Cygron's DataScope gives organizations the power to monitor, analyse and report on financial information and use that information to drive strategic decisions and minimize financial risk. In this case study we show how DataScope can be used to undertake the data analysis underpinning a complex investment strategy.

Fig. 1: Visualising the results of screening

Clear and insightful analysis of company fundamentals is essential for making strategic investment decisions and Cygron's DataScope provides a powerful and flexible environment for achieving this. Employing advanced technology to graphically portray complex, multi-dimensional information, DataScope brings powerful data analysis and decision support to the desktops of financial analysts. In this case study, we show how DataScope can be used by an analyst to implement a complex investment strategy which combines the principle of "Growth at a reasonable price", or GARP, with the analyst's own investment heuristics.

We take as our starting point an extensive list of the fundamentals of over 17,000 US-listed stocks. Our goal is to identify the top ten companies most suitable for investment.

GARP follows the principle that a Price:Earnings ratio is reasonable if it is less than or equal to the company's annual rate of earnings growth. In the first phase of our case study we apply this and three other screens to eliminate poor investment opportunities - at each step we utilize DataScope's powerful visualization to investigate and validate the results.

Screening for Profitability, Earnings Growth and Stock Valuation Ratios

We begin by looking for companies that are growing and have EPS (Earnings Per Share) growth of at least 20% over the last 3 years. Easily implemented using DataScope's data import wizard, this screen reduces the number of possible investments to a little over 1,700 companies. We continue by applying screens to eliminate further companies. Unlike the first screen, we implement these using DataScope's expression-builder and visual querying capability. Boolean expressions are built to indicate when the five-year average return on equity is less than 20% above the five-year return on investment, when EPS growth exceeds sales growth over the past 5 years, and when the Price:Earnings ratio is lower than the rate of earnings growth. Applying these screens shrinks the number of suitable companies to about 300.

Although investment screens are often applied in a mandatory fashion it is important to validate their applicability, this is where DataScope's extensive visualisation capabilities prove invaluable. Interactive 2D and 3D charts display the key attributes of all companies, accepted companies appear in red and rejected companies in green (see Fig.1). The charts are synchronized so that any company identified in one chart can be easily be tracked across all other charts.

Fuzzy investment heuristics

Beyond the screens described above, there are many other criteria to be considered in our stock selection process. These are somewhat fuzzy criteria and cannot be implemented as simple screens. We briefly describe three:

• Moderate institutional interest: We look for companies where institutional investors have begun to show interest, but there remains room for more.
• Volatility: We look for companies that are undervalued but historically have been volatile, this may indicate a greater likelihood of short term gains in share value.
• Recent short interest: We view companies subjected to moderate short-selling positively since an upswing in stock price is often magnified by investors buying high to cover their short positions.

Fig. 2:The ranked list of screened companies

These criteria are fuzzy since it is difficult to set absolute hard boundaries on their ideal values and it is also unlikely that we can simultaneously satisfy all of them. Datascope's multi-criteria decision support tool is ideally suited for this type of "soft" decision-making. Using this tool we build a decision model by specifying our decision-making criteria, specifying the desired range of values for each and assigning their relative importance. We can arrange the criteria into a hierarchy to facilitate more complex decision-making.

Fig. 3: Further visual analysis of ranked companies prior to the final investment decision

Applying the decision model to the remaining companies generates an aggregate score for each company according to our specified preferences. We can view the companies ranked according to this score and this ranked list can become the basis of our investment decisions (see Fig. 2). DataScope also provides visual tools to enable individual comparisons to be made between the ranked companies.

Saving the rankings or aggregate scores into a new database field allows them to be overlaid, as color, shape, size etc., onto the 2D and 3D charts (see Fig. 3). This enables us to see where our best stock picks lie in relation to the unselected companies. A further visual query allows us to select and visualise the ten top-ranked companies - we have now achieved our goal!

This case study describes just one of many analysis scenarios possible with DataScope. DataScope also contains many additional features, not described in this case study, which can further facilitate advanced data analysis, including powerful clustering and relationship finding algorithms.