T
Theo Hildyard
Guest
Imagine this scenario: you represent a firm doing business in capital markets whose business model depends on innovative and agile use of technology. You are being impacted by the recent regulatory initiatives requiring brokers-dealers to have better risk controls in place with respect to Automated Trading and HFT. These changes may have come from ESMA in Europe, SEBI in India, ASIC in Australia or any number of jurisdictions around the world – take your pick.
What is common to all of these initiatives is the need for real-time pre-trade risk management, and you need to build these additional controls into your system in such as way to not have the additional latency render your strategies useless and to preserve the flexibility that helps your firm maintain its edge. There are some critical choices ahead and each has an opportunity cost.
You could buy a vendor product to solve your problem— this would give you a fast time to market but it is potentially expensive and subject to the limitations of the vendor’s capabilities; not to mention a solution that is widely available on the market and therefore widely available to your competitors. You might also be concerned about the vendor’s ability to continue to meet your needs when the regulators announce their next big thing and it all hands to the pump again.
On the other hand, you could ask your capable IT team to build it in-house. But some of what is needed has a familiar, maybe even commodity, feel to it and if they are re-inventing the wheel when there’s a perfectly good wheel-maker at the ready, you may be taking them away from more strategic projects unnecessarily.
“Build or buy” is often posited as a simple dichotomy, but there is a third option that presents a continuum of choices—a sort of hybrid where you build critical and often customized features but integrate them with market-tested third-party solutions for more standard features.
Every firm in the financial industry faces a similar dilemma. There is no single right answer, but here is a high-level checklist to help you ask the right questions:
The argument to build:
The argument to buy:
The argument for a hybrid:
In order to resolve the original problem of how best to implement real-time pre-trade risk management controls, you’ll need to carefully evaluate your firm’s priorities, strengths and available resources.
For more information and discussion on this topic see our white paper “The Buy or Build Decision Continuum” at http://www.progress.com/buybuild
Continue reading...
What is common to all of these initiatives is the need for real-time pre-trade risk management, and you need to build these additional controls into your system in such as way to not have the additional latency render your strategies useless and to preserve the flexibility that helps your firm maintain its edge. There are some critical choices ahead and each has an opportunity cost.
You could buy a vendor product to solve your problem— this would give you a fast time to market but it is potentially expensive and subject to the limitations of the vendor’s capabilities; not to mention a solution that is widely available on the market and therefore widely available to your competitors. You might also be concerned about the vendor’s ability to continue to meet your needs when the regulators announce their next big thing and it all hands to the pump again.
On the other hand, you could ask your capable IT team to build it in-house. But some of what is needed has a familiar, maybe even commodity, feel to it and if they are re-inventing the wheel when there’s a perfectly good wheel-maker at the ready, you may be taking them away from more strategic projects unnecessarily.
“Build or buy” is often posited as a simple dichotomy, but there is a third option that presents a continuum of choices—a sort of hybrid where you build critical and often customized features but integrate them with market-tested third-party solutions for more standard features.
Every firm in the financial industry faces a similar dilemma. There is no single right answer, but here is a high-level checklist to help you ask the right questions:
The argument to build:
- From end to end, the system needs to contain strategically important functionality that must be kept close to the chest.
- Flexibility is more important than the go-live date. A fully-customized solution will always take longer, but might better accommodate important nuances unique to your business.
- You have the resources to both create and maintain the system over time.
The argument to buy:
- A well-regarded vendor service is readily available for a system that is outside of your internal team’s core competencies or key areas of focus.
- Performance is paramount. Some vendors invest heavily in quality control, especially for high performance systems. If something should go wrong, they will be highly motivated fix the problem for all of their clients.
The argument for a hybrid:
- You recognize that some of the functional aspects of your system contain precious intellectual property, but other functionality is commoditized, so vendor products could be leveraged. For example: market data handlers, FIX adapters CEP engines are all readily available, so it often makes more sense to buy them and assemble a system comprised of both vendor components and custom development.
- More change is coming and you can’t afford to blow your entire development budget crafting a solution for a single business need—you want flexible systems that can adapt and change quickly to respond to market shifts. Some vendor components will prove more flexible than in-house development, so the hybrid approach of buying components and building custom functionality around them may serve your needs more efficiently in the long run.
In order to resolve the original problem of how best to implement real-time pre-trade risk management controls, you’ll need to carefully evaluate your firm’s priorities, strengths and available resources.
For more information and discussion on this topic see our white paper “The Buy or Build Decision Continuum” at http://www.progress.com/buybuild
Continue reading...