A federated data model can efficiently bring together disparate data across all delegated
authorities, and unify the information from multiple risk, exposure and claims bordereaux
into the ultimate, consistent data format.
Coverholders have long been a critical component of the London market’s business
strategy, but underwriting through delegated authority is not without challenges.
Perhaps the greatest is the acquisition of data. Under the Solvency II regime, detailed
exposure calculations are essential, but the longstanding bordereaux system is
struggling to provide sufficient granularity in a consumable format.
The London Market Group (LMG) has therefore made straight-through processing for
delegated authority underwriting a priority and has begun to work on a solution, but
so far its attempts have not provided the panacea desired. However, a solution exists
which can unlock this critical exposure information.
Enter the federated model
A federated data model can efficiently bring together disparate data across all
delegated authorities, and unify the information from multiple risk, exposure and
claims bordereaux into the ultimate, consistent data format. It can then be harnessed
for company-level or central analysis. The cost, learning curve and daily effort
involved are minimal, the data reconfiguration automated and the lag between original
input on primary systems and population of the central database measured in seconds.
Solutions adopting this model have already been designed and implemented. Aegon
Group is a €36bn ($40.16bn) multinational composite insurance business. Its group
reinsurance company underwrites quota-share and excess-of-loss treaties for group
subsidiaries, comprising numerous independent companies brought into the Aegon
fold over decades of organic and acquisitive growth.
Each subsidiary possesses a unique information system and structure employing its
own data standards. A federated model allows the company effortlessly to combine the
risk information generated from the full range of group companies underwriting life,
property/casualty, commercial and consumer covers.
Aegon’s structure – and its consequent information challenge – are remarkably similar
to that faced by London companies and syndicates, which source business from
diverse agencies. On a macro level, the challenge matches that faced by the London
Aegon, like its London parallels, had two choices: change the underlying systems of its
subsidiary companies (a costly, high-risk solution which would have taken years) or
centrally harmonise the disparate datasets. The selected alternative was achieved in
just six months, creating an invaluable additional central resource.
For London and Lloyd’s, at the market level, adopting any solution is more
complicated. Central bodies desire a central service, but have no concrete rights over
data, and cannot force action. Businesses within the market possess vastly different
resources, and pursue incompatible business strategies. The level of control they wish
to hand off may be entire or entirely limited. A federated data model provides the
solution. Data from divergent systems is imported and reconfigured – without
rekeying – according to the schema used by the data recipient.
In the case of the syndicates this schema is defined by their internal exposure data
warehouse. For Lloyd’s and the Prudential Regulation Authority the schemas are
defined by the class specific data schemas drafted by the Target Operating Model
In a federated data model the syndicate schema and TOM class schema may be one
and the same, but all importantly they do not need to be as the transform takes place at
the point of communication between the data provider and recipient. This is a highly
flexible and hence incredibly effective method for achieving convergence of data
formats as data is harmonised into the correct, usable format at each node in the data
network. It is fed into the centrally managed meta-database and updated in real time.
Delegated Authority Entity Relationships and movement of data
The web-based system is completely scalable and almost effortless to maintain. It is
secure, since access to individual data blocks is controlled. Perhaps most importantly,
it requires no-one to change their existing policy and claims administration systems.
At the core of the federated model is a high-performance, expandable, “private
cloud” architecture which provides a single system for the London market. It can
house as many tenants as demand requires. It delivers controlled but instant data
availability, rapid processing speeds and massive capacity. Users can converse with
the data, and among themselves within the closed system. To ensure all the data joins,
mapping and transforms are correct for both data sender and recipient.
Federated model, meet TOM
In a recent briefing entitled Delegated authority straight-through processing, the
LMG’s TOM initiative set out five objectives and five strategies to meet them:
What the federated model offers
To tackle the lack of consistency in processes and
data requirements across the London market
The model’s standard schema accomplishes this by
harmonising disparate data structures, without the
need for intervention in existing underwriting
To meet the growing need for more detailed and
timely information, fuelling improved
management and reporting
A federated model, incorporating straight-through
processing, accomplishes this through real-time
updating. As soon as data is entered into the source
system, it can be interrogated at any time by anyone
with access to the specific data in question, to
produce a tailored output
To streamline processes to drive out costs and
make it easier to work with London
The implementation and maintenance costs are low
as the communications and transformations are
conducted by the agent and underwriter who have
intimate knowledge of their data and systems
To improve the timeliness, quality, and
availability of management information
A federated model ensures all data is available from
inception in a single, understandable place and can
be manipulated to meet any end
To address the lack of a centralised London data
A centrally managed federated model is the answer,
but it meets the objectives without forcing market
companies to adopt new systems, sign up to costly
new physical infrastructure, or change the way they
TOM’s strategies to achieve these objectives are met by a federated model. TOM
seeks: one-touch data entry at the coverholders; consistent data requirements; more
frequent submissions from coverholders, removing the need for monthly bordereaux;
tools for market participants creating a single, central place in London to submit data,
with the templates and standard processes that allow companies to continue to use
their own systems; and central London data services to ensure that underwriters,
exposure and claims managers, and those responsible for tax and regulatory
reporting have access to data. The federated data model has it covered.
Implementing a federated data model will not be effortless but its collaborative
framework will enable rapid adoption. Each underwriter and bordereaux manager
will have the ability to transform the data to the recipient’s format with coverholder
and underwriter working together to this end.
This is the beauty of a crowdsourced solution and in this case harnessing the power of
the 3,941-strong Lloyd’s network of coverholders and syndicates with the goal of
reducing to zero any errors in the data translation process. This approach shares the
workload, and as “many hands make light work”, the harmonised dataset will be
reached fast, as is the way of internet-powered solutions. Some costs will be involved,
but they will be much lower than creating a new physical central infrastructure and
much less than the cost of imposing a new system on every risk carrier and
coverholder in the market.
The federated model is no Kinnect. Lloyd’s ill-fated electronic trading platform can be
summed up by its single fateful flaw: it did not make anyone’s job easier. A federated
data model makes life easier for everyone. The approach is tried and tested. It can be
implemented rapidly, deployed with relative ease and the inevitable wrinkles ironed
out very swiftly. This is the occasion to embrace innovation.