Predicting accounting fraud using imbalanced ensemble learning classifiers – evidence from China

The current research aims to launch effective accounting fraud detection models using imbalanced ensemble learning algorithms for China A-Share listed firms. Based on a sample of 33,544 Chinese firm-year instances from 1998 to 2017, this research respectively established one logistic regression and four ensemble learning classifiers (AdaBoost, XGBoost, CUSBoost, and RUSBoost) by 12 financial ratios and 28 raw financial data. Additionally, we divided the sample into the train and test observations to evaluate the classifiers' out-of-sample performance. In detail, we applied two metrics, namely, Area under the ROC (receiver operating characteristic) curve (AUC) and Area under the Precision-Recall curve (AUPR), to evaluate classifiers' discriminability. In the supplement test, this study put forward an algebraic fused model on the basis of the four ensemble learning classifiers and introduced the sliding window technique. The empirical results showed that the ensemble learning classifiers can detect accounting fraud for the imbalanced China A-listed firms far more effectively than the logistic regression model. Moreover, imbalanced ensemble learning classifiers (CUSBoost and RUSBoost) effectively performed better than the common ensemble learning models (AdaBoost and XGBoost) in average. The algebraic fused model in the supplement test also obtained the highest average AUC and AUPR among all the employed algorithms. Our results offer firm support for the potential role of Machine Learning (ML)-based Artificial Intelligence (AI) approaches in reliably predicting accounting fraud with high accuracy. Similarly, for the Chinese settings, our ML-based AI offers utmost advantage in forecasting accounting fraud. Finally, this paper fills the research gap on the applications of imbalanced ensemble learning in accounting fraud detection for Chinese listed firms.     

Detecting Accounting Fraud in Publicly Traded U.S. Firms Using a Machine Learning Approach

We develop a state-of-the-art fraud prediction model using a machine learning approach. We demonstrate the value of combining domain knowledge and machine learning methods in model building. We select our model input based on existing accounting theories, but we differ from prior accounting research by using raw accounting numbers rather than financial ratios. We employ one of the most powerful machine learning methods, ensemble learning, rather than the commonly used method of logistic regression. To assess the performance of fraud prediction models, we introduce a new performance evaluation metric commonly used in ranking problems that is more appropriate for the fraud prediction task. Starting with an identical set of theory-motivated raw accounting numbers, we show that our new fraud prediction model outperforms two benchmark models by a large margin: the Dechow et al. logistic regression model based on financial ratios, and the Cecchini et al. support-vector-machine model with a financial kernel that maps raw accounting numbers into a broader set of ratios.     

The role of attribute selection in Deep ANNs learning framework for high‐frequency financial trading

In financial trading, technical and quantitative analysis tools are used for the development of decision support systems. Although these traditional tools are useful, new techniques in the field of machine learning have been developed for time‐series forecasting. This paper analyses the role of attribute selection on the development of a simple deep‐learning ANN (D‐ANN) multi‐agent framework to accomplish a profitable trading strategy in the course of a series of trading simulations in the foreign exchange market. The paper evaluates the performance of the D‐ANN multi‐agent framework over different time spans of high‐frequency (HF) intraday asset time‐series data and determines how a set of the framework attributes produces effective forecasting for profitable trading. The paper shows the existence of predictable short‐term price trends in the market time series, and an understanding of the probability of price movements may be useful to HF traders. The results of this paper can be used to further develop financial decision‐support systems and autonomous trading strategies for the financial market.     

DESIGNING AN IF–THEN RULES‐BASED ENSEMBLE OF HETEROGENEOUS BANKRUPTCY CLASSIFIERS: A GENETIC ALGORITHM APPROACH

This paper proposes a framework for an ensemble bankruptcy classifier that uses if–then rules to combine the outputs from a heterogeneous set of classifiers. A genetic algorithm (GA) induces the rules using an asymmetric, cost‐sensitive fitness function that includes accuracy and misclassification costs. The GA‐based ensemble classifier outperforms individual classifiers and ensemble classifiers generated by other methods. The results of the classifier are in the form of if–then rules. We apply the approach to a balanced dataset and an imbalanced dataset. Both are composed of firms subject to financial distress and cited in the US Securities and Exchange Commission's Accounting and Auditing Enforcement Releases. Copyright © 2014 John Wiley & Sons, Ltd.     

