The adoption of cloud technologies by hedge funds is growing but in this ever-changing space the myriad of options and software available can be complex to navigate. Whether you opt for a public, private or hybrid solution, transfer all business data to a cloud or just the essential information, there are numerous questions for firms to answer. HFMTechnology spoke to Ray Bricknell, managing director at Behind Every Cloud and Eze Castle Integration’s director of service Simon Eyre, to explore the key issues surrounding cloud computing.
HFMTechnology (HFMT): What are the most popular uses of cloud infrastructure?
Ray Bricknell (RB): I think you need to split the market up into three main groups – start-ups, smaller more established firms and large firms. I would say that for start-ups it is almost a ubiquitous adoption of community cloud services. They are almost exclusively using cloud out of the box and many are even
adopting technologies like Office365, which is much more of a public cloud type of solution. The main uses for start-ups is to run their entire business because they have not usually got complex businesses, but it is obviously driven by budgets and what is most cost-effective for that business.
For more established but smaller (sub-$1bn) funds, I estimate that about 25% of all those firms have implemented a cloud or community cloud type model and that is definitely increasing. Typically, such a decision is driven by a major event, either a significant tech refresh cost or an office move or something
else which supports making a move to a cloud structure beforehand. Those organisations typically use the cloud for running the whole business and are sometimes (knowingly or not) sharing a community cloud with other, similarly sized, hedge funds.
With large firms we are currently seeing quite a number of them beginning to use public or shared cloud services for testing and development. The early movers in this part of the market are also starting to seriously explore the strategic implications and viability of a move of all production systems to a full infrastructure-as-a-service type platform – albeit with a very carefully chosen vendor who will meet all of their wide and varied (and somewhat unique) needs.
Simon Eyre (SE): Start-up hedge funds are taking advantage and going full on with cloud technology and building much of their core infrastructure on the cloud. In this day and age there is a certain amount of uncertainty as to where businesses are going to go and for smaller firms, cloud technology gives them a chance to scale up or down based on how the business evolves. Some firms that are going through a hardware cycle re-fresh may adopt cloud technologies and other times if a firm is deploying a new system – perhaps an OMS – they may not want to commit to buying more hardware or they don’t have the space, and so cloud technology can be really useful bolted on to a company’s existing infrastructure and can help advance that.
HFMT: What are the technology benefits / burdens of using cloud technology?
SE: When you are looking at hedge funds with AuM below $1bn, cloud technology can offer great functionality in terms of remote connectivity and the level of disaster recovery in place for all systems, and cloud platforms can help these smaller funds implement the technology and security that they previously weren’t able to afford.
Larger funds can benefit by offloading a lot of their in-house IT. A larger fund doesn’t necessarily need to rely on having its own IT staff maintaining and managing a full infrastructure and also key for them is the up-time. None of these businesses work on a 9-5pm business model, their expectations are 24/7 and a cloud can really offer great service continuity.
On the difficulties that cloud technology brings, you cannot just look at your cloud provider once and assume it has met all requirements and then step back from it. It is extremely important to make sure that the
business providing cloud solutions is providing a high-level product, and there is burden on a hedge fund to do the necessary due diligence and to continually monitor it.
The regulations in this business are changing quite rapidly right now, particularly around the use of hosting platforms and service providers, and you need to be sure that you are keeping up with them and so is your cloud provider.
RB: In our opinion, the technology benefits are enormously wide ranging. Simply put, only a very small number of very large funds in the UK market can afford to invest sufficiently to match the quality, resilience and support / service models that a tens of millions of pounds p.a. IT vendor can deliver.
Over time, you should expect to see an attrition of most organisations towards some sort of infrastructure-as-a-service platform. The main technology benefits are primarily around quality, resilience, availability and flexibility of the platform and the levels of knowledge, toolsets and process maturity to support the service and support models that come with it.
But the real benefits of cloud are not only technical. Delivered well, the cloud model offers significant cost, functional and service level benefits – especially if you properly compare the alternatives on a like-for-like quality basis. Things like cost levelling, scalability and flexibility both for capacity and correlated to costs, orchestration and dev/ops tools responsiveness all contribute to enabling the infrastructure to respond to “quick demand” business projects. The “institutional quality” that we look for in vendors, supported by external accreditations that internal teams cannot match, often make the comparison a pretty one sided exercise.
