« Neural Networks in Econometric System Identification, Forecasting and Control | Blog home | Report on the Multilevel/mixed Models Using Stata workshop »

We are delighted to proclaim the conclusion of another very successful event at the Faculty of Economics and Business. Each year the workshop is growing and attracting more national and overseas contributions.

The Third Meeting on MEAFA, 21-22 Jan 2010, fulfilled the aspirations and had the honour to host multi-disciplinary debates over key issues in a vast range of specialties, including accounting, finance, bankruptcy probability, market micro-structure, portfolio management, statistics, economics, and political science.

Across the two days, the discussion comprised many contemporary questions, such as the relationship between accounting numbers and market value, a defense of positive accounting research, the disclosure incentives in market micro-structure, the use of logit models to forecast takeovers, etc.

MEAFA would like to thank everyone for the attendance and especially for their invaluable multi-disciplinary contributions that covered the broad spectrum of MEAFA's research area.

Special thanks extend to the six international speakers (in alphabetical order):

Prof Michael Bradbury, Massey University, NZ
Prof Paul Dunmore, Massey University, NZ
Prof Stephen Penman, Columbia University, US
Dr Ivana Raonic, City University, UK
Prof Mark Tippett, Loughborough University, UK
Prof Roger Willett, University of Otago, NZ

Prof Roger Willet started the workshop presenting two papers. The first discussed the consequences of using mis-specified additive models of the relationship between accounting numbers and market value and draw particular attention to the importance of the assumption of ‘homogeneous’ firm parameters in cross-section estimation. On his second paper, Prof Willet explained how a fixed effects panel analysis can be used to approximate the average data generating process of the firms in the sample, and pointed out the slightly systematic impact of accounting variables compared to the autoregressive component in market value. Following the presentation, Prof Mark Tippet reviewed the important implications of the paper and brought out the thorny issue of the efficiency or otherwise of their parameter estimation procedures. The discussant concluded suggesting that the model’s inability to account for negative earnings and/or book values is a significant issue and it still unclear from the paper how such companies can be handled.

Dr Maurice Peat initiated the presentation reviewing the bankruptcy forecasting literature in the past decades and afterwards deliberated about what makes a ‘good’ probability forecast and its economic value. Dr Peat presented the applications of the method of estimation probability prediction model to Ohlson Models and concluded that the probabilities produced by this style of models are apparently not of themselves “bankable” or worth their cost.

Dr Demetris Christodoulou offered an explanation of the dynamic relationship between accounting earnings and stock market returns, including the speed with which market returns are reflected in earnings within the current period and in the long run. Dr Christodoulou presented a unified analytical framework within which previous specifications are encompassed as special case models and can be used to evaluate the properties of earnings in a dynamic panel data context. Prof Penman discussion's report commented on the usefulness of generalised model and its econometric power but also highlighted certain issues that need to be addressed. The most notable include: (i) the integration of economic theory consistent with the generalised framework, (ii) demonstration of how the model can be reduced to reflect special case accounting approaches and not just to analytical models and (ii) mapping to the Easton, Harris and Ohlson (1992) framework.

Prof Paul Dunmore debated the positive accounting in the broader sense of a research program which aims at developing causal explanations of human behaviour in accounting settings. Prof Dunmore targeted the common problems related to positive research, such as the casual construction of theoretical models to be tested, undue reliance on the logic of hypothesis testing, a lack of interest in the numerical values of parameters, etc. Prof Dunmore’s concluding remarks emphasised how positive research seems largely incapable of achieving the scientific objectives and its currently practice in accounting.

Prof Mark Tippet employed an empirical analysis to show that the probability distributions implied by the Schrödinger model are strongly compatible with the returns earned by the individual components of the London Stock Exchange’s FTSE All Share Index. Prof Tippet explained how the Schrödinger equation can be used to determine the wave functions of investment portfolios, and finalised arguing about the inverse relationship between the variance associated with the momentum in portfolio returns and the variance of the portfolio returns process itself. Prof Paul Dunmore opened the discussion reflecting on the momentum of a quantum particle and its equation. Prof Dunmore recommended an expansion of the momentum of the stock return and defended that the momentum should be related to an observable property of returns. Prof Dunmore concluded the debate suggesting the use of a new model with asymmetric exponential, but with non-Gaussian tails, and questioning the progress using Fokker-Planck equation and conventional mean-variance arguments.

Dr Ivana Raonic introduced the concept of transparency associated with trading stocks and commented on the different interpretations of transparency in the literature. Dr Raonic identified that, in the case of thinly traded stocks, the incentives for firms to improve disclosure of their market position and prospects depends critically upon the market microstructure. Dr Raonic extended the question suggesting that the benefits from increased disclosure may only be captured by the firms if an appropriate trading mechanism is simultaneously put in place. The comments of Prof Roger Willet and Dr Michael Falta on the paper outlined some general points regarding the structure and development of the paper that need to be improved. Prof Willet were particularly concerned with the methods used and the statement of the hypothesis.

Dr Maxwell Stevenson started the presentation explaining the probabilistic regression models adopted to predict takeover targets, and in special those that use financial statement variables. Dr Stevenson analysed the takeover market in Australia and empirically demonstrated that investors who are able to accurately predict firms that will be the subject of a takeover attempt should be able to earn these excess returns. Dr Richard Gerlach discussed the methodology applied in the paper and warned about the consequences from multicolinearity in accounting variables (that is a matter of construct and not empirical sampling). The argument attracted the contribution of the audience that also noted the crisis and currency factors over the years that may affect Dr Stevenson's results.

Prof Graeme Dean wrapped the workshop with an interesting discussion about the ‘Unresolved Methodological Questions at the Cross-section of Accounting and Finance’ of Ray Chambers, which counted with the participation of the audience. Prof Dean pointed out that the methods of research into accounting are much broader than the narrow positivist allegiance, extending upon primarily the works of Chambers. Prof Dean concentrated in particular on the arguably vacuous attempts to categorise accounting theory and research as either positive or normative.

About the Blog

MEAFA is a cross-disciplinary research group that promotes methodical analytical work with empirical applications in financial analysis.