Early Signals
of disclosed accounting fraudsChinese firms’ evidence
(Discussion Draft)
Yi Wei, Jianguo Chen and Jing Chi
Introduction
• China has experienced a rapid change from a
planned economy to a market mechanism.
• 20% of listed firms had committed serious
fraud since the Chinese stock market was
established in the early 1990s (CSRC).
• China is rated at 71 out of 145 countries (Fan,
Rui& Zhao, 2008) in terms of Corruption
Perception Index of Transparency
International.
Management and Controlling shareholders
• Management may manipulate financial report
through tunnelling, insider trading, creative
accounting and false statements (Bai, Yen & Yang,
2008)
• Controlling shareholders in Chinese listed
companies is to maximize the proceeds from
investors, and they have less care of improving
corporate governance and financing subsequent
investment and growth (Shi, 2004).
One form of management frauds
• Called fraud financial statements (FFS)
• It mainly refers to intentional
misstatements of omission in financial
statements (Rezaee, 2010)
Common tricks
in the book of Financial Shenanigans:
• Recording revenue before it is earned;
• Creating fictitious revenue;
• Boosting profits with non-recurring
transactions;
• Shifting current expenses to a later period;
• Failing to record or disclose liabilities;
• Shifting current income to a later period and
• Shifting future expenses to an earlier period
This Research
• The study do not ask the specific types. It is a
general investigation of the chance of
corruption behaviour related to the reported
accounting balances.
• We do find some evidence that there exist a
strong relation between the disposed
corruption firms and the reported account
balance ratios.
Relative literature
• Conflicts of interest between majority and
minority shareholders (Berle &Means, 1932)
• “Tunneling” behaviour which refers to controlling
shareholders expropriating minority shareholder
in many ways when the corporate governance
mechanisms are poor (Johnson, La Porta, Lopezde-Silanes, and Shleifer, 2000).
• Fraudulent financial reporting (FFR) (Erickson,
Hanlon, Maydew, 2004)
Relative literature
Fei (2005) concluded four types of fraudulent
in China:
• non-monetary transactions;
• related party transactions;
• assets restructuring and
• change of accounting estimates.
Recent studies
• Logit regression analysis of 75 fraud and 75 no-fraud
firms have indicated that uncorrupt firms have boards
with significantly higher percentages of outside
members than fraud firms (Beasley, 1996)
• The Classification and Regression Tree (CART) was
employed by Bai, Yen and Yang (2008) to indentify and
predict the impacts of FFS. They included 24 FFS and
124 normal firms
• Fan, Rui and Zhao (2008) report that fraud and rent
seeking have an influence on firms’ behaviour of
leverage, especially obvious relation on long-term debt
• ….
Difference
In our research, we consider every single
account in balance sheet and income
statement to detect the relation with the
corruption.
Sample Sources (Web pages):
• China Security Regulation Committee
(CSRC),
• Shanghai Stock Exchange (SHSE) and
• Shenzhen Stock Exchange (SZSE)
• Law Yearbook of China (1998-2008)
• Others (Disciplinary Cases of the Communist Party of China, China
securities news, Yahoo Finance, Xinhua News, Google China and Baidu)
Our Sample
• We have identified 76 firms involved in
scandals in total.
• Sample period: 1990 through 2002.
Comparable firms
We subcategorize all the corruption firms into 8
industry area:
–
–
–
–
–
–
–
–
Communication and Cultural Industry;
Farming, Forestry, Animal Husbandary and Fishery;
Information Technology;
Manufacturing;
Multiple (some firms not just major in one area);
Real Estate; Transportation and Warehousing;
Utilities;
Wholesale and Retail Trade.
Comparable firms
• Then we collect 76 size matched controlled
firms in the same industry through GTA
database with total assets closest to the fraud
firm.
• We also include the data 5 years before and 5
years after the frauds (restricted by the listed
information and availability of the data).
Model
• We hypothesize that corrupted firms can
potentially manipulate any account.
• So we have tested each account in our model.
• Four types of account values are used: ratio
(to the base of total assets); change rate; ratio
deviation and absolute value of ratio
deviation.
Logistic formula
Y = F( a+b*X)=F(Z)
= 1/(1+exp(-2Z))
Where Y=1 for corrupted firms and 0 for
not-corrupted firms and
X are account variables and b are
coefficients
Every account is tried
• We have tried the regression separately with
Asset accounts, Liability accounts, Equity
accounts and Income Statement items.
• Only Assets and Liability accounts shows strong
consistent relation.
• Only Asset and Liability regressions are reported.
• And the significant accounts (From Assets and
Liabilities) are combined in one single regression
as the final model.