INTEGRATION OF RANDOM SAMPLE SELECTION,SUPPORT VECTOR MACHINES AND ENSEMBLES FOR FINANCIAL RISK FORECASTING WITH AN EMPIRICAL ANALYSIS ON THE NECESSITY OF FEATURE SELECTION

Financial risk forecasting (FRF) is an effective tool to help people forecast whether or not a company will fail in future. Among all techniques of FRF, the support vector machine (SVM) is the most newly developed, and one of the most accurate and effective techniques. This study is devoted to investigating an ensemble model of FRF by integrating bagging with an SVM to generate a data‐driven SVM ensemble. Bagging is used to produce diverse training datasets on which multiple SVM classifiers are trained to make FRF for a target company. Simple voting is employed to produce a final decision from the SVM model committee. The empirical study has two objectives. One is to verify whether the data‐driven SVM ensemble can produce a more dominating performance than the most frequently used techniques in the area of FRF, i.e. multivariate discriminant analysis, logistics regression and a single SVM. The other is to verify whether feature selection is necessary to help the SVM make more precise FRF, although the SVM can handle high‐dimensional data. The results indicate that the data‐driven SVM ensemble significantly improves the predictive ability of SVM‐based FRF. Meanwhile, feature selection can effectively help the SVM achieve better predictive performance, which means that use of feature selection is necessary in SVM‐based FRF. Copyright © 2012 John Wiley & Sons, Ltd.     

Evaluating interpretable machine learning predictions for cryptocurrencies

This study explores various machine learning and deep learning applications on financial data modelling, analysis and prediction processes. The main focus is to test the prediction accuracy of cryptocurrency hourly returns and to explore, analyse and showcase the various interpretability features of the ML models. The study considers the six most dominant cryptocurrencies in the market: Bitcoin, Ethereum, Binance Coin, Cardano, Ripple and Litecoin. The experimental settings explore the formation of the corresponding datasets from technical, fundamental and statistical analysis. The paper compares various existing and enhanced algorithms and explains their results, features and limitations. The algorithms include decision trees, random forests and ensemble methods, SVM, neural networks, single and multiple features N-BEATS, ARIMA and Google AutoML. From experimental results, we see that predicting cryptocurrency returns is possible. However, prediction algorithms may not generalise for different assets and markets over long periods. There is no clear winner that satisfies all requirements, and the main choice of algorithm will be tied to the user needs and provided resources.     

Corporate performance and firm perception: The British experience

Previous academic research has presented a theoretical basis for a relationship between attributes of a firm's reputation and its financial performance. For the United States, researchers have analysed the correspondence between market and accounting based measures of US firm performance and external evaluators' perceptions of the qualitative attributes of US firms. In this study, expert surveys on the qualitative performance of British firms conducted by the British publication, the Economist, which are similar in content to surveys conducted by Fortune magazine for US firms, are used to determine the correspondence between qualitative and quantitative measures of British firms' performance. Results indicate that differences may exist between US and Britain in the use of qualitative survey data on a firm's strategic attributes as a forecast of a firm's future quantitative performance measures. Results also indicate that for small firms, certain qualitative factors (e.g. capacity to innovate) may be of greater importance in forecasting accounting and security market returns.     

