246
The denitions of the variables are the same as in Equations (1) and (2).
b1 and b2 are multiplicative parameters that relate book value to
market value. If book value is equivalent to capitalized normal earnings,
then rm value is equivalent to the sum of capitalized normal earnings
and capitalized abnormal earnings, or simply capitalized earnings. b3, b4
and 5 are earnings capitalization factors that take into account crosscountry dierences in risk, expected growth and other factors. In the
valuation framework of the clean surplus model, book value is assumed
to provide information on normal earnings, and abnormal earnings is
captured in the linear information dynamics between accounting earnings and book value. More stringent banking regulations are expected to
have a positive impact on the linear information dynamic for commercial
banks because, holding everything else constant, current periods income
determines terminal wealth.
Recall that signicant dierence exists between France and Germany
(Germany, France and the United Kingdom) versus the United Kingdom and United States (the United States) in terms of capital requirement (earnings components) in H1 (H2). We expect signicant dierence
in the linear information dynamic, rst between the shareholder- and
creditor-oriented systems, and then between the United Kingdom in
addition to France and Germany versus the United States, because the
variation in earnings quality would aect that of the capital components.
3. Empirical Tests
247
France
Germany
UK
US
12.0083
11.5378
14.585nnn
39.5418
37.5563
2.680nnn
30.9732
30.2702
6.0511
5.0247
4.647nnn
5.2445
4.1929
2.662nnn
4.2014
5.8254
3.0345
3.9305
12.398nnn
11.6244
12.1004
3.672nnn
7.8950
8.3475
2.4770
2.2628
10.893nnn
6.8290
5.5222
4.416nnn
4.7667
4.1556
2.9579
2.7321
4.518nnn
1.9143
2.0520
0.1953
0.6745
4.938nnn
0.0834
0.4468
1.5458
2.1514
2.255nn
0.1370
0.5119
1.399
nn
and nnnDenote statistical signicance at level of 5% and 1%, respectively.
Note
All variables are scaled by square root of assets.
t-value is calculated for France, Germany and UK, respectively, as compared with US.
248
1.491
1.575
2.546
6.490
0.179
1.249
Signicance 5 .000
SE
t-statistic
1.283
0.076
0.048
1.064
0.153
0.156
1.162
20.633
53.256
6.099
1.170
7.998
Adjusted R2 5 0.789
Signicance
.245
.000
.000
.000
.242
.000
1.152
1.582
2.628
7.845
3.465
0.229
0.377
0.339
0.100
1.137
1.236
0.223
Signicance 5 .000
SE
t-statistic
1.302
0.079
0.056
1.325
1.533
1.764
0.222
0.210
0.308
0.204
0.260
0.117
0.885
20.145
47.135
5.919
2.261
0.130
1.697
1.612
0.323
5.585
4.750
1.903
Adjusted R2 5 0.792
Signicance
.376
.000
.000
.000
.024
.897
.090
.107
.746
.000
.000
.057
249
250
4.362
9.100
8.400
1.732
3.560
3.074
5.008
1.385
Signicance 5 .000
SE
t-statistic
1.145
0.138
0.276
0.747
0.950
0.393
0.482
1.216
3.811
65.900
30.436
2.307
3.747
7.824
10.393
1.139
Adjusted R2 5 0.827
Signicance
.000
.000
.000
.021
.000
.000
.000
.255
4.764
9.416
9.551
0.137
5.721
2.208
4.922
3.320
3.336
2.018
4.928
6.425
5.890
3.535
2.255
2.205
Signicance 5 .000
SE
t-statistic
1.132
0.149
0.297
0.920
1.144
1.296
1.529
0.495
0.568
0.343
0.614
0.651
0.716
1.483
1.829
1.754
4.209
63.044
32.168
0.149
5.002
1.705
3.220
6.712
5.871
5.879
8.031
9.868
8.230
2.384
1.233
1.257
Adjusted R2 5 0.838
Signicance
.000
.000
.000
.882
.000
.088
.001
.000
.000
.000
.000
.000
.000
.017
.218
.209
Dependent variable: market value of equity; NII 5 net interest income; NNII 5 net noninterest
income; ONI 5 other net income.
D 5 country dummy; D1 5 1 if France; D2 5 1 if Germany; D3 5 1 if the UK.
All variables are scaled by square root of assets.
251
252
dierence in ONI (NII and NNII) between France and the United States
(the United Kingdom and United States) again provide strong evidence
of the eect of banking regulation on earnings quality in addition to
reporting system (H2).
Table 5 provides the regression result of Model 3, which test the linear
information dynamic between capital and earnings compositions (H3).
Both the system and country test results have the highest adjusted R2
among all the three models, indicating that the clean surplus model
captures the linear information dynamic missing in Models 1 and 2.