Table 1: Balance Sheet Accounts
Assets
Liabilities
Equities
Account
Descriptions
Account
Descriptions
Account
Descriptions
A1
Cash & equivalents
L1
Accounts payable
E1
Share capital
A2
Short term investments
L2
Notes payable
E2
Retained earnings
A3
Receivables
L3
Short term debt
E3
Capital reserves
A4
Inventories
L4
Other payables
E4
A5
Net prepayments
L5
Other current liabilities
E5
Other stockholder
equity
Total stockholder
equity
A6
Other current assets
L6
Total current liabilities
A7
Total Current Assets
L7
Long term debt
A8
Long term Assets
L8
Other long term liabilities
A9
Goodwill
L9
Deferred tax liabilities charges
A10
Others
L10
Total non-current liabilities
A11
Long term investments
L11
Total liabilities
A12
Total Long term assets
A13
Total Assets
Table 2. Simple statistics comparison of corruption and non-corruption groups (A)
Cash &equivalents
Short term investments
Receivables
Inventories
Net prepayments
Other current assets
Goodwill
Others
Long term investments
Obs. #
Accounts payable
Notes payable
Short term debt
Other payables
Other current liabilities
Other long term liabilities
Obs. #
Non Corrupted
Corrupted
Diff Ratio
Mean
Median
Mean
Median
Mean
Median
0.14
0.12
0.14
0.10
-0.07
-0.15
0.01
0.00
0.01
0.00
-0.14
0.19
0.16
0.24
0.21
0.26
0.23
0.15
0.13
0.14
0.11
-0.12
-0.12
0.03
0.02
0.04
0.03
0.27
0.28
0.00
0.00
0.00
0.00
-0.27
-0.18
0.04
0.02
0.05
0.02
0.37
0.37
0.06
0.03
0.06
0.03
0.00
-0.13
0.08
0.04
0.08
0.04
-0.05
0.03
547
548
Non Corrupted
Corrupted
Diff Ratio
Mean
Median
Mean
Median
Mean
Median
0.19
0.15
0.12
0.09
-0.35
-0.36
0.04
0.01
0.04
0.00
0.16
-0.60
0.36
0.38
0.46
0.50
0.28
0.31
0.26
0.22
0.23
0.17
-0.11
-0.22
0.02
0.00
0.02
0.00
-0.02
0.48
0.01
547
0.00
0.03
548
0.00
3.39
#DIV/0!
Table 3: Regression Results: A: Coefficients for Asset variables
Variable types
Intercept
Cash &Equivalents
Short Term Investments
Receivables
Inventories
Net Prepayments
Other Current Assets
Goodwill
Others
Long Term Investment
No. of observations
Likelihood Ratio®
Ratios
Change rate
Deviation
-0.9536***
(-5.16)
0.6549*
(1.74)
1.0337
(0.76)
2.1958***
(6.93)
0.0147
(0.04)
4.1880***
(4.57)
-0.5555
0.0845
(1.08)
0.0355
(0.62)
-0.0001
(-0.72)
-0.0176
(-0.57)
-0.0939
(-1.15)
0.0175
(1.57)
0.0010
-0.2657***
(-4.02)
0.6549*
(1.74)
1.0337
(0.76)
2.1958***
(6.93)
0.0147
(0.04)
4.1880***
(4.57)
-0.5555
Absolute
deviation
-0.4654***
(-4.28)
1.1080
(2.55)
-0.0651
(-0.05)
1.6323***
(4.00)
-0.5527
(-1.26)
4.3950***
(4.05)
-0.2335
(-0.39)
2.9358***
(4.69)
1.0739*
(1.90)
0.4681
(1.22)
1094
86.191
(0.16)
0.0004
(0.57)
-0.0495*
(-1.67)
0.0000
(0.04)
311
12.16
(-0.39)
2.9358***
(4.69)
1.0739*
(1.90)
0.4681
(1.22)
1094
86.19
(-0.14)
2.1429***
(3.40)
0.8745
(1.43)
-0.1033
(-0.25)
1094
51.522
