An Evaluation of Fraud and Deposit Money Banks’ Profitability in Nigeria: (2009-2018)

This work evaluates Fraud and Profitability of Deposit Money Banks (DMB’s) in Nigeria for ten years (2009-2018); with the specific objective of assessing whether the rate at which fraud occurs, the number of persons involved in fraud, the amount targeted in fraud and the loss that the banks incurred to fraud has significant impact on Profitability for the period. Using Regression Analysis on historical data from Nigeria Deposit Insurance Corporation (NDIC) Reports, we found out that there exists a strong positive correlation between DMB’s Fraud and Profitability at more than 90%; and judging by our data analytics at 5% level of significance, the frequency of fraud, fraud amount and monies that could not be recovered from fraud proved to assert strong influence on the profitability of DMB’s in Nigeria with only Fraud Involvement proving insignificant. To mitigate fraud, we recommend that banks create fraud policies that are robust enough to prevent fraud perpetrators from committing fraud and a Whistleblowing System (WBS) that guarantees protection for a Whistle Blower.


Scope of the Study:
Profitability in this research is projected in the context of ROE while fraud is explained in the context of the frequency of occurrence, the number of persons involved, the amount involved in the fraud and the amount lost to the fraud. The study covers ten years (2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018).

Limitation of the Study:
Profitability in this study is limited to the average ROE of Deposit Money Banks and her fraud report as detailed in year-on-year NDIC Reports.

Operational Definition of Terms:
Fraud Frequency -The rate at which fraud occurs Fraud Amount -The amount involved per reported fraud Fraud Loss -The amount unrecovered owning to the fact that fraud had occurred Fraud Involvement -The number of persons involved in a fraud Profitability -The ability to generate profit Albrecht, & Albrecht 2008). Projecting a false representation is a fraud. Ogunleye, (2010), affirmed that fraud impact adversely on the assets of the bank and can be committed by (but not limited to): bank employees, bank staffs (at whatever capacity) and bank vendors. According to the Black's Law Dictionary as cited in (Ogunleye 2010); fraud is a perversion of truth. This oftentimes is done in intentionality to gain undue advantage. According to (Fakunle 2006;Ojeaga, Ikpefan & Odejimi 2014), Bank Management Fraud is the act of manipulation of records and account by management staff in pursuant of self-interest. There are aids to fraud which can be trailed to an institution and societal lapses. This is reflected in poor monitoring and evaluation owning to frail institutional structures by regulatory agencies, poor wages and societal decadence (Adewunmi, 1986;Cahill, Chen & Lambert 2002;Ojeaga, Ikpefan & Odejimi 2014).
Empirical Review: Ojeaga, Ikpefan, & Odejimi (2014), investigates factors that incite fraud in the banking sector in Nigeria, on times series for fraud obtained from CBN data from 1998 to 2010 using regression estimation method. It was found that high bank deposit was primarily responsible for a high-rise fraudulent occurrence in the Nigerian banking sector particularly management fraud, some other factors that were also jointly responsible for these occurrences include high-interest rates, low commercial bank lending and poor oversight function by the Central Bank and other financial regulatory agencies. Oloidi & Ajinaja (2014), examine the causes, types, detection, and prevention of frauds and forgeries in Nigeria's banking sector. A questionnaire was designed to collect data from 81 bank branches in South Western Nigeria by ranking causes, types, detection, and prevention order of viciousness with findings that the major factor in play was the problem of an effective internal control system and enforcement of strict adherence. It was recommended that banks should install an effective internal control system and enforce strict compliance with control measures. Adeusi, Kolapo, & Aluko (2014), examined the factors that influence the profitability of commercial banks in Nigeria based on data gathered from 2000 to 2013 on fourteen banks using time series and crosssectional data analysis. Profitability is measured with Return on Assets (RoA) while capital adequacy ratio; asset quality; management efficiency; liquidity ratio; inflation; and economic growth are all explanatory variable. The findings revealed that management efficiency, asset quality, and economic growth proved significant as determinant of commercial banks' profitability. Ani, Ugwunta, Ezeudu, & Ugwuanyi (2012), did an in-depth investigation of the determinants of the profitability of Deposit Money Banks in Nigeria with data set that is made up of 147 bank-level observations over ten years from 2001 to 2010 in respect of 15 banks. Data were obtained from the annual reports and accounts of the sampled banks. Multiple regression analysis was used to estimate the coefficients. Major outcomes of the analysis include that increase in size (higher total assets) may not necessarily lead to higher profits due to diseconomies of scale; higher capital-assets ratio and loans and advances contribute strongly to bank profitability. Overall, the paper suggests bank size, capital and asset composition as the major endogenous determinants of bank profitability in Nigeria. Nwankwo (2013), evaluated the impact of fraud on the performance of commercial banks in Nigeria. He also sought to ascertain the relationship between bank ATM Fraud, Forged Cheque, Clearing Cheque Fraud and bank performance using regression analysis. The outcome of the research revealed that there is a significant impact of fraud on the performance of commercial banks in Nigeria. The implication of this is that if the level of fraud in the commercial banks if not reduce to the barest minimum, may not allow commercial banks to perform well and contribute significantly to the growth of Nigeria's economy. Nwankwo recommends that there is an urgent need for effective monitoring of bank fraud through the use of ATM to allow for the growth of Nigeria's commercial banks' performance. Akinwumi, Michael & Raymond (2017), researched four Deposit Money Banks in Nigeria between 2007 and 2016, using Pearson correlation co-efficient technique. The empirical results revealed that there is a statistically significant relationship between banks' liquidity, return on asset and return on equity. However, the relationship is not all that statistically significant when the return on asset was used as a proxy for profitability. It was suggested that the banks should evaluate and redesign their liquidity management strategy so that it will optimize returns to shareholders' equity and also optimize the use of the assets. The study showed that good management and control of fact ors Volume IX Issue 3, September 2018 27 www.scholarshub.net influencing the liquidity of commercials banks in the country could improve the financial performance of banks. Adetoso & Akinselure (2016), examine fraud control and fraud prevention in selected Nigerian commercial banks in Nigeria. The methodology used was based on primary data, which were gotten through one hundred and fifty (150 questionnaire to selected respondents. This data was then analyzed using statistical tools such as; Ordinary least square, Durbin Watson and P-value in EViews, which were also used for the interpretation of the hypotheses in this study. The result of the analysis shows that there was a significant relationship between fraud control and fraud prevention because their proxies considered in the study such as; Management control, internal audit, and whistleblowing showed a P-values of 0.0004 and 0.0001, which were lower than the 5% critical value specified in Eviews for this analysis. Based on this result, the study concludes and recommends among other things, that Management policies must be able to strengthen both fraud control and fraud prevention of commercial banks based on the sample study since the proxies of both variables have a significant effect on each other.

