p–ISSN: 2723 - 6609 e-ISSN: 2745-5254
Vol. 5, No. 7 July 2024 http://jist.publikasiindonesia.id/
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3215
Analyzing the Impact of Indonesian Financial Accounting
Standards (IFAS) 71 on Allowance for Impairment Losses
Jauharotul Izzati
1
, Mulyadi Soetardjo
2
, Muchamad Irham Fathoni
3*
, Akbar
Saputra
4
Direktorat Jenderal Pajak, Indonesia
1,3,4
, Universitas Pelita Harapan, Indonesia
2
Email:
jauharotulizzati@kemenkeu.go.id
1
*, [email protected]
2
,
[email protected]
3
, akbar.saputra@kemenkeu.go.id
4
*Correspondence
ABSTRACT
Keywords: allowance for
impairment losses; ifas 71;
interest income; loans;
non-performing loans.
This article analyzes the factors that impact allowance for
impairment losses on banks in Indonesia, after the
implementation of Indonesian Financial Accounting
Standards (IFAS) 71. IFAS 71 became effective in 2020,
replacing IFAS 55. IFAS 71 introduced several new methods
for calculating the allowance for impairment losses. We
collected data from conventional banks listed on the
Indonesia Stock Exchange from 2016—2020. We created
two models: the first will test the impact of several key
factors like loans provided by banks, non-performing loans,
and interest income towards allowance for impairment
losses, while the second will test IFAS 71 implementation
for these factors towards allowance for impairment losses.
Out of these three factors, we concluded loans provided by
banks hurt impairment loss allowance while the other two
have a positive effect, regardless of IFAS 71
implementation. However, while this allowance is found to
be higher after IFAS 71, the three key factors do not have a
significantly stronger effect after IFAS 71 implementation.
Introduction
The banking industry is a significant sector influencing the economic development
of a country. The strategic role of banks is to provide funds to support financing activities
in the real sector (ARHAMI, 2022). Given the importance of financing activities as one
of the bank's revenue-generating activities, risk management is essential to mitigate the
risks banks face as creditors, specifically through the allowances for receivables (Prina,
Suparman, & Prina, 2023).
Allowance for impairment losses is a reserve for receivables based on the estimated
uncollectible value of receivables by the bank. The value of this allowance is evaluated
at each financial reporting date using the expected credit loss impairment model. This
model measures whether the credit risk of financial instruments has significantly
increased since initial recognition, using a fair and supported forward-looking approach
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3216
by the Indonesian Financial Accounting Standards (IFAS) 71 regarding Financial
Instruments.
IFAS 71, which adopts the International Financial Reporting Standard (IFRS) 9
regarding Financial Instruments, became effective on January 1, 2020. This new standard,
replacing IFAS 55 regarding Measurement and Recognition: Financial Instruments,
requires the calculation of loan allowances based on expected or non-payment by the
debtor. This approach takes into account the probability of future impairment due to
economic changes that induce credit risks. To recognize a decline in credit quality, this
approach does not require a specific event to record credit losses, as long as timely
information on each indicator suggesting potential credit losses is available. IFAS 71
requires the measurement and substantiation of expected credit loss through accurate
estimation of the expected amount, consideration of the time value of money, and
provision of documented and supported information based on past and current conditions,
as well as anticipated future scenarios (Dewi, 2021).
In theory, this would increase banks’ allowances for impairment losses. With a
more lenient way that IFAS 71 introduces, banks would probably become more
conservative in regards to recognising impairment losses, compared to the period when
IFAS 55 was still in effect.
Previous studies like those (Sultanoğlu, 2018) have confirmed that the
implementation of IFRS 9 will result in a significant increase in banks' impairment loss
allowances. (Abad & Suarez, 2017) also confirmed that the expected credit loss stipulated
in IFRS 9 is highly responsive to economic condition changes compared to the IAS 39
model. IFRS 9 governs the expected credit loss model for the timely recognition of credit
losses, calculated based on actual credit losses and future information related to the
current loan portfolio (Zaman Grof & Mörec, 2021). IFRS 9 also introduces new
principles for classifying and measuring financial instruments, managing the depreciation
of financial assets, and hedge accounting (Ercegovac, 2018). A study by (Blažeková,
2017) indicates that IFRS 9 is designed to enhance the integrity of the banking financial
system by increasing allowances for impairment loss compared to the situation before its
implementation.