Panel B: Coefficients for Liability variables
Variable types
Ratios
change rate
deviation
absolute
deviation
Intercept
-0.9146***
0.4419***
-0.1280**
0.0499
(-3.64)
(3.95)
(-2.18)
(0.59)
-1.2010***
-0.0048
-1.2010***
-2.6085***
(-3.10)
(-0.06)
(-3.1)
(-6.21)
2.3695***
0.0073
2.3695***
1.3553**
(3.86)
(0.66)
(3.86)
(2.35)
1.7533***
-0.1099
1.7533***
0.5390
(5.75)
(-1.08)
(5.75)
(1.52)
0.8853***
-0.0526
0.8853***
0.0539
(2.73)
(-1.12)
(2.73)
(0.18)
-0.3279
-0.0014
-0.3279
-1.0219
(-0.39)
(-0.72)
(-0.39)
(-1.27)
3.9770***
-0.0008
3.9770***
3.3689***
(5.46)
(-0.40)
(5.46)
(4.63)
No. of observations
1094
162
1094
1094
Likelihood Ratio®
137.9
6.928
137.91
74.7
Accounts Payable
Notes Payable
Short Term Debt
Other Payables
Other Current Liabilities
Other Long Term
Liabilities
Table 4: Final regression results
ratios
change rates
deviation
-1.5676***
0.3677***
-0.3429***
absolute
deviation
-0.4616***
0.8522**
0.0253
0.8522**
1.3678***
Receivables
2.0776***
0.2093
2.0776***
1.6803***
Net Prepayments
4.3843***
0.0185
4.3843***
4.5288***
Goodwill
2.2847***
-0.0670
2.2847***
1.8949***
1.3474**
-0.0013
1.3474**
1.1409*
-1.6702***
-0.0154
-1.6702***
-2.6990***
Notes Payable
2.2401***
-0.0008
2.2401***
1.6143***
Short Term Debt
1.4213***
-0.0767
1.4213***
0.4423
0.7244**
-0.0365
0.7244**
-0.2530
3.5774***
0.0001
3.5774***
2.8224***
No. of observations
1094
172
1094
1094
Likelihood Ratio®
208.74
8.979
208.74
119.65
Intercept
Cash &Equivalents
Others
Accounts Payable
Other Payables
Other Long Term Liabilities
Table 5. Prediction Results using Ratios model (Accuracy=68%)
Predicted
Predicted
Predicted
Predicted
Actual
1
0
1
384
191
575
0
164
548
356
547
520
1095
Actual
1
0
0.35
0.15
0.50
0.17
0.33
0.50
Actual
1
0
0.70
0.30
1.00
0.35
0.65
1.00
Actual
1
0
0.67
0.32
0.33
0.68
1
0
1
0
1
0
0.53
0.47
1.00
1.00
Table 6: Typical examples: A (firm code: 000603)
Years
A3
2.08
A5
4.38
A9
2.28
A10
1.35
L1
-1.67
L2
2.24
L3
1.42
L4
0.72
L8
3.58
Prob.
Coeff.
-1.57
A1
0.85
1999
1.00
0.00
1.00
0.14
0.00
0.00
-0.10
0.00
0.85
0.17
0.00
0.69
2000
1.00
0.01
1.06
0.12
0.00
0.00
-0.08
0.00
0.90
0.16
0.00
0.72
2001
1.00
0.00
1.09
0.09
0.00
0.00
-0.07
0.00
0.50
0.13
1.00
0.88
2002
1.00
0.05
0.23
0.16
0.00
0.00
-0.17
0.00
0.27
0.27
0.88
0.55
2003
1.00
0.07
0.40
0.70
0.83
0.00
-0.17
0.00
0.26
0.28
0.87
0.95
2004
1.00
0.00
0.40
0.47
0.82
0.00
-0.16
0.00
0.17
0.29
1.08
0.93
2005
1.00
0.00
0.10
0.31
0.99
0.00
-0.18
0.00
0.19
0.24
1.23
0.90
2006
1.00
0.01
0.41
0.10
1.00
0.00
-0.17
0.00
0.19
0.24
1.22
0.93
2007
1.00
0.17
0.95
0.45
0.00
0.10
-0.10
0.00
0.18
0.32
1.27
0.96
2008
1.00
0.25
0.29
0.19
0.93
0.03
-0.11
0.00
0.19
0.31
1.33
0.97
Table 6: Typical examples: A (firm code: 000779)
Years
A3
2.08
A5
4.38
A9
2.28
A10
1.35
L1
-1.67
L2
2.24
L3
1.42
L4
0.72
L8
3.58
Prob.
Coeff.
