Enterprise large language models: Knowledge characteristics,risks, and organizational activities

Since the release of OpenAI's ChatGPT, there has been substantial interest in and concern about generative AI systems. This paper investigates some of the characteristics, risks, and limitations with the enterprise use of enterprise large language models. In so doing, we study the organizational impact, continuing a long line of research on that topic. This paper examines the impact on expertise, the organizational implications of multiple correlated but different responses to the same query, the potential concerns associated with sensitive information and intellectual property, and some applications that likely would not be appropriate for large language models. We also investigate the possibility of agents potentially manipulating the content in these large language models for their own benefit. Finally, we investigate the emerging phenomenon of “ChatBot Enterprise” versions, including some of the implications and concerns of such enterprise large language models.     

Predicting Corporate Bankruptcy: An Evaluation of Alternative Statistical Frameworks

Corporate bankruptcy prediction has attracted significant research attention from business academics, regulators and financial economists over the past five decades. However, much of this literature has relied on quite simplistic classifiers such as logistic regression and linear discriminant analysis (LDA). Based on a large sample of US corporate bankruptcies, we examine the predictive performance of 16 classifiers, ranging from the most restrictive classifiers (such as logit, probit and linear discriminant analysis) to more advanced techniques such as neural networks, support vector machines (SVMs) and “new age” statistical learning models including generalised boosting, AdaBoost and random forests. Consistent with the findings of Jones et al. ( 2015 ), we show that quite simple classifiers such as logit and LDA perform reasonably well in bankruptcy prediction. However, we recommend the use of “new age” classifiers in corporate bankruptcy modelling because: (1) they predict significantly better than all other classifiers on both the cross‐sectional and longitudinal test samples; (2) the models may have considerable practical appeal because they are relatively easy to estimate and implement (for instance, they require minimal researcher intervention for data preparation, variable selection and model architecture specification); and (3) while the underlying model structures can be very complex, we demonstrate that “new age” classifiers have a reasonably good level of interpretability through such metrics as relative variable importances (RVIs).     

DNA Mining and genealogical information systems: Not just for finding family ethnicity

The primary expected use of DNA and genealogy sites has been their ability to help users find their family, find their ethnicity and to help them connect with distant relatives. In so doing such sites help users to “learn more about themselves.” Such systems have also been proposed to have the broader goals of helping connect mankind and show people how their similarities are greater than their differences. However, the use of DNA and genealogy information recently turned away from just finding family connections, ethnicity and origins. Recently it was announced that the “Golden State Killer” had been caught using information generated from using DNA and consumer genealogical websites. This paper investigates some of the questions and unanticipated consequences raised by this alternative use of these technologies and their impact on individuals, organizations and society. As part of that analysis we analyze some of the immediate consequences on the firm from which the DNA information was gathered, the new emerging approach used by law enforcement, some privacy concerns and provide a network game formulation as a means to model user behavior. Finally, we examine some potential emerging research issues.     

Economists in the 2008 financial crisis: Slow to see,fast to act

We study the economics- and finance-scholars’ reaction to the 2008 financial crisis using machine learning language analyses methods of Latent Dirichlet Allocation and dynamic topic modelling algorithms, to analyze the texts of 14,270 NBER working papers covering the 1999–2016 period. We find that academic scholars as a group were insufficiently engaged in crises’ studies before 2008. As the crisis unraveled, however, they switched their focus to studying the crisis, its causes, and consequences. Thus, the scholars were “slow-to-see,” but they were “fast-to-act.” Their initial response to the ongoing Covid-19 crisis is consistent with these conclusions.     

Reimagining design science and behavioral science AIS research through a business activity lens

In this paper, we present a novel approach for reimagining the scope and impact of design science and behavioral science accounting information systems (AIS) research. We do so by first explicitly considering the broad impact of accounting on business functions. The proliferation of information technology throughout the organization coupled with the blurring of the lines between “accounting” and “business” activities has spawned a world where (technology-enabled) accounting has truly become the language of (technology-driven) business. Leveraging the International Standards Organization model of the phases of business activity, we highlight how utilization of information systems artifacts in each business activity phase yields a broad array of AIS research questions. Second, we encourage design science and behavioral science AIS research to work synergistically, such that the outputs of each paradigm inform the research conducted in the other paradigm. We suggest that a more purposeful integration of design science and behavioral science AIS research over time can improve the rigor and relevance of AIS research to advance knowledge in the field, amplify the impact of AIS research for our colleagues in both accounting and information systems, and improve the practical applicability of the research findings.     

Enterprise knowledge graphs with applications to continuous monitoring systems

Continuous monitoring systems generate continuous business intelligence by gathering text from a range of sources and typically storing that text information using snippets of text. Further, continuous monitoring systems derive information from those snippets, such as sentiment and new concepts emerging in the text, beyond the existing ontology and dictionaries. However, it is difficult for users to gather much additional information from large quantities of text snippets. As a result, this paper investigates introducing the notion of “knowledge graphs” into such systems and analyzes some of the potential applications and capabilities in business intelligence settings. As part of that analysis, this paper extends current-generation continuous monitoring systems to include knowledge graphs.     

