Improving data capture through collaboration
Adrian Rands, Chairman, Quantemplate
When examining differences between the insurance technology arena and wider financial services, you need only look at their origins to begin to understand why the two are fundamentally different beasts.

In the late 1970s and early 1980s, financial services companies such as banks and hedge funds started investing in teams of internally hired programmers to create and manage vast databases. With the information accumulated, they applied innovative algorithms that could analyse data from the capital markets and give their employers an edge over the competition. Over time, many of these teams split away from the banks, either as founders of their own companies or subsidiaries to service the capital markets.

The nature of these providers' origins led to collaboration and internal investment between the technology and financial services industries. This cemented their place as an important fixture in the financial services landscape and an inherent part of the culture.

In contrast, catastrophe models, arguably the first complex fintech systems endorsed by the insurance markets, originated from academia, not the industry itself. Similarly, businesses like Xchanging and many of the London market software companies have been service providers to the insurance industry from the point of origin, rather than spin-offs from internal teams or subsidiaries.

A culture of collaboration between fintech firms and the companies they serve is critical to the successful development and integration of standardised, reliable data

A culture of collaboration between fintech firms and the companies they serve is critical to the successful development and integration of standardised, reliable data and there is no better way to guarantee a vested interest than investment. Bloomberg was financially backed by Merrill Lynch, which fostered an environment for iterative product evolution between the two parties for mutual benefit. Similarly, Markit, a younger financial services information company, was born by demand from the banking sector, which wanted more in-depth analysis on the less vanilla segments of the capital markets.


A consortium of the leading investment banks provided Markit's start-up capital and additionally contributed their proprietary data on over-the-counter derivatives and pockets of more unusual asset classes Bloomberg and Reuters were not addressing. The closest the insurance world has come to this type of collaboration is with Lloyd's- and London Market Group-driven technology initiatives such as Kinnect or Placing Platform Limited.

The fundamental difference between technology ventures such as Kinnect and Markit is the insurance initiatives have not been motivated by direct gains on investment, but rather funded through a levy, by members of the market for the collective benefit. Although this altruistic approach may be desirable to many, history has favoured businesses with a focus on profitable performance, particularly in the financial markets. This further demonstrates how the insurance market is far less developed than the wider financial services arena.

Without the central management of data,
there is no standard of data in the market.

However, while a culture of collaboration is important, so too is the central management of data, something the insurance markets are yet to develop. To process data, you need an algorithm. An algorithm in itself is simply a process or set of rules that incorporates logical arguments so there can be structured decision-making in the way it processes data. It is this algorithm and the data that make up the two core elements to creating well-managed, central data, and both rely on each other if software is to bring value to an end user. Without the central management of data, there is no standard of data in the market. Without having a history of storing consistent and uniform data it has been very challenging for technologies to provide meaningful, useful services to insurers.


The setting of data standards has been fraught in the insurance industry for a number of reasons. When you examine data in the capital markets, say foreign exchange or commodities trading, data variables include bid price, buy price, volume, buyer, seller, instrument and maybe a couple of others factors. The data is very simple, consistent and fully homogenised. Similarly, with equities, if you look at the data that is reported for financial statements of underlying companies, you have close to 400 variables that make up a financial statement, seven or eight that make up the market price and another 100 or 200 qualitative factors such as directors, industry and so on for a software product to process. This is still very simple thanks to the historic, standardised data captured by the financial services industry that has led to each factor being very well defined.

Capital Markets
Data standards across industries

Insurance is a different animal and the issues go right back to the legal principles of the transaction. Capital market investments operate under caveat emptor contracts, which put the onus of discovery on the buyer and the questions that the capital markets ask about an investment have been standardised over time. The insurance industry's use ofuberrima fides agreements make the disclosure of all material information by a broker or assured mandatory. The materiality of information is subjective and creates inherently complex and unique data requirements that cannot be standardised. This is why, from a data perspective, the insurance industry is where the capital markets were in the 1970s (that is, without standards).

Intermediaries and brokers are now getting heavily involved in technology, both by developing it themselves and acquiring businesses that have technological capabilities. Many assume a key driver in Aon's acquisition of Benfield was to get hold of ReMetrica. Equally, Willis's purchase of Towers Watson will give it Igloo, which probably drove Towers Watson original purchase of EMB, Igloo's creator, in the first place. The creation of Aon Grip and ReMetrica demonstrate the increased interest in technology on the sell side of insurance to enhance services delivered to clients.

The underlying challenges and requirements between the capital markets and insurance are so different, there is little crossover. Many tech companies that have dominated the banking sector have later attempted to move into insurance, but most have failed because they have come from a world where there's very good, clean, instantly available data, to one where it is patchy, hidden and often does not exist at all. In turn, the lack of technology buy-in among insurance professionals, due to poor quality software, expense or complexity has left the once enthusiastic view of insurance tech slightly dampened.

The right technology is available to the insurance industry, it has simply not hit the mainstream yet. With ever-increasing levels of data maturity within the market, when the tech firms that are offering genuine, innovative, insurance-specific solutions come to the fore, cracking the challenges of insurance technology will be a victory worth shouting about.

Originally published in Insurance Day on 25 September 2015

The Federated
Data Model
Taming the complexity of
regulatory reporting
Insurance models
for the technology age
Understanding changing
distribution channels
Making each business
line count
Brokers as
risk consultants
Four steps to
data-driven reinsurance
Portfolio steering
in the soft cycle
What is
responsive reserving?
The price is tight,
but is it right?
Improving data capture
through collaboration