-1.57
A1
0.85
1998
1.00
0.01
0.13
0.10
0.06
0.13
-0.53
0.00
0.18
0.40
0.00
0.14
1999
1.00
0.06
0.14
0.03
0.06
0.00
-0.39
0.00
0.20
0.45
0.00
0.15
2000
1.00
0.18
0.15
0.02
0.04
0.03
-0.25
0.00
0.20
0.51
0.00
0.24
2001
1.00
0.15
0.30
0.05
0.04
0.00
-0.27
0.00
0.20
0.50
0.00
0.28
2002
1.00
0.09
0.17
0.03
0.04
0.01
-0.21
0.00
0.31
0.48
0.00
0.26
2003
1.00
0.08
0.34
0.02
0.04
0.00
-0.17
0.00
0.00
0.65
0.00
0.27
2004
1.00
0.05
0.61
0.03
0.05
0.00
-0.06
0.29
0.06
0.18
1.94
0.94
2005
1.00
0.04
0.65
0.01
0.05
0.00
-0.11
0.01
0.00
0.18
2.41
0.95
2006
1.00
0.05
0.72
0.02
0.05
0.00
-0.13
0.00
0.00
0.17
2.44
0.96
2007
1.00
0.14
0.51
0.01
0.07
0.00
-0.07
0.00
0.00
0.19
2.48
0.96
Table 6: Typical examples: A (firm code: 600776)
Years
A3
2.08
A5
4.38
A9
2.28
A10
1.35
L1
-1.67
L2
2.24
L3
1.42
L4
0.72
L8
3.58
Prob.
Coeff.
-1.57
A1
0.85
1998
1.00
0.15
0.64
0.26
0.03
0.01
-0.90
0.04
0.38
0.07
0.00
0.39
1999
1.00
0.19
0.72
0.18
0.03
0.05
-0.61
0.06
0.54
0.13
0.00
0.17
2000
1.00
0.15
0.69
0.12
0.04
0.03
-0.82
0.16
0.09
0.17
0.00
0.29
2001
1.00
0.21
0.52
0.16
0.04
0.05
-0.74
0.57
0.12
0.09
0.00
0.11
2002
1.00
0.13
0.59
0.08
0.04
0.06
-0.90
0.04
0.15
0.16
0.00
0.19
2003
1.00
0.22
0.48
0.11
0.05
0.04
-0.78
0.03
0.42
0.10
0.00
0.10
2004
1.00
0.25
0.44
0.03
0.03
0.02
-1.00
0.28
0.11
0.09
0.00
0.03
2005
1.00
0.24
0.32
0.01
0.03
0.01
-1.24
0.15
0.10
0.07
0.00
0.02
2006
1.00
0.31
0.49
0.01
0.03
0.00
-1.40
0.01
0.01
0.09
0.00
0.00
Table 1: Balance Sheet Accounts
Assets
Liabilities
Equities
Account
Descriptions
Account
Descriptions
Account
Descriptions
A1
Cash & equivalents
L1
Accounts payable
E1
Share capital
A2
Short term investments
L2
Notes payable
E2
Retained earnings
A3
Receivables
L3
Short term debt
E3
Capital reserves
A4
Inventories
L4
Other payables
E4
A5
Net prepayments
L5
Other current liabilities
E5
Other stockholder
equity
Total stockholder
equity
A6
Other current assets
L6
Total current liabilities
A7
Total Current Assets
L7
Long term debt
A8
Long term Assets
L8
Other long term liabilities
A9
Goodwill
L9
Deferred tax liabilities charges
A10
Others
L10
Total non-current liabilities
A11
Long term investments
L11
Total liabilities
A12
Total Long term assets
A13
Total Assets
Conclusions
The results show that there exist a strong relation
between FFR and the account ratios:
–
–
–
–
–
Receivables
Net Prepayments
Goodwill
Accounts payable (negative)
Notes payable
– Short term debt;
– Other long terms liabilities
– We also get a high prediction accuracy of 68% with the
estimated model result.
Conclusions
• The corruption firms tend to use more intangible
assets (Goodwill) or paper records (A/R, and Net
Prepayments), and less physical records
(Inventory, fixed assets)
• It looks strange to find that the corruption firms
have more Cash & Equivalents. (Maybe that is the
reason for corruption, such as avoid dividends,
and do “tunneling”)
• The negative relation with the accounts payable is
consistent with Higher working capital level.
(Higher than normal is consistent with Higher NI)
Comments
Consistent with the literature (?)
Fei (2005) concluded four types of fraudulent
in China:
• non-monetary transactions;
• related party transactions;
• assets restructuring and
• change of accounting estimates.
Comments
• We hypothesize that corrupted firms can
potentially manipulate any account.
• Our regression results have eight significant
accounts, which is consistent with the
hypothesis that different company may use
different accounts as the form of corruption.
Future Research
• The relation between the types of corruption and
corresponding accounts used.
• the “tunnelling” type, the cash raised by the equity
market source is transferred to the other “related”
company.
• Some other types are about illegal account profit
adjustment so that the firm’s share becomes more
attractive to the general investors.
• The type of corruption should be related to the
profitability level or changes of the business.
• Another interesting direction is about the timing of the
accounting frauds. (such as the ages after IPC)
Your Comments
and Suggestions
Thanks !