Building a successful technology enabled MGA

Twitter icon
Facebook icon
LinkedIn icon
Podcast

The Managing General Agent (MGA) business model, by which companies can underwrite risks without retaining capital, has proved popular among insurtechs looking to use new technology to improve underwriting.

So how do you set up a successful MGA? Matthew was joined by leading figures from tech-enabled MGAs when he hosted a panel at the 2021 Virtual KBW European Financials & Insurtech Conference.

The panelists were Elizabeth Jenkin (Chief Commercial Officer at Nimbla), Graham Elliott (CEO of Azur), Alastair Speare-Cole (President of Insurance at QOMPLX) and Jonathan Spry (CEO of Envelop Risk).

Talking points include:

  • Why to set up an MGA
  • How to make better use of data
  • Innovating vs investing in innovators
  • Data sharing
  • The future of MGAs

If you like what you're hearing, please leave us a review on whichever platform you use, or contact Matthew Grant on LinkedIn

Sign up to our newsletter for a fresh view on the world every Wednesday morning.

Continuing Professional Development - Learning Objectives

InsTech London is accredited by The Chartered Insurance Institute (CII). By listening to an InsTech London podcast, or reading the accompanying transcript, you can claim up to 0.5 CPD hours towards the CII member CPD scheme.

  • Claim 0.5 hours for listening to Episode 142 of the InsTech London Podcast

Building a successful technology enabled MGA - Episode 142 Highlights

Matthew: Graham, what made you decide to build a tech-enabled MGA?

Graham: I first ran an MGA in 2012, RK Harrison. I loved the vehicle, its flexibility and the fact that it was capital-light. But in an industry that should be all about data, I was shocked by the technology.

The MGA I was running was on a legacy stack that was fifteen years old. If we wrote more than one type of insurance with a customer, we had no idea what we were doing with them systematically.

There was an opportunity to solve that problem with data. That required grasping the nettle of technology and becoming tech-enabled. This was a $4.3 trillion industry that was driving down the road at 90 miles an hour only using the rear-view mirror, with bordereaux that could be 60 days out of date.

The environment in insurance was changing to make product and service increasingly important. The user experience outside of the risk transfer itself was shoddy. MGAs were not serving their end insureds, brokers and capital providers well.

Talent will want to work in modern companies on modern technology stacks, not in companies with 30-year-old technology.

Matthew: What are you doing at Azur?

Graham: Azur started in high-net-worth personal lines insurance and now writes business in the UK and Europe. We have been asked by some insurers to build products for them.

We have deployments as a tech company in admitted and non-admitted lines across the US [1] in multiple classes and we are building a global OEM (original equipment manufacturer) product launching in France in the summer.

It is all done on a modern technology stack in the cloud, using what has already been built, where possible, to minimise costs.

Matthew: Alastair, QOMPLX now has a valuation of over a billion dollars. What is your strategy?

Alastair: QOMPLX’s thesis is that there needs to be a more interoperable approach to data extraction, normalisation, cleaning, semantification, storage, process management flows and analytics tools.

The company founders came from a cyber risk management background and they recognised that some of the fundamental problems around cyber were really data problems.

The immediate use cases for the company they would build were in cyber risk management and insurance.

We could have detected the SolarWinds attack and removed it from systems before it did much damage. And we have built an automated underwriting platform for various data modelling schemes like Oasis LMF.

Matthew: Jonathan, you’ve described Envelop as “the world’s most reluctant MGA.” Why did you set up an MGA?

Jonathan: We set up Envelop to improve the way that underwriting decisions were made and overcome heuristic underwriting. We choose to address complex risks which cannot be completely automated.

We wanted to bet on ourselves. We saw that the world is full of pricing signals, so we wanted to develop proprietary technology to observe those pricing signals and then trade on them.

We were not keen on developing technology for anyone else or licensing in technology from anyone else. We wanted to generate output by having proprietary toolkits and better data, and, more importantly, people that can actually use that data.

The MGA model is a nice way to get flexibility and build alignment with capital providers. We believe our role is capital allocation. So, we need to be heavily incentivised to have skin in the game.

An MGA takes us about halfway to what we would like, which is an aligned underwriting platform. That would give us independence to make decisions and develop proprietary technology in a data-driven environment, something that insurers and reinsurers struggle to develop in-house.

We are prepared to live or die by the underwriting decisions we make and have as much alignment as possible with the results of our capital allocation.

We will exceed 100 million in premium this year. The challenge now is to continue to embed that technology at scale.

Matthew: Elizabeth, Nimbla offers invoice insurance for small and medium enterprises. Why did Nimbla go down the MGA route?

Elizabeth: We are selling a type of insurance that people haven’t previously been buying. Only 4% of small and medium enterprises currently buy trade credit, because they are usually priced out of the market.