The non-performing loan ratio is a key performance indicator for banks to assess
the quality of their assets. This ratio indicates the risk of a bank failing to receive interest
and principal payments on loans. Therefore, to address this risk, banks need to adjust their
impairment loss allowance funds according to the risk of credit default. A high proportion
of non-performing loans is associated with an increase in a bank's impairment loss
allowance (Islam, 2018). Previous research by (Mohd Isa & Abdul Rashid, 2018) has
proven a positive and significant effect of non-performing loans on the impairment loss
allowance. A positive and significant influence implies that as non-performing loans
continue to increase, so too will the impairment loss allowance.
This study aims to examine the factors influencing the magnitude of the impairment loss
allowance in Indonesian banks listed on the Indonesia Stock Exchange and assess the
impact of IFAS 71 implementation on the impairment loss allowance. It is still unclear to
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3217
what extent the implementation of IFAS 71 affects the amount of impairment loss
allowance, its role in enhancing the capacity and efficiency of impairment loss allowance,
and in reducing the use of receivables related to impairment loss allowance by bank
administration to achieve specific objectives, such as signalling risk-taking and others.
Research Methods
Research Data
This study focuses on investigating the factors influencing the allowance for
impairment losses using three independent variables: loans provided, non-performing
loans, and total income, along with adding IFAS 71 as a moderating dummy variable to
understand the role of IFAS 71 in moderating the relationship between loans provided,
non-performing loans, and total income towards the allowance for impairment losses.
This study will further examine the impact of the first-time implementation of IFAS 71
in Indonesia based on empirical data reported by banks before and after the
implementation of IFAS 71 during the period 2016 to 2020. The study is conducted by
examining actual data from financial statements that have been prepared and published
by companies on the Indonesia Stock Exchange.
The study uses secondary data consisting of data processed by companies and made
public. This secondary data includes financial statements and annual reports. The data
source for this study is taken from the financial statements for the years 2016—2020,
published on the Indonesia Stock Exchange website. The sampling method used in this
research is purposive sampling based on predetermined criteria.
Table 1
Research Samples
Criteria
Amount
Conventional banks listed in the Indonesia Stock Exchange
from the years 2016—2020 consecutively
43
Conventional banks listed in the Indonesia Stock Exchange
from the years 2016—2020 non-consecutively
0
Incomplete data
(1)
Data used
42
Number of years observed
5
Total observation (42 x 5)
210
Outliers
(29)
Total samples
181
Research Models
This study uses the multiple regression analysis method because it consists of one
dependent variable and several independent variables. The regression equation of this
study is formulated in two empirical models because the study examines the factors
influencing the allowance for impairment losses before and after the implementation of
IFAS 71. The regression equation for model 1 used in this study is adopted from previous
research (Mohd Isa & Abdul Rashid, 2018) and model 2 is formulated as follows:
LLP
i,t
= β
0
+ β
1
LOANS
i,t
+ β
2
NPL
i,t
+ β
3
GI
i,t
+ β
4
PSAK71
i,t
+ β
5
SIZE
i,t
+ e .(Model 1)
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3218
LLP
i,t
= β
0
+ β
1
LOANS
i,t
+ β
2
NPL
i,t
+ β
3
GI
i,t
+ β
4
PSAK71
i,t
+ β
5
(LOANS×PSAK71)
i,t
+ β
6
(NPL×PSAK71)
i,t
+ β
7
(GI×PSAK71)
i,t
+ β
8
SIZE
i,t
+ e ................(Model 2)
with
LLP : Loan Loss Provision or Allowance for Impairment Losses
β
1
LOANS
i,t
: Beta variable for loan ratio
β
2
NPL
i,t
: Beta variable for non-performing loans
β
3
GI
i,t
: Beta variable for gross interest income
β
4
PSAK71
i,t
: Beta dummy variable for IFAS 71, where 1 represents the year after the
implementation of IFAS 71 and 0 for the year before its implementation
β
5
SIZE
i,t
: Beta variable for company size
β
0
: Intercept parameter
i,t : Indicator for company i and year t
e : Error term distributed with a mean of zero and variance
2
The regression equation for model 1 is used to test hypotheses H
1
, H
2
, H
3
, and H
7
.
The β
1
coefficient value in model 1 is the focus for testing hypothesis H
1
. The β
2
coefficient value in model 1 is the focus for testing hypothesis H
2
. The β
3
coefficient value
in model 1 is the focus for testing hypothesis H
3
. The β
4
coefficient value in model 1 is
the focus for testing hypothesis H
7
.
The regression equation for model 2 is used to test hypotheses H
4
, H
5
, and H
6
. The
β
5
coefficient value in model 2 is the focus for testing hypothesis H
4
. The β6 coefficient
value in model 2 is the focus for testing hypothesis H
5
. The β
7
coefficient value in model
1 is the focus for testing hypothesis H
6
.