A predictive system integrating intrinsic mode functions,artificial neural networks,and genetic algorithms for forecasting S&P500 intra‐day data

There is an abundant literature on the design of intelligent systems to forecast stock market indices. In general, the existing stock market price forecasting approaches can achieve good results. The goal of our study is to develop an effective intelligent predictive system to improve the forecasting accuracy. Therefore, our proposed predictive system integrates adaptive filtering, artificial neural networks (ANNs), and evolutionary optimization. Specifically, it is based on the empirical mode decomposition (EMD), which is a useful adaptive signal‐processing technique, and ANNs, which are powerful adaptive intelligent systems suitable for noisy data learning and prediction, such as stock market intra‐day data. Our system hybridizes intrinsic mode functions (IMFs) obtained from EMD and ANNs optimized by genetic algorithms (GAs) for the analysis and forecasting of S&P500 intra‐day price data. For comparison purposes, the performance of the EMD‐GA‐ANN presented is compared with that of a GA‐ANN trained with a wavelet transform's (WT's) resulting approximation and details coefficients, and a GA‐general regression neural network (GRNN) trained with price historical data. The mean absolute deviation, mean absolute error, and root‐mean‐squared errors show evidence of the superiority of EMD‐GA‐ANN over WT‐GA‐ANN and GA‐GRNN. In addition, it outperformed existing predictive systems tested on the same data set. Furthermore, our hybrid predictive system is relatively easy to implement and not highly time‐consuming to run. Furthermore, it was found that the Daubechies wavelet showed quite a higher prediction accuracy than the Haar wavelet. Moreover, prediction errors decrease with the level of decomposition.     

Financial clusters,industry groups,and stock return correlations

Industry classifications are used by investors, economists, and policy makers for a great variety of purposes. The traditional economic‐activity‐based systems (Global Industry Classification Standard, North American Industry Classification System, Standard Industrial Classification, and Fama–French) have been supplemented in recent years by alternative classification systems. Our purpose is to provide another alternative system that forms classification groups based on the structure of firm financial statements. Using cluster analysis, a multivariate tool that forms groups where their characteristics are similar within groups and distinct across groups, we form clusters of large U.S. firms based on their common‐size financial statements (percentage breakdowns of balance sheets and income statements). We characterize the financial clusters based on their industry classifications and other economic information and assess the ability of financial clusters and industry groups, separately and jointly, to explain stock return correlations of all pairs of firms. Our results demonstrate that using financial clusters and industry groups together proves advantageous relative to using either alone.     

How to classify failures when collecting data for safety-instrumented systems in the oil and gas industry

The oil and gas industry places a high value on achieving high reliability and availability on safety critical equipment. To achieve this, assessments of the reliability performances of such equipment are required, both before and during the production phase. The fact that the reliability data available to support the assessments is often sparse or insufficiently detailed presents a challenge. These assessments also typically require insights into the system in which the equipment is used and information about failure detection. However, this ensemble information is often difficult to achieve in the way the data are collected today. As a response to this challenge, one suggested option is to collect reliability data using one acknowledged failure mode classification specifically designed to assess the reliability of safety-instrumented systems. This is a classification adopted from the International Electrotechnical Commission standard 61508. In this article, we discuss the pros and cons of adopting this failure mode classification in generic reliability data collection in the oil and gas industry. One argument discussed is that the data may lack relevant information about the associated safety system and thus be valid for a specific system only, not for generic equipment and systems in general. Hence, should the classification be implemented, the collected data should be used with care.     

Comparing numerical data and text information from annual reports using self-organizing maps

More and more companies provide their accounting information in electronic form today. The accounting information in electronic form can be found in large commercial databases or on the web. This information is of great interest for different stakeholders, i.e., stockholders, creditors, auditors, financial analysts, and management. For the stakeholders it is important to be able to extract both quantitative and qualitative information concerning the companies they are interested in. The annual reports contain information both in numerical and symbolic form. So far, only the numerical information has been analyzed with help of computers. However, technology has evolved and in particular neural networks in the form of self-organizing maps (SOMs) provide a new tool for analyzing also text information. In this paper, we compare results on quantitative data with results on qualitative data from annual reports. We use smart encoding, SOMs, and document histograms for comparing the performance of forest companies worldwide. Firstly, we cluster the companies according to, on the one hand, quantitative information, and on the other hand, qualitative information. Secondly, we compare the results produced by the clustering methods. Our results of the comparison show that there is a difference between the results.     

