Tag Archives: policy admin

The End of Policy Admin Systems?

Will policy administration systems cease to exist as we know them? Insurance core systems have evolved from monolithic policy, billing and claims suites that were primarily for accounting functions to fully functional service offerings. Configuration capabilities were added next, reducing the size of the code base.

Solution providers are now aggressively working to offer headless (no user interface) core offerings with microservices and integration layers to address the legacy nature of the core systems. As a service-oriented architecture, microservices consist of independent, modular services that each run a unique process and communicate via a lightweight protocol. Microservices-based architectures allow for continuous delivery and deployment, and these offerings meet carriers’ needs for competitive differentiation of customer experiences, faster deployment of changes, integration with new analytics capabilities and deeper digitalization of processes.

What’s next? We see leading solution providers modernizing their systems by segmenting and extracting functionality from the remaining core components to create standalone services that can be hosted in the cloud and accessed directly from any platform. They will continue to dismantle these components and build microservices to enable more continuous change and flexibility. These services will leverage cloud capabilities and expand on demand to meet spikes in workload, then shrink to more appropriate deployment levels.

See also: Policy Administration: Ripe for Modernizing  

Orchestration of microservices for online interactions will eventually be controlled by the UI layer. Batch jobs have been modernized to background functions to increase availability. Will policy administration systems shrink into just the orchestration layer for bulk processes? Will they disappear entirely like a house built out of Lego blocks that has been dismantled? Will current solution providers adapt quickly enough, or will insurtech disrupt this market with core services built on more adapted architectural frameworks?

One thing is for sure: Carriers will need more sophisticated tools to monitor performance and manage microservices. By implementing the modern core systems of today, carriers can not only avoid needing to transform their existing systems, but can also focus on differentiation while outsourcing evolution to their solution providers. The key is to remember that the point of buying a system isn’t just for benefits today, it’s also for the value that comes tomorrow.

What’s in Store for Blockchain?

Blockchain, blockchain, blockchain! What does that mean for insurance? No one knows yet, but that doesn’t stop blockchain from being one of the hottest topics in the insurance industry right now. This week, I take a look at the direction this puck is heading.

Hype or reality?

Last September, the World Economic Forum published a report titled, Deep Shift – Technology Tipping Points and Societal Impact. The report is based on surveys with more than 800 executives and experts about new technologies and innovations. The point of the report is to identify deep shifts in society that result from new technologies. These include areas such as 3D printing, driverless cars, wearables and artificial intelligence.

I was drawn to shift No. 16, simply called “Bitcoin and the blockchain.” By 2025, 58% of these experts and executives believed we would hit the tipping point for Bitcoin and blockchain. This was defined as:

“10% of global gross domestic product will be stored on blockchain technology.”

To put that into context, the total worth of Bitcoin today in the blockchain is about 0.025% of today’s $80 trillion global GDP.

Also of interest, especially given that it looks like Tunisia will be the first country to issue a digital currency on a blockchain, shift No. 18 was called “Governments and the blockchain.” Here, almost three out of four in the survey group expected that “governments would collect tax via a blockchain by 2023.”

It’s a reality then!

It’s certainly looks that way. And $500 million of venture capital money in 2015 can’t be wrong, can it?

The prospect of a seismic shift on a par with the impact of the Internet is compelling. That explains all the attention, predictions and excitement about blockchain. But, if we use the evolution of the Internet as a benchmark, the development of blockchain today for commercial use is equivalent to the Internet in, say, the mid-1990s, at best.

The debates on Bitcoin, on whether private or public blockchains will be used, on Sybase vs Oracle (oops, wrong century) are yet to play out. The ability of the Bitcoin blockchain to scale to handle massive volumes at lightning speed remains unproven.

Now, just as it was in 1995, blockchain technology is at an embryonic stage. Still finding its way, it has yet to prove it is a viable, industrial-strength, large-scale technology capable of solving world hunger.

That is why I am going to focus on the use case for insurance rather than the technology itself. (For one explanation of how blockchain works, go to Wired.)

The smart insurance contract

This is getting the most attention right now. The notion of automating the insurance policy once it is written into a smart contract is compelling. The idea that it will pay out against the insurable event without the policyholder having to a make a claim or the insurer having to administer the claim has significant attractions.

First, the cost of claims processing simply goes away. Second, the opportunity for fraud largely goes away, too. (I hesitate here simply because it is theoretical and not yet proven.) Third, customer satisfaction must go up!

One example being used to illustrate how these might work came from the London Fintech Week Blockchain Hackathon last September. Here, a team called InsurETH built a flight insurance product over a weekend on the Ethereum platform.

