Schemata and categories in the corporate knowledge-bases of expert portfolio managers

Cocks, Neil John. (1990). Schemata and categories in the corporate knowledge-bases of expert portfolio managers PhD Thesis, School of Business, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
THE6937.pdf Full text application/pdf 12.31MB 0
Author Cocks, Neil John.
Thesis Title Schemata and categories in the corporate knowledge-bases of expert portfolio managers
School, Centre or Institute School of Business
Institution The University of Queensland
Publication date 1990-01-01
Thesis type PhD Thesis
Total pages 237
Language eng
Subjects 1503 Business and Management
1502 Banking, Finance and Investment
Formatted abstract
This thesis investigates the content and structure of twelve expert equity portfolio managers' knowledge-bases. Chapter 1 argues that these corporate knowledge-bases are important because they influence equity portfolio selection decisions. Corporate knowledge is the focus of this thesis. Bouwman [1984] proposes that part of a portfolio manager's corporate knowledge-base is a set of typical company profiles (templates). These templates are schemata. The cognitive psychology approach taken in this thesis regards schemata as a way to manage a complex information processing task.

Chapter 2 discusses both categorisation and memory structure theory. An exploratory study in each area provides empirical evidence about portfolio managers' schemata, and the experimental approaches of use to investigate these. Free category sorting and multi-trial free recall were chosen as the two approaches to be used. It was proposed that managers use schemata to evaluate companies for portfolio selection and structure their memory of corporate information.

Chapter 3 determines those companies with which managers are familiar. This is the knowledge domain of interest to this thesis. Each of the 12 managers rated each of the 815 industrial companies then listed on the Australian Stock Exchange (ASX) on a 1 to 5 scale of familiarity. They are highly familiar with only 14% of these companies. The agreement between subjects is high (81%). The knowledge-base for these managers is thus selective. They specialise in about 100 companies. This is the first stage of their information reduction strategy. They attend only to the information relating to the companies of interest to them.

Chapters 4 to 6 contain the theoretical propositions, experimental design, hypotheses, results, and discussion for the three experiments on schemata and categorisation. These experiments focus on the companies with which managers are familiar. Each experiment operationalises the theoretical propositions and tests them as hypotheses. The first, in Chapter 4, investigates the content of portfolio manager's schemata. It is based on categorisation theory. Part A of this experiment requires managers to sort 60 company names into categories according to six specified bases. These bases were; primary, industry, management quality, fundamentals, price correctness, and trading intention. The consensus between subjects as to the company membership of the categories was high for the industry basis and low for the other five bases. They then described the attributes that companies in these categories shared. These also differed substantially. Only the primary and industry categories shared more than one attribute. Only those categories are schemata. In part B of the experiment managers sorted highly famiUar and moderately familiar companies into primary categories and subcategories. They discriminated more closely among the companies they were more familiar with.

Chapter 5, investigated whether experts can categorise companies according to investment news, and how each news item affects the companies' share prices. Three pieces of investment news were presented to subjects. Experts categorised companies according to investment news. These categories were not schematic for two reasons. First, the categories for both news items shared fewer than one attribute in addition to the one on which they were categorised. Second, the effect of the news on share price across categories was only different for the first news item. The categories for the second news item did not contain 'unlike' companies with respect to share price.

The third and final experiment, in Chapter 6, involves multi-trial free recall. By examining the memory recall of 15 company names. Chapter 6 investigates the use of any schemata identified in Chapter 4. Novices were compared to experts. Both groups clustered their recall. It is concluded that experts have an established memory structure for their company knowledge bases and novices do not. Expert portfolio managers organize companies by industry. They do not, however, simply adopt the reigning ASX industry categories. They have their own conception of the industry in which a company operates. The industries in which portfolio managers place companies influences the coding system adopted by the ASX, rather than the reverse. The structure novices use is only weakly associated with the ASX industry categories.

Chapter 7 summarises and integrates the thesis. It draws together the implications of the three experiments for portfolio management, outlines the limitations of the research, and indicates further profitable avenues for research.
Keyword Portfolio management
Investment analysis

Document type: Thesis
Collection: UQ Theses (RHD) - UQ staff and students only
Citation counts: Google Scholar Search Google Scholar
Created: Thu, 27 Jan 2011, 23:47:45 EST by Mr Brendan Shand on behalf of Social Sciences and Humanities Library Service