Research Variables
Dependent Variable – Allowance for Impairment Losses
The dependent variable in this research is the allowance for impairment losses on
loans issued by conventional commercial banks. The allowance for impairment losses
used in this study is a contra account or a reduction to the loans issued by the bank,
presented in the financial position statement. The measurement of the allowances for
impairment losses in this study follows the methodology of (Casta, Lejard, & Paget-
Blanc, 2019), formulated as follows:
𝐿𝐿𝑃 =
𝐿𝑜𝑎𝑛 𝑙𝑜𝑠𝑠 𝑃𝑟𝑜𝑣𝑖𝑠𝑖𝑜𝑛
𝑇𝑜𝑡𝑎𝑙 𝐿𝑜𝑎𝑛
Moderating Variable – IFAS 71
This study employs IFAS 71 as a moderating variable concerning financial
instruments. This standard adopts IFRS 9 and replaces IFAS 55. The implementation of
IFAS 71 influences the accounting treatment for the recognition and measurement of
financial assets. The proxy for the IFAS 71 variable uses a dummy variable, where 1
represents the years following the implementation of IFAS 71, and 0 represents the years
before its implementation.
Independent Variable – Loans Provided by Banks (Loans)
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3219
Loans are represented by the total credit or loans issued by the bank divided by total
assets. A higher ratio of credit issued by a bank leads to greater losses due to higher credit
risk exposure. The formula for loans used in this study is based on the methodology of Al
(Casta et al., 2019).
ns =
Total Loans
Total Assets
Independent Variable – Non-Performing Loans (NPL)
Non-performing loans (NPL ratio) are the ratio of the total loans issued to the level
of doubtful, substandard, and non-performing loans, compared to the total loans issued
by the bank (Slamet Riyadi, 2006). Credit risk indicates a bank's failure to earn interest
and/or loan receivables, necessitating increased allowances for anticipated default losses.
The operationalization of NPL in this study follows previous research.
𝑁𝑃𝐿 =
𝑁𝑜𝑛𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑖𝑛𝑔 𝐿𝑜𝑎𝑛
𝑇𝑜𝑡𝑎𝑙 𝐿𝑜𝑎𝑛
Independent Variable – Gross Interest Income
Banks use loans to generate income. The larger the loans issued to customers, the
higher the bank's interest income. This study uses the gross interest income ratio (GI),
calculated as the bank's total income divided by total assets. Significant increases or
decreases in gross income lead to corresponding adjustments in the allowances for
impairment losses to normalize the rate of return on assets. The operationalization of GI
is formulated as follows:Cap𝐶𝑎𝐺𝐼 =
𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝐼𝑛𝑐𝑜𝑚𝑒
(
𝐿𝑜𝑎𝑛𝑠
)
𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠
Control Variable – Company Size
The size of a company positively affects the allowances for impairment losses, as
larger banks have higher business levels compared to smaller banks (Ozili, 2017). The
formula for company size in this study is operationalized as in previous research (Casta
et al., 2019):
𝑆𝑖𝑧𝑒 = ln (𝑡𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠)
Results and Discussion
Table 2 presents the summary of descriptive statistics for the research variables.
The average value of the allowance for impairment losses (LLP) on gross loans during
the 2016—2020 period is 0.02591 (approximately 2.6%). This generally reflects credit
risk management of 2.6% of the gross loans issued by banks. The 2.6% value is lower
than the average nonperforming loan rate of 3.6%. This lower rate may indicate
weaknesses in credit risk management and may also reflect the bank's administrative
interest in increasing profitability by reducing loan loss provisions. The allowance for
impairment losses ratio ranges from 0.07% to 8.6%, with a standard deviation of 1.7%.
Table 2
Descriptive Statistics Summary
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3220
Variable
N
Mean
Min
Max
Std. Dev.
LLP
181
0.02591
0.00074
0.08601
0.01718
LOANS
181
0.63852
0.34742
0.82054
0.09464
NPL
181
0.03552
0.00000
0.11678
0.02082
GI
181
0.08008
0.04386
0.12305
0.01432
PSAK71
181
0.17127
0.00000
1.00000
0.37779
SIZE
181
17.39434
13.55332
21.19954
1.08897
The variable for loans issued by banks (LOANS) has an average value of
0.6385229, meaning that on average, the credit facilities provided by the bank constitute
63.8% of its assets. This implicitly reflects the high exposure of the bank to the credit risk
emanating from these facilities. The loan ratio ranges from 34.7% to 82.05%, with a
standard deviation of 9.5%.