The use case is simple. In the 12 months leading up to May 2015, there were 558,000 passengers who did not file claims for delayed or canceled flights in and out of the UK. In fact, fewer than 40% of passengers claimed money from their insurance policy.

InsurETH built a smart contract where the policy conditions were held on blockchain. Using the Oraclize service to connect the blockchain with the Internet, publicly available data is used to trigger the insurance policy.

In this case, a delayed flight is a matter of fact and public record. It does not rely on anyone’s judgement or individual assessment. It is what it is. If a delayed flight occurs, the smart contract gets triggered, and the payout is made, automatically and immediately, with no claims processing costs for the insurer and to the satisfaction of the customer.

Building on this example and applying it to motor, smart contracts offer a solution for insurers to control claims costs after an accident. A trigger that there has been an accident would come to the blockchain via the Internet from a smartphone app or a connected car. Insurers are always frustrated when customers go a more expensive route for repairs, recovery and car hire. So, with a smart contract, insurers could code the policy conditions to only pay out to the designated third parties (see related article by Sia Partners).

So long as the policy conditions are clear and unambiguous and the conditions for paying are objective, insurance can be written in a smart contract. When the conditions are undeniably reached, the smart contract pays. As blockchain startup SmartContract put it, “Any data feed trusted by a counterparty to release payment or simply complete an agreement can power a smart contract.”

To understand this better, I asked Joshua Davis, the technical architect and co-founder at blockchain p2p InsurTech Dynamis, to explain. He said:

“You need well-qualified oracle(s) to establish what ‘conditions’ exist in the real world and when they have been ‘undeniably reached.’  An oracle is a bridge between the blockchain and the current state of places, people and things in the real world.  Without qualified oracles, there can be no insurance that has any relation to the world that we live in.

“As far as oracles go, you can use either a single trusted oracle, who puts up a large escrow that is lost if they feed you misinformation, or many different oracles who don’t rely on the same POV [point of view] or data sources to verify that events occurred.

“In the future, social networks will be the cheapest and most used decentralized data feeds for various different insurance applications.  Our social networks will validate and verify our statements as lies or facts.  We need to be able to reliably contact a large enough segment of a claimant’s social network to obtain the truth.  If the insurance policy can monitor the publishing or notification of our current status to these participants and their responses accurately confirm it, then social networks will make for the cheapest, most reliable oracles for all types of future claims validation efforts.”

Is this simply too good to be true?

Personally, I don’t think it is. Of course, a smart contract doesn’t have to be on the blockchain to deliver this use case.

However, what the blockchain offers is trust. And it offers provenance. The blockchain provides an immutable record and audit trail of an agreement. The policyholder does not have to rely on the insurer’s decision to pay damages because the insurer has broken its promise to keep the client safe from harm. As the WEF report states, this is an “unbreakable escrow.” The insurer will pay before it even knows what happened.

There’s another reason for going with the blockchain: cybersecurity!

With the blockchain sitting outside the corporate firewall and being managed by many different and unconnected parties, the cyber criminal no longer has a single target to attack. As far as I’m aware, blockchain is immune to all of the conventional cyber threats that corporations are scared of.

What happens when you put blockchain and P2P insurance together?

In December, I published a two-part article on Peer 2 Peer Insurance (here are Part 1 and Part 2). When you put the P2P model together with the blockchain, this creates the potential for a near-autonomous, self-regulated insurance business model for managing policy and claims.

Last year, Joshua Davis wrote an interesting white paper called “Peer to Peer Insurance on the Ethereum Blockchain.” He presents the theory behind blockchain and the creation of decentralized autonomous organizations (DAO). These are corporate entities with no human employees.

The DAOs would be created for groups of policyholders, similar to the P2P group model with the likes of Guevara and Friendsurance. No single body or organization would control the DAO; it would be equally “controlled” by policyholders within each group. All premiums paid would create a pool of capital to pay claims.

And because this is a self-governing group with little or no overhead, any float at the end of the year would be distributed back among the policyholders. Arguably, this makes the DAO a non-profit organization and materially increases the capital reserve for claims costs.

The big question mark for this model is regulation. There still is no answer to who will maintain the blockchain code within each DAO when regulations change. But, what does seem a dead certainty is that someone, somewhere is figuring out how to solve this.

Blockchain offers the potential for new products and services in a P2P insurance model. It should also open insurance to new markets, especially those on or near the poverty line.

For now, we must watch to see what comes from the likes of Dynamis, which is using smart contracts to provide supplementary employment insurance cover on Ethereum.

