Operational risk is intrinsic to financial institutions and it is commonly agreed that it is much harder to quantify and model than either market or credit risks.  Interface design as an operational risk issue, holds this assertion true when configuring screens during integration just as much as when designing screens for a new build.

The opportunity to design completely new systems is a rare privilege and the responsibility is acute because the consequences of design decisions continue to be felt for the lifetime of the application. But even when configuring a screen layout for the integration of an established tool, the consequences are lasting. So the operational risk issue bundled up with interface design has real duration and is worth more consideration than it is usually afforded.

The specifics of this issue are relatively straight-forward. Good interface design makes it hard for a system operator to make mistakes without delaying operations. Bad interface design makes it easy to make mistakes. Truly awful design results in an interface that seems to guide the user into a mistake.

In financial systems, interface design revolves around two pervasive manifestations. The first is the table or grid—row upon row financial data, flashing with updates. The design pressure is to get us much on the screen as possible, but it is a pressure worth resisting. Figure 1 shows the result of capitulating; a screen over-filled with text that is too small to read. Grid data does not need to be presented so horribly. Figure 2 shows a similar grid structure, with far larger text, laid out with space and clarity. The contrast between these two is not simply a matter of aesthetics. In the first example, the chances are far too great of choosing the wrong line, or of not noticing something. If the overlooked line is an order that should be executed quickly, there will be compensation to pay for the error.

The second most common interface element is the screen form, and the most risk-laden instance of the form is the order ticket. With the order ticket, you have a simple way of assessing the design of any financial system. In far too many instances you have to really look hard to determine whether a given order is a buy or a sell. It is a crucially important piece of information that many systems obscure. Beyond very basic failings such as that, the frequently repeated error is to cram too much into too little space with insufficient thought for the order of entry. Figure 3 shows a simple example of a basic, workable screen form, but for forms to be really elegant, the logical flow of information needs to align with the thinking of the user. The classic example of an order that could be called across from investment to dealing desk, "Buy 10,000 shares in IBM" is so rarely implemented like that on screen. To read the average order management screen you would be forgiven for assuming that the portfolio manager would shout "10000 ordinary shares on the New York Stock Exchange with no limit and no conditions, at better than the average price for the day, buy from whichever broker you choose." All that padding just makes it easier for an error to slip through.

Of course, there are circumstances when users require large amounts of information, but they are far rarer than occurrences of users asking for more than they really need.

In an ideal world, users are presented with fields or data items in the precise order that is needed, within the appropriate context; and this points at the crux of the development issues around front office interface design.

To get truly effective design, the input needs to come from a development team whose insight into investment processes is greater than that of a single user-representative. Ideally, system design should come from a team that has a vision of what is possible beyond the immediate requirements of a single reference user group. When that occurs, the resulting system exceeds the expectations of the user base and offers leadership in how they can operate safely, quickly and effectively.

Editors Information

Andy Charman is an Independent Consultant operating through Cumulus White Ltd. He has been working with Buy-Side front office systems for past ten years. He worked on the design of BidRoute and is currently designing a middle office system for processing OTC Derivatives, among other projects.

BidRoute (www.BidRoute.com) recently won Most Innovative Project of the Year at the ICFA Global Awards 2008.

Figure 1

Figure 2

Figure 3