QuanTemplate – reducing friction & democratizing data for the reinsurers

By Rick Huckstep

Reinsurance is a fascinating world with its unfathomable jargon, highly complex trading arrangements and a heritage going back nearly 300 years. ‘My word is my bond” remains the business ethos in this disruptive, digital age.

If you want to get a taste of this world, buy a €200 ticket and go to the annual industry meet-up in Monte Carlo. Held every September, this is called Rendez-Vous and this annual meeting of insurers, reinsurers and brokers signals the start of the renewal season for reinsurance treaties. During the day, the great and the good debate and discuss the important issues for the industry. But more importantly, they are all there to do business.

Then in the evening, they hold lavish cocktail parties in the grandest hotels in the Principality. Drinking champagne in the grounds of the Hermitage or the Fairmont, business and pleasure are combined for the reinsurance industry.

But seriously, reinsurance is a critical component in the insurance ecosystem, which simply wouldn’t function with out it. Simplistically, reinsurers (or “assumed insurers”) assume liability for insurance policies written by insurers (or “cedents’), which means that the reinsurer will cover the cost of claims should they be made. This enables primary insurers to transfer all or part of their risk to the reinsurer and in so doing, they are hedging their own exposure.

Reinsurance is a way for insurance companies to protect themselves from unexpected or catastrophic losses, such as in the case of a natural disaster. The earthquake and tsunami that hit Japan in 2011 is estimated by the World Bank to have cost $235bn.

But is not just all about catastrophic events, as reinsurance is applied to all forms of insurance in both life and non-life classes. Through the vehicle of reinsurance, an insurer can improve its balance sheet and capital adequacy, create capacity to take on more new business, limit and protect itself against losses and stabilize cash flow.

Because reinsurers sit behind the market, they are able to see information from across the network of insurance carriers. This enables them to have access to a large pool of data for assessing risks. It is ironic then, that this industry is relatively low tech and ripe for innovation, arguably with the exception of catastrophic reinsurance.

For mainstream reinsurance, there is still a great reliance on spreadsheets, gut instinct and basic aggregate management to price and assess risk. It is, therefore, no surprise, that InsuranceTech startup QuanTemplate are gaining the attention of some of the biggest names in the industry.

Earlier this week, I skyped with Adrian Rands, CEO and co-founder of QuanTemplate to find out why the reinsurance industry is so excited by these newcomers.QT-logo-image-email@2x

Adrian started with the back-story behind the business. Having dropped out of university, Adrian went straight into a job as a casualty reinsurance broker in Lloyd’s of London. Having got to grips with the role, Adrian set about building his own stochastic pricing models based on historic pricing performance.

(A stochastic model is a tool that considers the random nature of losses to enable better predictions.)

Adrian taught himself to program but he knew his limitations, so he contacted Marek Nelken to help him make his models run faster. Marek was an old friend with a computer science degree who was working for a hedge fund. And it was at this point that they started to talk about working together and building a tech business.

It was also at this point around 2009 that the markets collapsed. So the two friends started to play the markets and returned a 250% gain over 18 months that enabled them to bootstrap QuanTemplate!

Originally, they started to build a full suite of apps to support a reinsurance business, which Adrian described as “our Maximum Viable Product’. But they quickly learnt that to be successful as a startup, they needed to focus on a niche area first, and then build out from there. Taking the business advice given to them as they went through the Techstars accelerator, they set their sights on data analysis and dashboards.

By bringing together the capabilities of high frequency trading with their knowledge of the reinsurance market, they set out to reduce the inefficiency on in-house solutions to provide access to data.

Reinsurers typically have a data warehouse or business intelligence solution sitting over their numerous core business applications. These solutions work by building massive central reports from which extracts and subsets of data can be taken. However the limitations are that (a) these are one-size fits all, centralized reports which don’t always accommodate the subtle differences in exposure and quantification factors effecting each class of business, and (b) can take days to produce!

As Adrian was explaining the sheer scale of the improvement in processing times and flexibility that QuanTemplate has achieved, I challenged him with a “sounds too good to be true!” comment.

To which he replied, “well, let me show you”!

I watched online as Adrian opened up a report with 1,048,576 rows of data and endless data fields. He dragged and dropped fields into the report and in real time seconds, new pivot table after new pivot table was created.

The largest test QuanTemplate have done is to analyze 50 million rows of data in about 900 milliseconds.

This quite staggering performance is achieved through their own software. Running on AWS, QuanTemplate have created a solution that can be located at major geographies around the globe to suit any regulatory or corporate requirements for data location.

The data processing is performed by X-Stitch, a Scala built backend platform to create in-memory SQL datasets that has “unlimited scaling, drawing on the techniques of high frequency trading”.

The presentation of the data is performed by mafic, to provide “a high performance GUI framework for their tablet ready interface”.

Having seen QuanTemplate in action, it was no surprise when Adrian listed the large name insurers who have subscribed to their solution. The demands for significantly quicker analysis of data are not only being driven by the need for better risk forecasting, underwriting and pricing.

Solvency II has upped the ante to the point where in-house solutions are going to struggle to meet the demand for data analysis. Any hesitancy that existed in the market about external cloud based solutions is being abated. Insurers and reinsurers alike have little choice but to turn to innovative tech players like QuanTemplate to meet their business needs.

And yet, QuanTemplate is NOT a disruptive play! This solution does not replace the existing IT within the reinsurer’s estate; it sits on top of it. Nobody has to make painful decisions to write off the $millions invested in the corporate data warehouse. QuanTemplate just sits right on top of whatever is there today.

This is a key selling point that Adrian was proud to explain. The solution is intuitive and users are self-taught (QuanTemplate have made a conscious decision to make it a “no training system”). QuanTemplate avoid any prolonged buying process by making their solution a non-intrusive, try before you buy purchase.

Unlike many of the InsuranceTech startups focused on new ways to distribute insurance products, QuanTemplate are innovating the underwriting of these products and touching the very core of insurance.

Backed by VC fund Anthemis Group, QuanTemplate’s vision is the democratization of data. Where today, the cost of administering the reinsurance treaty is as much as 45% of the premium because it is manual and clunky, tomorrow, it won’t be…

…and I can say this because I’ve seen the future. I’ve seen QuanTemplate!