Because we do tactical invoice insurance where businesses can buy insurance for one or even 100 invoices, our clients are able to make dynamic decisions about what they want to insure. We can quote and bind in under three minutes, where, in the traditional market, it takes days if not weeks.

By getting real-time data from the cloud accounting systems, we make intelligent real-time algorithmic credit decisions for those businesses, who can decide whether they want to make the trade and then insure it at the click of a button. New startups who use cloud accounting don’t have any fear about plugging that into us.

To be profitable in this space, companies need to educate their customers digitally. Bigger insurance companies and brokers did not want to make that investment and Nimbla did.

Having worked in big companies, I was frustrated by the amount of data and customer information that big companies don’t use to set their strategy.

Matthew: Why don’t insurance companies use that data?

Elizabeth: The big companies use legacy systems. It is difficult to extract the data or retrieve it from different places. When I worked at Aon, there were discussions about the reasons why we couldn’t do it. But I like to consider the opportunities.

We are a technology-led company, not an MGA-led company. So we want bigger companies to see how Nimbla’s technology could make their customers stickier and make them more profitable.

Matthew: Graham, how would you advise large companies to see the value of data?

Graham: Everybody understands how important data is to making decisions, and what a good user experience looks like at the front end.

But it’s not the toolkit that matters, it’s the mindset. Large companies can solve this problem, but they must do a few things.

Firstly, capital can only write what it has an appetite for. Companies cannot produce a product that copes with every customer need, because they may not have the appetite to write it.

Secondly, where companies have multiple systems and data all over the place, they have an orgy of rekeying. This produces data verification and validation issues that slow them down and put them in a batch process world rather than a live world.

Thirdly, on the underwriting side in the traditional insurance industry, the underwriter rules the roost and the tech people crawl around under the desks and make their systems work. Tech people need to be given equality of primacy with underwriters. That is hugely culturally difficult.

If companies can solve those issues—getting their appetite right, sorting out core legacy problems and giving technology people equality with underwriters—they have solved the problem. But it is not easy to do.

Matthew: Because it is so difficult to do those things, should incumbents give up on innovating, and invest in tech-enabled MGAs instead?

Graham: Companies have to solve this problem because in 20 years’ time, there won’t be insurance companies and insurtech companies. They will either be tech-enabled insurance companies or they won’t exist.

Partnership is one way to solve the problem. The opportunity for agile players like us is to lead that conversation.

But the CEO has to understand what multi-tenant cloud computing is. If they don’t, or can’t articulate it, how are they going to make the right decision on tech stacks?

Multi-tenant cloud computing involves using archetypes that exist already and configuring where possible.

For example, Facebook has one version of the code and 2.7 billion users, and each person has their own individual Facebook page. It means going from data pools to individual one-on-one relationships with millions of people.

The only way for companies to do that is not to build their own customer relationship management system but to use someone else’s customer relationship management system. They can break out to code when they need to get more complex but otherwise stick within the guidelines.

Alastair: There is a cultural challenge. In a big company, there are decision-makers who don’t understand the technology but have internal champions for their own solutions.

There are conflicting users: the underwriters, the people doing capital modelling, those who do aggregate control and those who do activities for portfolio optimisation. All these people need the same data to do different things.

Big companies have to get the data out of their different systems without sinking it into a data lake. They need to organise the data as it comes out of systems, track and register it, put the data into a multi-user environment and augment it with external data.

Matthew: Jonathan, how does Envelop use data for underwriting?

Jonathan: Data is a challenge. So we do not underwrite deals unless we have real visibility on the underlying risk. That requires trust from our clients, so we handle their data in an environment of integrity.

We do not pass on data to any of our carriers. We have a data science team that handles data and we don’t pass it around the market. We want to be known as custodians of data.

We also reward our clients for giving us data by giving them insight into their risk.

Matthew: Elizabeth, what is your perspective on data sharing?

Elizabeth: I think we can move faster with partnerships. MGAs and technology represent an opportunity rather than a threat. Everybody can win if we apply technology.

We can get the funding out of the door faster, get the capital allocation in real-time and the broker gets a stickier customer.

Matthew: Graham, what is the direction of travel for MGAs? Will they evolve into insurance companies, be swallowed up by insurance companies or carry on as MGAs forever?

Graham: I think MGAs are here to stay. There will be some massive global MGAs. It will be a choice for the management team and shareholders if they want to sell out before that happens or become a carrier. But there is a space at the top table for this vehicle. It is not just a soft market phenomenon.

 

[1] 'Admitted lines' refers to the mainstream regulated US insurance companies that are required to have their insurance rates approved by the state regulators. ‘Non-admitted’ business is often the type of complicated or non-standard cover that mainstream insurers won’t underwrite and commonly gets picked up by the MGA market.