At the end of January, Descartes Underwriting announced the roll out of its first-ever cyber parametrics solution. It chose its home market France for the initial launch, but if everything goes well similar products could also start appearing elsewhere as well, including in APAC.
“We are certainly very excited with the launch [and] we will definitely look to replicate in other geographies once the French pilot is proven to be successful,” says Ben Qin, Head of North-East Asia for Descartes Underwriting.
Since its establishment in 2018, the managing general agent has made a name for itself around the world with the launch of a series of parametric solutions for dealing with complex nat cat risks.
This is an area of the insurance market that tends to have an abundance of well-understood data associated with it, making it a perfect candidate for applying parametric triggers.
Ben Qin
Head of North-East Asia at Descartes Underwriting.The same is not true of cyber cover, where a myriad of data providers and securities firms mean that data is often patchy and poorly understood.
“The basic idea is to tackle the same insurability challenge as we have done successfully on climate risks , but for emerging risk like cyber,” says Qin. “Via use of sophisticated technology and AI algorithms, we are trying to provide a more cost-effective and sustainable solution on this risk topic.”
Qin says that “Apac is [one of the] smallest but fastest growing cyber insurance markets” in the world, making parametrics solutions particularly relevant to the region.
Singapore is in the process of implementing an amendment to its existing cyber law, which could create further impetus for new parametric solutions.
Benefits and challenges
The benefits of parametric solutions, within both the fields of nat cat and cyber cover, are clear.
They offer greater certainty for the insured in terms of the money that will be paid out when things go wrong. They also allow for swifter payouts than more traditional solutions.
However, realising the full potential of parametric solutions requires the right trigger mechanism to be put in place.
“A good parametric solution is one in which neither the buyer nor the seller can have any influence over the source of information,” says Henri Winand, chief executive officer of AkinovA, an electronic marketplace for trading digital risk.
In 2020, AkinovA facilitated the purchase of a parametric cloud outage product for an Asia-based technology company, working closely with Guy Carpenter who represented the buyer.
Henri Winand
Chief Executive Officer at AkinovABuilding robust and immutable indices requires having access to good and reliable data – something that may not always be possible in certain APAC countries that do not have such a long history of cyber cover.
“Given the fact that cyber insurance has been a common purchase for many businesses in the US and segments of Europe, combined with the regulatory landscape of these territories, loss data is more readily available for insurers to make determinations in these markets,” says Struan Todd, co-founder and managing director of Pandamatics Underwriting. “Although the cyber insurance markets across Asia are growing, the same volume and quality of local loss data is not available.”
Todd says that many insurers active in the region still base their assumptions on their experiences elsewhere, which can create a basis risk. Pandametics Underwriting is currently working with technology partners to develop a more localised pricing methodology.
Struan Todd
Co-Founder and Managing Director at Pandamatics UnderwritingIt’s not nat cat
However, cyber exposure is not the same as nat cat risk, which means that the data problem is not just about the availability of data. It is about the interpretation of that data.
“For Cyber parametric, (Re)insurers don’t necessarily need decades of years of data like you do in nat cat,” says Alex Mican, Director of Property Claim Services (PCS) at Verisk.
“Cyber parametric deals do not have to just be done on a loss basis. They can be done on other factors or indices that can influence a computer system. This all comes down to how and what data is being considered.”
Verisk’s PCS unit provides third-party industry loss estimates for a range of risk events, including cyber.
Alex Mican
Director of Property Claim Services (PCS) at VeriskMican says that shifting the focus away from pure loss data requires a more targeted approach to cyber parametrics. This is why so many cyber parametric providers have chosen to look at cloud outages and business interruption in the first instance.
Both the Descartes Underwriting product launched in France at the start of the year and the solution that AkinovA facilitated in 2020 offered compensation for business interruption as a result of a cyber attack.
“Cloud outage is one of those areas where historical data is not as critical. What matters is the understanding what qualifies as an out-of-ordinary event,” says Mican. “On that basis a cedent can determine the most appropriate threshold for triggering. In other words, how long does the system have to be down before compensation is paid?”
However, given the wildly different ways in which cyber data can be analysed and interpreted, constructing a robust and immutable data index can be challenging.
“The nature of the perils is completely different, the fact that cyber is a man-made peril brings a completely different set of considerations,” says Juan Marcano, Principal Client Account Manager at CyberCube.
Juan Marcano
Principal Client Account Manager at CyberCubeWinand says that, with nat cat, the natural environment – be that the weather or the shifting of tectonic plates – provides “a clear and certain” source of information.
“It’s very difficult for human beings to go in and change the data from an earthquake or a landslide,” he says.
This is not the case with cyber.
“With cyber, if you are close to the source – the cloud infrastructure provider, for example – then it is possible that you end up creating an index that can be changed by the buyer. This is why having an independent umpire is so important,” says Winand.
It is true that a lack of robust data in APAC continually causes a problem for the development of many markets, but in the case of the kinds of parametric solutions that insurance clients want in the region – chiefly to guard against business interruption – the absence of data isn’t as big a problem as a failure to properly understand that data.
Henri Winand
Chief Executive Officer at AkinovA“The availability of data within these sectors and within these companies is improving and catching up to global markets,” says Mican. “Because of technology and its daily usage, and how a lot of these countries are associated in the area, you have an abundance of data now. It’s all about sifting through that data to determine what is needed for properly assessing your risk.”
Alex Mican
Director of Property Claim Services (PCS) at VeriskBut, as Todd from Pandamatics Underwriting points out, constantly-evolving technologies and attack types means that curating “local and relevant data” is not always straight-forward.
“With new solutions that risk managers have limited exposure to such as parametric insurance, adoption can take a long time,” he says.