From comfort to stretch zones: A field study of two multinational companies applying “beyond budgeting” ideas

This paper explores how change in the design principles of management control systems (MCSs) based on implementing the beyond budgeting (BB) ideas has influenced the transition of decision-makers from “comfort” to “stretch” zones and how this transition changed the supply of and demand for managerial information. This paper's starting point is based on the research evidence showing that there are many organizational problems associated with using budgets. Thus, this paper puts forth a previously neglected research context of companies that claim to have abandoned budgeting. In two cases, we illustrate how changes in the design of the MCS can create new management practices based on new ideas of information needed for decision-making. In particular, this paper illustrates how the use of new information provided by the MCS design, which is based on new principles, move decision-makers into the “stretch zone” characterized by new characteristics of decision-makers’ mindset and behavior. We also demonstrate how unbundling target setting, forecasting, and dynamic resource allocation enables better forward-looking and strategy-oriented decisions in situations requiring negotiation and learning.     

On the relationship between citations and appearances on “top 25” download lists in the International Journal of Accounting Information Systems

Citations from existing sources like the Social Science Citation Index (SSCI) have long been used to evaluate impact and importance of research. However, in addition to SSCI becoming digital, additional citation sources have been developed, such as SCOPUS and Google Scholar. Further, information now is available regarding papers that are among the most frequently downloaded, providing a new potential measure of impact and importance. This paper analyzes the use of digital and web citations, and newly available digital download information in the form of “top 25” most downloaded paper lists for the International Journal of Accounting Information Systems.This paper finds that any of the number of three types of “downloads” (abstract, paper and denied accesses) are highly correlated with each other. Further, in an analysis of the International Journal of Accounting Information Systems papers in “top 25” downloads, this paper finds that the number of citations and the number of times that a paper is in a “top 25” of downloaded papers are statistically significantly correlated. Finally, this paper finds that the set of “top 25” downloaded papers has a disproportionate number of citations.     

An analysis of the literature on systemic financial risk: A survey

This article presents an analysis of the literature on systemic financial risk. To that end, we analyze and classify 266 articles that were published no later than September 2016 in the databases Scopus and Web of Knowledge; these articles were identified using the keywords “systemic risk”, “financial stability”, “financial”, “measure”, “indicator”, and “index”. They were evaluated based on 10 categories, namely, type of study, type of approach, object of study, method, spatial scope, temporal scope, context, focus, type of data used, and results. The analysis and classification of this literature made it possible to identify the remaining gaps in the literature on systemic risk; this contributes to a future research agenda on the topic. Moreover, the most influential articles in this field of research and the articles that compose the mainstream research on systemic financial risk were identified.     

Accounting and environmentalism: An exploration of the challenge of gently accounting for accountability,transparency and sustainability

Concern for the natural environment has not occupied a prominent role in accounting scholarship and practice. This paper attempts to redress this omission by investigating the implications for accounting of placing the environment at the centre of the analysis. The paper introduces the principles of this “deep green” position and explores how accounting might articulate them. Whilst there may be no place for what we currently consider to be conventional financial accounting in any “green utopia” the paper does not attempt to operationalise, within a non-green world, certain of the principles of the green position. Emphasis is placed on an accounting's potential for contribution to accountability and transparency in participative democracy, the potential for non-financial accounts of the biosphere and, perhaps most contentiously, the use of current accounting techniques for the operationalisation of an accounting for sustainability.     

AN ALTERNATIVE TO HEDGING FOR THE CORPORATION: MANAGING A PORTFOLIO OF RISKS

Financial Risk Management for most corporations has now become a topic of discussion, if not an issue for active implementation. The issues of “Should One Hedge” and “How Should One Hedge” are often segregated and have tended to be so in the literature. Articles dealing with the hedge decision often ignore the implementation issues and vice versa. This article explores the use of Mean Variance portfolio optimization techniques in hedging corporate risks. It links the various sets of literature together to argue that for some firms using modern portfolio management techniques to manage corporate risks can prove advantageous. The article concludes with an analysis of a synthetic European firm exhibiting many of the risks and other qualities apropos a firm deciding to utilize MEAN-VARIANCE techniques to manage financial risks.     

Exploring corporate governance research in accounting journals through latent semantic and topic analyses

The literature on corporate governance (CG) has been expanding at an unprecedented rate since major corporate scandals surfaced, such as Enron, WorldCom, and HealthSouth. Corresponding with accounting's important role in CG, accounting scholars increasingly have investigated CG in recent years, so the body of literature is growing. Although previous attempts have been made to summarize extant literature on CG via reviews, none of these attempts has utilized recent developments in text analyses and natural language processing. This study uses latent semantic and topic analyses to address this research gap by analysing abstracts from 1,399 articles in all accounting journals that the Australian Business Deans Council (ABDC) has rated A and A*. The ABDC journal list is widely recognized as a journal‐quality indicator across many universities worldwide. The analyses revealed 10 distinct research topics on CG in the ABDC's top accounting journals. The results presented include the five most representative articles for each topic, as distinguished by topic scores. This study carries important practice and policy implications, as it reveals major research streams and exhibits how researchers respond to various CG problems.     