The non-performing loans (NPL) variable has an average value of 0.0355218. A
value of 3.5% reflects a high-quality credit portfolio in conventional commercial banks,
yet remains within a globally safe level (not exceeding 10%). The proportion of non-
performing loans ranges from 0% to 11.7%, with a standard deviation of 2.1%, indicating
a reasonable ratio convergence within the research year range.
The variable for the interest income ratio (GI) is calculated based on total interest
income divided by the total assets of the bank. The average value of the interest income
ratio during the study period is 0.0800789, ranging from 4.4% to 12.3% with a standard
deviation of 1.4%. This also indicates relative stability in the interest income of
Indonesian commercial banks during the study period.
The variable for the implementation of IFAS 71 (PSAK71) is a dummy variable.
The value is 0 for the years before the implementation of IFAS 71 and 1 for the years of
implementation of PSAK 71 in the study period. The standard deviation value of the IFAS
71 implementation variable is 37.78%.
SIZE is a control variable for company size. The average value of company size is
17.39434, meaning that the average size of the company based on the assets owned is
17.39434. The minimum value of company size is 13.55332, while the maximum value
of company size is 21.19954. The standard deviation of the company size is 1.88971.
Correlation Analysis
Correlation analysis is a method used to determine the presence or absence of a
linear relationship between two variables. If the correlation coefficient is statistically
significant, it indicates that the two variables are correlated. However, if the correlation
coefficient is not statistically significant, then the two variables are not correlated.
Table 3
Correlation Analysis
Variable
LLP
LOA
NS
NPL
GI
PSAK71
LOANS
×
PSAK71
NPL×
PSAK71
GI
×
PS
AK
71
SIZ
E
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3221
LLP
1
LOANS
-
0.00
27
0.97
17
1
NPL
0.51
19*
0.00
00
0.026
4
0.724
5
1
GI
-
0.08
13
0.27
64
0.468
3*
0.000
0
0.09
04
0.22
63
1
PSAK71
0.29
43*
0.00
01
-
0.350
3*
0.000
0
0.03
05
0.68
32
-
0.45
56*
0.00
00
1
LOANS×
PSAK71
0.29
82*
0.00
00
-
0.283
5*
0.000
1
0.03
69
0.62
22
-
0.43
40*
0.00
00
0.9864*
0.0000
1
NPL×
PSAK71
0.38
01*
0.00
00
-
0.286
3*
0.000
1
0.20
63*
0.00
53
-
0.38
77*
0.00
00
0.8749*
0.0000
0.8717*
0.0000
1
GI×
PSAK71
0.29
11*
0.00
01
-
0.325
5*
0.000
0
0.03
49
0.64
12
-
0.40
13*
0.00
00
0.9862*
0.0000
0.9815*
0.0000
0.8690*
0.0000
1
SIZE
0.40
67*
0.00
00
0.265
6*
0.000
3
-
0.12
75
0.08
71
-
0.12
77
0.08
67
0.0291
0.6976
0.0429
0.5663
0.0609
0.4158
0.0
351
0.6
386
1
The correlation analysis shows that the LLP variable has a significant 5%
correlation with the NPL variable with a coefficient of 0.0000, PSAK71 with a coefficient
of 0.0001, LOANS×PSAK71 with a coefficient of 0.0000, NPL×PSAK71 with a
coefficient of 0.0000, GI×PSAK71 with a coefficient of 0.0001, and SIZE with a
coefficient of 0.000. Meanwhile, LOANS and GI do not have a correlation relationship
with LLP as they have significant values above 5%. Almost all variables have a
correlation coefficient below 0.8, indicating no signs of multicollinearity problems,
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3222
except for LOANS with a correlation coefficient of 0.9717. Furthermore,
multicollinearity issues will be analyzed in the multicollinearity test.
Classical Assumption Test
Research can be proven using multiple linear regression methods, provided that all
independent variable data are normally distributed. To determine whether the data used
in the study are normally distributed, BLUE (Best Linear Unbiased Estimates) Gauss-
Markov divides the classical assumption test into four types: normality test,
multicollinearity test, heteroskedasticity test, and model specification test.
Normality Test
A regression model is considered good if it has a normal or near-normal
distribution. This study uses the Shapiro-Wilk normality test, Shapiro-Francia normality
test, and skewness/kurtosis test for normality. If the probability value (prob>z) is more
than 0.05, then the data is normally distributed. Conversely, if the probability value is less
than 0.05, then the data is not normally distributed. The results of the normality test are
presented in Table 4.