Gap in Literature:
Some literature on fraud used the total deposit as a proxy for the performance of Banks but nothing should explain performance in this regard than profit/profitability since these Banks are profit -driven. Proxies like Net Interest Margin, Return on Asset, Return on Equity among other profitability ratios should be used to proxy profitability while Profit Before/After Tax should be used to proxy profit. However, a blend of these parameters can be used as a dependent variable proxy for measuring fraud against performance. This work used ROE as a proxy for profitability as not many literatures examined it that way.
Theoretical Framework: Cressey Fraud Theory: Figure 1: Fraud Triangle Source: Kassem & Higson (2012) Every fraud transcends three factors were a non-shareable financial problem, opportunity to commit the trust violation, and rationalization by the trust violator (Kassem & Higson, 2012). As regards nonshareable financial problem, Cressey stated that: "[p]ersons become trust violators when they conceive of themselves as having incurred financial obligations which are considered as non-socially sanctionable and which, consequently, must be satisfied by a private or secret means" (Cressey, 1950). Cressey emphasized that window for perpetration of fraud is opened when fraudsters sees a hole via which fraud can be committed are damn sure that they will not be caught. (Kassem & Higson, 2012). As per rationalization, Cressey believed that most fraudsters are new to it; most times they do not have former criminal record but are pressured to commit fraud and would like to use personal expediencies to justify (Kassem & Higson, 2012 (2019) In 2004, Wolfe and Hermanson introduced the "Fraud Diamond Model" Wolfe and Hermanson believed many frauds would not ordinarily have occurred without the right person that possesses the right capabilities for implementing the details of the fraud. They put forward four observable signals to fraud which transcends; authoritative position/function within the organization, capacity to understand and exploit accounting systems and internal control weaknesses, confidence that she/he will not be detected when and if fraud is perpetuated and capability to deal with the stress created within a seemingly good person when she commits bad acts.  (Kassem & Higson, 2012).