Management control as a system or a package? Conceptual and empirical issues

Despite the trend in investigating combinations of MC practices that form packages or systems, there is ambiguity about what is meant by a “control package” or “control system”. In this paper, we address questions related to MC as a package and as a system, both from a conceptual and empirical perspective. The purpose of our study is to clarify a number of issues, so as to guide future research in this area. We do not claim to have all the answers. Rather we make a number of assumptions explicit to show when it is important to address multiple MC practices simultaneously, why this is the case, and what the empirical implications are. Our research directly relates to the systems approach to contingency theory and the literature’s struggle to clearly define its most fundamental concept, i.e., “internal consistency”. Our analysis provides a formal definition of “internal consistency” and shows how it can be used to conceptualize MC systems, thereby giving contingency theorists the necessary ingredient to move forward.     

ACCOUNTING FOR INSECURITY

This paper argues that modern systems of budgetary control are implicated in the exploitation and production of insecure forms of employment. The flexibility of direct labour is assumed at a very basic level in some of the core techniques of costing and budgetary control. Previous historical studies of the development of these forms of control, moreover, have shown that they were used to shift the costs of economic fluctuation from capital to labour as well as to encourage the efficient utilisation of human effort. In the light of these observations, it is to be expected that the use of budgetary targets which incorporate direct labour costs will: (1) be more prevalent where the workers are least able to resist the various forms of “flexibility”; and (2) encourage recourse to redundancies where the performance of business units within a company falls below expectation. The paper then tests these hypotheses against data from a recent survey of industrial relations practice in large UK companies. The results show, firstly, that there is a strong and positive association between the proportions of females and part-timers within the workforce and the use of unit labour costs and the direct labour cost/sales ratio as performance targets. Secondly, the use of return-on-investment (R.O.I.) targets is associated with the declaration of redundancies in business units which have failed to perform satisfactorily. For those to whom insecurity of employment constitutes a social problem rather than a managerial convenience to be celebrated as “flexibility”, these findings indicate that the accounting control systems typical of the modern company constitute part of the problem. The achievement of the long-standing trade union aim of security of income and employment will depend, in part, on changing these systems of control.     

半强制股利政策与股权融资成本

在一些国家,强制股利支付是改善公司治理和弥补法律保护不足的重要手段,我国自2001年起陆续出台了类似的半强制股利系列政策。然而现有部分研究却发现,半强制股利政策可能会产生监管“悖论”。那么,事实是否如此？以往这些研究主要从监管成本角度来分析,可能忽视了监管带来的收益,我们认为虽然半强制股利政策提高了融资门槛,但也可能实现股东之间的利益共享,并有利于投资者形成稳定的股利预期,从而实现治理的“溢价”。本文利用2008年监管政策提供的良好自然实验机会,主要从半强制股利政策的治理效应角度来评估政策产生的经济后果。研究发现,从总体平均意义上看,半强制股利政策有助于降低受影响公司的股权融资成本。进一步研究发现,在代理成本高的公司,半强制股利政策的治理作用更为明显,存在一定的治理“溢价”。当然,半强制股利政策也存在一定的局限性,在公司的信息披露质量差和外部融资约束较大的公司,半强制股利政策的治理效应被削弱。     

Financial Systems, Corporate Finance and Corporate Governance

Financial systems of industrialised countries are usually classified as being either capital market- or bank-dominated. This paper tries to shed some light on the validity of this distinction. In order to do this, it analyses two specific roles of the financial sector vis-à-vis the enterprise sector: the role of financing or funding business and the role of capital markets and banks, respectively, in the context of corporate governance. Our analysis of the relevant findings in the literature shows that financing patterns are quite similar between countries, while corporate governance systems are quite dissimilar. The paper discusses how these divergent findings could be made compatible with existing theories and what implications they may have for the question of whether there is a tendency for corporate governance systems in different countries to become as similar as corporate financing patterns seem to be.     

Strategic complementarities and money market fund liquidity management

I use a unique institutional feature of money market funds to identify whether funds hold additional liquidity to guard against and prevent potential investor runs. Specifically, some funds are used as a cash management vehicle for related entities, such as other funds in the fund family. These “internal” funds should experience less outflows during market stress, and should thus have less need to hold this additional liquidity. Indeed, these “internal” prime money market funds do hold lower liquidity than other prime funds. This effect is most pronounced at quarter ends, when there is an exogenous reduction in cash demand from non-US bank dealers.     

基于EGARCH模型的黄金周对中国股市影响研究

现行国内外学者在股票市场发现“周末效应”、“节日效应”及“休市效应”等现象,但并未对这一现象给出合理性解释。本文针对中国独有的黄金周长假,以上证指数为例进行实证研究,发现中国股市存在效应为正的黄金周节前效应。这一结论验证了股票价格的异常波动与休市有关。结合中国股市的实际情况,得到需要“提高市场信息公开的透明程度”和“提升投资者自身的投资素质”两个启示。