The FSA and other regulators are increasingly driving businesses to be able to demonstrate due diligence in the selection of any outsourcing provider (not just IT), and this can be a burden for some firms when compared to the “in-house” model. The other issue is that there is a tendency with the cloud model for the client to be forced to buy a “one-size-fits-all” solution – or otherwise risk breaking the cost model benefits, and so the challenge is to go out and find a vendor who meets all of your requirements ‘off the shelf’ rather than trying to make a vendor build something bespoke to meet your needs.
HFMT: What are the main technology-related questions that CTOs will pose to you about the cloud? In relation to security, what type of protection do hedge funds require a cloud to be surrounded by?
SE: A lot of the CTOs are driven by investor expectations, so a lot of what CTOs ask us are coming through on investor DDQs. A key question when a firm is looking to the cloud, centres on data ownership. The old school thinking that you are losing control of data and are no longer in control of your environment by moving to the cloud is most definitely not true but queries surrounding data security are still prevalent among CTOs.
Another important question that we get from CTOs is around exit strategy. If the cloud provider goes down or perhaps the service isn’t quite what they expected, how do you get off the cloud? How can we close up and move on? Depending on the structure that you are using, that can be quite a complicated process – some providers are very proprietary and getting that data out and usable in a format that you can transfer somewhere else is often quite complicated.
Security is always a major concern and not just relating to the cloud provider but also with regards security within the business itself, who has access to things? Who can get access to the data centre? What are the risk between intra-cloud security?
RB: Believe it or not, in our experience most CTOs and COOs really don’t understand fully yet what the term “cloud” really means to them. So the discussions we are having are centred on – “what’s available out there in the market place? And who are the best vendors to meet the needs of London’s
Cloud is often used in two main contexts which boil down to i) More traditional managed services outsourcing on either a dedicated or shared platform and ii) The full automated GUI driven orchestration concept for delivery of new services. The technology discussions tend to centre on security, DR/BCP, the underlying tech stack, data centre topology, network implications, higher value services such as SQL DBA, licensing, high availability designs and the implications of multi-tenancy.
A lot of conversation also revolves around what is out there in the marketplace today, who are the best vendors to meet the needs of small clients and what everybody else is doing.
The key attributes with regards security are predominantly good practices and processes, external accreditations like ISO27001, things like good quality staff vetting and regular external penetration testing, encryption, logical client separation and ensuring that physical and electronic access to the environment is restricted. A key aspect of good due diligence is complete transparency, and clients should ensure they engage
directly with key staff to assess the design and architecture and operational service management processes.
HFMTechnology (HFMT): When you speak to hedge funds, do they generally want to use a private, public, or hybrid cloud and what are their reasons for picking one of these three choices?
RB: Smaller organisations are much more cost-focused and tend to have less demanding technical needs. They also typically focus a little less on what they are doing from a security and availability perspective than larger firms – partly because their investors are in turn less demanding. They usually have less in-house expertise to support a very rigorous assessment and decision process, so usually end up just “following the herd”. Those organisations are even becoming more comfortable with using public cloud environments – primarily at this stage for email and office functionality. However, public cloud
solutions often do not provide a high level of recoverability, disaster recovery or data protection and this becomes a major issue as the fund size grows.
With bigger organisations, there is more inclination towards using dedicated or private environments. Such firms have (or can afford to buy-in) the technical prowess to really “get under the hood” of the solutions on offer, tend to have more demanding investors and pay more attention to regulatory compliance and operational risk, and ultimately have the budget to create or search for more tailored solutions. They also tend to have more legacy equipment and thus often need to identify more customisable or hybrid solutions.
SE: I think we will see more penetration of public cloud services in the future, but I think with the hedge fund industry specifically, people want to have a say on how things are carried out and so the use of private cloud software is often much more prevalent. We find that public platforms – Office365,
Dropbox, Google Docs – are generally not well established with regulators and it can be quite difficult for firms to know if they are going to be compliant by using them, but such offerings are still popular among start-up and smaller firms. The hybrid and private cloud solutions can offer a level of customisation and plans and procedures that are required by many in the industry, and ultimately that is what separates them from the more standardised public cloud offerings.
While the number of businesses that actually own cloud infrastructure is actually dropping off, there is only a handful of the very large, multi-national firms that are still operating private cloud that are set-up with their own infrastructure.