A comparison of the test results of Model 3 with those of Models 1
and 2 for all the similar items reect that the value relevance of book
value and earnings depends on whether they are assessed separately
(Models 1 and 2) or in a linear information dynamic frame (Model 3).
We rst compare Models 2 and 3 for earnings components. NII D,
NNII D and ONI D are identical at both the system and country
levels between Models 2 and 3, except for that ONI D for France is
(not) signicant in Model 2 (3). A comparison of Models 1 and 3 for
capital composition reveals that CC D is insignicantly negative at the
system level in Model 1, however, it becomes signicantly positive in
Model 3. This change occurs because in Model 1, the value relevance of
book value is estimated without reference to earnings as a source of
capital; but in Model 3, it also reects the linear information dynamic
between book value and earnings. At the country level, the most
noticeable distinction is that while RC D is signicantly dierent for
France, Germany and the United Kingdom in Model 1, it becomes
insignicant in Model 3. Such changes arise because the capital and
earnings components are segregated in Model 3, but in Model 1 the
earnings is part of capital components as being contained into RC. By
combining the three models together, we can discern the value relevance
of book value (earnings) in a clearer perspective by holding earnings
(book value) constant. In the context of commercial banks, the test
results of Model 3 can also be interpreted as the joint eects of shareholder- versus creditor-oriented system (which aects book value) and
banking regulations (which aect earnings components) on the properties of accounting information contents.4
In summary, a comparison of all the three models sheds light on the
joint eect of shareholder- versus creditor-oriented system and banking
regulations on the accounting information content of commercial banks.
Models 1 and 2 demonstrate that book value and earnings components
have higher value relevance in shareholder-oriented countries where
r 2008 Blackwell Publishing Ltd.
253
1.957
0.800
1.020
6.571
7.461
2.364
5.196
0.375
0.989
2.405
4.070
0.970
Signicance 5 .000
SE
t-statistic
1.101
0.093
0.093
0.255
0.278
0.722
0.920
0.180
0.093
0.494
0.467
1.159
1.777
8.585
10.996
25.724
26.848
3.275
5.649
2.073
10.583
4.868
8.717
.836
Adjusted R2 5 0.845
Signicance
.076
.000
.000
.000
.000
.001
.000
.038
.000
.000
.000
.403
3.888
0.782
0.936
6.632
7.946
1.280
7.330
3.723
3.746
0.264
0.286
0.606
0.359
0.018
0.034
3.298
4.350
1.534
4.563
5.971
5.234
1.351
0.684
0.219
Signicance 5 .000
SE
t-statistic
0.591
0.105
0.117
0.359
0.353
0.913
1.164
1.310
1.556
0.227
0.318
0.330
0.311
0.292
0.203
0.995
0.999
0.622
0.815
0.717
0.740
1.541
1.788
1.731
2.052
7.473
7.992
18.459
22.504
1.401
6.298
2.843
2.407
1.163
.902
1.837
1.154
.063
.167
3.316
4.354
2.466
5.598
8.329
7.070
.877
.383
.126
Adjusted R2 5 0.852
Signicance
.040
.000
.000
.000
.000
.161
.000
.005
.016
.245
.367
.066
.249
.950
.868
.001
.000
.014
.000
.000
.000
.381
.702
.899
254
banking regulations are also stricter. The test results of Model 3 show
that bank accounting information in shareholder-oriented countries are
more sensitive to linear information dynamic between book value and
earnings components than in creditor-oriented countries. For example,
both the book value and earnings components are more positive for UK
than for French and German banks in Models 1 and 2; however, the
negative implication of earnings components for book value is only
discernible for the United Kingdom in Model 3.
France
Germany
UK
US
0.0919
2.2976
0.0091
0.0273
0.0379
0.2785
0.0527
0.8385
0.0005
0.0147
0.0302
0.1553
0.0890
1.8165
0.0143
0.0514
0.0491
0.2941
0.0854
0.8796
0.0127
0.0412
0.0341
0.2388
0.032
0.033
0.034
0.036
0.003
0.003
0.005
0.003
0.062
0.070
0.199
0.066
0.040
0.042
0.042
0.046
0.005
0.008
0.009
0.011
0.297
0.329
0.109
0.001
0.325
0.256
0.190
0.114
0.090
0.353
0.003
0.155
SE
0.440
0.166
0.457
0.044
0.357
0.318
0.121
0.053
Coecient
3.47
5.53
2.08
0.017
3.04
3.49
2.40
2.10
2.33
6.91
0.075
2.99
3.97
1.71
6.34
0.68
0.49
0.48
1.59
0.15
t-stat
.001
.000
.037
.986
.002
.000
.016
.036
.020
.000
.940
.003
.000
.086
.000
.496
.621
.626
.111
.875
Signicance
.120
.296
.261
.051
.303
0.763
0.091
0.164
0.017
0.883
0.794
0.195
0.581
1.623
0.299
1.14
0.095
0.514
0.393
0.035
0.039
0.812
0.203
0.227
0.032
0.227
0.444
0.199
0.071
Coecient
Dependent variables: net income and market value, scaled by square root of assets, respectively.