Table 4
Normality Test Result
Variables
Shapiro-
Wilk Test
Shapiro-
Francia
Test
Skewness/
Kurtosis
Test
Conclusion
LLP
0.00000
0.00001
0.0000
Not
significant
LOANS
0.00007
0.00021
0.0021
Not
significant
NPL
0.00000
0.00001
0.0000
Not
significant
GI
0.25791
0.22849
0.1060
Significant
PSAK71
0.00002
1.00000
0.0000
Not
significant
LOANS×PSA
K71
0.00000
0.00717
0.0000
Not
significant
NPL×PSAK7
1
0.00000
0.00003
0.0000
Not
significant
GI×PSAK71
0.00000
0.03309
0.0000
Not
significant
SIZE
0.00192
0.00746
0.0074
Not
significant
Based on Table 4, each variable is not significant except for the GI variable. It can
be concluded that the data is not normally distributed except for the GI variable. The
Central Limit Theorem states that the larger the sample data, the more normally the data
is distributed. According to Gujarati (2012), data is considered large if the number of
observations exceeds 100 data points. The data observed in this study, totalling 181 data
points, can be assumed to be normally distributed based on the Central Limit Theorem.
Multicollinearity Test
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3223
To detect correlation, one can use the variance inflation factor (VIF) or tolerance =
1/VIF (TOL). If the VIF is less than 10, then the sample does not have a multicollinearity
problem.
Table 5
Multicollinearity Test Result
Model
Mean VIF
Conclusion
1
1.33
No multicollinearity exists
2
22.00
Multicollinearity exists
Based on Table 5, the test results for multicollinearity issues show that the VIF
value in model 1 is 1.33, indicating that there is no multicollinearity problem in model 1.
However, in model 2, the VIF value is 22.00, which means that a multicollinearity
problem among variables is found because the VIF is more than 10.
Heteroskedasticity Test
Table 6
Heteroskedasticity Test Result
Model
Breusch-Pagan Test
White Test
Prob>chi
2
Conclusion
Prob>chi
2
Conclusion
1
0.0000
Significant
0.4169
Not significant
2
0.0000
Significant
0.5157
Not significant
Based on Table 6, the heteroskedasticity test results for model 1 and model 2 using
the Breusch-Pagan test indicate significant results, which means there is a problem of
heteroskedasticity. In the testing using the White test for model 1 and model 2, the results
show insignificance, which means there is no problem of heteroskedasticity. To perform
statistical inference, this study applies robust standard error to correct the standard error
without changing the regression coefficients.
Coefficient of Determination Analysis (R2 Test)
Table 7
Coefficient of Determination Analysis Result
Model
Dependent
Var
Independent Var
R
2
Value
1
LLP
LOANS + NPL + GI + PSAK71
+ SIZE
0.5684
2
LLP
LOANS + NPL + GI + PSAK71 +
LOANS×PSAK71 + NPL×PSAK71 +
GI×PSAK71 + SIZE
0.5725
Based on Table 8, it is known that the R-squared value for model 1 is 0.5684. This
indicates that the independent variables can explain 56.84% of the allowance for credit
losses, and the remaining 43.16% is explained by other factors outside those used in
model 1 of this research. The R-squared for model 2 is 0.5725, which indicates that the
independent variables can explain 57.25% of the allowance for credit losses. Meanwhile,
the remaining 42.75% is explained by other factors outside those used in model 2 of this
research.
Model Specification Test (F-Test)
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3224
The test is conducted by comparing the significance level of the F-statistic from the
test results with the significance level determined in this research, which is 5%. If the F-
statistic from the test results is below 5%, then all the independent variables
simultaneously affect the dependent variable. Based on the F-test results presented in
Table 9, the F-statistic for model 1 and model 2 is 0.0000 and 0.0000 respectively, or
below the 5% significance level. It can be concluded that the independent variables of
model 1 and model 2 have a simultaneous effect on the allowance for credit losses.
Tabel 8
F-Test Result
Model
Dependen
t Var
Independent Var
Prob>F
1
LLP
LOANS + NPL + GI + PSAK71
+ SIZE
0.0000
2
LLP
LOANS + NPL + GI + PSAK71 +
LOANS×PSAK71 + NPL×PSAK71 +
GI×PSAK71 + SIZE
0.0000
Hypothesis Test (t-Test)
The hypothesis test in model 1 aims to determine whether the independent variables
(loans provided by the bank, non-performing loans, income, and the implementation of
IFAS 71) individually affect the allowance for credit losses. The hypothesis test in model
2 aims to determine whether the implementation of PSAK71 as a moderating variable
strengthens or weakens the influence of the independent variables (loans provided by the
bank, non-performing loans, income, and the implementation of IFAS 71) on the
allowance for credit losses. The analysis of the t-test results is summarized in Table 10,
and the research model tested is as follows.