Integrating Sustainability Into Capital Markets: Bloomberg LP And ESG's Quantitative Legitimacy

The rise in prominence of environmental, social, and governance data has been driven in large part by a growing interest among investors who seek to gain an edge through the incorporation of such data in their investment decision‐making. There are, however, several significant obstacles to the integration of ESG data into mainstream investing analysis. Perhaps most important, while finance today is a fundamentally quantitative discipline, ESG is often qualitative. Moreover, the ESG data that is available is incomplete and inconsistent, due largely to a reliance on voluntary reporting by individual companies. In short, ESG has not yet earned its quantitative legitimacy in the eyes of the investor community. Nevertheless, recent work in the area of stranded asset values has provided Bloomberg LP, a leading provider of financial data and analytics, an opportunity to “bridge theory and practice” by translating the stranded assets framework into a first‐cut valuation tool designed for mainstream financial analysts. The tool offers a quantitative introduction to an ESG issue that the authors believe will eventually become an important focus of many investment decision‐makers' analysis. While the tool continues to evolve in analytical sophistication, the authors “preview” it here in its early form as one step towards Bloomberg's broader vision of “sustainable finance,” and the company's role in supporting the quantitative maturation of ESG through the twin engines of standardization and disclosure.     

INSOLVENCY PREDICTION IN THE PRESENCE OF DATA INCONSISTENCIES

In this paper we use data inconsistencies as an indicator of financial distress. Traditional models for insolvency prediction normally ignore inconsistent data, either by removing or replacing it. Instead of removing that information, we propose a new variable to capture it; using it together with traditional accounting variables (based on financial ratios) for the purpose of insolvency prediction. Computational tests use three datasets based on the financial results of 2033 Brazilian Health Maintenance Organizations over 7 years (2001 to 2007). Sixteen classification methods were used to evaluate whether or not the new variable impacted solvency prediction. Tests show a statistically significant improvement in classification accuracy – average results improve 1.3 (p = 0.003) and 1.8 (p = 0.006) percentage points, for 10‐fold and leave‐one‐out cross‐validations respectively. In addition, the analysis of false positives and false negatives shows that the new variable reduces the potentially harmful misclassification of false negatives (i.e. financially distressed companies being classified as financially healthy) and also reduces the estimated overall error rate. Regarding the extensibility of the results, even though this work uses data from Brazilian companies only, the calculation of the financial ratios variables, as well as the inconsistencies, could be extended to most companies worldwide subject to governmental accounting regulations aligned with the International Financial Reporting Standards. Copyright © 2014 John Wiley & Sons, Ltd.     

Retail Financial Advice: Does One Size Fit All?

Using unique data on Canadian households, we show that financial advisors exert substantial influence over their clients' asset allocation, but provide limited customization. Advisor fixed effects explain considerably more variation in portfolio risk and home bias than a broad set of investor attributes that includes risk tolerance, age, investment horizon, and financial sophistication. Advisor effects remain important even when controlling flexibly for unobserved heterogeneity through investor fixed effects. An advisor's own asset allocation strongly predicts the allocations chosen on clients' behalf. This one‐size‐fits‐all advice does not come cheap: advised portfolios cost 2.5% per year, or 1.5% more than life cycle funds.     

Empirical Evidence on the Value of Group‐Level Financial Information in Insurer Solvency Surveillance

The existing empirical research on insurer insolvency relies almost exclusively upon individual insurance company financial data, even though the insurance industry is dominated by group‐affiliated firms. This is the first study to evaluate the benefit of using group‐level data to predict insurer insolvencies for group‐affiliated insurers. The study uses financial ratios from the NAIC FAST scoring system, measured at both the company level and group level, as potential predictor variables. The results indicate that group‐level financial information substantially improves the predictive power of an insolvency prediction model relative to a model that uses only the analogous company‐level variables. In fact, the group‐level variables are found to often be substantially more powerful than company‐level variables in predicting individual insurer insolvencies. These results suggest that future insolvency analysis should, whenever feasible, include group‐level information to obtain higher predictive accuracy.     