HFMTechnology (HFMT): Does employing cloud infrastructure reduce
costs at a firm? If so, what costs are impacted?
RB: Done properly, and by an IT service provider that has the financial reserves to really run a large scale platform at the required critical mass, cloud does offer enormous cost saving potential. We are seeing hard evidence in the market that the costs for procurement for the back-end infrastructure platform from a larger vendor are significantly less than if you were to build inhouse or buy those same services from a smaller vendor – and with the bonus of better service availability and resilience too.
As a general rule though, most firms in our niche at this stage are not looking at cloud as an outright cost saving exercise, they’re more looking at it as a mechanism for them to extract more value from their internal IT resources with greater efficiency, and as a way of being more responsive to the changing needs of the business within the same budgetary structure.
SE: Very often companies think that there is going to be very significant savings by going to the cloud over buying traditional hardware. I think particularly over the first three years the costs move around and you certainly don’t have the capital expenditure that you do when buying servers but it does take your operational costs and expense up quite significantly.
In the models that we draw up over three years there are some savings there but it is not like 50% or 25% saving on traditional hardware – it is not as drastic as that.
There also comes a lot of hidden costs with having your own infrastructure. Mayfair is notoriously expensive per square metre and if you are building a communications room just to house PCs and servers that is quite a significant outlay that you can do away with. Obviously when it comes around to a firm’s next hardware re-fresh – usually within three to five years – then the scale changes and the cost benefits of cloud technology become more tangible.
HFMTechnology (HFMT): How does the cloud function fit into disaster recovery and business continuity plans? Are firms today using cloud as an impromptu data warehouse?
SE: One of the first products that Eze Castle did as a shared platform offered disaster recovery to the masses. Offices having to rent space or maybe hot seats can incur significant costs if they sit cold for all but two tests a year and so from that aspect, the benefits to our solution were clear. Businesses can’t be down for three or four days anymore. If you have a disaster overnight, they need to be up the next day and adopting cloud technologies for company infrastructure can aid the continuity of a business.
RB: Again, if a cloud platform is properly built, high availability (i.e. no single points of failure), high spec DR RPOs and RTOs and good supporting BCP management, and even active-active multi-DC split load and clustering for an “always on” platform is baked into the design. Big banks have been building such systems for years – and a good cloud platform simply takes that really well built platform and slices the costs across lots and lots of smaller clients. No small firm can afford to match that with an in-house built (and I include small vendors in that group).
HFMTechnology (HFMT): What are the main functions of the cloud capability
on a daily basis?
RB: I would say the main functions of cloud technology today are to a) Provide an extremely reliable, stable, resilient infrastructure platform and b) Provide flexibility and scalability and
toolsets that support and automate the hugely dynamic workload that goes with software development.
SE: A cloud should almost be transparent to the end-user, so if you are using a cloud on a daily basis it mustn’t feel like you are using a cloud on a daily basis. It must act as good as if not better than traditional hardware and what it should also do is reduce your IT needs and help streamline the IT efficiency of the firm.
HFMTechnology (HFMT): Is it a mission critical function or
can firms operate without it?
SE: Yes I think it is. If all a firm’s core systems are on the cloud – communications, OMS, risk management systems – it is a huge amount of the business and whether or not you can operate if those systems are down, it is possible that you could get through, but the impact to your reputation and the cloud provider’s reputation is very significant.
It is mission critical because I think investors are looking at you to provide a stable business for their money and not put them through any undue risk and so the use of cloud technology is becoming vitally important across the industry.
RB: My view is that in three to five years cloud will be ubiquitous in our niche. There will be very few organisations in this market who will be able to substantiate continue a strategy to build and operate their own environment – and except in perhaps for algoythmic/systemic/low latency funds, it won’t differentiate them in any way if they do. Any delay in cloud adoption won’t be for functional, performance or cost reasons, they will primarily be for perception reasons. Like all such trends in our market, when investors get fully comfortable with the idea, adoption will accelerate.
It was only a few years ago that many funds didn’t outsource at all for the same reason. Today, I think that it is true that for start-ups and that 25% of small established funds their cloud platform is absolutely mission critical because that is where their entire business is running and I think what we will see is an increasing level of use of Cloud by larger funds for more and more critical services in the near future.