Citi Prime Finance’s 2011 IT Trends & Benchmarks Survey
I
21
In creating their native capabilities, managers launching in the
Hedge Fund 1.0 model built locally hosted data centers on their
own premises. These data centers were often established and
maintained by third-party network engineering and support
frms. These frms, such as Eze Castle Integration and Richard
Fleischman&Associates in theU.S., andMatsco in theU.K., would
charge time and materials for the network build-out. Support
would be bought in blocks, with frms pre-paying for an allotted
amount of support hours. Hardware would often be provisioned
through this service provider, with a cost mark-up applied, based
on the wholesale price received from the manufacturer.
Earlier-vintage managers would license software from a vendor
and then have that vendor work with their external network
engineering and support provider to install the software on their
local network. If the manager had hired external consultants
or internal developers to program proprietary software, the
resulting product would likewise be hosted within the manager’s
internal data center.
The fund’s data, therefore, would be housed within these locally
hosted internal systems. Disaster recovery typically entailed
tape backups, occasionally taken off site—often to the home of a
fund manager employee.
The creation of these locally hosted infrastructures required a
fairly substantial outlay of capital. Even if the manager opted to
lease rather than buy some of their equipment, they nonetheless
had to provision their data center with suffcient cooling and
power, build their own network connectivity and factor in the
cost of support. This was a barrier to entry for many frms and
helped give rise to the model whereby the majority of emerging
frms in the early 2000s looked to their prime broker to provide
technology to help them cover their foundational functions.
The high cost and scale of investment to build such platforms
also help explainwhymany of the hedge fund pioneers discussed
in Section 1 were found among the largest hedge funds of the
time. The costs of having a locally hosted infrastructure were
substantial and identifying opportunities that helped them
differentiate their fund through this investment helped to justify
and leverage the costs sunk into these platforms.
Proliferation of Data Centers Help Drive
Hybrid 2.0 Model
The explosion of the Internet and Web technologies in the
early part of the last decade helped drive changes in the
options available to managers launching by the mid-2000s.
As bandwidth increased and became cheaper, a new business
model emerged. Third parties would build data centers where a
number of tenants could place their hardware, take advantage
of shared lines and reduce their costs in terms of ensuring
power and back-up capabilities.
Hedge Fund’s Data
Held Off Premises in
Provider’s Data
Centers
Housed in Off-Premises
Data Center
Housed in Off-Premises
Data Center
“Software as a Service”
Hosted on Manager’s
Selected Infrastructure
Provider
“Infrastructure as a
Service” with Hardware
Maintained Off Premises
HEDGE FUND 3.0
Hedge Fund’s Data
Held Off Premises in
Provider’s Data
Centers
Housed in Off-Premises
Data Center
Purchased or Built
Onsite
Vendor
Hosted
"Software as a Service"
Off
Premises
Data Center
On Site in Systems
or Data
Center
Installed on Site --
Managed by
Internal Teams
HEDGE FUND 2.0
Hedge Fund’s Data Held
Off Premises in Provider’s
Data Centers
Installed On Site--
Created & Maintained by
External Providers
Purchased &
Installed or Built
On Site
Maintained On Site in
Systems or In Personal
Data Center
Primarily Accommodated
by Onsite Tape
Back Ups
HEDGE FUND 1.0
Hardware
Software
Fund Data
Disaster
Recovery
PB & HFA
Service
Providers
Proliferation of
Data Centers &
Cheaper
Bandwidth
Proliferation of
Managed
Service
Providers
Chart 12: Infrastructure Evolution