Maurice Peat

Detecting and modelling nonlinearity in flexible exchange rate time series

The aim of this paper is to examine the appropriateness of nonlinear time series analysis as a framework in which to model the dynamics of exchange rates. This aim has been motivated by the questioning of the power of classical unit root tests, the accumulating amount of evidence which suggests that exchange rates follow some kind of nonlinear process, and the fact that standard asset pricing theories do not explain well the empirical observations of exchange rate movements. The paper has three major objectives. First, to test for the presence of unit roots in nominal exchange rate time series. Second, for those nominal exchange rate time series found to be stationary, to test for nonlinearity using both tests derived without a specific nonlinear alternative in mind and tests against a specific nonlinear model. Finally, we motivate the types of nonlinearity for which we test by examining a recently proposed nonlinear model of exchange rate dynamics.

Using Neural Nets To Combine Information Sets In Corporate Bankruptcy Prediction

We demonstrate that the use of a neural network (NN) model to combine information from corporate financial statements and equity markets provides improved predictive estimates of the probability of corporate bankruptcy. Using performance measures, based on the receiver operating characteristic curve, the forecast combinations from the NN models are demonstrated to outperform the forecasts derived from a forecast combination generated using a logistic regression approach. This result provides support for the use of forecast combinations generated from NN models in the estimation of corporate bankruptcy probabilities as it outperforms the standard approach of forming a hybrid forecasting model which includes all the explanatory variables. Copyright

The impact of auctions on residential property prices

This study assesses whether the sale method in residential real estate markets – auction versus private treaty – is a determinant of sale price. Utilising a larger and richer dataset than previous research, we test for a price effect in auction sales in Sydney and Christchurch. When self‐selection biases are corrected for, using two‐stage hedonic regression analysis and a matched sampling procedure, we find no significant difference between prices of properties sold at auction to those sold by private treaty. This conflicts with the conclusions of previous research in the Australian and New Zealand housing markets, which have documented a price premium associated with auction sales.

Does Technical Analysis Beat the Market? – Evidence from High Frequency Trading in Gold and Silver

Previous research has identified that investors place more emphasis on technical analysis than fundamental analysis, however the research has largely been confined to daily data and stock market indices. This paper studies whether intraday technical trading rules produce significant payoffs in the gold and silver market using three popular moving average rules. We find that using the standard parameters previously used in the literature, technical trading rules offer are not profitable. However after utilising a universe of parameters, we find a number of parameter combinations offer significant profits in the gold market, but there remains no significant payoff in the silver market. Our results show that parameters that use longer histories are more successful than the traditional parameters chosen in the literature. Intraday technical trading rules can be profitable in the gold market but offer no significant profit in the silver market.

A Case Study in Using ADAGE for Compute-Intensive Financial Analysis Processes

The Ad hoc DAta Grid Environment (ADAGE) has been proposed as a framework to support analysis processes for large repositories of ad hoc data. Its use of a service-oriented architecture (SOA) brings the promise of flexibility, as well as enabling domain experts to define their own analysis processes at a high level of abstraction. However, these claims have not been verified empirically and the performance penalty of using additional abstract software layers has not been assessed on complex problems. This chapter describes a case study involving a realistic analysis process conducted by an expert user. It assesses the benefits and drawbacks of using the ADAGE approach versus conventional manual analysis processes. This chapter also outlines some avenues for future research to address existing limitations.

Data + Information Systems = Financial Innovation

Finance and Computing are linked in a symbiotic relationship, in this paper this relationship and its implications for both finance and computing will be explored. The outcome of the exploration is a suggested agenda for the development of financial computing, which will support a broad view of financial innovation. This agenda involves the support and enhancement of standards setting exercises and the adoption of services computing as the approach for a wide range of process innovation.

Stylized facts of intraday precious metals

This paper examines the stylized facts, correlation and interaction between volatility and returns at the 5-minute frequency for gold, silver, platinum and palladium from May 2000 to April 2015. We study the full sample period, as well as three subsamples to determine how high-frequency data of precious metals have developed over time. We find that over the full sample, the number of trades has increased substantially over time for each precious metal, while the bid-ask spread has narrowed over time, indicating an increase in liquidity and price efficiency. We also find strong evidence of periodicity in returns, volatility, volume and bid-ask spread. Returns and volume both experience strong intraday periodicity linked to the opening and closing of major markets around the world while the bid-ask spread is at its lowest when European markets are open. We also show a bilateral Granger causality between returns and volatility of each precious metal, which holds for the vast majority subsamples.

A Software Architecture for Enabling Statistical Learning on Big Data

Most big data analytics research is scattered across multiple disciplines such as applied statistics, machine learning, language technology or databases. Little attention has been paid to aligning big data solutions with end-user’s mental models for conducting exploratory and predictive data analysis. We are particularly interested in the way domain experts perform big data analysis by applying statistics to big data with a focus on statistical learning. In this paper we compare and contrast the different views about data between the fields of statistics and computer science. We review popular analysis techniques and tools within a defined analytics stack. We then propose a model-driven architecture that uses semantic and event processing technologies to achieve a separation of concerns between expressing the mathematical model and the computational requirements. The paper also describes an implementation of the proposed architecture with a case study in funds management.

A Proposed Framework for Evaluating the Effectiveness of Financial News Sentiment Scoring Datasets

The impact of financial news on financial markets has been studied extensively. A number of news sentiment scoring techniques are being widely used in research and industry. However, results from sentiment studies are hard to interpret contextual and sentiment related parameters change. Sometimes, the conditions which lead to the results are not fully documented and the results are not repeatable. Based on service-oriented computing principles, this paper proposes a framework that automates the process of incorporating different contextual parameters when running news sentiment impact studies. The framework also preserves the set of parameters/dataset and conditions for the end user to enable them to reproduce their results. This is demonstrated using a case study that shows how end users can flexibly select different contextual and sentiment related parameters and conduct news impact studies on daily stock prices.

Deteriorating Complexity in Gold Returns: Evidence from the Compass Rose

The compass rose pattern in financial data may indicate the presence of a nonlinear, possibly chaotic, data generating mechanism. Analysis reveals that over four equivalent subperiods, from 1996 to 2015, the compass rose pattern in gold returns fades. This feature provides an opportunity to establish which of several data characteristics is consistent with the observed deterioration in the compass rose pattern. The observed monotonic change in pattern quality, the number of different price moves and the number of zero returns are all consistent with a changing trading environment that has led to improvements in financial market efficiency. We also conclude that it is unlikely that the compass rose pattern is the product of an underlying nonlinear structure, since there is evidence of nonlinearity in all time periods, even those where the compass rose pattern in not evident.