Choice and change of measures in performance measurement models

This paper uses management control, resource-based, systems-based and contingency-based strategy theories to describe a large U.S. manufacturing company's efforts to improve profitability by designing and using a performance measurement model (PMM). This PMM includes multiple performance measures relevant to its distribution channel for products, repair parts and maintenance services. The PMM is intended to reflect the company's understanding of performance relations among strategic resources, operational capabilities, and desired financial outcomes. The PMM also reflects its intended distribution strategy, the types of performance necessary to achieve that strategy by its distributors, and its desired financial outcomes. Furthermore, the company uses the model to evaluate its North American distributors and intends to use these evaluations as a partial basis for annual and long-term rewards. Thus, the PMM embodies the measurable portion of the firm's management control system of its distribution channel.The study addresses four research questions: (1) Are measure attributes important considerations for performance measure choice? (2) Does the importance of attributes differ according to firm strategy? (3) Does the importance of attributes for design and use differ according to firm strategy? (4) Does a company trade-off some individual attributes for others? The questions are investigated using qualitative and quantitative analyses of archival documents and interviews with top managers and distributors. Principal findings are that measure attributes are important considerations for choice and change of performance measures, design attributes are more important than use attributes, the importance of attributes does not appear to differ according to strategy, and some individual attributes are traded-off for other attributes.     

投资项目绩效审计评价指标体系与框架设计研究

由于投资项目绩效审计的对象千差万别,目前尚没有一套科学统一的投资项目绩效审计评价指标体系及评价标准。现有的投资项目绩效审计评价指标体系,在定性指标与定量指标之间的转化上以及对投资项目滞后性绩效评价的关注尚存一些不足。由此本文充分考虑定性与定量指标的一致性及滞后性绩效评价等问题,在问卷调查和统计分析的基础上,以5E为核心设计了一个投资项目绩效审计评价的基本框架,由四部分组成:5E属性层、11个一级指标、43个二级指标层和标准层,并就指标体系的程序及测评方法,结合一个建筑投资项目绩效审计评价的具体实例进行了说明。通过对该建筑项目的绩效审计评价证明了所设计的评价指标体系具有较强的可操作性。     

Intelligent Order Matching Systems for Commodity Markets

Automatic order matching systems have emerged as an electronic alternative to traditional markets. In current automatic order matching systems, price and quantity are the only product dimensions used for the order matching. However, a single-commodity market is made up of many heterogeneous goods which are close to each other but different in qualities and delivery conditions. Price and quantity are important but represent only parts of product attributes that commodity traders want to take into account. This study aims to extend current automatic order matching systems by diversifying product dimensions. An intelligent order matching system not only maximizes the total transaction volume based on the price and quantity but also satisfies traders' qualitative preferences over attributes other than price and quantity. The intelligent order matching mechanism combines an economic model with a preference model to incorporate both quantitative and qualitative utility of market participants. Constraint logic programming is investigated as a new information technology to structure and implement the intelligent order matching system.     

BEYOND THE NUMBERS: MINING THE ANNUAL REPORTS FOR HIDDEN CUES INDICATIVE OF FINANCIAL STATEMENT FRAUD

Unlike previous fraud detection research, a vast majority of which has focused primarily on the use of quantitative financial information to predict fraud, in this study we examine qualitative textual content in annual reports to predict fraud and see whether there are discernible differences in the writing and presentation style between companies that committed fraud and those that did not. We believe that while numeric financial information in the annual reports can hide details of fraud, textual information relating to writing and presentation styles in such reports provides valuable clues pertaining to the existence of fraud. In this study we use the chi‐square test to analyse our data and test hypotheses about predictors of fraud that may explain linguistic feature variations in fraudulent and nonfraudulent annual reports. We provide new results on the usefulness of the qualitative content of annual reports in detecting fraud. Copyright © 2012 John Wiley & Sons, Ltd.