The economic case for cloud services is more complex than first meets the eye. Comparing any in-house supported technology to one delivered as a cloud service is inherently fraught with subjectivity; depending on your point-of-view and assumptions one could argue the business case for either side.
At first glance, cloud economics appear very compelling. Whether it’s a provider of infrastructure, platform, software or data – most cloud service providers can legitimately claim to offer many, if not all, of the following benefits:
- Economies of scale. Providing anything on a shared basis can help reduce the unit cost of service provision. For large infrastructure-as-a-service providers, economies-of-scale are not only derived from bulk purchasing (of assets like hardware, facilities, power etc), but also bypassing traditional vendors and building bespoke technology stacks from component manufacturers. For platform- and software-as-a-service providers, cloud delivery models can also streamline administrative overhead, allowing software updates and patches to be pushed out broadly rather than adopted on a piecemeal basis – also simplifying support by ensuring everyone is always on the latest version release.
- Increased utilisation of IT assets. Many businesses experience uneven patterns of demand for IT with resources required principally during the working day. That means any dedicated hardware can remain idle and underutilised for much of its productive life. Sharing a common pool of infrastructure can result in better utilisation by re-provisioning IT resources across geographies or consumers with different usage profiles.
- Specialisation. Cloud services enable greater concentration of skillsets, as those most familiar with a particular piece of technology (whether that be infrastructure, platform, software or data) take on greater responsibility for its management and support.
- Elastic, Pay-Per-Use pricing. The benefit of elastic, pay-per-use pricing models are typically mentioned in relation to start-up IT companies that need to scale up rapidly. However, when it comes to financial services – a mature, cyclical industry – elasticity can be a benefit in both directions. Being able to cut IT costs in line with revenues during a downturn could be just as important as scaling up in good times.
- Cost transparency. Cloud services are more transparent because their commercial models are based on modular or atomic provision of shared services. Whether you are charging for applications on a per user model, charging for infrastructure based on CPU cycles or charging for data and analytics on a per item basis, there is invariably clearer insight into the total cost of ownership (TCO) of service provision.
Although the arguments above seem extremely compelling, naysayers could probably come up with counterpoints to dispute many promised benefits. Some may argue that cloud service providers run sufficiently high margins, such that the benefits of economies-of-scale, increased utilisation and specialisation are not passed on to customers. Others may contest that elastic pricing models and increased hardware utilisation cannot always be realised—particularly when the use case calls for dedicated IT resources. Equally, one could claim that it is not possible to cut fixed IT costs in the short term (people, facilities, hardware and enterprise software licenses), so cloud services fail to replace legacy technologies and simply add to the bill.
Ultimately, the view of whether cloud services offer an economic benefit can depend on one’s assumptions – including scope, timelines and service level requirements. That said, there is one characteristic of cloud that is difficult to dispute.
Cost Transparency: A Clear Benefit
Ascribing accurate total cost of ownership (TCO) figures to different technologies hosted and managed in-house is not an easy task. Large complex institutions that have centralised the provision of IT across business lines may have already started down this path. By operating a show-back or charge-back model, costs can be allocated in a way that helps reflect the resources consumed by each line of business. But such models are always fraught with challenges – particularly because of the complexity in apportioning fixed costs.
Invariably, cloud services offer a clearer view of TCO because of the way their commercial models are structured. Being able to accurately budget what one spends makes for a more predictable environment. And that predictability and transparency is an asset in itself. No hidden costs if a server packs in. Or if your system administrator needs to take extended sick leave. Your service provider assumes those responsibilities. It’s almost like having a built-in insurance contract.
Even More Compelling: The Gravity of Data
When it comes to procuring data, analytics and data management technologies as-a-service, the long-term argument for cloud makes even more intuitive sense. The benefits of a multi-tenant environment can be magnified when it comes to storing large volumes of historic market data – with no clear advantage in each consumer storing their own copy. Similarly, overlaying EDM systems and processes over source data means value-added functions—such as validation, cleansing and the derivation of analytics—can all enjoy the same benefits of scale and transparency.
However, data carries a lot of gravity when it comes to the architecture of financial systems. Not only can the cost of data transfer be prohibitive, but speed of data transfer is often a key requirement – so applications will want to reside close to data sources.
As more exchanges and data vendors start offering points-of-presence within the cloud, the economics will become more compelling and we may eventually reach a true ‘tipping point’ where cloud is the only logical option.
In fact, this ‘tipping point’ may already have begun to occur. Since this blog was first published, both of the leading global data aggregators – Bloomberg and Refinitiv – have pushed forward with their cloud distribution strategies, completing global rollouts of their Points-of-Presence (PoPs) in the cloud. The next step in this evolution would be for exchanges and trading venues to host more of their infrastructure in the cloud. Given that there is a bid on the table for the London Stock Exchange Group to acquire Refinitiv and there are rumours that the Intercontinental Exchange could bid for Bloomberg, this evolution could be accelerated through industry consolidation.
We at Xenomorph were pioneers in offering cloud data management solutions (we launched our TimeScape platform on Microsoft Azure in 2013), but have seen much greater levels of pull from our clients to host our technology in the cloud. Earlier this year we successfully migrated our core platform onto Microsoft Azure on behalf of one of our global bank clients – Mizuho. Given the scope and mission critical nature of this implementation, the project rightly earned us the Best EDM Initiative at this year’s Inside Reference Data Awards ceremony.
Overall, we are seeing a clear acceleration in the uptake of cloud services over the last couple of years across the financial services industry. That said, there will invariably be hiccups along the way – one case in point being the recent breach effecting Capital One’s AWS infrastructure. Any major hiccup of this kind will heighten concerns relating to information security, as well as identify and access management. Ultimately, technical constraints in customising solutions, uncertainty around regulators’ attitude towards cloud and evolving data sovereignty and privacy legislation – can all still act as blockers. Yet one undisputable benefit remains: cloud certainly offers greater cost transparency.