Table 9
t-Test Result
Variables
Model 1
Model 2
coefficient
t-value
sig.
coefficient
t-value
sig.
Intercept
-0.071742
-9.06
0.000
-0.073339
-9.41
0.000
LOANS
-0.019796
-1.68
0.095
-0.023265
-2.05
0.042
NPL
0.463796
10.59
0.000
0.459151
9.58
0.000
GI
0.131112
1.81
0.072
0.167016
2.40
0.018
PSAK71
0.012456
4.18
0.000
0.020606
1.10
0.273
SIZE
0.004667
11.70
0.000
0.004728
12.09
0.000
LOANS×PSA
K71
0.017625
0.44
0.664
NPL×PSAK71
0.027354
0.21
0.835
GI×PSAK71
-0.286115
-0.99
0.325
The t-test results from the research model, conducted using STATA v.16, present
two-tailed probabilities. Thus, for testing the hypotheses of this study, which use one-
sided tests, the two-tailed probability values are divided by two.
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3225
H1: Loans provided by banks have a positive effect on the allowance for impairment
losses.
According to the t-test results presented in Table 10, the regression coefficient for
loans provided (LOANS) is -0.019796, with a one-tailed probability value of 0.0475
(calculated by dividing the significance in Table 10 by two). This value is less than the
5% significance level, implying that loans granted negatively affect the allowance for
impairment losses. Therefore, the first hypothesis, which states that loans provided by
banks positively affect the allowance for impairment losses, is not supported and is
rejected.
The first hypothesis test result indicates that loans provided by banks negatively
impact the allowance for impairment losses, leading to the rejection of the hypothesis.
This outcome suggests that the credit risk exposure borne by the bank from its loan
portfolio inversely affects the formation of the allowance for impairment losses. As the
volume of loans provided by banks increases, the allowance for loan losses set aside
decreases. The high lending activity of the bank is inversely proportional to the size of
the allowance for losses established by the bank.
The results contradict the first hypothesis due to two factors: first, a lack of evidence
supporting the impact of credit from the loan portfolio on the formation of the allowance
for impairment losses. Second, the allowance for impairment losses is formed based on
the credit risk exposure of the loans granted by the bank. When the majority of loans are
estimated not to have significant credit risk from the initial recognition (stage 1) or when
loans improve from previously having significant credit risk, the provision for impairment
losses does not significantly increase or decrease. This is predicted to influence the
absence of a positive effect of bank-issued loans on the allowance for impairment losses.
H2: Non-performing loans have a positive effect on the allowance for impairment
losses.
Based on the t-test results presented in Table 10, the regression coefficient for non-
performing loans (NPL) is 0.463796, with a one-tailed probability value of 0.000. This
value is less than the 5% significance level, indicating that non-performing loans
positively affect the allowance for impairment losses. Consequently, the second
hypothesis stating that non-performing loans positively affect the allowance for
impairment losses is supported and accepted.
The second hypothesis test result shows that non-performing loans positively
impact the allowance for impairment losses, leading to the acceptance of the hypothesis.
The influence of non-performing loans on the formation of the allowance for impairment
losses can be explained by the fact that an increase in the nonperforming loan ratio drives
the formation of the allowance for losses due to a change in the credit quality to non-
performing, doubtful, and less than satisfactory. This change in credit quality is assessed
based on the business prospects, performance of the debtor, and payment ability
supported by objective evidence. As non-performing loans increase, the formation of the
allowance for losses, which is a contra account to the loans granted by the bank in the
financial position statement, also increases. Similarly, the position of the allowance for
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3226
impairment losses as an expense reducing earnings before tax and provisions decreases
the accounting profit.
H3: Interest income has a positive effect on the allowance for impairment losses.
Based on the t-test results presented in Table 10, the regression coefficient for total
income (GI) is 0.131112, with a one-tailed probability value of 0.036. This value is less
than the 5% significance level, indicating that total income positively affects the
allowance for impairment losses. Thus, the third hypothesis stating that total income
positively affects the allowance for impairment losses is supported and accepted.
The positive relationship indicates that when the bank anticipates high income, it
increases the allowance for impairment losses. The link between income and loans (as a
means to generate income) is also related to credit risk due to the increased capacity of
borrowers corresponding to the rising loan portfolio.
The bank's asset management performance in generating income becomes one of
the financial performance indicators that attract public attention. Banks maintain a normal
return on assets level by increasing the allowance for impairment losses when interest
income rises or decreasing the allowance for impairment losses when interest income
falls. The allowance for impairment losses represents each bank's discretion in estimating
the value of the allowance formed. This estimation nature is used to absorb losses arising
from loan defaults by debtors.