Innovation will come from new players

It has been my belief for some time that, in the main, incumbent insurance firms will not be able to materially innovate from within. As with Fintech, the innovation that will radically change this industry will come from new entrants and start-up players, such as:

Dynamis

SmartContract

Rootstock

Everledger (see previous article on Daily Fintech)

Tradle

Ethereum Frontier

Codius (Ripple Labs) (update: Codius discontinued)

This is particularly true with blockchain in insurance. These new age pioneers are unencumbered by corporate process, finance committees, bureaucracy and organizational resistance to change.

Besides, the incumbent insurance CIOs have heard this all before. For decades, software vendors have promised nirvana with new policy administration, claims and product engines. So, why should they listen to the claims that blockchain is the panacea for their legacy IT issues? But,  that is a subject for another post … watch this space!

Getting Beyond the Policy Admin System

As SMA’s Karen Furtado wrote in last month’s blog post about core systems, “Now that the insurance industry recognizes modernization as an indispensable tool for remaining competitive, it is worthwhile to take a step back and look at the technical capabilities that insurers really need.” With underwriting, this requires a platform that extends beyond the policy administration system and makes optimal use of the expertise of the underwriters themselves.

Today’s environment is full of infinite possibilities for the future of underwriting. Advances in the electronic exchange of information have benefited the insurance industry in major ways. One example is apparent with the portals and exchanges that are making it easier for agents to submit business opportunities. Given the ease, more submissions are coming in the door. This increased workload coupled with new data sources for validation and verification leaves underwriters at a tipping point. With increased demand and increasingly more complex variables, they need a solution that gives them enhanced capabilities that extend beyond the same old way of doing things.

In today’s competitive market, the ability to issue a quote for every desired risk is critical. The power literally has shifted to the palm of the consumers’ hands, where they get instant gratification via their mobile devices. For some insurers, not being able to handle the volume of quotes that are being submitted to them means leaving significant money on the table.

Therefore, a modern policy admin system is necessary for its ability to automate the processes that are performed by the underwriting department. These systems automate the data capture, base rating and rules and final pricing, and they manage formulas and document production for all risks. They process transactions for new business, renewals, endorsements, cancellations, reinstatement, etc. But, for complex risks, the risk analyses and evaluations that are determined based on information about credit, hazards, financials and loss experience are made outside the policy admin system. Automation supporting these decision-making processes takes place outside the policy admin system. SMA research shows just 37% of the entire underwriting process is managed via the policy admin system.

Before that harsh reality sets in, realize that the modern underwriting platform is not, should not be and cannot be a standalone system. Nor is the modern policy admin system a standalone solution. Now, the two (underwriting platform and policy admin system) should be connected, with the ability to perform the complex functions mentioned above.

One of our SMA imperatives is: “Interconnect Intelligence for Underwriting.” Nothing in modern insurance can happen in isolation, in a traditional silo. Those days are over, but, fortunately, the technology is available to support current and future needs. The key is finding the right connection points, the right technology and the right fit for your organization. Today’s real-time, big-data, high-volume market dictates the same from your company’s system, and that is why modern support for underwriting requires more than just a policy admin system.

Why Implementations of Core Systems Fail

As an engineer (at least that’s what my university degree says), I must say I like to solve problems. Big, ugly, complex problems can a great challenge.

We all know what has been happening with insurers’ core systems over the past several years. To respond to the challenging needs for product agility, customer-centricity and operational effectiveness, insurance companies are moving toward new core systems and away from the constraints of their legacy systems. And there are oodles of problems to be solved. Product modeling and patterns, configurability and customization options, integration and connectivity, external data sources, testing automation…it’s a tasty list, my friend.

And yet…and yet.

Even if these complex problems are nicely solved, many insurance companies fail to achieve the anticipated returns with their new core systems.

Over the past years of these types of projects, when we at Wipfli analyze the root causes, we find that the following risks have not been properly managed or mitigated:

1. Expectation risk – Are we all looking for the same things?
2. Acceptance risk – What could prevent us from leveraging this investment?
3. Alignment risk – What could prevent us from achieving the value we expect?
4. Execution risk – Are we getting things done effectively and efficiently?
5. Solution risk – Will this solution deliver on its potential?
6. Resource risk – Have we accounted for the total investment required for success?

What’s most enlightening about these risks is that five of them are about people and not technology. Only solution risk encompasses technology. As the engineer once said, “This project would have been a roaring success except for the people!” Don’t be that guy….

The desired future state following an implementation is only achieved when individual contributors do their jobs differently.

So, yes, systems projects are about the people.