NIEA: noninterest expense to assets.
NIITI: noninterest income to total income.
D 5 Country dummy; D1 5 1 if France; D2 5 1 if Germany; D3 5 1 if the UK.
Capital ratio
D1
D2
D3
Loan-to-deposit
D1
D2
D3
Return on assets
D1
D2
D3
Interest Margin
D1
D2
D3
NIEA
D1
D2
D3
NIITI
D1
D2
D3
Variable
Net Income
0.231
0.240
0.244
0.260
0.032
0.032
0.049
0.032
0.915
0.957
0.937
1.100
0.477
0.536
1.52
0.505
0.221
0.233
0.230
0.254
0.043
0.068
0.076
0.094
SE
9.29
1.27
3.06
0.36
1.26
1.26
2.65
1.78
17.38
5.38
15.75
2.65
7.65
8.43
0.85
0.60
13.53
4.93
5.13
1.05
6.70
9.94
6.22
1.56
t-stat
.000
.202
.002
.716
.205
.207
.008
.075
.000
.000
.000
.008
.000
.000
.392
.545
.000
.000
.000
.291
.000
.000
.000
.117
Signicance
Market Value
.327
.769
.547
.626
.117
.616
R2
256
Constant
D1
D2
D3
Capital ratio
D1
D2
D3
Loan-to-deposit
D1
D2
D3
Return on assets
D1
D2
D3
Interest margin
D1
D2
D3
NIEA
D1
D2
D3
NIITI
D1
D2
Variable
SE
0.004
0.004
0.006
0.005
0.029
0.033
0.035
0.035
0.002
0.002
0.004
0.002
0.072
0.087
0.172
0.105
0.067
0.072
0.070
0.111
0.007
0.009
0.012
Coecient
0.007
0.017
0.141
0.025
0.289
0.026
1.27
0.033
0.463
0.393
0.215
0.203
0.288
0.040
0.045
0.068
0.175
0.276
0.848
0.050
0.130
0.108
0.374
2.91
0.542
0.995
0.517
0.974
2.21
6.39
0.385
2.37
1.81
7.75
1.80
2.31
1.49
0.372
2.90
0.270
17.07
0.523
0.873
0.821
3.44
0.817
t-stat
Net Income
.004
.588
.320
.605
.330
.027
.000
.701
.018
.070
.000
.072
.818
.134
.710
.004
.787
.000
.601
.383
.412
.001
.414
Signicance
0.058
0.200
0.238
0.041
0.103
0.075
0.026
0.028
1.614
1.484
0.169
0.749
1.476
0.662
1.041
0.322
0.097
0.068
0.139
0.387
0.441
0.305
0.008
0.196
0.029
0.019
0.050
Coecient
0.020
0.022
0.031
0.026
0.159
0.182
0.204
0.201
0.013
0.013
0.022
0.013
0.733
0.795
0.749
1.20
0.400
0.477
0.899
0.582
0.351
0.376
0.371
0.577
0.039
0.049
0.064
SE
2.92
4.88
4.71
1.11
1.81
1.38
0.58
0.77
5.73
5.83
5.08
5.66
19.72
14.37
17.94
8.20
1.71
1.63
5.78
5.22
4.61
4.59
0.11
2.84
0.94
0.57
1.86
t-stat
Market Value
.003
.000
.000
.264
.070
.166
.562
.442
.000
.000
.000
.000
.000
.000
.000
.000
.086
.102
.000
.000
.000
.000
.910
.004
.345
.564
.063
Signicnce
0.022
74.46
0.000
0.525
Coecient
0.015
SE
0.302
t-stat
.763
Signicance
0.026
374.13
0.000
0.868
Coecient
Dependent variables: net income and market value, scaled by square root of assets, respectively.
NIEA: noninterest expense to assets.
NIITA: noninterest income to total income.
D 5 country dummy; D1 5 1 if France; D2 5 1 if Germany; D3 5 1 if the UK.
D3
F Value
Signicance
Adjusted R2
Variable
Net Income
Table 8. (Continued.)
0.081
SE
0.67
t-stat
Market Value
.502
Signicnce
258
Ronald Zhao and Yihong He
259
4. Conclusion
Previous literature has studied the value relevance of accounting earnings
and book value in relation to managerial discretion based on industrial/
commercial rms. This study investigates how nancial system and
banking regulations aect international bank accounting information.