H4: The amount of the allowance for impairment losses will be higher in the period
after the implementation of IFAS 71.
Based on the t-test results presented in Table 10, the regression coefficient for the
implementation of IFAS 71 (PSAK71) is 0.0124561, with a one-tailed probability value
of 0.000. This value is less than the 5% significance level, meaning that the
implementation of IFAS 71 positively affects the increase in the amount of allowance for
impairment losses formed by the bank. Therefore, the fourth hypothesis stating that the
amount of the allowance for impairment losses will be higher in the period after the
implementation of IFAS 71 is supported and accepted.
From these results, it is concluded that after the implementation of IFAS 71, the
amount of the allowance for impairment losses is higher. This finding aligns with the
research of (Mohd Isa & Abdul Rashid, 2018), which demonstrated that the
implementation of IFRS 9 led to an increase in the formation of the allowance for
impairment losses. A higher allowance for impairment losses indicates that banks are
aware that setting aside an allowance for impairment losses is a preemptive step against
future credit risks that may disrupt banking performance, financial system stability, and
economic growth.
H5: The positive influence of loans provided on the allowance for impairment losses
will be stronger in the period after the implementation of IFAS 71.
Based on the t-test results presented in Table 10, the regression coefficient for the
variable of loans granted after the implementation of IFAS 71 (LOANS×PSAK71) is
0.017625, with a one-tailed probability value of 0.332. This value is greater than the 10%
significance level, indicating no difference in the effect of loans granted by banks on the
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3227
allowance for impairment losses before and after the implementation of IFAS 71. Thus,
hypothesis H5 is rejected.
From these results, it is concluded that there is no difference in the effect of loans
granted by banks on the allowance for impairment losses before and after the
implementation of IFAS 71. The addition of an expectation factor in calculating losses
from the impairment of financial assets as part of IFAS 71's implementation does not
prove to strengthen or weaken the effect of loans granted by banks on the allowance for
losses. No correlation is found between the credit risk borne by the bank and the provision
for the allowance after the implementation of IFAS 71.
The absence of a moderating effect of the implementation of IFAS 71 on the
relationship between loans provided by banks and the formation of the allowance for
impairment losses is estimated to be due to banks not significantly increasing their
allowance for impairment losses after the implementation of IFAS 71.
H6: The positive influence of non-performing loans on the allowance for impairment
losses will be stronger in the period after the implementation of IFAS 71.
Based on the t-test results presented in Table 10, the regression coefficient for the
variable of non-performing loans after the implementation of IFAS 71 (NPL×PSAK71)
is 0.0273543, with a one-tailed probability value of 0.4175 (significance in Table 10,
0.835 divided by two). This value is greater than the 10% significance level. Previously,
the t-test results for the non-performing loans (NPL) variable showed a one-tailed
probability value of 0.000 and a regression coefficient of 0.463796, meaning that non-
performing loans positively affected the allowance for impairment losses before the
implementation of IFAS 71. However, after the implementation of IFAS 71, non-
performing loans do not significantly affect the allowance for impairment losses. In other
words, there is no difference in the effect of non-performing loans on the allowance for
impairment losses before and after the implementation of IFAS 71. Therefore, hypothesis
H6 is rejected.
The absence of a moderating effect of the implementation of IFAS 71 on the
relationship between non-performing loans and the formation of the allowance for
impairment losses is estimated to be due to banks not significantly increasing their
allowance for impairment losses after the implementation of IFAS 71. This is concerning
because it may imply that banks are not aware of forming an allowance for impairment
losses. The lack of consideration in forming an allowance for impairment losses by banks
could be due to two reasons. First, banks are optimistic about the collectability of loans
granted. Lastly, estimating the allowance for impairment losses based on IFAS 71 is very
complex and incurs significant costs. The substantial cost of estimating the allowance for
impairment losses, especially in terms of macroeconomic forecasts and forward-looking
information, requires economic expertise.
H7: The positive influence of interest income on the allowance for impairment losses
will be stronger in the period after the implementation of IFAS 71.
Based on the t-test results presented in Table 10, the regression coefficient for the
variable of total income after the implementation of IFAS 71 (GI×PSAK71) is -0.286115,
Jauharotul Izzati, Mulyadi Soetardjo, Muchamad Irham Fathoni, Akbar Saputra
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3228
with a one-tailed probability value of 0.1625. This value is greater than the 10%
significance level, indicating no difference in the effect of interest income on the
allowance for impairment losses before and after the implementation of IFAS 71. Thus,
hypothesis H7 is rejected.
