December 2015: Agile: Beneficial But Not Without Its Pitfalls

Like many aspects of the capital markets, agile software development means different things to different organizations, depending on where they sit on the agile continuum. A large investment bank, for example, might dabble in agile for non-mission-critical, peripheral development, but consider the methodology largely suitable only to small, nimble organizations with modest IT budgets, embraced mostly by necessity as opposed to choice. For small organizations, however, agile might constitute the optimal development methodology for their needs, allowing them to tweak software already in a production environment with the minimum of disruption along its evolutionary path.      

Barney Dalton, CTO of London-based systematic hedge fund Aspect Capital, and the subject of this month’s cover story on page 20, is a big agile proponent. “We operate on much faster cycles, so the releases we put out are much smaller, which might represent anything from hours to two weeks’ work,” Dalton explained in the interview he did recently with John Brazier. “You get much better visibility of what is in that release; you can understand all of the moving parts,” he said. 

For a buy-side firm like Aspect, which places a premium on its largely proprietary technology stack, and which it constantly tinkers with in pursuit of optimization, agile makes perfect sense: There is no finish line, projects can be simplified down to a slither of the size typically associated with the waterfall model, and goals and directions can be set and reset as often as deemed necessary. 

Yes, agile, through its scrums and sprints, has the potential to deliver significant benefits to capital markets firms, and thus has appreciable numbers of advocates. But it also has its shortcomings, chief among which is its tendency to lead to scope creep in the absence of a well-organized, disciplined program. Given that agile deadlines are soft at best and that one of the central tenets underpinning it is the ability to adapt to changing circumstances or requirements as projects are rolled out, there is a very real danger of not only losing sight of the original objective, but that there is no well-defined objective whatsoever, rendering the discrete bits of development being regularly churned out to something resembling an assembly line. 

Clearly, agile is not for everyone, but it does deliver in areas where the waterfall method has its Achilles heel: regular, possibly even daily-delivered development, allowing end-users to benefit far earlier in the cycle than they ordinarily would, illustrating the tangible business gains to firms’ upper echelons. And that, from a CIO’s perspective, is invaluable.