Test results provide empirical evidence that the joint eects of nancial
system and banking regulations account for, to a signicant degree,
international variation in the information content of bank earnings and
book value across France, Germany, the United Kingdom and United
States. Capital requirements have an impact on the association between
book value components and market valuation. Regulations on bank
operating activities are found to have a signicant eect on the association of equity price with earnings components. The joint eect of
nancial system and banking regulations on the linear information
dynamic for commercial banks is explainable by the objective bank
function model, which shows that earnings of the period determines the
terminal book value, thus consistent with the clean surplus approach. An
understanding of the impact of cross-country institutional dierence on
the information content of accounting variables also increases the value
relevance of analytical ratios in comparing bank performance in an
international context.
Notes
1. The central bank governors of the Group of 10 countries established the Basle
Committee on Banking Supervision.
2. The acronym CAMEL stands for capital adequacy, asset quality, management
competence, earnings stability and liquidity.
3. UK banks have more items in ONI. However, as there are more US than UK banks
in the sample, the test results may be weighted toward US banks.
r 2008 Blackwell Publishing Ltd.
260
4. The dierence between book value (CC1RC) in Models 1 and 3 can also be
statistically tested. It is (not) statistically dierent from 1 in Model 1 (3) in accordance to
the theoretical constraints of the clean surplus approach.
References
Ali, A. and L. Hwang, Country-Specic Factors Related to Financial Reporting and the
Value Relevance of Accounting Data, Journal of Accounting Research 38 (2000), pp.
121.
Ball, R., S. Kothari and A. Robin, The Eect of International Institutional Factors on
Properties of Accounting Earnings, Journal of Accounting and Economics 29 (2000),
pp. 152.
Barth, J.R., D.E. Nolle and T.N. Rice Commercial Banking Structure, Regulation, and
Performance: An International Comparison, Economic Working Paper 97-6, Oce of
the Comptroller of the Currency, 1997.
Berger, A.N. and D.B. Humphrey, Eciency of Financial Institutions: international
Survey and Directions for Future Research, European Journal of Operational Research
98 (1997), pp. 175212.
Bhattacharya, U., H. Daouk and M. Welker, The World Price of Earnings Opacity, The
Accounting Review 78 (2003), pp. 641679.
Brown, L.D. and H.N. Higgins, Managing Earnings Surprises in the US versus 12 Other
Countries, Journal of Accounting and Public Policy 20 (2001), pp. 373398.
Edwards, F. and F. Mishkin The Decline of Traditional Banking: Implications for
Financial Stability and Regulatory Policy, Economic Policy Review, Federal Reserve
Bank of New York, July, 1995, pp. 2745.
Feltham, G.A. and J.A. Ohlson, Valuation and Clean Surplus Accounting for Operating
and Financial Activities, Contemporary Accounting Research 11 (1995), pp. 689731.
Johnson, S.A., The Eect of Bank Reputation on the Value of Bank Loan Agreements,
Journal of Accounting Auditing and Finance 12 (1997), pp. 83100.
Leuz, C., D. Nanda and P.D. Wysocki, Investor Protection and Earnings Management:
an International Comparison, Journal of Financial Economics 69 (2003), pp. 505527.
Moyer, S., Capital Adequacy Ratio Regulations and Accounting Choices in Commercial
Banks, Journal of Accounting and Economics 13 (1990), pp. 123154.
Ohlson, J.A., Earnings, Book Value, and Dividends in Security Valuation, Contemporary Accounting Research 11 (1995), pp. 661687.
Pope, P. and M. Walker, International Dierences in the Timeliness, Conservatism, and
Classication of Earnings, Journal of Accounting Research 37 (1999), pp. 5387.
Rose, J.T. and J.D. Walken Statistical Cost Accounting Models in Banking: A
Reexamination and an Application, Sta Study 150, Board of Governors of the
Federal Reserve System, 1986.
Santomero, A.M., Modeling the Banking Firm, Journal of Money, Credit and Banking 4
(1984), pp. 576602.
Stigum, M. and R. Branch Jr., Managing Bank Assets and Liabilities: Strategies for Risk
Control and Prot (Homewood, IL: Dow-Jones Irwin, 1983).
Trueman, B., S. Titman and P. Newman, An Explanation for Accounting Income
Smoothing: discussion, Journal of Accounting Research 26 (1988), pp. 127143.
Wahlen, J.M., The Nature of Information in Commercial Bank Loan Loss Disclosures,
The Accounting Review 69 (1994), pp. 455478.
Zhao, R. and Y. He, The Impact of SFAS No. 114 on the Linear Information Dynamic
for Commercial Banks, Review of Quantitative Finance and Accounting 23 (2004), pp.
313328.
r 2008 Blackwell Publishing Ltd.