Kranacher, Riley & Wells
Here, motivation is the inducements of fraud committers, and it is wrapped in Money, Ideology, Coercion and Ego. Motivation wrapped in MICE is what replaces the pressure angle of Cressey's Fraud Triangle Model; thus, before fraudsters commit fraud, there must be a motivation that is induced by MICE and an opportunity that makes the perpetrator to rationalize that by committing the act, the likelihood of being caught is slim or not there at all (Kassem & Higson, 2012). Also, (Dorminey, Fleming, Kranacher & Riley 2012) argued that two sides of the fraud triangle, pressure, and rationalization, cannot be easily observed; thus, the model cannot solve the fraud problem alone.  (2012) According to the New Fraud Triangle Model as presented by (Kassem & Higson 2012), the fraudsters capability is central to any fraud occurrence. They opined that for fraud to occur, there must be an incentive/inducement/motivator in the form of MICE and an opportunity aided by a weak control system; and a ready perpetrator whose personal integrity has been compromised (Kassem & Higson). Obviously, all fraud theories explained in this work takes their bearing from Cressey's Fraud Triangle Model, but, the New Fraud Triangle Model is upheld for this work as it centres on fraudsters capabilities.

METHODOLOGY:
Research Design: The data used are secondary and is quantitative by design as it requires rigorous data analysis.

Population of the Study:
The population of this study is based on the Fraud Frequency, Fraud Involvement, Fraud Loss, Fraud Amount and Return on Equity estimates of all Deposit Money Banks in Nigeria.

Sample Size and Sampling Techniques:
The sampling technique used in this work is purposive as it covers 10 years (2009 -18).

Technique of Data Analysis:
This study adopted the Regression Analysis using E-Views 9.     (Peck, Olsen, & Devore 2008), posits that; the clustering confirm if there is deviance from normality in data set, the statistics were between the ranges of (-3 and +3) which in itself suggest that the data emanated from a normally distributed series. In reinforcing this, the Jarque-Bera Statistics (J.B.S.) for normality test was not significant for all variables at 5%, thereby affirming that the distribution of the series is normal (Studenmund, 2000).

Durbin Watson (D.W.) Test:
For D.W. test, values in the range of 1.50 to 2.50 are relatively normal. What would be a cause for concern is value outside this range. According to (Field, 2009), values less than 1 or more than 3 are a definite cause for concern. Since DW Stat. is 2.4875 is between the range, it means the model is verified not to have autocorrelation and can be said to have satisfied the no serial correlation assumption. From the Breusch-Godfrey (LM) test for Serial Corelation, the null hypothesis stating that there is no serial correlation was accepted, F (2,3) = 0.0304, p > 0.05 (Studenmund, 2000). To test for serial correlation of the error term, it is said to be satisfactory when the hypothesis of no constant residual error in the model is rejected. The Breusch-Pagan-Godfrey test of heteroskecdacity was conducted to test the serial correlation of the error term. The result of the analysis revealed the absence of heteroskedasticity, F (4,5) = 0.5307, p > 0.05 (Studenmund, 2000). This implies that the residual error is constant in the series.  05. This is the basis for which the hypothesis that FL does not impact significantly on ROE will be accepted. Likewise, Fraud Amount (FA) has a t-stat. (-3.792060, -5.189156) = 0.012; p-value <0.05. This proves significant and prompts the rejection of the hypothesis that Fraud Amount (FA) does not impact significantly on the profitability of Deposit Money Banks (DMB's). In contrast; for Fraud Involvement (FI) on ROE: such that, DMB's t-Stat. (-1.135557, -0.078050) = 0.3076, p-value >0.05. This is the basis for which the hypothesis that FA does not impact significantly on ROE will be accepted. The coefficient of determination (R 2 ) of 0.914 implies that 91.4% of the variation in the dependent variable (ROE) is explicated by the independent variable (FF, FI, FL, and FA) while 8.6% were triggered by variables not included in the model. This affirms that the model is robust enough for quality decision modelling for fraud and profitability of Nigeria's Deposit Money Bank's for the period in focus (2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018). Similarly, the adjusted coefficient of determination ̅ 2 at 0.845 projects that 84.5% of the variation in the dependent variable is caused by variation in the independent variables while 15.5% of the variation in dependent variables is explained by other variables not captured in the model but explained by error term. The F-Stat. for the overall goodness-of-fit stood at F (13.22690) = 0.007195, p-value >0.05. Thus; one can posit that with the exemption of the intercept, it is not all the parameter in the model are statistically significant at 5% level.

DISCUSSION OF FINDINGS:
Our findings show that there exists a positive and significant relationship between Fraud and ROE during the period under focus. This study is in agreement with the works of (Nwankwo 2013;Adetoso & Akinselure 2016;Akinwumi Michael & Raymond (2017). They all project the impact of Fraud on Profitability/Performance of Banks to be significant.