The use of derivatives by hedge funds is well known. In a survey of 5,551 hedge funds from the Trading Advisor Selection System hedge fund database, 71% of the funds analysed used at least one type of derivative product from equity, fixed income, currency and commodities. Correspondingly, the use of third-party software for pricing and risk managing derivatives positions is not uncommon on the buy-side, although the tendency is to use these in conjunction with in-house systems.
While such an approach is perfectly sensible, there is still a massive opportunity waiting to be harnessed from the public cloud, enabling hedge funds to leverage industry-standard analytics and vast computing power, while paying a small fraction of the cost of a traditional in-house solution.
For many the cloud is still a nebulous technology concept. They are unaware that the cloud represents a shift in the economics of IT. At its heart the cloud is an anonymous set of compute and data storage resources. These can be in the form of a dedicated web server, as a long-running background process or a large relational database or big data store.
The compute resources can be scaled up to large numbers of machines handling a large job queue and so form the basis of massively parallel grid computation. The beauty of the cloud approach is that you get billed for what you use – it is a utility. Rather than investing in a grid of, say, 1,000 machines and then using them only when you need to run a large risk analysis, you only pay for when you run that analysis.
You can use the inherent elasticity of the cloud to reduce time-to-insight from an analysis, as using 2,000 machines for one hour costs exactly the same as using 1,000 machines for two hours. Cloud users find that access to this powerful set of capabilities is straightforward. You can be up and running in the time it takes you to set up an account – a matter of minutes, not months.
The personnel considerations are no less significant. On-premise grids require on-premise maintenance, with a corresponding head count. For a cloud solution, the whole cost of managing the machines is included in the commoditised price. So, no need to employ IT staff to help users with installation and upgrades – they receive that automatically.
So, should everything go on the cloud? Microsoft believes that the long-term steady state will be a hybrid of on-premise IT and public cloud, as demonstrated by the recent launch of Windows Server 2012, which provides integration with the Windows Azure public cloud.
This hybrid approach is used by Derivitec. Market data, whether in raw or derived form, is stored in the cloud and may be used directly for computation or streamed to the client side for analysis. Trade data, however, sits with the client. This works on the basis of the following example: if you trade an at-the-money call on the S&P 500, the notional of the option is never sent to the cloud, only the underlying option characteristics are sent.
Once the message is received by a worker, the results are calculated and sent back, where they are then combined with the notional sitting on the client’s machine to give the actual results for the trade. This is an accepted pattern now in hybrid risk management solutions – messages passed to the cloud from the client never contain any sensitive customer data.
The cloud allows Derivitec to provide full analytics for pricing the value and risks on a hedge fund trading book at a fraction of the cost of an in-house solution for a fraction of the effort. Derivitec takes advantage of the democratisation of access to scale and economies of scale that Windows Azure provides. This, in turn, allows hedge funds to focus on core capabilities.
As of July 21, 2010, the Dodd-Frank Act now imposes much more stringent compliance and risk management reporting requirements on hedge funds than before. Such requirements will present a significant burden on technology and infrastructure, and a corresponding cost to those persisting with an in-house approach.
Those hedge funds willing to embrace the cloud can, however, continue to concentrate on the profits from trading rather than the cost of technology. With the impact of regulation growing daily, not being in the cloud will soon no longer be an option.