The absence of a significant moderating effect of the implementation of IFAS 71
between total interest income and the formation of the allowance for impairment losses
is due to a lack of supporting evidence. Management discretion in estimating the
likelihood of credit losses by charging the allowance for impairment losses in the income
statement is not utilized by banks. The measurement of the allowance for impairment
losses using the expected credit loss approach in IFAS 71 gives management discretion
to measure losses on loans granted by banks. The determination of forward-looking
factors as part of the expected loss provisioning requires management's judgment in
estimating the likelihood of credit losses based on macroeconomic condition projections.
Conclusion
Through a series of statistical tests, it was found that loans provided by banks had
a negative impact on the impairment loss reserve; The larger the loan amount given, the
smaller the loss reserve set aside. In contrast, non-performing loans have a positive impact
on impairment loss reserves; An increase in the non-performing loan ratio leads to an
increase in loss reserves due to a deterioration in credit quality to substandard, doubtful,
and problematic. Total revenue also has a positive impact on impairment loss reserves;
The larger the loan that the bank gives to the customer, the higher the interest income
earned by the bank, thus increasing the risk of payment default by the borrower. The
implementation of IFAS 71 increases the amount of impairment loss reserves, indicating
that banks recognize the importance of setting aside reserves as a preemptive measure
against future credit risks that could disrupt banking performance, financial system
stability, and economic growth.
In addition, there was no significant difference in the effect of loans provided by
banks on impairment loss reserves before and after the implementation of IFAS 71, which
may indicate that banks are not fully aware of the need to establish impairment loss
reserves due to optimism about loan collection capabilities and the complexity and high
cost of estimating losses. The same applies to non-performing loan securities and interest
income against impairment loss reserves before and after the implementation of IFAS 71;
No significant differences were found. Based on the limitations of this study, researchers
are further advised to consider other factors that affect impairment loss reserves, such as
the capital adequacy ratio, the CET 1 ratio, and the income tax rate as suggested by Molla
(2021). Future research may also include longer study periods to generate more data,
especially data from the period following the implementation of IFAS 71.
Analyzing the Impact of Indonesian Financial Accounting Standards (IFAS) 71 on Allowance
for Impairment Losses
Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 7, July 2024 3229
Bibliography
Abad, Jorge, & Suarez, Javier. (2017). Assessing the cyclical implications of IFRS 9-a
recursive model. ESRB Occasional Paper Series.
ARHAMI, ARHAMI. (2022). Strategi Penanganan Wanprestasi Kredit Usaha Rakyat
(Kur) Mikro Pada Bank Rakyat Indonesia Unit Salobulo Kota Palopo. Institut
Agama Islam Negeri (IAIN) Palopo.
Blažeková, Petra. (2017). The impact of IFRS 9 (increase in credit risk provisioning) on
banks’ regulatory capital. Comenius Management Review, 11(2), 27–42.
Casta, Jean François, Lejard, Christophe, & Paget-Blanc, Eric. (2019). The
implementation of the IFRS 9 in banking industry. EUFIN 2019: The 15th
Workshop on European Financial Reporting.
Dewi, Putri Fahlinda. (2021). Analisis Strategi Penyelesaian Penurunan Jumlah Nasabah
Pada Produk Pembiayaan Murabahah di PT. BPRS Al Washliyah Medan.
Universitas Islam Negeri Sumatera Utara.
Ercegovac, Roberto. (2018). IFRS 9 impact on bank landing policy and structural risk
management. Eurasian Journal of Business and Management, 6(3), 53–60.
Mohd Isa, Mohd Yaziz, & Abdul Rashid, Md Zabid Hj. (2018). Regulatory capital funds
and risk-sharing behavior in distressed financial conditions: An empirical analysis
on Islamic banks in Malaysia. Journal of Financial Reporting and Accounting,
16(1), 197–216.
Molla, Simeneh Almaw. (2021). The cyclicality of loan loss provision and income
smoothing behavior of commercial banks pre and post IFRS: Evidence from
Ethiopia. Research Journal of Finance and Accounting, 12(3), 13–22.
Prina, Elyn, Suparman, Yusep, & Prina, Gumgum. (2023). The Implementation of Latent
Gaussian Model in the Forecasting Process. Jurnal Penelitian Pendidikan IPA,
9(12), 12155–12165.
Sultanoğlu, Banu. (2018). Expected credit loss model by IFRS 9 and its possible early
impacts on European and Turkish banking sector. Muhasebe Bilim Dünyası
Dergisi, 20(3), 476–506.
Zaman Grof, Maja, & Mörec, Barbara. (2021). IFRS 9 transition effect on equity in a post
bank recovery environment: the case of Slovenia. Economic Research-Ekonomska
Istraživanja, 34(1), 670–686.